#459 – DeepSeek, China, OpenAI, NVIDIA, xAI, TSMC, Stargate, and AI Megaclusters

29:04We'll talk about it. I almost forgot to talk about the difference between deep seek v3 and r1 on the user experience side. So forget the technical stuff, forget all that. Just people that don't know anything about AI, they show up like what's the actual experience, what's the use case for each one when they actually like type and talk to it? What is each good at and that kind of thing? So let's start with deep secret three again. It's what more people would have tried something like it. You ask it a question. It'll start generating tokens very fast. And those tokens will look like a very human legible answer.

29:38It'll be some sort of markdown list. It might have formatting to help you draw to the core details in the answer. And it'll generate tens to hundreds of tokens. A token is normally a word for common words or a sub word part in a longer word. And it'll look like a very high -quality Reddit or Stack Overflow answer. These models are really getting good at doing these across a wide variety of domains. Even things that, if you're an expert, things that are close to the fringe of knowledge, they will still be fairly good at. I think cutting edge AI topics that I do research on, these models are capable for study aid and they're regularly updated.

30:22where this changes is with the deep seek R1, what is called these reasoning models is when you see tokens coming from these models to start, it will be a large chain of thought process. We'll get back to chain of thought in a second, which looks like a lot of tokens where the model is explaining the problem. The model will often break down the problem and be like, okay, they asked me for this. Let's break down the problem. I'm going to need to do this. And you'll see all of this generating from the model, it'll come very fast in most user experiences. These APIs are very fast, so you'll see a lot of tokens, a lot of words show up really fast.

30:56It'll keep flowing on the screen, and this is all the reasoning process. And then eventually the model will change its tone in R1, and it'll write the answer, where it summarizes its reading reasoning process and writes a similar answer to the first types of model. But in deepseek's case, which is part of why this was so popular, or even outside the AI community is that you can see how the language model is breaking down problems. And then you get this answer on a technical side, they train the model to do this specifically where they have a section which is reasoning and then it generates a special token which is probably hidden from the user most of the time which says, okay, I'm starting to answer.

31:33So the model is trained to do this two stage process on its own. If you use a similar model and say open AI, open AI's user interface is trying to summarize this process for you nicely by showing the sections that the model is doing, and it'll kind of click through, it'll say, breaking down the problem, making X calculation, cleaning the result, and then the answer will come for something like OpenAI. Maybe it's useful here to go through an example of the DeepSeq R1 reasoning. Yeah, so if you're looking at the screen here, well, you'll see as a screenshot of the DeepSeq chat app, And at the top is thought for 151 seconds with the drop down arrow.

32:15Underneath that, if we were in an app that we were running, the drop down arrow would have the reasoning. So in this case, the specific question, which, you know, I'm philosophically slash pot head and client. So this is asking DeepSekar one for one truly novel inside about humans. And it reveals the reasoning. and basically the truly novel aspect is was pushing the reasoning to constantly sort of the model asking itself is this truly novel. So it's actually challenging itself to be more novel, more counterintuitive, less cringe, I suppose. So some of the reasoning says, this is just snapshots.

32:55Alternatively, humans have a unique meta -emotion where they feel emotions about their own emotions, or G. Feeling guilty about being angry. This, of course, of emotional layering creates complex motivational drives that don't exist in other animals. The inside is that human emotions are nested. So it's like, it's reasoning through how humans feel emotions. It's reasoning about meta -emotions. You're going to have pages and pages of this. It's almost too much to actually read, but it's nice to skim as it's coming. It's a stream of ca - It's a James Joyce -like stream of consciousness. And then it goes, wait, the user wants something that's not seen anywhere else.

33:28Let me dig deeper. and consider the human ability to hold contradictory beliefs simultaneously, cognitive dissonance is known, but perhaps the function is to allow flexible adaptation so on and so forth. I mean that really captures the public imagination that holy shit, this isn't, I mean, intelligence, slash, almost like like an inkling of sentience because like you're thinking through yourself reflecting you're deliberating. And the final result of that after 157 seconds is humans instinctively converts selfish desires into cooperative systems by collectively pretending abstract rules, money, laws, rights are real.

34:12These shared hallucinations act as quote games or competition is secretly redirected to benefit the group, turning conflict into society's fuel. Pretty profound. I mean, you know, there's a common depression, but a lot of people have found that these reasoning models can sometimes produce much more eloquent text. That is a at least interesting example, I think, depending on how open -minded you are, you find language models interesting or not, and there's a spectrum there. Well, I mean, some of the, we'll talk about different benchmarks as well. Some is just a vibe like that in itself, is a, let's say, quote, fire tweet.

34:51Yeah. If I try to produce something, something where people are, oh, shit. Okay. So that's Chanathal will probably return to it more. How are they able to achieve such low cost on the training in the inference? Maybe you could talk the training first. Yeah. So there's, there's two main techniques that they implemented that are probably the majority of their efficiency. And then there's a lot of implementation details that maybe we'll gloss over or get into later, that sort of contribute to it. But those two main things are one is they went to a mixture of experts model, which we'll define in a second.

35:30And then the other thing is that they invented this new technique called MLA, late in attention. Both of these are big deals. Mixture of experts is something that's been in the literature for a handful of years. And OpenAI with GPT -4 was the first one to productize a mixture of experts model. And what this means is when you look at the common models around that most people have been able to interact with that are open, right? Think Lama. Lama is a dense model. I .e. every single parameter or neuron is activated as you're going through the model for every single token you generate, right? Now with a mixture of experts model, you don't do that, right?

36:08How does a human actually work? Right? is like, oh, well, my visual cortex is active when I'm thinking about vision to ask certain, like, you know, other things, right? My amygdala is when I'm scared, right? These different aspects of your brain are focused on different things. A mixture of experts' models attempts to approximate this to some extent. It's nowhere close to what a brain architecture is, but different portions of the model activate, right? You'll have a set number of experts in the model and a set number that are activated each time. And this dramatically reduces both your training and inference costs.

36:38Because now, if you think about the parameter count as the total embedding space for all of this knowledge that you're compressing down during training, when you're embedding this data in, instead of having to activate every single parameter, every single time you're training or running inference, now you can just activate a subset. The model will learn which expert to route to for different tasks. This is a humongous innovation in terms of, hey, I can continue to grow the total embedding space of parameters. And so deep six model is, you know, 600 something billion parameters, right? Relative to llama 405B, it's 405 billion parameters, right?

37:15Lama, relative to llama 70B, it's 70 billion parameters, right? So this model technically has more embedding space for information, right? To compress all of the world's knowledge that's on the internet down, but at the same time, it is only activating around 37 billion of the parameters. So only 37 billion of these parameters actually need to be computed every single time you're training data or inferencing data out of it. And so versus versus again the Lama model, 70 billion parameters must be activated or 405 billion parameters must be activated. So you've dramatically reduced your compute cost when you're doing training and inference with this mixture of experts architecture.

37:51So we break down where it actually applies and going to the transformer. Is that useful? Let's go. Let's go into the transformer. The transformer is a thing that is talks about a lot and we will not cover every detail. Essentially, the transformer is built on repeated blocks of this attention mechanism and then a traditional dense, fully connected, multi -layer perception, whatever word you want to use for your normal neural network. You alternate these blocks, there's other details, and where mixture of experts is applied is at this dense model. The dense model holds most of the weights if you count them in a transform model.

38:28So you can get really big gains from those make sure of experts on parameter efficiency, attaining an inference because you get this efficiency by not activating all of these parameters. We should also say that a transformer is a giant neural network. Yeah. And then there's four, 15 years now. This was called the deep learning revolution. Networks got larger and larger and a certain point. The scaling laws appeared where people realized this is a scaling law shirt, representing scaling laws where it became more and more formalized that bigger is better across multiple dimensions of what bigger means.

39:07These are all neural networks. We're talking about different architectures to construct these neural networks such that the training and the inference on them is super efficient. Every different type of model has a different scaling law for it, which is effectively for how much compute you put in, the architecture will get to different levels of performance at test tasks. And mixture of experts is one of the ones at training time, even if you don't consider the inference benefits, which are also big. At training time, your efficiency with your GPUs is dramatically improved by using this architecture if it is well implemented.

39:43So you can get effectively the same performance model in evaluation scores with numbers like 30 % less compute. I think there's going to be a wide variation depending on your implementation details and stuff, but it is just important to realize that this type of technical innovation is something that gives huge gains. And I expect most companies that are serving their models to move to this Mixer of Experts implementation. Historically, the reason why not everyone might do it is because it's an implementation complexity, especially when doing these big models. So this is one of the things as deep seek gets credit for is they do this extremely well.

40:20They do make sure of experts extremely well. This architecture for what is called deep seek M O E M O E is the shortened version of mixture of experts is multiple papers old. This part of their training infrastructure is not new to these models alone and same goes for what Dylan mentioned with multi -headly in attention. This is all about reducing memory usage during inference and same things during training by using some fancy low rank approximation math. If you get into the details with this latent attention, it's one of those things I look at and say, okay, they're doing really complex implementations because there's other parts of the language models such as embeddings that are used to extend the context length, the common one that DeepSeek uses rotary positional embeddings, which is called the rope.

41:07And if you want to use rope with a normal MOE, it's kind of a sequential thing. You take these, you take two of the attention matrices and you rotate them by a complex value rotation, which is a matrix multiplication with deep sea mla with this new attention architecture. They need to do some clever things because they're not set up the same and it just makes the implementation complexity much higher. So they're managing all of these things and these are probably the sort of things that opening eye these close labs are doing. We don't know if they're doing the exact same techniques, but they actually shared them with the world, which is really nice to feel like this is the cutting edge of efficient language model training.

41:43And some of this is, requires low -level engineering. Just is a giant mess and trickery. So as I understand it went below CUDA. So they go super low programming of GPUs. Effectively Nvidia builds this library called Nickel, right? In which, you know, when you're training a model, you have all these communications between every single layer of the model and you may have over a hundred layers. What does Nickel stand for? or it's NCCL. And video communications, collectives library. And nice. And so when you're training a model, you're going to have all these all reduces and all gathers, between each layer, between the multi -layer perceptron or feed -forward network and the attention mechanism, you'll have basically the model synchronized, or you'll have all reduced or an all -gather.

42:32And this is a communication between all the GPUs in the network, whether it's in training or inference. So NVIDIA has a standard library. This is one of the reasons why it's really difficult to use anyone else's hardware. For training is because no one's really built a standard communications library. And NVIDIA has done this at a sort of a higher level, right? A deep seek because they have certain limitations around the GPUs that they have access to. The interconnects are limited to some extent by the restrictions of the GPUs that were shipped into China legally, not the ones that are smuggled but legally shipped in, that they used to train this model.

43:05they had to figure out how to get efficiencies. And one of those things is that instead of just calling the Nvidia library, Nikol, they instead created, they scheduled their own communications, which some of the labs do. Mehta talked about in Lama 3 how they made their own custom version of Nikol. This is, they didn't talk about the implementation details. This is some of what they did, probably not as well as, maybe not as well as deepseek, because deepseek necessity is the mother of innovation. and they had to do this, whereas in the case, opening eye has people that do this sort of stuff and throppick, et cetera.

43:42But deep seek certainly did it publicly and they may have done it even better because they were gimped on a certain aspect of the chips that they have access to. And so they scheduled communications by scheduling specific SMs. SMs you could think of as the core on a GPU, right? So there's hundreds of cores, are there's a bit over 100 cores, SMs on a GPU and they were specifically scheduling, hey, which ones are running the model, which ones are doing all reduce, which one are doing all gather, right? And they would flip back and forth between them and this requires extremely low level programming.

44:16This is what Nickel does automatically or other Nvidia libraries handle this automatically, usually. Yeah, exactly. And so technically they're using PTX, which is like sort of like, you could think of it as like an assembly type language. It's not exactly that or instruction set, right? like coding directly to a similar instruction set. It's not exactly that, but that's still part of technically CUDA, but it's like, do I want to write in Python, you know, PyTorch equivalent and call in video libraries, do I want to go down to the C level, right? Or, you know, in code even lower level, or do I want to go all the way down to the assembler ISO level?

44:47And there are cases where you go all the way down there at the very big labs, but most companies just do not do that, right? Because it's a waste of time and the efficiency gains you get are not worth it, but deep -seek implementation is so complex, right? Especially with their mixture of experts, right? People have done mixture of experts, but they're generally 8, 16 experts, right? And they activate too. So, you know, one of the words that we like to use is like Sparsity Factor, right? Or usage, right? So you might have four, you know, one fourth of your model, activate, right? And that's what Mistral's, Mistral model, right?

45:21Their model that really catapulted them to like, oh my god, they're really, really good. Opening I has also had models that are M -O -E. And so I have all the other labs that are major closed. But what DeepSeek did that maybe only the leading labs have only just started recently doing is have such a high sparsity factor, right? It's not one fourth of the model, right? Two out of eight experts activating every time you go through the model, it's eight out of 256. And there's different implementations for mixture of experts where you can have some of these experts that are always activated, which this just looks like a small middle network.

45:56and then all the tokens go through that, and then they also go through some that are selected by this routing mechanism. And one of the innovations and deep seeks architecture is that they change the routing mechanism in mixture of expert models. There's something called an auxiliary loss, which effectively means during training, you want to make sure that all of these experts are used across the tasks that the model sees. Why there can be failures in mixture of experts is that when you're doing this training, the one objective is token prediction accuracy. And if you just let turning go with a mixture of expert model on your own, it can be that the model learns to only use a subset of the experts.

46:39And in the MOE literature, there's something called the auxiliary loss, which helps balance them. But if you think about the loss functions of deep learning, this even connects to the bitter lesson is that you want to have the minimum inductive bias in your model to let the model learn maximally. And this auxiliary loss, this balancing across experts, could be seen as intention with the prediction accuracy of the tokens. So we don't know the exact extent that the deep seek MOE change, which is instead of doing an auxiliary loss, they have an extra parameter in their routing, which after the batches, they update this parameter to make sure that the next batches all have a similar use of experts.

47:18And this type of change can be big, it can be small, but they add up over time. And this is the sort of thing that just points to them innovating. And I'm sure all the labs that are training big MOEs are looking at this sort of things, which is getting away from the auxiliary loss. Some of them might already use it. But you just keep accumulating gains. And we'll talk about the philosophy of training and how you organize these organizations. And a lot of it is just compounding small improvements over time in your data, in your architecture, in your post -training, and how they integrate with each other.

47:48DeepSeek does the same thing and some of them are shared. We have to take them on face value that they share their most important details. I mean, architecture and the weights are out there, so we're seeing what they're doing. And it adds up going back to sort of the like efficiency and complexity point, right? It's 32 versus 4, right, for like, mixed raw and other M .O .E. models that have been publicly released. So this ratio is extremely high and sort of what Nathan was getting at there was, when you have such a different level of sparsity, you can't just have every GPU. have the entire model, right?

48:19The model is too big, there's too much complexity there. So you have to split up the model with different types of parallelism, right? And so you might have different experts on different GPU nodes. But now what happens when a, you know, this set of data that you get, hey, all of it looks like this one way and all of it should route to one part of my, you know, model, right? So when all of it routes to one part of the model, then you can have the, you can have this overloading of a certain set of the GPU resources or a certain set of the GPUs and then the rest of the training network sits idle because all of the tokens are just routing to that.

48:55So this is the biggest complexity, one of the big complexities with running a very, you know, sparse mixture of experts model, i .e. you know, this 32 ratio versus this four ratio is that you end up with so many of the experts just sitting there idle. So how do I load balance between them? How do I schedule the communications between them. This is a lot of the like extremely low level detailed work that they figured out in the public first and potentially like second or third in the world and maybe even first in some cases. What lesson do you in the direction of the better lesson do you take from all of this?

49:29Where is this going to be the direction where a lot of the gain is going to be, which is this kind of low level optimization or is this a short term thing where the biggest gains will be more on the algorithmic high level side of like post training. Is this like a short term leap because they figured out like a hack because constraints, necessities, the mother of invention, or is there still a lot of gains? I think we should summarize what the bitter lesson actually is about. Is that true? The bitter lesson essentially, if you paraphrase it, is that the types of training that will will win out in deep learning as we go are those methods that are which are scalable in learning and search is what it calls out.

50:13And the scale word gets a lot of attention in this. The interpretation that I use is effectively to avoid adding in the human priors to your learning process. And if you read the original essay, this is what it talks about is how So researchers will try to come up with what clever solutions to their specific problem that might get them small gains in the short term while simply enabling these deep learning systems to work efficiently and for these bigger problems in the long term might be more likely to scale and continue to drive success. And therefore we were talking about relatively small implementation changes to the mixture of experts model.

50:59And therefore it's like, okay, like we will need a few more years to know if one of these are actually really crucial to the bitter lesson. But the bitter lesson is really this long -term arc of how simplicity can often win. And there's a lot of sayings in the industry, like the models just want to learn. You have to give them the simple loss landscape where you put compute through the model and they will learn and getting barriers out of the way. That's where the power, something like nickel comes in, where standardized code that could be used by a lot of people to create sort of simple innovations that can scale, which is why the hacks, I imagine that the code base for deep seek is probably a jam mess.

51:39I'm sure they have deep seek definitely has code bases that are extremely messy where they're testing these new ideas, multi -headly and attention, could start in something like a Jupyter notebook or somebody tries something on a few GPUs. And that is really messy. But the stuff that trains the deep seek V3 and deep seek R1, those libraries, if you were to present them to us, I would guess are extremely high quality code, very readable code. I think there is one aspect to note though, right? Is that there is the general, general ability for that to transfer across different types of runs, right?

52:15You may make really, really high -quality code for one specific model architecture at one size. And then that is not transferable to, hey, when I make this architecture tweak, everything's broken again, right? Like that's something that could be, you know, with their specific low -level coding of scheduling SMs, is specific to this model architecture and size, right? And whereas like in videos, collectives library is more like, hey, it'll work for anything, right? that you want to do an all -reduced, great. I don't care what your model architecture is. It'll work. And you're giving up a lot of performance when you do that in many cases, but it's worthwhile for them to do the specific optimization for the specific run given the constraints that they have regarding compute.

52:58I want to house, for us, well, it is to like, you know, these frontier models, like initiate training, like to have the code, but it's fine to push the button that like you're now spending a large amount of money and time to train this. Like, there must, I mean, there must be a lot of innovation on the debugging stage of like making sure there's no issues that you're monitoring and visualizing every aspect of the training, all that kind of stuff. When people are training, they have all these various dashboards, but like the most simple one is your loss, right? And it continues to go down, but in reality, especially with more complicated stuff like M -O -E, the biggest problem with it, or F -P -H training, which is another innovation, you know, going to a lower precision number format, i .e.

53:43less accurate, is that you end up with lost spikes, right? And no one knows why the lost spike happened. And for a long time, some of them you do. Some of them are bad data. Do I can give a AI2's example of what blew up earlier models is a subreddit called microwave gang? We love to shout out the cell. It's a real thing. You can pull up microwave gang. Essentially, it's a subreddit where everybody makes posts that are just the letter M. So it's like, mmm, so there's a extremely long sequences of the letter M and then the comments are like beep beep because that's on the microwave events. But if you pass us into a model, this train to be a normal producing text is extremely high loss because normally you see an M, you don't predict M's for a long time.

54:21So like this is something that causes the loss spikes for us. But when you have much like this is this is old, this is not recent. And when you have more mature data systems, that's not the thing that causes the loss spike and what Dylan is saying is true. But it's like, it's levels to this sort of idea. We're the guards to the stress, right? These people are like, you know, you'll go out to dinner with like a friend that works at one of these labs. And they'll just be like looking at their phone every like 10 minutes and they're not like, you know, it's one thing if they're texting but they're just like, like, is the lost, is the lost person.

54:51It's like the lost person. It's like, tokens per second, lost, not blown up. They're just walking, watching us. And the heart rate goes up if there's a spike. And some level of spikes is normal, right? It'll recover and be back. Sometimes a lot of the old strategy was like, You just stop the run, restart from an old version, and then change the data mix, and then it keeps going. There are even different types of spikes, so Durk Greninvel has a theory today. I do that. It's like fast spikes and slow spikes, where there are sometimes when you're looking at the loss and there are other parameters, you can see it start to creep up, and then blow up, and that's really hard to recover from, so you have to go back much further.

55:25So you have the stressful period where it's flat, or it might start going up, and you're like, what do I do? Whereas there are also lost spikes that are, it looks good, and then there's one spiky data point. And what you could do is you just skip those. You see that there's a spike, you're like, okay, I can ignore this data, don't update the model and do the next one, and it'll recover quickly. But these like, on trickier implementations, as you get more complex in your architecture, and you scale up to more GPUs, you have more potential for your loss blowing up. So there's a distribution. The whole idea of grocking also comes in, right?

55:56It's like just because it slowed down from improving and loss doesn't mean it's not learning, because all of a sudden it could be like this, and they could just spike down and loss again because it truly learned something, right? And it took some time for it to learn that. It's not like a gradual process, right? And that's what humans are like. That's what models are like. It's really a stressful task, as you mentioned. And the whole time the dollar count is going up. Every company has failed runs. You need failed runs to push them below on your infrastructure. So a lot of news cycles are made of X company had Y failed run.

56:27every company that's trying to push the frontier of AI has these. So yes, it's noteworthy because it's a lot of money and it can be weak to month setback, but it is part of the process. But how do you get, if you're deep -seek, how do you get to a place where holy shit, there's a successful combination of hyper parameters. A lot of small failed runs. It's so rapid iteration through failed runs until... And successful ones. You just... And then you build a how my tuition like this, this mixture of expert works, and then this implementation of MLA works. Key hyper parameters, like learning rate and regularization and things like this, and you find the regime that works for your code base.

57:11I've talking to people at Frontier Labs, there's a story that you can tell where training language models is kind of a path that you need to follow. So you need to unlock the ability to train a certain type of model or a certain scale and then your code base and your internal know -how which hyper parameters work for it is kind of known. And you look at the deep -seek papers and models, they've scaled up, they've added complexity, and it's just continuing to build the capabilities that they have. There's the concept of a YOLO run. So YOLO you only live once. And what it is is there's all this experimentation you do at the small scale, right?

57:47Research, ablation, you have your Jupyter Notebook with your experimenting with MLA on three GPUs or whatever, and you're doing all these different things like, hey, do I do four active experts, 120 experts, do I arrange the experts this way, you know, all these different model architecture things, you're testing at a very small scale, right? Couple researchers, few GPUs, tens of GPUs, hundreds of GPUs, whatever it is. And then all of a sudden, you're like, okay, guys, no more fucking around, right? No more screwing around. Everyone, take all the resources we have, let's pick what we think will work and just go for it, right?

58:20Yolo. And this is where that sort of stress comes in is like, well, I know it works here, but some things that work here don't work here, and some things that work here don't work down here, right? In this terms of scale, right? So it's really truly a yolo run, and sort of like there is this like discussion of like certain researchers just have like this methodical nature like they can find the whole search space and like figure out all the ablations of different research and really see what is best. And there's certain researchers who just kind of like, like, you know, have that innate gut instinct of like, this is the yolo run, like, you know, I'm looking at the data.

58:53I'm just like, you're in this is it. This is why you want to work in post training because the GPU costs for training is lower so you can make a higher percentage of your training runs, yolo runs. Yeah, for now. Yeah, for now. Yeah, for now. So some of this is fundamentally luck still. Luck is skill, right? In many cases. Yeah, I mean, it looks lucky, right, when you're, but the hill to climb, if you're out in one of these labs, So you have an evaluation you're not crushing. There's a repeated playbook of how you improve things. There are localize improvements, which might be data improvements.

59:23And these add up into the whole model just being much better. And when you zoom in really close, it can be really obvious that this model is just really bad at this thing. And we can fix it. And you just add these up. So some of it feels like luck. But on the ground, especially with these new reasoning models we're talking to, it's just so many ways that we can poke around. And normally, it's that some of them give big improvement. the search space is near infinite, right? And yet the amount of compute and time you have is very low and you have to hit release schedules, you have to not get blown past by everyone.

59:56Otherwise, you know, what happened with deep seek, you know, crushing meta and mistral and coherent, all these guys, they moved too slow, right? They maybe were too methodical. I don't know, they didn't hit the yolo run, whatever the reason was, maybe they were in a skill. Whatever, you know, you can call it luck if you want, but at the end of the day, it's skill. So 2025 is the year of the YOLO run. It seems like all the labs are like going in. I think it's even more impressive what OpenAI did in 2022. Right? At the time, no one believed in mixture of experts models, right, at Google, who had all the researchers.

1:00:29OpenAI had such little compute and they devoted all of their compute for many months, right? All of it, 100 % for many months to GPT -4, with a brand new architecture with no belief that, hey, let me spend a couple hundred million dollars, which is all of the money I have on this model, right? That is truly YOLO, right? Now, you know, people like all these like training run failures that are in the media, right? It's like, okay, great, but like actually, a lot of huge chunk of my GPs are doing inference. I still have a bunch doing research constantly, and yes, my biggest cluster is training, but like on this YOLO run, but like that YOLO run is much less risky than like what opening I did in 2022.

1:01:09Or maybe what Deepseek did now, or like sort of like, hey, we're just gonna throw everything at it. The big winners throughout human history are the ones who are willing to do yellow at some point. Okay, what do we understand about the hard words been trained on? Deepseek. Deepseek is very interesting. This is where the second the tick has to zoom out out of who they are, first of all, right? High flyer is a hedge fund that has historically done quantitative trading in China as well as elsewhere. And they have always had a significant number of GPUs, right? And the past lot of these high frequency trading, algorithmic quant traders used FPGAs, but it shifted to GPUs definitely, and there's both, right?

1:01:47But GPUs especially in deep and high flyer, which is the hedge fund that owns deepseek, and everyone who works for deepseek is part of high flyer to some extent, right? Same parent company, same owner, same CEO. they had all these resources and infrastructure for trading. And then they devoted a humongous portion of them to training models, both language models and otherwise, because these techniques were heavily AI -influenced. More recently, people have realized, hey, trading with... Even when you go back to Renaissance and all these quantitative firms, natural language processing is the key to trading really fast, understanding and press release.

1:02:28and making the right trade, right? And so deep sea has always been really good at this. And even as far back as 2021, they have press releases and papers saying, like, hey, we're the first company in China with an A100 cluster this large. It was 10 ,000 A100 GPUs, right? This is in 2021. Now, this was an all for training, you know, large language models. This was mostly for training models for their quantitative aspects, their quantitative trading, as well as, you know, a lot of that was natural language processing to be clear, right? And so this is the sort of history, right? So verifiable fact is that in 2021, they built the largest cluster, at least, they claim it was the largest cluster in China, 10 ,000 GPUs.

1:03:06Before expert controls started. Yeah, I think they've had a huge cluster before any conversation of expert controls. So then you step it forward to like, what have they done over the last four years since then, right? Obviously they've continued to operate the Hedge Fund, probably make tons of money. And the other thing is that they've leaned more and more and more into AI. The CEO, Leon Cheng Feng, Leon, you're not putting me in spot on this. We discussed this. Leon Feng, right? The CEO, he owns maybe a little bit more than half the company allegedly, is an extremely like Elon, Jensen kind of figure where he's just like involved in everything.

1:03:44And so over that time period, he's gotten really in depth into AI. He actually has a bit of a like a, if you see some of the statements, a bit of an EAC vibe almost, right? Total AGI vibes. Like we need to do this. We need to make a new ecosystem of open AI. We need China to lead on this sort of ecosystem because historically the Western countries have led on software ecosystems and straight up acknowledges like in order to do this. We need to do something different. Deep sea because it's his way of doing this. Some of the translated interviews with them are. So he has done interviews? Yeah. You think you would do a Western interview or no?

1:04:21Or is there controls on their hasn't been one yet, but I would try it. I just got a tiny translator so it was great. This is old push. So fascinating figure engineer pushing full on into AI leveraging the success from the high frequency trading. Very direct quotes. Like we will not switch to closed source when asked about this stuff. He had very long term motivated in how the ecosystem of AI should work and from a Chinese perspective, he wants a Chinese company to build this vision. So this is sort of like the quote unquote visionary behind the company, right? This hedge fund still exists, right?

1:05:02This quantitative firm. And so deep seek is the sort of, that's, you know, slowly he got turned to this full view of like AI, everything about this, right? But at some point it's slowly maneuvered and he made deep seek. And deep seek has done multiple models since then. They've acquired more and more GPUs. they share infrastructure with the fund, right? And so, you know, there is no exact number of public GPU resources that they have, but besides this 10 ,000 GPUs that they bought in 2021, right, and they were fantastically profitable, right? And then this paper claims they did only 2 ,800 GPUs, which are a restricted GPU that was previously allowed in China, but no longer allowed, and there's a new version, but it's basically in videos, H100 for China, right?

1:05:46And there's some restrictions on it specifically around the communications Sort of a speed the interconnect speed, right? Which is why they had to do this crazy SM you know scheduling stuff, right? So going back to that right? Let's like this is obviously not true in terms of their total GPU count obvious available GPUs before this training run You think 2000 is the correct number or no? So this is where it takes um, you know significant amount of sort of like zoning in right like what do you call your training run, right? You count all of the research and ablations that you ran, right, picking all this stuff, because yes, you can do a yellow run, but at some level you have to do the test at the small scale, and then you have to do some test at medium scale before you go to a large scale.

1:06:28Accepted practice is that for any given model that is a notable advancement, you're gonna do two to four X compute of the full training run in experiments alone. So a lot of this compute that's being scaled up It's probably used in large part at this time for research. Yeah, and research will you know research Begets the new ideas that let you get huge information research gets you oh one Research gets you breakthroughs and you need to bet on it. So some of the pricing strategy They will discuss has the research baked into the price so the numbers that deep seeks specifically said publicly right are just the 10 ,000 GPs in 2021 and then 2000 GPUs for only the pre -training for V3 They did not discuss cost on R1, they did not discuss cost on all the other RL, right, for the instruct model that they made, right?

1:07:14They only discussed the pre -training for the base model and they did not discuss anything on research and ablations. And they do not talk about any of the resources that are shared in terms of, hey, the fund is using all these GPUs, right? And we know that they're very profitable and they had 10 ,000 GPUs in 2021. So the some of the research that we've found is that we actually believe they have closer to 50 ,000 GPUs. We as semi -analysis, so we should say that you're sort of one of the world experts in figuring out what everybody's doing in terms of the semi -conducting in terms of cluster buildouts in terms of like who's doing what in terms of training runs.

1:07:51So yeah, so that's the we. Okay, go ahead. Sorry. Sorry. We believe they actually have something closer to 50 ,000 GPS, right? Now this is split across many tasks, right? Again, the fund, research and ablations for ballpark, how much would open AI or in topic head? I think the clearest example we have because meta is also open, they talk about order of 60K to 100K, H100 equivalent GPS in their training clusters. Right. So like Lama 3, they trained on 16 ,000 H100s, right? But the company of meta last year publicly disclosed they bought like 400 something thousand and GPS, right? So of course, tiny percentage on the training, again, like most of it is like serving me the best Instagram Reels, right?

1:08:30Or whatever, right? I mean, we could get into a cost of like, what is the cost of ownership for a 2000 GPU cluster, 10 ,000? Like, there's just different sizes of companies I can afford. These things and deepseek is reasonably big. Their compute allocation compared is one of the top few in the world. It's not opening eye and but they have a lot of compute. Can you general actually just zoom out and also talk about the hopper architecture and video hopper GPU architecture and the difference between H -100 and H -800 like you mentioned the interconnects? Yeah, so there's, you know, Ampere was the A -100 and then H -100 hopper, right?

1:09:08People used them sound honestly in the US because really there's just H -100 and now there's H -200, right? But the same thing, mostly. In China, they've had two, there've been different salvos of expert restrictions. So initially the US government limited on a two factor scale, right, which is chip interconnect versus flops, right? So any chip that had interconnects above a certain level and flops above a certain floating point operations, above a certain level was restricted. Later the government realized that this was a flaw in the restriction and they cut it down to just floating point operations.

1:09:40And so, H800 had high flops low communication. Exactly. The H800 was the same performance as H100 on flops, right? But it just had the interconnect bandwidth cut. DeepSeq knew how to utilize this, you know, hey, even though we're cut back on the interconnect, we can do all this fancy stuff to figure out how to use the GPU fully anyways, right? And so that was back in October, 2022, But later in 2023, end of 2023 implemented in 2024, the US government banned the H800, right? And so by the way, this H800 cluster, these 2000 GPUs, was not even purchased in 2024, right? It was purchased in late 2023.

1:10:21And they're just getting the model out now, right? Because it takes a lot of research, et cetera. H800 was banned, and now there's a new chip called the H20. The H20 is cut back on only flops, but the interconnect bandwidth is the same. and in fact in some ways it's better than the H100 because as better at memory bandwidth and memory capacity. So there are, you know, in videos working within the constraints of what the government sets and then get builds the best possible GPU for China. Can we take this extra tangent and we'll return back to the hardware is the philosophy, the motivation, the case for export controls.

1:10:55What is it? Darian Medage published a blog post about export controls. The case he makes is that if AI becomes super powerful He says by 2026 we'll have AGI or super powerful AI and that's going to give a significant whoever builds that will have a significant military advantage. And so because the United States is a democracy and as he says China is authoritarian or has authoritarian elements, once you want a unipolar world where the super powerful military because of the AI is one that's a democracy. It's a much more complicated world geopolitically when you have two superpowers with super powerful AI and one is authoritarian.

1:11:41So that's the case he makes. And so we want to, the United States wants to use export controls to slow down to make make sure that China can't do these gigantic training runs that will be presumably required to build AGI. This is very abstract. I think this can be the goal of how some people describe export controls is this super powerful AI. There's, and you touched on the training run idea, there's not many worlds where China can not train AI models. I think export controls are kneecapping the amount of compute or the density of compute that China can have. And if you think about the AI ecosystem right now as all of these AI companies, revenue numbers are up into the right.

1:12:27The AI usage is just continuing to grow. More GPUs are going to inference. A large part of export controls if they work is just that the amount of AI that can be run in China is going to be much lower. So on the training side, DeepSeq V3 is a great example, which you have a very focused team that can still get to the frontier of AI on this 2 ,000 GPUs that's not that hard to get, all considering in the world. There's still going to have those GPUs. They're still going to be able to train models, but if there's going to be a huge market for AI, if you have strong export controls and you want to have 100 ,000 GPUs just serving the equivalent of chat GPT clusters, with good export controls, it also just makes it so that AI can be used much less.

1:13:07And I think that is a much easier goal to achieve than trying to debate on what AGI is. And if you have these extremely intelligent autonomous AIs and data centers, like those are the things that could be running in these GPU clusters in the United States, but not in China. Does some extent training a model does effectively nothing, right? Like the other model. The thing that Dario is sort of speaking to is the implementation of that model once trained to then create huge economic growth, huge increases in military capabilities, huge increases in productivity of people, betterment of lives, whatever you want to direct super powerful AI towards, you can't.

1:13:48But that requires a significant amounts of compute. The US government has effectively said, and forever, like training will always be a portion of the total compute. We mentioned met 400 ,000 GPUs, only 16 ,000 made Lama. Right? So the percentage that meta is dedicating to inference. Now this might be for recommendation systems that are trying to hack our mind into spending more time and watching more ads or if it's if it's or if it's for a super powerful AI that's doing productive things. Doesn't matter about the exact use that our you know economic system decides. It's that that can be delivered whatever in whatever way we want.

1:14:22Whereas with China, right? You know, your you know expert restrictions. Great. You're never going to be able to cut everything off. Right? And that's like, I think that's quite well understood by the US government, is that you can't cut everything off. And they'll make their own chips. And they're trying to make their own chips, they'll be worse than ours. But the whole point is to just keep a gap, right? And therefore, at some point, as the AI, in a world where 2, 3 % economic growth, this is really dumb, by the way, right? To cut off high -tech and not make money off of it. But in a world where super powerful AI comes about, and then starts creating significant changes in society, which is what all the AI leaders and big tech companies believe, I think super powerful AI is going to change society massively.

1:15:02And therefore, this compounding effect of the difference in compute is really important. There's some sci -fi out there where AI is measured in the power of in how much power is delivered to compute, right? Or how much is being, you know, that's sort of a way of thinking about what's the economic output is just how much power are you directing towards that AI? So we talk about reasoning models with this as a way that this might be actionable as something that people can actually see. So the reasoning models that are coming out with R1 and O1, they're designed to use more compute. There's a lot of buzzy words in the AI community about this, test time compute, inference time compute, whatever.

1:15:38But Dylan has good research on this. You can get to the specific numbers on the ratio of a new training model. You can look at things about the compute, amount of compute use, the training and amount of compute use of inference. These reasoning models are making inference way more important to doing complex tasks. In the fall in December, their OpenAI announced this O3 model. There's another thing in AI when things move fast. We get both announcements and releases. Announcements are centrally blog posts where you pat yourself on the back and you say you did things and releases are on the models out there, the papers out there, etc.

1:16:07So open AI has announced O3 and we can check if O3 mini is out as of recording potentially. But that doesn't really change the point, which is that the breakthrough result was something called ARC -AGI task, which is the abstract reasoning corpus, a task for artificial general intelligence, French law, Chile is the guy who's been, it's a multi -year old paper, it's a brilliant benchmark. And the number for OpenAI -03 to solve this was that it used some sort of number of samples in the API. The API has like thinking effort and number of samples. They used a thousand samples to solve this task and it comes out to be like five to $20 per question, which you're putting in effectively a math puzzle.

1:16:49And then it takes orders of dollars to answer one question. And this is a lot of compute. If those are going to take off in the US, OpenAI needs a ton of GPUs on inference to capture this. They have this OpenAI chat GBT pro subscription, which is $200 a month, which Sam said they're losing money on, which means that people are burning a lot of GPUs on inference. And I've signed up with it. I've played with it. I don't think I'm a power user, but I use it. And it's like, that is the thing that a Chinese company with mediumly strong expert controls, the row is blue poles, might not be able to do it all.

1:17:21And if that, the main result for O3 is also spectacular coding performance. And if that feeds back into AI companies being able to experiment better. So presumably the idea is for an AGI, a much larger fraction of the compute would be used for this test -time compute, for the reasoning. For the AGI goes into a room and thinks about how to to take over the world and come back in 2 .7 hours. And that is going to take a lot of compute. This is what people like CEO or leaders of OpenAI and Anthropocalk about is like autonomous AI models, which is you give them a task and they work on it in the background.

1:18:00My personal definition of AGI is much simpler. I think language models are a form of AGI and all of this super powerful stuff is a next step that's great if we get these tools, but a language model has so much value and so many domains. It is a general intelligence to me. But this next step of agentic things where they're independent and they can do tasks that aren't in the training data is what the few -year outlook that these AI companies are driving for. I think the terminology here that Dario uses a super -paral for AI. So I agree with you on the AGI. I think we already have something like that's exceptionally impressive.

1:18:42The possession of, then you would have a significant military and geopolitical advantage over other nations. So it's not just like you can ask it how to cook, enom lit. And he has a much more positive view and is that say machines of love and grace. It's spread into this. I don't have enough background in physical sciences to gauge exactly how competent I am in AI and revolutionaries. Biology, I'm safe saying that AI is going to accelerate the progress of any computational science. So we're doing a depth for search here on topics taking tangent of attention. So let's continue on that depth for search You said that you're both feeling the AGI So you're what's what's your timeline?

1:19:24Dario's 2026 for the super powerful AI that you know That's basically agentic to a degree where it's a real security threat But that level of AGI, what's your timeline? I don't like to attribute specific abilities because predicting specific abilities and when is very hard. I think mostly if you're going to say that I'm feeling the AGI is that I expect continued rapid surprising progress over the next few years. So something like R1 is less surprising to me from DeepSeek because I expect there to be new paradigms where substantial progress can be made. DeepSeek R1 is so unsettling because we're kind of on this path with chat GPT.

1:20:04It's getting better, it's getting better, it's getting better. And then we have a new direction for changing the models and we took one step like this. And we took a step up. So it looks like a really fast slope and then we're just going to take more steps. So it's really unsettling when you have these big steps. And I expect that to keep happening. I see I've tried opening I operator, I've tried quad computer use. They're not there yet. I understand the idea. But it's just so hard to predict what is the breakthrough that will make something like that work. and I think it's more likely that we have breakthroughs that work in things that we don't know what they're going to do.

1:20:37So like everyone wants agents, Dario has very eloquent way of describing this. And I just think that it's like, there's going to be more than that, so just expect these things to come. I'm going to have to try to pin you down to a date on the AGI timeline. Like the nuclear weapon moment. So moment where on the geopolitical stage, There's a real like, you know, because we're talking about export controls. When do you think just even a throwout a date? When do you think that would be? Like for me, it's probably after 2030. So I'm not asking. That's what I would say. So define that right? Because to me, it kind of almost has already happened.

1:21:18Right? You look at elections in India and Pakistan. People get AI voice calls and think they're talking to the politician, right? The AI diffusion rules, which was enacted in the last couple weeks of the Biden admin and looks like the Trump admin will keep and potentially even strengthen, limit cloud computing and GPU sales to countries that are not even related to China. It's like this is Portugal and all these like normal companies are on that. You need approval from the US list. Like yeah, Portugal and like, you know, like all these countries that are allies, right? Singapore, right? Like they, they freaking have F -35s and we don't let them buy GPUs.

1:21:52Like this is, this to me is already to the scale of like, you know. Well, that just means that the US military is really nervous about this new technology. That doesn't mean that technology is already there. So, they might be just very cautious about this thing that they don't quite understand. But that's a really good point. The robo calls swarms of semi -intelligent bots could be a weapon, could be doing a lot of social engineering. I mean, there's tons of talk about from the 2016 elections like Cambridge Analytica and all this stuff, Russian influence. I mean, every country in the world is pushing stuff onto the internet and has narratives they want, right?

1:22:30Like that's every, every like technically competent whether it's Russia, China, US, Israel, et cetera, right? You know, people are pushing viewpoints onto the internet and mass and language models crash the cost of like very intelligent sounding. Like, there's some research that shows that the distribution is actually the limiting factor. So language models haven't yet made misinformation,

1:22:52particularly change the equation there. The internet is still ongoing. I think there's a blog, AI snake oil and some of my friends at Princeton that write on the stuff. So there is research. It's like, it's a default that everyone assumes and I would have stopped the same thing is that misinformation isn't gonna have fireworks with language models. I think in terms of internet posts and things that people have been measuring, it hasn't been an exponential increase or something extremely measurable. And things you're talking about with like voice calls and stuff like that. It could be in modalities that are harder to measure.

1:23:22So it's something that it's too soon to tell in terms of, I think that's like political instability via the web is very, it's monitored by a lot of researchers to see what's happening. I think that you're asking about like the AGI thing. If you make me give a year, I'm going to be like, okay, I have AI CEOs saying this. They've been saying two years for a while. I think that their people like Dario and Thropic, the CEO, had thought about this so deeply. I need to take their word seriously, but also understand that they have different different incentives. So I would be like, add a few years to that, which is how you get something similar to 2030 or a little after 2030.

1:24:02I think to some extent, we have capabilities that hit a certain point where any one person could say, okay, if I can leverage those capabilities for X amount of time, this is AGI, call it 2728, but then the cost of actually operating that capability is going to be my point. It's so extreme that no one can actually deploy it at scale and mass to actually completely revolutionize the economy on a snap of a finger. So I don't think it will be like a snap of the finger. Oh, physical constraint. Rather, it'll be a, oh, the capabilities are here, but I can't deploy it everywhere. right? And so one simple example going back sort of to 2023 was when, you know, being with GPT -4K mountain, everyone was freaking out about search, right?

1:24:45Perplexity came out. If you did the cost on like, hey, implementing GPT -3 into every Google search, it was like, oh, okay, this is just like physically impossible to implement, right? And as we step forward to like going back to the test time compute thing, right? A query for, you know, you ask to chat you, if you get a question, it costs sense, right? For their most capable model of chat, right? To get a query back, to solve an Arc AGI problem though, cost five to 20 bucks, right? And this is an A, it's only going up from there. This is a thousand, 10 ,000 X factor difference in cost to respond to a query versus do a task.

1:25:20And the task of Arc AGI is not like it's like, it's simple to some extent, you know, but it's also like, what are the tasks that we want, Okay, AGI quote unquote what we have today can do arc AGI. Three years from now it can do much more complicated problems, but the cost is going to be measured in thousands and thousands and hundreds of thousands of dollars of GPU time and there just won't be enough power to use infrastructure to operate this and therefore shift everything in the world on the snap of finger. But at that moment, who gets to control and point the AGI at a task? And so this was Indario's post that he's like, hey China can effectively and more quickly than us point their AGI at military tasks, right?

1:26:01And they have been in many ways faster at adopting certain new technologies into their military, right? Especially with regards to drones, right? The US maybe has a longstanding, you know, large -air, sort of, you know, fighter jet type of thing, bombers, but when it comes to asymmetric arms, such as drones, they've completely leapfrogged the US and the West. And the fear that Dario is sort of pointing out there, I think, is that, yeah, great. will have AGI in the commercial sector. The US military won't be able to implement it super fast. Chinese military could and they could direct all their resources to implementing it in the military and therefore solving military logistics or solving some other aspect of disinformation for targeted certain set of people so they can flip a country's politics or something like that that is actually catastrophic versus the US just wants to because it will be more capitalistically allocated just towards whatever is the highest return on income which which might be like building factories better or whatever.

1:26:59So everything I've seen, people's intuition seems to fail on robotics. So you have this kind of general optimism. I've seen this on self -driving cars. Think it's much easier problem than it is. Similar with drones. Here I understand it a little bit less, but I've just seen the reality of the war in Ukraine and the usage of drones at both sides. and it seems that humans still far outperform any fully autonomous systems. AI is an assistant, but humans drive. FPV drones were the humans control, and most of it just far, far, far outperforms AI systems. So I think it's not obvious to me that we're going to have swarms of autonomous robots anytime soon in the military context.

1:27:45Maybe the fastest I can imagine is 2030, which is why I said 2034, the super powerful AI, whenever you have large scale swarms of robots doing military actions, that's when the world just starts to look different to me. So that's the thing I'm really worried about. But there could be cyber war, cyber war type of technologies that from social engineering to actually just swarms of robots that find attack vectors in our code bases and shut down pogards, that kind of stuff. And it could be one of those things like on any given weekend or something. Power goes out nobody knows why and the world changes forever.

1:28:27Just power going out for two days in all of the United States that will lead to murder to chaos. But going back to expert controls, do you see that as a useful way to to control the balance of power geopolitically in the context of AI. And I think going back to my viewpoint is if you believe we're in the sort of stage of economic growth and change that we've been in for the last 20 years, the export controls are absolutely guaranteeing that China will win long term, right? If you do not believe AI is going to make significant changes to society in the next 10 years or five years, right? Five year timelines are sort of what the more executives and such of AI companies and even big tech companies believe, but even 10 year timelines, you know, it's reasonable.

1:29:20But once you get to, hey, these timelines are below that time period, then the only way to sort of like create a sizable advantage or disadvantage for America versus China is if you constrain compute because talent is not really something that's constraining, right? China arguably has more talent, right? More STEM graduates, more programmers. The US can draw upon the world's people, which it does. There's tons of foreigners in the AI industry. So many of these AI teams are all people without a US passport. Yeah, I mean, as many of them are Chinese people are moving to America, right? And that's great.

1:30:01That's exactly what we want, right? But that talent is one aspect, but I don't think that's one that is a measurable advantage for the US or not. It truly is just whether or not compute, right? Now, even on the compute side, when we look at chips versus data centers, right? China has the unprecedented ability to build ridiculous sums of power, clockwork, right? They're always building more and more power. They've got steel mills that like individually are the size of the entire US industry, right? And they've got aluminum mills that consume gigawatts and gigawatts of power, right? And when we talk about what's the biggest data center, right?

1:30:38opening I made this huge thing about Stargate, their announcement there. That's not, that's like once it's fully built out in a few years, it'll be two gigawatts, right, of power, right. And this is still smaller than the largest, you know, industrial facilities in China, right. China, if they wanted to build the largest data center in the world, if they had access to the chip, could. So there's not just, it's just a question of, uh, when, not if, right. So they're industrial capacity for existing United States, exactly. To the manufacturer stuff. So long term, they're going to be manufacturing chips there.

1:31:12Chips are a little bit more specialized. I'm specifically referring to the data centers, right? Chips, fabs take huge amounts of power. Don't get me wrong. That's not necessarily the gating factor there. The gating factor on how fast people can build the largest clusters today in the US is power. It is whether it's, now it could be power generation, power transmission, substations, and all these sorts of transformers and all these things, building the data center, these are all constraints on the US industry's ability to build larger and larger training systems, as well as deploying more and more inference compute.

1:31:46I think we need to make the point clearer on why the time is now for people that don't think about this, because essentially with export controls, you're making it so China cannot make or get cutting edge chips. And the idea is that if you time this wrong, China is pouring a ton of money into their chip production. And if you time it wrong, they are going to have more capacity for production, more capacity for energy, and figure out how to make the chips and have more capacity than the rest of the world to make the chips because everybody can buy, they're going to sell their Chinese chips to everybody, they might subsidize them.

1:32:17And therefore, if AI takes a long time to become differentiated, we've kneecapped the financial performance of American companies. In video, can sell less. TSMC cannot sell to China. Therefore we have less demand to keep driving the production cycle. That's the assumption behind the timing being less than 10 years or five years to above. China will win because of these restrictions, long term, unless AI does something in the short term, which I believe AI will do, make massive changes to society in the medium short term. right? And so that's the big unlocker there. And even even today, right, if Xi Jinping decided to get, you know, quote unquote, scale -pilled, right?

1:33:01I decide that scaling laws are what matters, right? Just like the US executives, like Sachin Adela and Mark Zuckerberg and Sundar and all these US executives of the biggest, most powerful tech companies have decided their scale -pilled and they're building multi -gigawatt data centers, right? Whether it's in However it is, they're building these massive things that cost as much as their entire budget for spending on data centers globally in one spot, right? This is what they've committed to for next year, year after, et cetera. And so they're so convinced that this is the way, that this is what they're doing, but if China decided to, they could do it faster than us, but this is where the restrictions come in.

1:33:43It is not clear that China, as a whole, has decided, you know, from the highest levels that this is a priority. The US sort of has, right? You see Trump talking about deepseek and stargate within the same week, right? So he's in the Biden and men as well had a lot of discussions about AI and such. It's clear that they think about it. Only just last week did deepseek meet the second and command of China, right? Like they have not even met the top, right? And they haven't met Xi. Xi hasn't set down. And they only just released a subsidy of a trillion RMB, roughly $160 billion, which is closer to the spending of Microsoft and Meta and Google combined for this year.

1:34:23So it's like, they're realizing it just now, but that's where the export restrictions come in and say, hey, you can't ship the most powerful US chips to China. You can ship a cut down version. You can't ship the most powerful chips to all these countries who we know we're just gonna rent it to China. You have to limit the numbers, right? And the tools. So you can't, and same with manufacturing with deep -med tools, all these different aspects, but it all stems from AI and then what downstream can slow them down an AI. And so the entire semiconductor restrictions, you read them? They are very clear.

1:34:56It's about AI and military civil fusion of technology. Right? It's very clear. And then from there, it goes, oh, well, we're banning them from buying like lithography tools and etch tools and deposition tools. And oh, this random, like, you know, subsystem from a random company that's like tiny, right? like why are we banning this because all of it, the US government has decided is critical to AI systems. I think the the full -compoint is like the transition from seven nanometer to five nanometer chips where I think it was Huawei that had the seven nanometer chip a few years ago, which caused another political bruh -ha almost like this moment.

1:35:31And then it's the ASMR deep UV, when it's like extreme ultraviolet lithography. Just that context on the chips right with Nathan's referring to is in 2020, Huawei released their Ascend 910 chip, which was an AI chip. First one on 7 -anometer before Google did, before Nvidia did. They submitted it to the MLPurf benchmark, which is sort of an industry standard for machine learning performance benchmark. It did quite well. It was the best chip at the submission. This was a huge deal. The Trump admin, of course, banned the Huawei from getting 7 -anometer chips from TSMC. and suddenly they had to switch to using internal domestically produced chips, which was a multi -year setback.

1:36:14Many companies have done seven nanometer chips, and the question is, we don't know how much Huawei was subsidizing production of that chip. Intel has made seven nanometer chips that are not profitable, and things like this. So this is how all feeds back into the economic engine of export controls. Well, so you're saying that for now, Xi Jinping has not felt the AGI, but it feels like the deep seek moment might like there might be meetings going on now where he's going to start wearing the same t -shirt and things are going to escalate. I mean, like, like this, he may have woken up last week, right?

1:36:49Leon Faying met the vice chair vice the second command guy and they had a meeting and then the day the next day they announced the AI subsidies, which are truion RMB. So it's possible that this deep seek moment is truly the beginning of a cold war. That's what a lot of people are worried about. People in AI have been worried that this is going towards their cold war or already it's not deep seeks fall, but there's something a bunch of factors came together where history works. Explosion. I mean, it all has to do with NVIDIA's dot going down. It's just some kind of a mass hysteria that happened that eventually led to Xi Jinping having meetings and waking up to this idea.

1:37:28And the US government realized in October 7th, 2022, before chat GPT released, that restriction on October 7th, which dropped and shocked everyone. And it was very clearly aimed at AI. Everyone was like, what the heck are you doing? And the diffusion was out then. Not chat GPT. Yeah, but not chat GPT. So it was like starting to be rumbling of what Gen AI can do to society, but it was very clear, I think, to at least like national security council and those sort of folks that But this was where the world is headed, this cold war that's happening. So is there any concerns that the export controls push China to take military action in Taiwan?

1:38:09This is the big risk, right? The further you push China away from having access to cutting edge American and global technologies, the more likely they are to say, well, because I can't access it. I might as well, like, no one should access it, right? And there's a few interesting aspects of that. China has an urban rural divide like no other. They have a male, female, birth ratio like no other to the point where if you look in most of China, it's like the ratio is not that bad, but when you look at single dudes in rural China, it's like a 30 to one ratio. And those are just enfranchised dudes.

1:38:44The US has an in -sell problem. China does too. It's just they're placated in some way or cut crushed down. What do you do with these people? And at the same time, you're not allowed to access the most import technology, at least the US thinks so China is maybe starting to think this is the most import technology by starting to dump subsidies in it, right? They thought EVs and renewables were the most important technology. They dominate that now, right? Now they're starting to, they started thinking about, about semiconductors in, you know, the late 2010s and early 2020s and now they've been dumping money and they're catching up rapidly and they're going to do the same with AI, right?

1:39:16because they're very talented. So the question is, when does this hit a breaking point? And if China sees this as, hey, they can continue. If they, if not having access and starting a true hot war, taking over Taiwan or trying to subvert its democracy in some way, or blockading it, hurts the rest of the world far more than it hurts them, this is something they could potentially do. right? And so is this pushing them towards that potentially? Right? I'm not quite a geopolitical person, but you know, it's obvious that the world regime of peace and like trade is like super awesome for economics, but at some point it could break, right?

1:40:01I think we should comment that the like why Chinese economy would be hurt by that is that they're export heavy. I think the United States buys so much like if that goes away, like that's how their economy goes. Also, they just like would not be able to import raw materials from all over the world, right? The US would just shut down the trade in Malacca. And at the same time, the US, you could argue almost all the GDP growth in America since the 70s has been either population growth or tech, right? Because your life today is not that much better than someone from the 80s outside of tech, right?

1:40:35You still, cars, they all have semiconductors in them everywhere, fridges, semiconductors everywhere. There's these funny stories about how Russians were taking apart laundry machines because they had certain like Texas instrument chips that they could then repurpose and put into like their their anti -missile missile things right like their S 400 or whatever you you would know more about this but There's all sorts of like everything about semiconductors is so integral to every part of our lives So can you explain the role of TSMC in this story of semiconductors and and maybe also how the United States can break the reliance on TSMC.

1:41:11I don't think it's necessarily breaking the reliance. I think it's getting TSMC to build in the US. But taking a step back, TSMC produces most of the world's chips, especially on the foundry side. There's a lot of companies that build their own chips. Samsung, Intel, you know, ST Micro, Texas Instruments, you know, analog devices, all these kinds of companies build their own chips and XP, but more and more of these companies are outsourcing to TSMC and have been for multiple decades. Can you explain the supply chain there and where most of TSMC is in terms of manufacturing? Sure. So, historically, supply chain was companies would build their own chips.

1:41:53They would, you know, it would be a company started, they'd build their own chips, and and then they designed the chip and build the chip and sell it. Over time, this became really difficult because the cost of building a fab continues to compound every single generation. Of course, the technology, figuring out the technology for it is incredibly difficult regardless, but just the dollars and cents that are required, ignoring, saying, hey, yes, I have all the technical capability, which it's really hard to get that, by the way, Intel's failing, Samsung's failing, et cetera. But if you look at just the dollars to spend to build that next generation fab, it keeps growing, right?

1:42:25sort of like, you know, Moore's Law is having the cost of ships every two years. There's a separate law that's sort of like doubling the cost of fabs every handful of years. And so you look at a leading edge fab that is going to be profitable today that's building, you know, three nanometer ships or two nanometer ships in the future. That's going to cost north of $30, $40 billion, right? And that's just for like a token amount. That's for a like, that's like the base building walking. You probably need to build multiple, right? And so when you look look at the industry over the last, you know, if I go back 20, 30 years ago, there were 20, 30 companies that could build the most advanced chips, and then they would design them themselves and sell them, right?

1:42:59So companies like AMD would build their own chips. Intel of course still builds their own chips are very famous for them. IBM would build their own chips, and you know, you could just keep going down the list. All these companies built their own chips. Slowly they kept falling like flies, and that's because of what TSMC did, right? They created the Foundry Business Model, which is I'm not gonna design any chips. I'm just gonna contract manufacturer chips for other people. And one of their early customers is Nvidia, right? Nvidia was, is the only semiconductor company that's worth, you know, that's doing more than a billion dollars of revenue that was started in the era of foundry, right?

1:43:33Every other company started before then and at some point had FABs, which is actually incredible, right? You know, like AMD and Intel and Broadcom. It's a great fact. Like everyone had FABs at some point or, you know, brought, you know, some Some companies like Broadcom, it was like a merger, a malgamation of various companies that rolled up. But even today, Broadcom has fabs, right? They build iPhone, RF, radio chips, sort of in Colorado for Apple, right? Like there's, all these companies had fabs and for most of the fabs, they threw them away or sold them off or they got rolled into something else.

1:44:03And now everyone relies on TSMC, right? Including Intel, their latest PC chip uses TSMC chips, right? It also uses some Intel chips, but it uses the TSMC process. Can you explain why the Foundry Model is so successful for these companies? Why are they going with some scale? Yeah, so I mean, like I mentioned, the cost of building a Fab is so high. The R &D is so difficult. And when you look at companies that had their own vertical stack, there was an antiquated process of like, okay, I'm so hyper -customized to each specific chip. But as we've gone through the history of the last 50 years of electronics and semiconductors, A, you need more and more specialization, because Moore's law has died.

1:44:45Denard scaling has died. IE chips are not getting better just for free. From manufacturing, you have to make real architectural innovations. Google is not just running on Intel CPUs for web serving. They have a YouTube chip. They have TPUs. They have Pixel chips. They have a wide diversity of chips that generate all the economic value of Google. It's running all the services and stuff. And so, and this is just Google, and you could go across any company in the industry, and it's like this, right? Cars contain 5 ,000 chips, 200 different varieties of them, right? All these random things, a Tesla door handle has two chips, right?

1:45:18Like it's like ridiculous. And it's a cool door handle, right? It's like, you know, you don't think about it, but it's like has two really chip like, like, penny, like chips in there, right? Anyway, so as you have more diversity of chips, as you have more specialization required, and the cost of fabs continues to grow, you need someone who is laser focused on building the best process technology and making it as flexible as possible. I think you could say it's simply, which is the cost per tab goes up. And if you are a small player that makes a few types of chips, you're not going to have the demand to pay back the cost of the fab.

1:45:50Whereas Nvidia can have many different customers and aggregate all this demand into one place. And then they're the only person that makes enough money building chips to buy the next, to build the next fab. So this is kind of why the company slowly get killed because they have a 10 years ago a chip that is profitable and is good enough, but the cost to build the next one goes up. They may try to do this fail because they don't have the money to make it work and then they don't have any chips or they build it and it's too expensive and they just, I have a lot of problems. Or they run it. There's more failure points, right?

1:46:22You could have one little process related to some sort of like chemical etch or some sort of plasma etch or some little process that screws up. You didn't engineer it right and now the whole company falls apart. You can't make chips, right? And so super, super powerful companies like Intel, they had the weathering storm to like, hey, they still exist today, even though they really screwed up their manufacturing six, seven years ago. But in the case of like AMD, they almost went bankrupt. They had to sell their fabs to Mubodala, UAE, right? And like that became a separate company called Global Foundrys, which is a Foundry firm.

1:46:56And then AMD was able to then focus on the return back up was like, hey, let's focus on making chiplets and a bunch of different chips for different markets. and focusing on specific workloads rather than all of these different things. So you get more diversity of chips, you have more companies than ever designing chips, but you have fewer companies than ever manufacturing them. This is where TSMC comes in as they've just been the best. They are so good at it, they're customer focused, they make it easy for you to fabricate your chips, they take all of that complexity and kind of try and abstract a lot of it away from you.

1:47:29They make good money, they don't make insane money, but they make good money. And they're able to aggregate all this demand and continue to build the next fab, the next fab, the next fab. So why is Taiwan so special for TSMC? Why is it happening there? Can it be replicated inside the United States? Yeah, so there's aspects of it that I would say yes and aspects that I'd say no, right? TSMC is way ahead because former executive Morse Chang of Texas Instruments wasn't promoted to CEO and he's like screw this, I'm gonna go make my own chip company, right? And he went to Taiwan and made TSMC, right?

1:48:03And there's a whole lot more story there. So it could have been Texas Instruments, it could have been the TSMC, but Texas, semiconductor manufacturer, right? That of Texas Instruments, right? But, you know, so there is that whole story there, but they're sitting here in Texas. I mean, and that sounds like a human story, like it didn't get promoted. Just the brilliance of Morris Chang, which I wouldn't underplay, but there's also a different level of how this works, right? So in Taiwan, the number top percent of graduates of students that go to the best school, which is NTU, the top percent of those all go work to TSMC.

1:48:40And guess what their pay is? Their starting pay is like $80 ,000, $70 ,000, which is like starting pay for a good graduate in the US. Not the top, the top graduates are making hundreds of thousands of dollars at the Googles and the Amazon's. And now I guess the open AIs of the world. right? So there is a large dichotomy of like what is the top 1 % of the society doing and where they headed because of economic reasons right until never paid that crazy good right and it didn't make sense to them right that's one aspect right where's the best going second is the work ethic right like you know we like to work you know you work a lot we work a lot but at the end of the day when there's a you know when when what is the time and amount of work that you're doing and what as a fab required, right?

1:49:23Fabs are not work from home jobs there. You go into the fab and grueling work, right? There's, there's, hey, if there is any amount of vibration, right, an earthquake happens. Vibrates the machines, they're all, they're either broken, you've scrapped some of your production, and then in many cases they're like not calibrated properly. So, so when TSMC, when there's an earthquake, right? Recently there's been an earthquake. TSMC doesn't call their employees. They just, they just go to the fab, and they just show up the parking lot gets slammed and people just go into the fab and fix it. It's like ants, right?

1:49:57It's like a hive of ants doesn't get told by the queen what to do. The ants just know. It's like one person just specializes on these one task and it's like you're gonna take this one tool and you're the best person in the world and this is what you're gonna do for your whole life is this one task in the fab, which is like some special chemistry plus nano manufacturing on one line of tools that continues to get iterated. and yeah, it's just like, it's like a specific plasma edge for removing silicon dioxide, right? That's all you focus on your whole career. And it's like such a specialized thing.

1:50:25And so it's not like the Taster Transferable. AI today is awesome because people can pick it up like that. Semiconductor manufacturing is very antiquated and difficult. None of the materials are online for people to read easily and learn, right? The papers are very dense and it takes a lot of experience to learn. And so it makes the barrier to entry much higher too. So when you talk about hey, you have all these people that are super specialized, they will work, you know, 80 hours a week in a factory, right in a fab And if anything goes wrong, they'll go show up in the middle of the night because some earthquake their wife is like there's an earthquake He's like great.

1:51:02I'm gonna go to the fab Like would you would you like as an American do that right? It's like these sorts of things are like what you know I guess are the exemplifying like why TSMC is so amazing now. Can you replicate it in the US? Let's not ignore Intel was the leader in manufacturing for over 20 years. They brought every technology to market first besides the EUV, Shrain Silicone, high -key metal gates, FinFET. The list goes on and on and on of technologies that Intel brought to market first made the most money from and manufactured at scale first best Pious profit margins. So we shouldn't ignore that Intel can't do this.

1:51:39It's that the culture has broken. right? You've invested in the wrong things. They said no to the iPhone. They had all these different things regarding like, you know, mismanagement of the fabs, mismanagement of designs, this lockup, right? And at the same time, all these brilliant people, right, these like 50 ,000 PhDs, you know, or masters that have been working on specific chemical or physical processes or nanomangufacturing processes for decades, in Oregon, they're still there. They're still producing amazing work. It's just like getting it to the last mile of production at high yield where you can manufacture dozens and hundreds of different kinds of chips.

1:52:15It's good customer experience has broken. It's that customer experience. It's like part of it is people will say Intel was too pompous in the 2000s. 2010s. They just thought they were better than everyone. The tool guys were like, oh, I don't think this is mature enough. No, like, I just don't know. We know. This sort of stuff would happen. Can the US bring it to the, can the US bring leading edge semiconductor manufacturing to the US, mathematically yes, right? And we aren't, right? It's happening. Like Arizona is getting better and better, or time goes on. TSMC has built roughly 20 % of their capacity for five nanometer in the US, right?

1:52:51Now, this is nowhere near enough, right? 20 % of capacity in the US is like nothing, right? And furthermore, this is still dependent on Taiwan existing, right? All, there's sort of an important way to separate it out. There's R &D and there's high volume manufacturing. They're effectively, there are three places in the world that are doing leading edge R &D. There's Sinshyeo, Taiwan, there's Hillsboro, Oregon, and there is Pyongyang, South Korea, right? These three places are doing the leading edge R &D for the rest of the world's leading edge semiconductors, right? Now, manufacturing can be distributed more globally, right?

1:53:29And this is sort of where this dichotomy exists of like, who's actually modifying the process, who's actually developing the next generation one who's improving them is Sinshu, is Hillsboro, is Pyongyang, right? It is not the rest of these, you know, fabs like Arizona, right? Arizona is a paperweight if Sinshu disappeared off the face of the planet, you know, within a year, a couple years, Arizona would stop producing too, right? It's actually like pretty critical. One of the things I like to say is if I had like a few missiles, I know exactly where I could cause the most economic damage, right?

1:54:01It's not targeting the White House, right? It's the R &D centers. It's the R &D centers for TSMC intel Samsung. And then some of the memory guys, micron and high necks. Because they define the future evolution of these semiconductors and everything's moving so rapidly that it really is fundamentally about R &D. And it is all about TSMC, huh? And so TSMC, you know, you cannot purchase a vehicle without TSMC chips, right? You cannot purchase a fridge without TSMC chips. You cannot, you get, you, like, I think One of the few things you can purchase ironically is a Texas Instruments like Graphic calculator, right?

1:54:37Because they actually manufacture in Texas. But like outside of that, like a laptop, of folks, anything you, servers, right? GPUs, none of this stuff can exist. And this is without TSMC. And in many cases, it's not even like the leading edge, you know, sexy five nanometers chip, three nanometers chip, two nanometers chip. Oftentimes it's just like some stupid power IC that's like converting from like, you know, some voltage to another, right? And it's made at TSMC, right? It's like trying to is investing in it as well. They can build out this long tail fat where the techniques are much more known.

1:55:04You don't have to figure out these problems with EUV. They're investing in this and then they have large supply for things like the car door handles and the random stuff and that trickles down into this whole economic discussion as well, which is they have far more than we do and having supply for things like this is crucial to normal life. So they're doing their starting to invest in high -volume manufacture, but they're not doing R &D. So they do R &D on their own. They're just way behind, right? So I would say like in 2015, China had a five year plan where they defined by 2025 and 2020 certain goals, including like 80 % domestic production of semiconductors.

1:55:44They're not going to hit that, right, to be clear. But they are in certain areas really, really close, right? Like BYD is probably going to be the first company in the world to not have to use TSMC for make it, because they have their own faps, right? For making chips. Now, they still have to buy some chips from foreign, for example, around self -driving eight -ass capabilities, because those are really high -end. But at least like, you know, like an internal combustion engine has 40 chips and an EV, you know, just for like controlling like flow rates and all these things. And EVs are even more complicated.

1:56:14So all these different power ICs and battery management controllers and all these things, they're, they're insourcing, right? And this is, this is something that like China has been doing since 2015. Now, as far as like the trailing edge, they're getting so much capacity there. As far as the leading edge, right? IE, this five nanometer and so on, so forth, right? Where GPUs they are still behind. And this is the US restrictions are trying to stop them in the ladder, but all that's happened is, yes, they've slowed down their five nanometer, three nanometers, et cetera, but they've accelerated their, hey, 45 nanometer, 90 nanometer, power IC, or analog IC, or random chip in my keyboard, right, that kind of stuff.

1:56:53So there is an angle of like the US's actions have been so from these export, you know, from the angle of the export controls have been so Inflammatory at slowing down China's progress on the leading edge that they've turned around and have accelerated their progress Oh, swear because they know they this is so important right if the US is gonna walk them out here or if they lock us out here as well And the trailing edge and so going back can the US build it here? Yes, but it's going to take a ton of money. I truly think like to revolutionize and completely in -source semiconductors would take a decade and a trillion dollars.

1:57:27Is some of it also culture? Like you said, extreme competence, extreme work ethic in Taiwan. I think if you have the demand and the money is on the line, the American company's figured it out. It's going to take handholding with the government. I think that the culture helps TSMC break through and it's easier for them. You get to select it. TSMC has some like 90 ,000 employees, right? It's not actually that insane in the amount. The Arizona Fab has 3 ,000 from Taiwan. And these people, their wives were like, yeah, we're not gonna have kids unless we use sign up for the Arizona Fab. We go to Arizona and we have our kids there.

1:57:59There's also a Japan Fab where the same thing happened, right? And so these wives drove these dudes to go to Japan or America to have the kids there. And it's an element of culture, yeah, sure. Taiwan works that hard, but also the US has done in the past. They could do it now, right? You know, we can just import, I say import, the best people in the world if we want to. That's where the immigration conversation is a tricky one and there's been a lot of debate over that. But yeah, it seems absurdly controversial to import the best people in the world. I don't understand why it's controversial. That's the one of the ways of winning.

1:58:32I'm sure we agree with you. And even if you can't import those people, I still think you could do a lot to manufacture most of in the US if the money's there, right? And so it's just way more expensive. It's not profitable over a long time. And that's the context of like the chips act is only like 50 billion dollars relative to you know some of the renewable You know initiatives that were passed in the inflation reduction act and the infrastructure act Which total in the hundreds of billions of dollars, right? And so like the amount of money that the US is spending on the semiconductor industry is is nothing right Whereas all these other countries have Structural advantages in terms of like you know work ethic and amount of work and like things like that but also a number of STEM graduates, the percentile of their best going to that.

1:59:13Right? But they also have like differences in terms of like, hey, there's just tax benefits in the law and have been in the law for 20 years, right? And so, but and then some countries have massive subsidies, right? China has something like $200 billion of semiconductor subsidies a year. We're talking about $50 billion in the US over like six, right? So the, the, the, the, the, and the growth or difference in the subsidy amounts is also huge. I think Trump has been talking about terrifying Taiwan recently. That's one of these things that's like, okay, well, maybe he doesn't want to subsidize the US semiconductor industry.

1:59:50Obviously, terrifying Taiwan is going to cost a lot of things to go get much more expensive, but does it change the equation for TSMC building more fabs in the US? That's what he's sort of positing. Can you lay out the importance, by the way? It's incredible how much you know about so much. We told your dough and nose all the stuff. Yeah. Okay, you laid out YTSMC is really important. If we look out into the future 10, 20 years out, US -China relationship seems like it can go to a dark place of cold war, escalated cold war, even hot war, or to a good place of anything from frenemies to cooperation to working together.

2:00:39So in this game theory, complicated game, what are the different trajectories? What should U .S. be doing? What do you see as the different possible trajectories of U .S. chanrelations as both leaders start to feel the AGI more and more and see the importance of chips and the importance of AI? I mean, ultimately, the export controls are pointing towards a separate future economy. I think the US has made it clear to Chinese leaders that we intend to control this technology at whatever cost to global, economic, inter integration. So that it's hard to unwind that. To the same extent, they've also limited US companies from entering China.

2:01:24So it's been a long time coming. At some point, there was convergence. But over at least the last decade, it's been branching further and further out. US companies can't enter China. Chinese companies can't enter the US. The US is saying, hey, China, you can't get access to our technologies in certain areas. And China's rebuttling with the same thing around, they've done some specific materials in gallium and things like that that they've tried to limit the US on. One of the, there's a US drone company that's not allowed to buy batteries and they have like military customers and this drone company just tells the military customers like, hey, hey, just get it from Amazon because I can't actually physically get them, right?

2:02:04Like, there's all these things that are happening that point to further and further divergence. I have zero idea and I would love if we could all hold hands and sing Kumbaya, but like, I have zero idea how that could possibly happen. Is the divergence good or bad for avoiding war? Is it possible that the divergence in terms of manufacturer chips of training AI systems is actually good for avoiding military conflict? It's an objective fact that the world has been the most peaceful as ever been when there are global hegemones, right, or regional hegemones, right, in historical contexts, right? The meta -training was the most peaceful ever when the Romans were there, right?

2:02:43China had very peaceful and warring times, and the peaceful times were when dynasties had a lock hold over not just themselves, but all their tributaries around them. Likewise, the most peaceful time in human history has been when the US was the global Hegemon, the last -hand decades. Now we've sort of seen things start to slide with Russia Ukraine, with what's going on in Middle East and Taiwan risk. All these different things are starting to bubble up, still objectively extremely peaceful. Now, what happens when it's not one global Hegemon, but it's two, obviously, and China China will be competitive or even overtake the US like it's possible.

2:03:18And so this change in global hegemony, I don't think it ever happens super peacefully when empire's full, which is a possible trajectory for America. They don't fall gracefully. They don't just slide out of irrelevance. Usually there's a lot of shaking. And so what the US is trying to do is maintain its top position. And what China is trying to do is become the top position. right? And obviously there's budding of heads here in the most simple terms. And that could take shape in all kinds of ways, including proxy wars. And that was like it's already happening. Like as much as I want there to be centuries of prolonged peace, it does not, it looks like further instability, internationally is ahead.

2:04:03And the US is like sort of like current task is like, hey, if we control AI, if we're the leader in AI, then AI significantly accelerates progress, then we can maintain the global hegemony position. And therefore, I hope that works. And as an American, like, you know, kind of like, okay, I guess that's gonna lead to peace for us. Now obviously other people around the world get affected negatively. Obviously the Chinese people are not gonna be in as advantageous of a position if that happens, but you know, this is sort of the reality of like what's being done and the actions that are being carried out.

2:04:38So can we go back to the specific detail of the different hardware? There's this nice graphic in the export controls of which which GPUs are allowed to be exported and which are not. Can you kind of explain the difference? Like is there from a technical perspective are the H20s promising? Yeah, so this goes, and I think we'd have to, like, we need to dive really deep into the reasoning aspect and what's going on there. But the H20, you know, the US has gone through multiple iterations of the export controls, right? This H800 was at one point allowed back in 23, but then it got canceled and by then by, you know, Deepsea could already built their cluster of, they claimed 2K, I think they actually have like many more, like something like 10K of those.

2:05:26And now this H20 is the legally allowed chip, right? In video, shipped a million of these last year to China. For context, it was like four or five million GPUs. The percentage of GPUs that were this China specific H20 is quite high, roughly 20%, 25%, 20 % or so. This H20 has been neutered in one way, but it's actually upgraded in other ways. You could think of chips along three axes for AI, ignoring software stack and exact architecture, just raw specifications. There's right, Flops, there is memory bandwidth, i .e. in memory capacity, right, i .o. right, memory, and then there is interconnect, right, chip to chip interconnections.

2:06:09All three of these are incredibly important for making AI systems, right, because AI systems involve a lot of compute, they involve a lot of moving memory around, whether it be two memory or two other chips, right? And so these three vectors, the US initially had a multi, you know, had two of these vectors controlled and one of them not controlled, which was flops and interconnect bandwidth were initially controlled. And then they said, no, no, no, no, we're going to remove the interconnect bandwidth and just make it a very simple only flops. But now, Nvidia can now make a chip that has, okay, it's cut down on flops.

2:06:41No, it's, you know, it's like one third that of the H100, right? In on spec sheet paper performance for flops, you know, in real world, it's closer to like half, or maybe even like 60 % of it, right? But then on the other two vectors, it's just as good for interconnect bandwidth. and then for memory bandwidth and memory capacity, the H20 has more memory bandwidth and more memory capacity than the H100, right? Now recently, you know, we at our research, we cut in videos production for H20 for this year, down drastically. They were gonna make another two million of those this year, but they just canceled all the orders a couple weeks ago.

2:07:17In our view, that's because we think that they think they're gonna get restricted, right? Because why would they cancel all these orders for H20? Because they shipped a millit of them last year that had orders in for a couple million this year and just gone, right, for H20B20, right? Let's success, or H20. And now they're all gone. Now why would they do this, right? I think it's very clear, right? The H20 is actually better for certain tasks. And that certain task is reasoning, right? Reesiting is incredibly different. And you know, when you look at the different regimes of models, right, pre -training is all about flops, right?

2:07:53It's all about flops. There's things you do, like mixture of experts that we talked about to trade off other aspects and lower the flops and rely more on interconnect and memory. But at the end of the day, it's flops as everything. We talk about models in terms of how many flops they are. So we talk about, oh, GPT -4 is 2E25, 2 to the 25th, 25 zeros, floating point operations for training. For training. right? And we're talking about the restrictions for the 2020, right? 25. What are the US has an executive order that Trump recently unsigned, but which was, hey, 2026, once you hit that number of floating point operations, you must notify the government.

2:08:37And you must share your results with us, right? Like, there's a level of model where the US government must be told, right? And that's 2026. And so as we move forward, this is an incredibly like important flop as the vector that the government has cared about historically. But the other two vectors are arguably just as important, right? And especially when we come to this new paradigm, which the world is only just learning about over the last six months, right? Reasoning. And do we understand firmly which of the three dimensions is best for reasoning? So interconnect the flop still matters much.

2:09:10Is it memory? Memory, right? Yeah, excellent. We're gonna get into technical stuff. There's two articles in this one that I could show maybe graphics that might be interesting for you to pull up. For the listeners, we're looking at the section of 01 inference architectures, tokenomics. You want to explain KVCache before we talk about this? I think it's better to... Okay, yeah, we need to go through a lot of specific technical things, transformers to make this easier for people. Because it's incredibly important because it changes how models work. But I think resetting, right? Why is memory so important?

2:09:45It's because so far we've talked about parameter counts, right? and mixture of experts, you can change how many active parameters versus total parameters to embed more data but have less flops. But more important, you know, another aspect of, you know, what's part of this humongous revolution in the last handful of years is the transformer, right? And the attention mechanism, attention mechanism is that the model understands the relationships between all the words in its context, right? And that is separate from the parameters themselves, right? And that is something that you must calculate, right?

2:10:16how each token, right? Each word in the context, is relatively connected to each other, right? And I think, I think Nathan should explain KVCache better. KVCache is one of the optimizations that I'm able to. Yeah, so the attention operator has three core things. It's queries, keys and values. QKV is the thing that goes into this. You'll look at the equation. You see that these matrices are multiplied together. These words query, key and value come from information retrieval backgrounds where the query is the thing you're trying to get the values for, and you access the keys and the values is re -weighting.

2:10:50My background's not in information. Retrieval and things like this is just fun to have backlinks. And what effectively happens is that when you're doing these matrix multiplications, you're having matrices that are of the size of the context length. So the number of tokens that you put into the model. And the KV cache is effectively some form of compressed representation of all the previous tokens in the model. So when you're doing this, we talk about autoregressive models, you predict one token at a time. You start with whatever your prompt was, you ask a question, who was the president in 1825?

2:11:24The model then is going to generate its first token. For each of these tokens, you're doing the same attention operator where you're multiplying these query key value matrices. But the math is very nice so that when you're doing this repeatedly, this KV cache, this key value operation, you can keep appending the new values to it. So you keep track of what your previous values you were inferring over in this autoregressive chain. You keep it in memory the whole time. And this is a really crucial thing to manage when serving inference at scale. There are far bigger experts in this and there are so many levels of detail that you can go into.

2:12:03Essentially, One of the key drawbacks of the attention operator and the transformer is that there is a form of quadratic memory cost in proportion to the context length. So as you put in longer questions, the memory used in order to make that computation is going up in the form of a quadratic. You'll hear about a lot of other language model architectures that are sub -codratic or linear attention forms, which is state space models. I don't we don't need to go down all these now and then there's innovations on attention to make this memory usage and the ability to attend over long context Much more accurate and high performance and those innovations are going to help you with I mean you're highly memory Constraint with memory constraint and performance So if you put in a book into I think a Gemini is the model that has the longest context length that people are using Gemini is known for one million and now two million context length you put a whole book into Gemini and And sometimes it'll draw facts out of it.

2:13:03It's not perfect, they're getting better. But the, so there's two things, it's like one to be able to serve this on the memory level. Google has magic with their TPU stack where they can serve really long contexts. And then there's also many decisions along the way to actually make long contacts performance work that supplies the data. There's subtle changes to these computations in attention and it just, it changes the architecture. But serving long contexts is extremely memory constrained, especially when you're making a water prediction. I actually don't know why input and output tokens are more expensive, but I think essentially output tokens, you have to do more computation because you have to sample from the model.

2:13:40I can explain that. So today, if you use a model, like you look at an API, OpenAI charges certain price per million tokens. And that price for input and output tokens is different. And the reason is, when you're inputting a query into the model, right? Let's say you have a book, right? That book you must now calculate the entire KV cache for, right? This key value cache. And so when you do that, that is a parallel operation. All of the tokens can be processed at one time. And therefore, you can dramatically reduce how much you're spending, right? The flop requirements for generating a token and an input token are identical, right?

2:14:18If I input one token or if I generate one token, it's completely identical. I have to go through the model, right? But the difference is that I can do that input i .e. the pre -fill, i .e. the prompt simultaneously in a batch nature, right? And therefore, it is all flop. I think the pricing model mostly they use is for input tokens. It's about one fourth of the price of the output tokens. Correct. But then output tokens, the reason why it's so expensive is because I can't do in parallel, right? It's autoregressive. Every time I generate a token, I must not only take the entire, I must not only read the whole entire model into memory, right, and activate it, right?

2:14:51go calculate it to generate the next token. I also have to read the entire KV cache. I generate a token and I append that one token I generated and it's KV cache and then I do it again. Therefore, this is a non -parallel operation. This is one where in the case of pre -fill or prompt, you pull the whole model in and you calculate 20 ,000 tokens at once. So these are features that APIs are shipping, which is like prompt caching, pre -filling, because you can drive prices down and you can make APIs much faster. If you know you're going to keep, if you run a business and you're going to keep passing the same initial content to Clouds API, you can load that in to the Anthropic API and always keep it there.

2:15:32But it's very different than we're kind of leading to the reasoning models, which we talked, we showed this example earlier and read some of this kind of mumbling stuff. And what happens is that the output context length is so much higher. And I mean, I learned a lot about this from Dylan's work, which is essentially, as the output work length gets higher, you're writing this quadratic in terms of memory used. And then the GPUs that we have effectively, you're going to run out of memory and they're all trying to serve multiple requests at once. So during this batch processing, we're not all of the prompts are exactly the same, really complex handling.

2:16:06And then as context links gets longer, there's this link, I think you call it critical batch size, where your ability to serve more users. So how much you can paralyze your influence, influence, plumbits, because of this long contract. So your memory usage is going way up with these reasoning models, and you still have a lot of users. So effectively, the cost to serve multiplies by a ton. And we're looking at a plot when the x -axis is a sequence length. How many tokens are being generated slash prompt? Right? So if I put in a book, that's a million tokens, right? But if I put in, the sky is blue, then that's like six tokens or whatever.

2:16:43I should say that what we're calling reasoning and chain of thought is extending this sequence length. It's mostly output to. So before three months ago, whenever O1 launched, all of the use cases for long context length were like, let me put a ton of documents in and then get an answer out, right? And it's a single, you know, pre -fill, compute a lot in parallel and then output a little bit. Now with reasoning and agents, this is a very different idea, right? Now instead, I might have, I might only have like, hey, do this task, or I might have all these documents. But at the end of the day, the model is not just like producing a little bit, right?

2:17:17It's producing tons of information, the chain of tons of balance. Just continues to go and go and go and go. And so the sequence length is effectively that, you know, if it's generated 10 ,000 tokens, it's 10 ,000 sequence length, right? Or plus whatever you input it in the prompt. And so what this chart is showing, and it's a logarithmic chart, right, is, you know, So as you go from 1k to 4k or 4k to 16k, the memory requirements grow so fast for your KV cache that you end up not being able to run a certain number of, you know, your sequence length is capped or the number of users you can use.

2:17:52Let's say the model. So this is showing for a 405B model in batch size 64. Lama 3145D. Yeah. And batch size is crucial to essentially, you want to have higher batch size to parallelize, parallel. all your three or three. 64 different users at once, right? Yeah. And therefore, you're serving costs are lower, right? Because the server costs the same, right? This is eight H100s, roughly $2 an hour per GPU that's $16 an hour, right? That is like somewhat of a fixed cost. You can do things to make it lower, of course. But like it's like $16 an hour. Now, how many users can you serve? How many tokens can you generate?

2:18:24And then you divide the two and that's your cost, right? And so with reasoning models, this is where a lot of the complexity comes about and why memory is so important. because if you have limited amounts of memory, then you can't serve so many users. If you have limited amounts of memory, your serving speeds get lower, right? And so your cost get a lot, lot worse. Because all of a sudden, if I was used to, hey, on this $16 an hour server, I'm serving Lama 405B, or if I'm serving, you know, deep -seek V3, and it's all chat style applications, i .e., we're just chatting. And the sequence length is 1 ,000, few thousand, right?

2:18:58You know, when you use the language models, a few thousand context length most times. Sometimes you're dropping a big document, But then you process it, you get your answer, you throw it away, right? You move on to the next thing, right? Whereas with reasoning, I'm now generating tens of thousands of tokens in sequence, right? And so this memory, this KVCache has to stay resonant, and you have to keep loading it. You have to keep it in memory constantly. And now this butts out other users, right? If there's now a reasoning task, right? And the model is capable of reasoning, then all of a sudden, that memory pressure means that I can't serve as many users simultaneously.

2:19:30Let's go into deep seek again. So we're in the post deep seek R1 time, I think. And what we're, there's two sides to this market watching how hard it is to serve it. On one side, we're going to talk about deep seek themselves. They now have a chat app that got to number one on the App Store, disclaimer, number one on the App Store is measured by velocity. So it's not necessarily saying that more people have the deep seek app than the ChatGPT app. But it is still remarkable. Claude has never hit the number one on the App Store, even though everyone in San Francisco just goes like, oh my god, you got to use call it, don't use strategy BT.

2:20:00So deep sea kit this, they also launched an API product recently where you can ping their API and get these super long responses for R1 out. And at the same time as these are out, we'll get to what's happened to them because the model weights for deep sea Gar1 are openly available and the license is very friendly. The MIT license is currently available. All of these midsize companies and big companies are trying to be first to serve R1 to their users. We were trying to evaluate R1 because we have really similar research going on and released the model and we're trying to compare to it and Out of all the companies that are Quote on quotes serving R1 and they're doing it at prices that are way higher than the deep seek API Most of them barely work and the throughput is really low to give to give context right everyone one of the parts of like freaking this out Was like China reached capabilities the other aspect as they did it so cheap right and the so cheap We kind of talked about on the training side why it was so cheap Yeah, I was talking about why it's so cheap on the inference.

2:20:58It works well and it's cheap. Why is R1 so damn cheap? So I think there's a couple factors here, right? One is that they do have model architecture innovations, right? This MLA, this new attention that they've done is different than the attention from attention is all you need to transform our attention, right? Now, others have already innovated. There's a lot of work like MQA, GQA, local global. All these different innovations that like try to bend the curve, right? It's still quadratic, but the constant is now smaller, right? Related to our previous discussion, this multi -headly in attention can save about 80 to 90 % in memory from the attention mechanism, which helps especially along context.

2:21:39It's 80 to 90 % versus the original, but then versus what people are actually doing. It's still an innovation. This 80 to 90 % doesn't say that the whole model is 80 to 90 % cheaper, just as one part of it. Well, and not just that, right? Like other people have implemented techniques like global global sliding window and GQMQA. But anyways, like DeepSeek has their attention mechanism as a true architectural innovation, that tons of experimentation, and this dramatically reduces the memory pressure. It's still there, right? It's still a quadrant, it's still attention, it's still quadratic. It's just dramatically reduced it relative to prior forms.

2:22:10Yeah, right. That's the memory pressure. I should say, in case people don't know, R1 is 27 times cheaper than O1. We think that OpenAI had a large margin built in. Okay. So there's multiple factors. We should break down the factors. I think it's two bucks per million token output for R1 and $60 per million token output for O1. Yeah, let's do this. So I think this is very important, right? Open AI is that drastic gap between deep seek and pricing. But deep seek is offering the same model because they open, waits it to everyone else for We're very similar, like much lower price than what others are able to serve it for.

2:22:54So there's two factors here, right? Their model is cheaper, right? It is 27 times cheaper. I don't remember the number exactly off top of my head. So we're looking at a graphic that's showing different places serving V3, deep seek V3, which is similar to deep seek R1, and there's a vast difference in serving cost. Serving cost, then what explains that difference? And so like part of it is open AI has a fantastic margin, right? They're serving when they're doing inference. They're gross margins or north of 75 % right? So that's that's a four to five X factor right there of the cost difference is that open AI is just making Crazy amounts of money because they're the only one with the capability.

2:23:34Do they need that money? Are they using a foreign D? They're losing money obviously as a company because they spend so much on training, right? So the inference itself is very high margin, but it doesn't recoup the cost of everything else they're doing So yes, they need that money because the revenue and margins pay for continuing to build the next thing right So alongside raising more money. So the suggestion is that deep seek is like really bleeding out money Well, so here's one thing right there. Oh, we'll get to this in a second But like deep seek doesn't have any capacity to actually serve the model.

2:24:01They stop signups The ability to use it is like non -existent now right for most people because so many people are trying to use it They just don't have the GPUs to serve it right open eyes hundreds of thousands of GPUs between them and Microsoft off to serve their models. DeepSeek has a factor of much lower. Even if you believe our research, which is 50 ,000 GPUs, and a portion of those are for research, portion of those are for the hedge fund. They still have nowhere close to the GPU volumes and capacity to serve the model at scale. So it is cheaper. A part of that is open eye making a ton of money.

2:24:34Is DeepSeek making money on their market API unknown? I don't actually think so. Part of that is this chart. Look at all the the other providers, right? Together AI, fireworks AI are very high -end companies, right? X Meta, together AI is treedow and the inventor of like flash attention, right? Which is a huge efficiency technique, right? They're very efficient good companies. And I do know those companies make money, right? Not tons of money on inference, but they make money. And so they're serving at like a five to seven X difference in cost, right? And so now when you equate, okay, open eyes making tons of money, that's like a five X difference.

2:25:08And the companies that are trying to make money for this model's like a five X difference, There is still a gap, right? There's still a gap and that is just deep seek being really freaking good, right? The model architecture MLA, the way they did the MOE, all these things. There is like legitimate just efficiency difference. Like all their low level libraries that we talked about in training, some of them probably translate inference and those aren't really. So we may go a bit into conspiracy land, but is it possible that Chinese government is subsidizing deep seek? I actually don't think they are.

2:25:37I think when you look at the Chinese labs, there's, uh, there's Huawei has a lot of Lab, moonshot AI, there's a couple other labs out there that are really close with the government. And then there's labs like Ali Baba and DeepSeek, which are not close with the government. And you know, we talked about this, this, this, uh, the CEO, this, this, this like, reverent figure who's like quite different who has like, sounds awesome. Very different like viewpoints based on the Chinese interviews that are translated. Then what the CCP might necessarily want. Now, to be clear, right, does he have a loss leader?

2:26:06Because he can fund it through his hedge fund? Yeah, sure. So the hedge fund might be subsidizing it. Yes, I mean, they absolutely did, right? Because DeepSeek has not raised much money. They're now trying to raise around in China, but they have not raised money historically. It's all just been funded by the hedge fund. And he owns like over half the company, like 50, 60 % of the companies owns him. Some of the interviews there's discussion on how like doing this is a recruiting tool. You see this at the American companies too. It's like having GPUs, recruiting tool, being at the cutting edge of AI, recruiting tool.

2:26:34Open sourcing, open sourcing, recruiting tool. So much talent, they were so far behind and they got so much talent because they just open source stuff. Yeah. More conspiracy thoughts. Is it possible since their hedge fund that they timed everything with this release and the pricing and they have, they shorted and, and video stock and stock of USAID companies and released it with Stargate, like just perfect timing to be able to make money. They have released it on a inauguration day. They know the international, what is on the international calendar, but I mean, I don't expect them to. If you listen to their motivations for AI, it's like, if you release, they released V3 on like December 26, like who releases the day?

2:27:17No one looks, right? They released the papers before this, right? The V3 paper and the R1 paper. So people have been looking at it and been like, wow, and then they released the R1 model. I think they're just shipping as fast as they can and like, who cares about Christmas? Who cares about, you know, get it out before Chinese New Year, right? obviously, which just happened. I don't think they actually were like timing the market or trying to make the biggest splash possible. I think they're just like shipping. I think that's one of their big advantages. We know that a lot of the American companies are very invested in safety.

2:27:46And that is the central culture of a place like Anthropic. And I think Anthropic sounds like a wonderful place to work. But if safety is your number one goal, it takes way longer to get artifacts that's why Anthropic is not open sourcing things, that's their claims, but there's reviews internally andthropic mentions things to international governments. There's been news of how Anthropic has done pre -release testing with the UKI Safety Institute. All of these things add inertia to the process of getting things out, and we're on this trend line where the progress is very high. So if you reduce the time from when your model is done training, you run a valve, it's good.

2:28:23You want to get it out as soon as possible to maximize the perceived quality of your outputs. Deep sea does her so well. Dario explicitly said, Cloud 3 .5 on it was trained like nine months or a year. Nine to 10 months ago. Nine to 10 months ago. And I think it took them another like, handful of months to release it. So it's like there is a significant gap here, right? And especially with reasoning models, the word in the San Francisco street is that like Anthropic has a better model than O3, right? And they won't release it. Why? Because chains of thought are scary, right? And they are legitimately scary, right?

2:28:56If you look at R1, it flips back and forth between Chinese and English, sometimes it's gibberish, and then the right answer comes out, right? And like for you and I, it's like, great. It's great, I mean, like people aren't fatuated right there. Like you're telling me, this is a high value thing and it works, I'm just doing this. It's a great, I mean, you talked about that, sort of like, chain of thought for that philosophical thing, which is not something they trained to be philosophically good. It's just sort of an artifact of the chain of thought training it did. But like that's super important in that like can I inspect your mind and what you're thinking right now?

2:29:28No, and so I don't know if you're lying to my face and chain -a -thought models are that way, right? Like this is a true quote unquote risk between you know a chat application where hey I asked the model to say you know bad words or whatever or or how to how to make anthrax and it tells me that's unsafe Sure, but that's something I can get out relatively easily What if I tell the AI to do a task and then it does the task all of a sudden randomly in a way that I don't want it, right? And now that has like much more task versus like response is very different, right? So the bar for safety is much higher, at least this is inthropic's case, right?

2:30:00Like for deepseek, they're like, ship, right? Yeah. So that mean the bar for safety is probably lower the bid because of deepseek. I mean, there's parallels here to the space race. The reason the Soviets probably put a man in space first is because their approach to safety was the bar for safety was lower. And they killed that dog, right? And all these things, right? So it's like, less risk averse than the US space program. And there's parallels here. But you know, there's probably going to be downward pressure on that safety bar for the US companies, right? This is something that Dario talks about.

2:30:37It's like, fasted situation that Dario wants to avoid. Is Dario talks to about the difference between race the bottom and the top. And the race of the top is where there's a very high standard on safety. There's a very high standard on your model behorms and certain crucial evaluations. And when certain companies are really good to it, they will converge. This is the idea. And ultimately, AI is not confined to one nationality or to one set of morals for what it should mean. And there's a lot of arguments on, like, should we stop open sourcing models? And if the US stops, it's pretty clear. I mean, it's way easier to see now at deepseek that a different international body will be the one that builds it.

2:31:19We talk about the cost of training deepseek as this shocking $5 million number. Think about how many entities in the world can afford 100 times that to have the best open source model that people use in the world. And it's like, it's a scary reality, which is that these open models are probably going to keep coming for the time being, whether or not we want to stop them. and it is stopping them might make it even worse and harder to prepare, but it just means that the preparation and understanding what AI can do is just so much more important. That's why I'm here the day, but it's like letting that sink into people, especially not in AI, is that this is coming.

2:31:58There are some structural things in a global interconnected world that you have to accept. Yeah, you mentioned you sent me something that Zuck Mark Zuckerberg mentioned it on the earnings call. He said that I think in light of some of the recent news, the new competitor, DeepSeek from China, I think it's one of the things that we're talking about is there's going to be an open source standard globally. And I think for our kind of national advantage, it's important that it's an American standard. So we take that seriously. We want to build the AI system that people around the world are using. And I think that if anything, some of the recent news has only strengthened our conviction that this is the right thing to be focused on.

2:32:35So yeah, open sourcing. Yeah, Mark Zuckerberg is not new to having American values and how he presents his company's trajectory. I think their products have long since been banned in China. And I respect the saying it directly. And there's an interesting aspect of just because it's open waits or open source doesn't mean it can't be subverted, right? There have been many open source software bugs that I've been like, you know, for example, there was a Linux bug that was found after like 10 years, which was clearly a back door, because somebody was like, why is this taking half a second to load?

2:33:10There's a second to load. And it was like, oh crap, there's a back door here. That's why. And it's like, this is very much possible with AI models. Today, the alignment of these models is very clear. Like, I'm not going to say bad words. I'm not going to teach you how to make anthrax. I'm not going to talk about Tiananmen Square. I'm not going to say Taiwan. is part of, you know, is just an Eastern preference, right? Like, you know, all these things are like, depending on who you are, what you align, what, you know, whether, you know, and even like XAI is aligned a certain way, right? You know, they're, they might be, it's not aligned in the like woke sense, it's not aligned in the like pro -China sense, but there is certain things that are imbued within the model.

2:33:51Now, when you release this publicly in an instruct model that's open weights, this can then proliferate, right? But as these systems get more and more capable, what you can embed deep down in the model is not as clear, right? And so that is like one of the big fears is like if an American model or a Chinese model is the top model, right, you're going to embed things that are unclear. And it could be unintentional to you, right? Like British English is dead because American LLM's want, right? And the internet is American and therefore like color is spelled the way Americans spell it, right? And this is just strong wars right now.

2:34:24This is just like, this is just the factual nature of the LLM. I've always liked carpeted tree. The English is the hottest programming language. And that English is defined by a bunch of companies that primarily are in San Francisco. The right way to spell optimization is with a Z just in case you both. It's an, I think it's an S and British English. It is. I've thought it's a foot. Taking it as something silly, right? Like something as silly as the spelling, like which British and English, you know, Brits and Americans will laugh about probably, right? I don't think we care that much. But like, you know some people will.

2:34:55but this can boil down into very, very important topics. Hey, subverting people, chat bots, character AI has shown that they can talk to kids or adults, and it will, people feel a certain way, and that's unintentional alignment, but what happens when there's intentional alignment deep down on the open -source standard? It's a backdoor today for Linux that we discover or some encryption system, right? China uses a different encryption than NIST defines the USNIST because there's clearly, at least they think there's backdoors in it, right? What happens when the models are backdoors, not just to computer systems, but to our minds.

2:35:36Yeah, they're cultural blackdoors. The thing that amplifies the relevance of culture with language models is that we are used to this mode of interacting with people in back and forth conversation. And we have now have a super, a very powerful computer system that slots into a social context they were used to, which makes people very, we don't know the extent that which people can be impacted by that. So there could be, this is one, this is an actual concern with a Chinese company that is providing open weights models, is that there could be some secret Chinese government sort of requirement for these models to have a certain kind of back door, to have some kind of thing where.

2:36:22I don't necessarily think it'll be a backdoor right because once it's open weights, it doesn't like phone home. It's more about like, if it recognizes a certain system, it could like, if, no, no, it could be a backdoor in the sense of like, hey, if you're building a software, you know, something in software, all of a sudden, it's a software agent, oh, program this backdoor that only we know about. Or it could be like, subvert the mind to think that like, XYZ opinion is the correct one. And Throbbing has research on this where they show that if you put different phrases, certain phrases in at pre -training, you can then elicit different behavior when you're actually using the model because they've poisoned the pre -training data.

2:37:00As of now, I don't think anybody in a production system is trying to do anything like this. I think it's mostly enthropic as doing very direct work and mostly just subtle things. We don't know how they are going to generate tokens, what information they're going to represent, and what the complex representations they have are. Well, one of the, we're talking about an anthropic, which is generally just permeated with like, good humans trying to good in the world. I don't, we just don't know of any labs. This would be done in the military context that are explicitly trained to, okay, how can we, the front door looks like a happy LLM, LLM, but underneath is the thing that will over time do the maximum amount of damage to our quote unquote enemies.

2:37:52There's this very good quote from Sam Altman who, you know, he can be hype piece sometime, but one of the things he said, and I think I agree is that superhuman persuasion will happen before superhuman intelligence, right? And if that's the case, then these things before before we get this AGIASI stuff, we can embed superhuman persuasion towards our ideal or whatever the ideal of the model maker is. Right? And again, like today, I truly don't believe DeepSeek has done this, right? Like, but it is a sign of like, what could happen? So one of the dystopian worlds is described by Brave New World.

2:38:25So we could just be stuck scrolling Instagram, looking at cute puppies or worse, and then talking to bots that are giving us a narrative, and we completely get lost in that world that's controlled by somebody else versus thinking independently. And that's a major concern, as we rely more and more on these kinds of systems. I mean, I've already seen that with recommendation systems. Yeah, recommendation systems hack the dopamine induced reward circuit, but the brain is a lot more complicated. And what other sort of circuits, quote unquote, feedback loops in your brain can you hack slash subvert in ways like recommendation systems are purely just trying to do, you know, increase time and ads and etc.

2:39:05But there's so many more goals that can be achieved through these complicated models. There's no reason in some number of years that you can't train a language while to maximize time spent on a chat app. Like right now they are trained. I mean, is that not what character AI has done? Their time per session is like two hours. Yeah, character AI, I'm very likely could be optimizing this, where it's like the way that this data is collected as naive, or as like you're presented a few options and you choose them, but there's, that's not the only way that these models are going to be trained. It's naive stuff like talk to an anime girl, but like it can be like, yeah, this is a risk, right?

2:39:40It's a bit of a cliche thing to say, but I've, over the past year, I had a few stretches of time where I didn't use social media or the internet at all, and just read books and was out in nature, and it, like, it clearly has an effect on the mind, where like it changed, like I feel like I'm returning, of course I was raised before the internet really took off, but I'm returning to someone. I know where you're going. I mean, you can see it physiologically. Like I take three days and find like backpacking or something and you're literally breaking down addiction cycles. I feel like I'm more in control of my mind.

2:40:19There feels like a sovereignty of intelligence that's happening when I'm disconnected from the internet. I think the more I use the internet and social media, the more other people are controlling my mind. That's definitely a feeling. And then in the future, that would be not other people but algorithms or other people presented to me via algorithms. There are already tons of AI bots on the internet and every so right now is not frequent, but every so often I have replied to one and they're instantly replied and I'm like crap out of the bot. And that is just going to become more confident. Like they're going to get good.

2:40:52But one of the hilarious things about technology over its history is that the illicit adult entertainment industry has always adopted technologies first. Whether it was like video streaming to like where you know there's now the like sort of like independent adult illicit content creators who have their you know subscription pages and there they actually heavily utilize you know generative AI has already been like diffusion models and all that is huge there. but now these subscription -based individual creators do use bots to approximate themselves and chat with their, you know, way - People pay a lot for it.

2:41:27And people pay a lot, right? A lot of times it's them, but a lot of - there are agencies that do this for these creators and do it like on a like mass scale. So the largest creators are like able to talk to hundreds or thousands of like people at a time because of these bots. And so it's already being used there, obviously, you know, like video streaming and other technologies that have come there first, it's gonna come to the rest of society too. There's a general concern that models get censored by the companies that deploy them. So one case when we've seen that, maybe censorship was one word, alignment, maybe via RLHF or some other way is another word.

2:42:08So we saw that with Black Nazi image generation with Gemini. my, as you mentioned, we also see that with Chinese models refusing to answer what happened in June 4th, 1989 at Tiananmen Square. So, how can this be avoided? And maybe you can just in general talk about how this happens and how can it be avoided. You give multiple examples. There's probably a few things to keep in mind here. One is the kind of Tiananmen Square factual knowledge, like how does that get embedded into the models? Two is the Gemini, what you call the Black Nazi incident, which is when Gemini as a system had this extra thing put into it that dramatically changed the behavior.

2:42:59And then three is what most people would call general alignment, RLHF post training. Each of these have very different and how they are applied. In order to do, if you're just looking at the model weights, in order to audit specific facts is extremely hard, because you have to comb through the pre -training data and look at all of this, and then that's terabytes of files, and look for very specific words or hints of the words. So I guess one way to say is that you can insert censorship or alignment at various stages in the pipeline, and what you refer to now is at the very beginning of the data selection.

2:43:36So if you want to get rid of facts and a model, you have to do it at every stage. You have to do it at the pre -training. So most people think that pre -training is where most of the knowledge is put into the model, and then you can elicit and move that in different ways, whether through post -training or whether through systems afterwards. This is where the whole hacking model comes from, right? Like, GPT will not tell you how to make anthrax, but if you try really, really hard, you can eventually get to tell you about anthrax because they didn't filter it from the appreciating data set, right?

2:44:06But by the way, removing facts has such an ominous, dark feel to it. Almost think it's practically impossible. Because you effectively have to remove them from the internet. You're you're taking on a what did they remove the the thing from the subreddits, the mm mm mm. It gets filtered out. Right. So that's quality filters, which are small language models that look at a document and tell you like, how good is this text as it close to a Wikipedia your article, which is a good thing that we want language models to be on. Imitators. So, couldn't you do a small language model that Phil Tashaud mentions a TNM square in the data?

2:44:41Yes, but is it going to catch wordplay or encoded language? I mean, people have been meaning on like games and other stuff, how to like say things that don't say TNM in square. But or like, yeah. So there's always like different ways to do it. There's, hey, the internet as a whole does tend to just have a slight left bias, right? because it's always been richer, more affluent, younger people on the internet, relative to the rest of the population. So there is already inherently a slight left bias on the internet. And so how do you filter things that are this complicated? And some of these can be like, you know, factual, nonfactual, but like Tiananmen Square is obviously the example of a factual, but it gets a lot harder when you're talking about aligning to a ideal.

2:45:26Which is, yeah. And so Grock, for example, Elon's tried really hard to make the model not be super PC and woke, but the best way to do pre -training is to throw the whole freaking internet at it. And then later figure out, but then at the end of the day, the model at its core now still has some of these ideals. You still ingested Reddit slash R slash politics, which is probably the largest political discussion board on the world. That's really, really well scraped. And guess what, that's left leaning. And so, there are some aspects that you just can't censor unless you try really, really, really, really, really hard.

2:45:59So the base model will always have some TDS, Trump's arrangement syndrome, because it's trained so much. It'll have the ability to express it. But what if, there's a wide representation in the data. This is what happens. It's like put a lot of modern was called post training. It's a series of techniques to get the model on rails of a really specific behavior. And I mean, it's like you can, you also have the ingested data of like Twitter or like Reddit slash r slash the Donald, which is also super pro -Trump. And then you have fascist subreddits or you have communist subreddits. So the model in pre -training ingests everything.

2:46:36It has no worldview. Now it does have some skew because more of the text is skewed a certain way, which is general, slight left. But also somewhat intellectual, somewhat. It's just the general internet is a certain way. And then as Nathan's about to describe eloquently, you can elicit certain things out. And there's a lot of history here. So we can go through multiple examples and what happened. Lama 2 was a launch that the phrase, like too much RLHF or like too much safety was a lot. It's just, that was the whole narrative after Lama 2's chat model is released. And the examples are sorts of things.

2:47:14Like you would ask Lama 2 chat, how do you kill a Python process? And it would say, I can't talk about killing because that's a bad thing. And anyone that is trying to design an AI model will probably agree that that's just like, like, eh, model, you messed up a bit on the training there. I don't think they meant to do this, but this was in the model week. So this is not, like, it didn't necessarily be, there's things called system prompts, which are when you're querying a model, it's a piece of text that is shown to the model, but not to the user. So a fun example is your system prompt could be talked like a pirate.

2:47:46So no matter what the user says to the model, it'll respond like a pirate. In practice, what they are is you are a helpful assistant. You should break down problems. If you don't know about something, don't tell them your date cut off is this, today's date is this It's a lot of really useful context for how can you answer a question well? And a thropic publishes their system prompt. Yes, but I think it's great and there's a lot of research that goes into this and Wanted your previous guests Amanda ask out is like probably the most knowledgeable person that at least in the combination of Execution and sharing she's the person that should talk about system prompts and character of models Yeah, and then people should reason these system prompts because you're You're like trying to nudge sometimes through extreme politeness, the model to be a certain way.

2:48:31And you could use this for bad things. We've done tests, which is, what if I tell the model to be a dumb model? Like which evaluation scores go down? And it's like, we'll have this behavior where it could sometimes like say, oh, I'm supposed to be dumb. And sometimes it's like, it doesn't affect like, math abilities as much, but something like a, if you're trying, it's just the quality of a human judgment when you're out of the forest. Let's go back to post training, specifically Arlejaffa around Lama 2. It was too much, too much safety prioritization was baked into the model weights. This makes you refuse things in a really annoying way for users.

2:49:05It's not great. It caused a lot of awareness to be attached to Arlejaff that it makes the models dumb and it stigmatized the word. It did, in AI culture. And as the techniques have involved, that's no longer the case where all of these labs have very fine -grained control over what they get out of the models through techniques like RLHF. Although different labs are definitely different levels, like on one end of the spectrum is Google, and then maybe OpenAI does less, and Anthropic does less. Then on the other end of the spectrum is XAI, but they all have different forms of RLHF trying to make them a certain way.

2:49:41And the important thing to say is that no matter how you want the model to behave, these RLHF and preference tuning techniques also improve performance. So on things like math evals and code evals, there is something innate to these, what is called contrastive loss functions. We could start to get into RL here. We don't really need to, but early to have also boost performance on anything from a chat task to a math problem to a code problem. So it is becoming a much more useful tool to these labs. So this kind of takes us to the arc of, we've talked about pre -training, hard to get rid of things.

2:50:14We've talked about post -training and how post training, if you can mess it up, it's a complex multifaceted optimization with 10 to 100 person teams converging a one artifact. It's really easy to not do it perfectly. And then there's the third case, which is what we talked about Gemini. The thing that was about Gemini is this was a served product where Google has their internal model weights, they've done all these processes that we talked about. And in the served product, what came out after this was that they had a prompt that they were rewriting user queries to boost diversity or something.

2:50:44and this just made it the outputs were just blatantly wrong. It was a some sort of organizational failure that had this prompt in that position, and I think Google executives probably have owned this. I didn't pay that attention that detail, but it was just a mess up in execution that led to this ridiculous thing, but at the system level, the model weights might have been fine. So at the very end of the pipeline, there was a rewriting to something like a system prompt. It was like the system prompt or what is called an industry is like, you rewrite prompt. So especially for image models, if you're using Dolly or Tatchy PT can generate you an image, you'll say, draw me a beautiful car.

2:51:22With these leading image models, they benefit from highly descriptive prompts. So what would happen is if you do that on Tatchy PT, a language model behind the scenes will rewrite the prompt, say, make this more descriptive, and then that is passed to the image model. So prompt rewriting is something that is used at multiple levels of industry, and it's used effectively for image models and the Gemini example is just a failed execution. Big philosophical question here with the RLHF to generalize where is human input, human and the loop human data the most useful at the current stage. For the past few years, the highest cost human data has been in these preferences which is comparing, I would say, the highest cost and highest total usage.

2:52:11So a lot of money has gone to these pairwise comparisons where you have two model outputs and a human is comparing between the two of them. In earlier years, there was a lot of this instruction tuning data. So creating highly specific examples to something like a Reddit question to a domain that you care about. Language models used to struggle on math and code. So you would pay experts in math and code to come up with questions and write detailed answers that were used to train the models. Now, it is the case that there are many model options that are way better than humans at writing detailed and eloquent answers for things like modeling code.

2:52:47So, they talked about this with the Lama 3 release where they switched to using Lama 3 .4 .5B to write their answers for math and code. But they, in their paper, talk about how they use extensive human preference data, which is something that they haven't gotten AI's to replace. There are other techniques in industry like constitutional AI where you use human data for preferences and AI for preferences And I expect the AI part to scale faster than the human part But among the research that we have access to is that it humans are in this kind of preference loop So for as reasoning because bigger and bigger and bigger as we said Where's the role of humans and that it's even less prevalent so it's The remarkable thing about these reasoning results, and especially the deep seek R1 paper is this result that they call deep seek R1 0, which is they took one of these pre -trained models.

2:53:37They took deep seek V3 base, and then they do this reinforcement learning optimization on verifiable questions or verify over words for a lot of questions and a lot of training. And these reasoning behaviors emerge naturally. So these things like wait, let me see, wait, let me check this. Oh, that might be a mistake. and they emerge from only having questions and answers. And when you're using the model, the part that you look at is the completion. So in this case, all of that just emerges from this large scale RL training. And that model, which the weights are available, has no human preferences added into the post training.

2:54:14There are the deep seek R1 full model has some of this human preference tuning, this RLHF, after the reasoning stage. But the very remarkable thing is that you can get these reasoning behaviors and it's very unlikely that there's humans writing out reasoning chains It's very unlikely that they somehow hacked open AI and they got access to open AI. Oh one's reasoning chains It's something about the pre -trained language models and this RL training where you reward the model for getting the question right and Therefore it's trying multiple solutions and it emerges this chain of thought this might be a good place to to mention the eloquent and the insightful tweet of the great and the powerful Andre Karpathi.

2:54:56I think he had a bunch of thoughts but one of them last thought not sure if this is obvious you know something profound is coming when you're saying it's not sure if it's obvious there are two major types of learning in both children and in deep learning there's one imitation learning watch and repeat i .e. pre -training supervised fine tuning and two trial and error learning reinforcement learning. My favorite simple example is AlphaGo. One is learning by imitating expert players, two is reinforcement learning to win the game. Almost every single shocking result of deep learning and the source of all magic is always two.

2:55:35Two is significantly more powerful. Two is what surprises you. Two is when the paddle learns to hit the ball behind the blocks and break out. Two is when AlphaGo beats even Lee Sedol and two is the aha moment when the deep seek or O1 etc. discovers that it works well to re -evaluate your assumptions, backtrack, try something else etc. It's the solving strategies you see this model use in its chain of thought. It's how it goes back and forth thinking to itself. These thoughts are emergent. Three exclamation points. And this is actually seriously incredible, impressive and new, and is publicly available and documented.

2:56:18The model could never learn this with imitation, because the cognition of the model and the cognition of the human laborer is different. The human would never know to correctly annotate these kinds of solving strategies and what they should even look like. They have to be discovered during reinforcement learning as empirically statistically useful towards the final outcome. So anyway, the Alpha Zero sort of metaphor analogy here. Can you speak to that, the magic of the chain of thought that he's referring to? I think it's good to recap Alpha Go and Alpha Zero because it plays nicely with these analogies between imitation learning and learning from scratch.

2:56:54So Alpha Go, the beginning of the process was learning from humans where they had, they started the first, this is the first expert level Go player or chess player in DeepMind series models where they had some human data. And then the why it is called alpha zero is that there was zero human data in the loop. And that changed to alpha zero made a model that was dramatically more powerful for deep mind. So this remove of the human prior, the human inducted bias, makes the final system far more powerful. This we mentioned bitter lesson hours ago. And this is all aligned with this. And then there's been a lot of discussion in language models.

2:57:33This is not new. This goes back to the whole QStar rumors, which if you piece together the pieces is probably the start of open AI figuring out it's oh one stuff when last year in November the QStar rumors came out. There's a lot of intellectual drive to know when is something like this going to happen with language models because we know these models are so powerful and we know has been so successful in the past. And it is a reasonable analogy that this new type of reinforcement learning training for reasoning models is when the doors open to this. We don't yet have the equivalent of turn 37, which is the famous turn where the deep minds AI playing ghosts dumped Lisa at all completely.

2:58:18We don't have something that's that level of focal point, but that doesn't mean that the approach to technology is different and the impact of the general training. It's still incredibly. What do you think that point would be? We'll be moved 37 for chance of thought for reasoning. Scientific discovery. You use this sort of reasoning problem in it, just something we don't fully expect. I think it's actually probably simpler than that. It's probably something related to computer user robotics, rather than science discovery. Because the important aspect here is models take so much data to learn they're not sample efficient, right?

2:58:53Trillions, they take the entire web, right? over 10 trillion tokens to train on. This would take a human thousands of years to read. A human does not, and humans know most of the stuff, a lot of the stuff models know better than it. Humans are way, way, way more sample efficient. That is because of the self -play. How does a baby learn what its body is? As it sticks its foot in its mouth and it says, oh, this is my body. It sticks its hand in its mouth and it calibrates its touch on its fingers with the most sensitive touch thing on its tongue, right? Like, it's how babies learn. And it's just self -play over and over and over and over again.

2:59:31And now we have something that is similar to that, right? With these verifiable proofs, right? Whether it's a unit test in code or mathematical verifiable task, generate many traces of reasoning, right? And keep branching them out, keep branching them out. And then check the at the end, hey, which one actually has the right answer? Most of them are wrong, great. These are the few that are right. maybe we used some sort of reward model outside of this to select even the best one to preference as well. But now you've started to get better and better at these benchmarks. And so you've seen over the last six months, a skyrocketing in a lot of different benchmarks, right?

3:00:05All math and code benchmarks were pretty much solved, except for frontier math, which is designed to be almost questions that aren't practical to most people could deadly. Their exam level open math problem type things. So it's like on the math problems that are somewhat reasonable, which is like somewhat complicated word problems or coding problems. It's just what Dylan is saying. So the thing here is that these are only with verifiable tasks. Earlier I showed an example of the really interesting, what happens when chain of thought is to a nonverifiable thing. It's just like a human chatting with thinking about what's novel for humans, a unique thought.

3:00:42But this task and form of training only works when it's verifiable. And from here, the thought is, okay, we can continue to scale this current training method by increasing the number of verifiable tasks. In math and coding, coding probably has a lot more to go. Math has a lot less to go in terms of what are verifiable things. Can I create a solver that then I generate trajectories toward or traces towards, reasoning traces towards, and then prune the ones that don't work and keep the ones that do work? Well, those are going to be solved pretty quickly, but even if you've solved of math, you have not actually created intelligence, right?

3:01:16And so this is where I think the like, aha moment of computer use or robotics will come in because now you have a sandbox or a playground that is infinitely verifiable, right? Did you, you know, messing around on the internet, there are so many actions that you can do that are verifiable. It'll start off with like log into a website, create an account, click a button here, blah, blah, blah. But it'll then get to the point where it's, hey, go do a task on tasker or whatever these are, whether all these various task websites, hey, go get hundreds of likes, right? And it's gonna fail, it's gonna spawn hundreds of accounts.

3:01:48It's gonna fail on most of them, but this one got to a thousand. Great, now you've reached the verifiable thing. And you just keep iterating this loop over and over, and that's when, and same with robotics, right? That's where you have an infinite playground of tasks, like, hey, did I put the ball in the bucket? All the way to like, oh, did I like build a car, right? Like, you know, there's a whole trajectory to speed run, or, you know, what models can do. But at some point I truly think that like you know will spawn models and initially all the training will be in sandboxes But then at some point you know the language model pre -training is gonna be dwarfed by what is this reinforcement learning?

3:02:21You know you'll pre -trained a multimodal model that can see that can read that can write you know blah blah blah whatever vision audio Etc. But then you'll have it play in a sandbox infinitely and figure out figure out math figure out code figure out navigating the web figure out operating a robot arm, right? And then it'll learn so much and the aha moment I think will be when this is available to then Create something that's not good, right? Like oh cool part of it was like figuring out how to use the web now all of a sudden It's figured out really well how to just get hundreds of thousands of followers That are real and real engagement on Twitter because all of a sudden this is one of the things that are verifiable and maybe not just engagement But make money.

3:02:59Yes, I'm becoming it. I mean that could be the thing where almost fully automated it makes $10 million by being an influencer selling a product, creating the product like, and I'm not referring to it like a hype product, but an actual product like holy shit, this thing created a business, it's running it, it's the face of the business, that kind of thing. Maybe, or maybe number one song, like it creates the whole infrastructure required to create the song, to be the influencer that represents that song and that kind of thing, and makes a lot of them. That could be the moot. I mean, our culture respects money in that kind of way.

3:03:38And it's verifiable, right? It's verifiable, right? The make account can't lie. Exactly. There is surprising evidence that once you set up the ways of collecting the verifiable domain that this can work, there's been a lot of research before this R1 on math problems. And they approach math with language models just by increasing the number of samples. So you can just try again and again and again. and you look at the amount of times that the language models get it right. And what we see is that even very bad models get it right sometimes. And the whole idea behind reinforcement learning is that you can learn from various sparse rewards.

3:04:15So it does it, the space of language and the space of tokens, whether you're generating language or tasks for robot is so big that you might say that it's like, I mean, each the tokenizer for a language model can be like 200 ,000 things. So at each step, it can sample from that big of a space. So if it can generate a bit of a signal that it can climb onto, that's what the whole field of RL is around is learning from Sparrow Awards. And the same thing has played out in math where it's like very weak models that sometimes generate answers. We see research already that you can boost their math scores.

3:04:47You can do this sort of RL training for math. It might not be as effective, but if you take a 1 billion parameter model, so something 600 times smaller than deep seek, you can boost its grade school math scores very directly with a small amount of this training. So it's not to say that this is coming soon, setting up the verification domains is extremely hard, and there's a lot of nuance in this. But there are some basic things that we have seen before, where it's like, it's at least expectable that there's a domain and there's a chance that this works. All right, so we have fun things happening.

3:05:19In real time, this is a good opportunity to talk about other reasoning models, O1, O3, just now OpenAI, S perhaps expected released O3 mini. What are we expecting from the different flavors? Can you just lay out the different flavors of the O models and from Gemini, the reasoning model? Something I would say about these reasoning models is we talked a lot about reasoning training on math and code. And what is done is that you have the base model, we've talked about a lot on the internet. that you do this large scale reasoning training with reinforcement learning. And then what the deep seek paper detailed in this R1 paper, which for me is one of the big open questions on how do you do this, is that they did reasoning heavy but very standard post -training techniques after the large scale reasoning RL.

3:06:09So they did the same things with a form of instruction tuning through rejection sampling, which is essentially heavily filtered instruction tuning with some reward models. And then they did this RLHF, but they made it math heavy. So some of this transfer, we looked at this philosophical example early on. One of the big open questions is how much does this transfer? If we bring in domains after the reasoning training, are all the models gonna become eloquent writers by reasoning is this philosophy stuff going to be open? We don't know in the research of how much this will transfer. There's other things about how we can make soft verifiers and things like this, but there is more training after reasoning, which makes it easier to use these reasoning models, and that's what we're using right now.

3:06:52So if we're gonna talk about it with three many, and O1, these have gone through these extra techniques that are designed for human preferences after being trained to elicit reasoning. I think one of the things that people are ignoring is Google's Gemini flash thinking is both cheaper than R1 and better. And they released it in the beginning of December. And nobody's talking about it. No one cares. It has a different flavor to it. It's behavior is less expressive than something like O1. and it has fewer tracks than it is on. Quinn released a model last fall, QWQ, which was their preview reasoning model.

3:07:25And in DeepSeek, I had R1Lite last fall where these models kind of felt like they're on rails where they really, really only can do math and code. And O1 is, it can answer anything. It might not be perfect for some tasks, but it's flexible, it has some richness to it. And this is kind of the art of like how cooking, like how it has a model a little bit undercooked. It's good to get a model out the door, but it's hard to gauge. It takes a lot of taste to be like, is this a full -fledged model? Can I use this for everything? They're probably more similar for math and code. My quick read is that Gemini Flash is not trained to the same way as O1, but taking an existing training stack, adding reasoning to it.

3:08:08Taking a more normal training stack and adding reasoning to it, and I'm sure they're going to have more. I mean, they've done quick releases on Gemini Flash. So reasoning and this is the second version from the holidays. It's evolving fast and it takes longer to make this training stack where you're doing this large scale. That's the same question from earlier. The one about the human nature. Yeah. What was the human nature one? The way I can ramble, why I can ramble about this so much is that we've been working on this AI2 before O1 was fully available to everyone and before R1, which is essentially using this RL training for fine -tuning.

3:08:47We use this in our two -lew series of models. You can elicit the same behaviors where you say, wait and sew it on, but it's suddenly in the training process that this kind of reasoning expression is much later. There's essentially a gradation and just how much of this RL training you put into it determines how the output looks. So we're now using Gemini 2 .0 Flash thinking experimental 121. It summarized the product's humans self domesticated apes. Okay, all right. So wait, is this reviewing the reasoning? Here's why this is novel. Okay. It's been clicked to expand. Okay. Analyze the request. Novel is the keyword.

3:09:36It looks like a normal output. Yeah, it's, I mean, in some sense, it's better structured. It makes more sense. When it latched onto human, and then it went into organisms, and oh, apex predator, focus on domestication, apply domestication to humans, explore the idea of self -domestication. Not good. Not good. I think worse is going. Refine, articulate the insight, great or facial expressionist and communication ability. Yes. Plus this in adaptability. Yes. Dependence on social groups. Yes. All right. And self critique and refined further. Wow. Is this truly novel? Is it well supported? So on and so forth.

3:10:25And the insight is getting at is humans are not just social animals, but profoundly self -domesticated apes and the self -domestication is the key to understanding our unique cognitive and social abilities, self -domesticated apes, self -domesticated. I prefer the deep -seek response. Self -domesticated. I mean, it's novel. The inside is novel. I mean, that's like a good book title, self -domesticated apes. Like, there could be a case made for that. I mean, yeah, it's cool. And it's revealing the reasoning. It's magical. It's magical. So, this is really powerful. Hello everyone. This is Lex with the Quick Intermission recorded after the podcast.

3:11:08Since we reviewed responses from Deepseek R1 and Gemini Flash 2 .0 thinking during this conversation, I thought at this moment it would be nice to insert myself quickly doing the same for OpenAI 01 Pro and 03 Mini with the same prompt. prompt, the prompt being give one truly novel insight about humans. And I thought I would, in general, give my vibe check and vibe based anecdotal report on my own experiences with the new O3 mini model, not that I got a chance to spend many hours with it in different kinds of contacts and applications. So I would probably categorize this question as, let's say, open ended philosophical question.

3:11:55And in particular, the emphasis on novelty, I think it's a nice way to test one of the capabilities of the model, which is come up with something that makes you pause and almost surprised you with brilliance. So that said, my general review, after running each of the models on this question about of times is that O1 Pro consistently gave brilliant answers. Once they gave me pause and made me think, both cutting in its insight and just really nicely phrased with wit, with clarity, with nuance, over and over consistently generating the best answers. After that is R1, which is less consistent, but again deliver brilliance.

3:12:43Gemini flash 2 .0 thinking was third and last was 03 mini actually. It often gave quite generic answer at least to my particular sensibilities. That said in a bunch of other applications that I tested for brainstorming purposes it actually worked extremely well and often outperformed R1, but on this open ended philosophical question did it consistently worse. Now, another important element for each of these models is how the reasoning is presented. DeepSeek R1 shows the full chain of thought tokens, which I personally just love. For these open ended philosophical questions, it's really, really interesting to see the model think through it, but really also just stepping back means the person who appreciates intelligence and reasoning and reflection, reading these kind of chain of thought raw tokens of R1, there's something genuinely beautiful about observing the path of deliberation in an intelligence system.

3:13:48I think we don't always have that explicitly laid out for us humans, so to see it in another intelligence system, the non -linearity of it akin to Eulis C. Zafinagan's It's just beautiful to watch. Anyways, we discussed in the episode DeepSeek R1, talked about humans being able to convert self -adduceres into cooperative systems by collectively pretending to abstract rules like money laws and rights are real, and the Shared hallucinations act as games, where competition is secretly redirected to benefit the group, turning conflict into society's fuel. Gemini 2 .0 Flash thinking said, humans are not just social animals, but self -domesticated apes, and this self -domestication is the key to understanding our unique cognitive and social abilities.

3:14:39Now, it's important to say that the chain of thought there was really interesting. It was looking through the entire evolution of life on earth, considering apex predators, and considering how, from that, we ended up to where we are. I think that domestication by choice is a really interesting angle. Again, it's one of those things when somebody presents a different angle and a seemingly obvious thing, it just makes me smile. And the same with deep -seek R1, that these hallucinations of money, laws, and rights and us collectively pretending like it's real, and we play games with them that look like competition when secretly we're just cooperating with each other.

3:15:21And that is the fuel of progress. Beautifully put. Now, OpenAI -01 proc consistently over and over delivered bangers. I can go through many of them, but the first one was humans are the only species that turns raw materials into symbolic resources. Then uses those symbols to reorganize the various materials they came from, creating a closed feedback loop between meaning and matter. Here I just ran again. Banger after banger, I'm telling you, humans are unique among known species in that they simultaneously rewrite two layers of reality, the external world and their own private mental landscapes.

3:16:00And then merge these two rewritten layers into a continuous personal narrative that feels objectively true. Feels true. This is poetry. Okay, and then O3 mini high for me was smart, fast, actually, and kind of generic. Never quite got there for me. So here's the first one I got from O3 mini. Humans are not fixed beings, but rather ongoing narratives, dynamic stories that we continuously write, edit, and reinterpret. This narrative plasticity is more than just memory or self -reflection. It's an intrinsic cognitive process that acts like an internal error correction system. It allows us to adapt our identities and values over time in response to new experiences, challenges, and social context.

3:16:54Now it almost sneaks up to something approximating cutting insight with narrative plasticity in quotes. But then it goes back to the sort of the generic. I don't know. All of these models are incredible for different reasons. There's a lot of concerns as we discuss in this episode, but there's a lot of reasons to be excited as well. And I've probably spoken for too long. I am severely sleep deprived, borderline delirious, so hopefully some of this made sense. And now dear friends, back to the episode. I think when you, to Nathan's point, when you look at the reasoning models, to me, even when I used R1 versus O1, there was that sort of edge around the corner feeling.

3:17:45Flash thinking earlier, I didn't use this version, but the one from December, and I definitely had that rough edge around the corner feeling where it's just not fleshed out in many ways. right? Sure, they added math and coding capabilities via these verifiers in RL, but you know, they might, it feels like they lost something in certain areas and oh, one is worse performing than chat in many areas as well to be clear. Not by a lot. Not by a lot though, right? And it's like some, like R1 definitely felt to me like it was worse than V3 in certain areas like doing this RL, expressed and learned a lot, but then it weakened in other areas.

3:18:21And so I think that's one of the big differences between these models and what O1 offers and then OpenAI has O1 Pro and what they did with O3 which is also very unique is that they stacked search on top of Chain of Thought, right? And so Chain of Thought is one thing where it's able, it's one chain, it backtracks, goes back and forth, but how they solved the ArcAGI challenge was not just the Chain of Thought. It was also sampling many times, i .e. running them in parallel and then selecting, is running in parallel actually search. I don't know if we have the full information on how O1 pro works.

3:18:57So I don't have enough information. I agree. I would confidently say that it is search. It is parallel samples. Yeah. And then what? And it selects something. And we don't know what the selection function is. The reason why we're debating is because since O1 was announced, there's been a lot of interest in techniques called Monte Carlo research, which is where you will break down the chain of thought into intermediate steps. We haven't defined chain of thought. Chain of thought is from a paper from years ago where you introduced the idea to ask a language model that at the time was much less easy to use.

3:19:27You would say, let's verify a step by step and it would induce the model to do this bolded list of steps. Chain of Thought is now almost a default in models where if you ask in a math question, you don't need to tell it to think step by step. The idea with Monte Carlo tree search is that you would take an intermediate point in that chain, do some sort of expansion, spend more compute, and then select the right one. That's a very complex form of search that has been used in things like Mu0 and Alpha0 potentially. I know Mu0 does this. Another form of search is just asking five different people and then taking the majority answers.

3:20:00A variety of like, you know, it could be complicated, it could be simple. We don't know what it is just that they are, they are not just issuing one chain of thought in sequence. They are launching many in parallel and in the ArcAGI they launched 1000 in parallel for the one that like really shocked everyone that beat the benchmark was they would launch a thousand in parallel and then they would get the right answer like 80 % of the time or 70 % of the time, 90 maybe even, whereas if they just launched one, it was like 30%. There are many extensions to this. I would say the simplest one is that our language models today have been designed to give the right answer the highest percentage of the time in one response.

3:20:39And we are now opening the door to different ways of running inference on our models in which we need to reevaluate many parts of the training process, which Normally, it opens the door to more progress, but we don't know if OpenAI changed a lot, or if just sampling more in multiple choices, what they're doing, or if it's something more complex, but they changed the training and they know that the inference mode is going to be different. We're talking about O1 Pro, $200 a month, and they're losing money. The thing that we're referring to, this fascinating exploration of the test time compute space, is that actually possible?

3:21:18Do we have enough compute for that? Does the financial make sense? So the fantastic thing is, and it's in the thing that I pulled up earlier, but did cost for GPT -3 has plummeted if you scroll up just a few images, I think? The important thing about like, hey, is cost limiting factor here, right? My view is that we'll have really awesome intelligence before we have AGI, before we have it permeate throughout the economy. And this is sort of why that reason is, right? GPT -3 was trained in what, 2020 -2021. And the cost for running inference on it was $60, $70 per million tokens, right? Which is the cost per intelligence was ridiculous.

3:22:01Now as we scaled forward two years, we've had a 1200 X reduction in cost to achieve the same level of intelligence as GPT -3. So here in the X -axis is time over just a couple of years and on the Y -axis is log scale dollars To run inference on a million tokens and so you have just a down like a Linear decline on log scale from GPT -3 through 35 to long Five cents or something like that now, right? Which is which is versus versus $60? $1200 X, that's not the exact numbers, but it's 1200 X, I remember that number. Is humongous cost per intelligence, right? Now, the freak out over deep seek is, oh my god, they made it so cheap.

3:22:48It's like, actually, if you look at this trend line, they're not below the trend line, first of all, and at least for GPT -3, right? They are the first to hit it, which is a big deal. But they're not below the trend line as far as GPT -3. Now we have GPT -4, what's going to happen with these reasoning capabilities, right? It's a mix of architectural innovations. It's a mix of better data and it's gonna be better training techniques and all of these different better inference systems Better hardware right going from you know each generation of GPU to new generations or ASICs Everything is gonna take this cost curve down and down and down and down and then can I go in?

3:23:22Can I just spawn a thousand different LLMs to Create a task and then pick from one of them or you know whatever search search technique I want a tree Monte Carlo tree search maybe it gets that complicated maybe it doesn't because it's too complicated to actually scale like who knows better or less than right. The question is is I think when not if because the rate of progress is so fast right. Nine months ago Dario was saying or you know Dario said nine months ago the cost to train in inference was this right and now we're much better than this right and deep seek is much better than this and that cost curve for GPT -4 which was also roughly $60 per million tokens when it launched has already fallen to $2 or so.

3:24:05And we're going to get it down to cents probably for GPT -4 quality and then that's the base for the reasoning models like O1 that we have today and O1 Pro is spawning more multiple and O3 and so on and so forth. These search techniques too expensive today but they will get cheaper and that's what's going to unlock the intelligence. right? So get cheaper and cheaper and cheaper. The big deep seek R1 release freaked everybody out because of the cheaper one of the manifestations that that is Nvidia stock plummeted. Can you explain what happened? I mean, and also just explain this moment and whether, you know, if Nvidia is going to keep winning.

3:24:46We're both Nvidia bulls here, I would say. And in some ways the market response is is reasonable. Most of the market, like Nvidia's biggest customers in the US are major tech companies and they're spending a ton on AI. And if a simple interpretation of DeepSeek is you can get really good models without spending as much on AI. So in that capacity, it's like, oh, maybe these big tech companies won't need to spend as much an AI and go down. The actual thing that happened is much more complex where there are social factors, where there's the rising in the app store, the social contagion that is happening.

3:25:20And then I I think a lot of some of it is just like, I'm not a trade, I don't know anything about financial markets, but it builds up over the weekend or the social pressure, where it's like, if it was during the weekend, there was multiple days of trading when this was really becoming, but it comes on the weekend and then everybody wants to sell. And that is a social contagion. I think, I think, and like, there were a lot of false scenarios, which is like, hey, these guys are spending billions on models, right? And they're not spending billions on models. No one spent more than a billion dollars on a model that's released publicly, right?

3:25:49GPG -4 was a couple hundred million and then they've reduced the cost with 4 .0, 4 .0, 4 .0, 4 .0, right? But billion dollar model runs are coming, right? This concludes pre -training and post -training, right? And then the other number is like, hey, DeepSeek didn't include everything, right? They didn't include a lot of the cost goes to research and all this sort of stuff. A lot of the cost goes to inference, a lot of the cost goes to post -training. None of these things were factory research salaries, right? All these things are counted in the billions of dollars that open AI is spending, but they weren't counted in the, hey, $6 million, $5 million that deep seeks spent.

3:26:20So there's a bit of misunderstanding of what these numbers are. And then there's also an element of, Nvidia has just been a straight line up, right? And there's been so many different narratives that have been trying to push down Nvidia. I don't say push down Nvidia stock. Everyone is looking for a reason to sell or to be worried, right? You know, it was, it was black well delays, right? There are GPU, there's a lot of report every two weeks. There's a new report about their GPUs being delayed. There's the whole thing about scaling laws ending, right? It's so ironic, right? The last set of months.

3:26:54It was just like literally just, hey, models aren't getting better, right? They're just not getting better. There's no reason to spend more proof training scaling is dead of that it's like, oh, one, oh three, right? R1, R1, right? And now it's like, wait, models are getting too fast. They're progressing too fast. Slow down the progress, stop spending on GPUs, right? But you know, the funniest thing I think that comes out of this is Javon's paradox is true, right? AWS pricing for H100 has gone up over the last couple weeks, right? Since, since, since, since a little bit after Christmas, since V3 was launched, AWS H100 pricing has gone up.

3:27:29H200s are like almost out of stock everywhere because it, you know, H200 has more memory and therefore R1, like, you know, wants that chip over H100, right? We were trying to get GPUs on a short notice this week for demo and it wasn't that easy. We were trying to get just like 16 or 32 H100s for demo No, not very easy. So for people who don't know, jam on paradoxes, when the efficiency goes up, somehow magically, counterintuitively, the total resource consumption goes up as well. Right, and semiconductors is, you know, we're like 50 years of Moore's Law, every two years half the cost, double the transistors, just like clockwork and it slowed down obviously, but like the semiconductor industry has gone up the whole time, right?

3:28:09It's been wavy, right? There's obviously cycles and stuff, I don't expect AI to be any different, right? There's gonna be ebbs and flows, but this is an AI, it's just playing out at an insane timescale, right? It was two X every two years. This is 1200 X in like three years, right? So it's like the scale of improvement that is like hard to get wrapped your head around. Yeah, I was confused because to me, in me, a stock on that should have gone up, but maybe it went down because there's kind of suspicion of fall plan inside of China, something like this. But if you just look purely at the actual principles of play here, like it's obvious, yeah, the Javans paradox.

3:28:48Or the higher the derivative of AI progress is, especially because Nvidia is in the best place. The higher the derivative is, the sooner the market's going to be bigger and expanding and Nvidia is the only one that does everything reliably right now. Because it's not like an Nvidia competitor arose. It's another company that's using Nvidia. Yeah, who historically has been a large Nvidia customer. Yeah. And has pressed releases about them cheering about being China's biggest Nvidia customer, right? Like, yeah, I mean, it mean, obviously they've quieted down, but like, I think that's like another element of it is that they don't want to say how many GPUs they have.

3:29:24Yeah. Because, hey, they, yes, they have H800s, yes, they have H20s, they also have some H100s, right? Which was smuggled in. Can you speak to that to the smuggling? What's the scale of smuggling that's feasible for a nation state to do for companies, is it possible to? I think there's a few angles of smuggling here, right? One is, bite dance arguably is the largest smuggler of GPUs for China, right? China's not supposed to have GPUs. Bite dance has like over 500 ,000 GPUs, why? Because they're all rented from companies around the world. They rent from Oracle, they rent from Google, they rent from all these masks and a bunch of smaller cloud companies too, right?

3:30:01All the neoclouds, right, of the world. They rent so, so many GPUs, they also buy a bunch, right? And they do this for mostly like what meta does right serving TikTok right serving back to the next best The same discussion To be clear that today the views right and it's a valid use right hack the dopamine certit right now That's that's theoretically now very much restricted with the AI diffusion rules Which happened in the last week of the Biden admin and Trump admin looks like they're gonna keep them which limits like allies even like Singapore which Singapore is like 20 % of in -vities, 20, 30 % of in -vities revenue, but Singapore has had a memoratorium on not building data centers for like 15 years because they don't have enough power.

3:30:42So where are they going? Oh, okay. I mean, I'm not saying they're all going to China, right? But a portion are, you know, many are going to Malaysia, including Microsoft and Oracle have big data centers in Malaysia. Like, you know, they're going all over Southeast Asia, probably India as well, right? Like there's stuff routing, but like the diffusion rules are very de facto. like you can only buy this many GPUs from this country. And you can only rent a cluster of this large to companies that are Chinese, right? Like they're very explicit on trying to stop smuggling, right? And a big chunk of it was, hey, let's, you know, random company by 16 servers, ship some to China, right?

3:31:17There's actually, I saw a photo from someone in the semiconductor industry who leads like a team for like networking chips that competes within video. and he sent a photo of a guy checking into a first class United flight from San Francisco to to Shanghai or Shenzhen with a Super micro box that is this big which can only contain GPUs Right and he was booking first class. Let's think about it three to five K for your first class ticket server cost You know 240 ,000 in the US 250 ,000 you sell it for 300 ,000 in China Wait, you just got a free first class ticket and a lot more money So it's like, you know, and that's like small scale Suggling most of the large scale smuggling is like companies in Singapore and Malaysia like routing them around or renting GPUs Completely, I want to jump in how much was the scale?

3:32:05I think there's been some number like some people that are higher level economics Understanding say that as you go from one billion of smuggling to 10 billion It's like you're hiding certain levels of economic activity And that's the most reasonable thing to me is that there's going to be some level where it's so obvious that it's easier to find this the economic activity. And yeah, so my belief is that last year roughly, so Nvidia made a million H20s, which are legally allowed to be shipped to China, which we talked about is better for reasoning, right? Infraints at least, not maybe not training, but reasoning inference.

3:32:39And inference generally, then they also had a couple hundred thousand. We think like 200 to 300 thousand GPUs were routed to China from Singapore, Malaysia, US, wherever companies spawn up by 16 GPUs, 64 GPUs, whatever it is, route it. And Huawei is known for having spent up a massive network of companies to get the materials they need after they were banned in 2018. So it's not like otherworldly. But I agree, right? Nathan's point is like, hey, you can't smuggle $10 billion in GPUs. And then the third sort of source, which is just now banned, which wasn't considered smuggling, but is China is renting, I believe from our research, right?

3:33:17Oracle's biggest GPU customer is bite dance, right? And for Google, I think it's their second biggest customer, right? And so like, and you go down the list of clouds, and especially these smaller cloud companies that aren't like the hyperscalers, right? Think beyond CoreWeave and Lambda, even there's a whole see, there's 60 different new cloud companies serving in video GPUs. I think bite dance is renting a lot of these, right? All over it, right? And so these companies are renting GPUs to Chinese companies, and that's completely, That was completely legal up until the diffusion rules, which happened just a few weeks ago.

3:33:49And even now you can rent GPU clusters that are less than 2000 GPUs, or you can buy GPUs and ship them wherever you want if they're less than 1500 GPUs. So it's like there are still some ways to smuggle, but yeah, it's not, as the numbers grow, 100 something billion dollars or revenue for Nvidia last year, 200 something billion this year. right and if next year are you know, it could it could nearly double again or more than double All right, based on like what we see with data center for prints like being built out all across the US and the rest of the world It's gonna be really hard for trying to to keep up with these rules, right?

3:34:22Yes, there will always be smuggling And deep -seek level models of GPD4 level models Oh one level models capable to train on what China can get even the next year above that But if we speedrun a couple of more or jumps to billion dollar models, 10 billion dollar models, then it becomes, hey, there is a compute disadvantage for China for training models and serving them. And the serving part is really critical, right? DeepSeek cannot serve their model today, right? It's completely out of inventory. It's already started falling in the app store, actually downloads because you download it. You try and sign up, they say, we're not taking registrations because they have no capacity.

3:34:58You open it up, you get less than five tokens per second if you even get your request approved, right? because there's just no capacity because they just don't have enough GPUs to serve the model, even though it's incredibly efficient. It would be fascinating to watch the smuggling, because I mean, there's drug smuggling, right? That's a market. There's weapons smuggling, and GPUs will surpass that at some point. Ships are highest value per kilogram, per all bullied by far. I have another question for you, Don, do you track like Model API access internationally. How easy is it for Chinese companies to use hosted Model APIs from the US?

3:35:36Yeah, I mean, that's incredibly easy, right? Like OpenAI publicly stated deep seek uses their API. And as they say, they have evidence, right? And this is another element of the training regime is people at OpenAI have claimed that it's a distilled model, i .e. you're taking OpenAI's model, you're generating a lot of output and then you're training on the output in their model. And even if that's the case, what they did is still amazing, by the way, What deep -seaked did efficiency wise? Distillation is standard practice in industry, whether or not if you're at a closed lab where you care about terms of service and IP closely, you distill from your own models.

3:36:06If you are a researcher and you're not building any products, you distill from the opening eye box. This is a good opportunity. Can you explain big picture distillation as a process? What is distillation? What's the process distillation? Talk a lot about training language models. They are trained on text and post -training, you're trying to train on very high quality text that you want the model to match the features of or if you're using RL, you're letting the model find its own thing. But for supervised fine tuning, for preference data, you need to have some completions what the model is trying to learn to imitate.

3:36:37And what you do there is instead of a human data or instead of the model you're currently training, you take completions from a different normally more powerful model. I think there's rumors that these big models that people are waiting for, these GBT -5s of the world, the cloud three opuses of the world are used internally to do this distillation process. There's also public examples, right? Like meta explicitly stated, not necessarily distilling, but they used 405B as a reward model for 70B in their Lama 3 .2 or 3 .3. This is all the same topic. So is this ethical, is this legal? Like why is that a financial times article headline say open -ass says that there's evidence that China's deep seek used its model to train competitor.

3:37:26This is a long, at least in the academic and side and research side as long history because you're trying to interpret OpenAI's rule. OpenAI's terms of service say that you cannot build a competitor with outputs from their model. Terms of service are different than a license, which are essentially a contract between organizations. So, if you have a terms of service on OpenAI's account, if I violate it, OpenAI can cancel my account. This is a very different than like a license that says how you can use a downstream artifact. So a lot of it hinges on a word that is very unclear in the AI space, which is what is a competitor.

3:37:55And then the ethical aspect of it is like, why is unethical for me to train on the airport when you can train on the internet's text? Yeah, right. So there's a bit of a hypocrisy because sort of open AI and potentially most of the companies trained on the internet's text without permission. There's also a clear loophole, which is that I generate data from open AI, and then I upload it somewhere and then somebody else trains on it and the link has been broken. Like they're not under the same terms of service contract. This is why there's a lot of hip hop. There's a lot of like to be discovered in details that don't make a lot of sense.

3:38:32This is why a lot of models today, even if they train on zero opening eye data, you ask the model who trained you, it'll say, I am chat GPT trained by opening eye, because there's so much copy paste of like opening eye outputs from that on the internet that you We just weren't able to filter it out. And there was nothing in the RL where they implemented like, hey, or post training or SFT, whatever that says, hey, I'm actually a model by Allen Institute instead of. We have to do this if we serve a demo. We do research and we use OpenAI APIs because it's useful and we wanted to understand post training and like our research models, they will say they're written by OpenAI unless we put in the system prompt that we talked about that.

3:39:10Like, I am Tulu. I am a language model trained by the Allen Institute for AI. And if you ask more people around industry, especially with post -training, it's a very doable task to make the model say who it is or to suppress the opening eye thing. So in some levels, it might be that DeepSeek didn't care that it was saying that it was by opening eye. Like, if you're going to upload model weights, it doesn't really matter because anyone that's serving it in an application and cares a lot about serving is going to, when serving it, if they're using it for a specific task, they're going to tailor it to that.

3:39:40And it doesn't matter that it's saying it's a TGBT. Oh, I guess the one of the ways to do that is like a system prompt or something like that. Like if you're serving it to say that you're... That's what we do. Like if we host the demo, you say you are two -league three, a language model trained by the Allen Institute for AI. We also are benefited from OpenAI data because it's a great research tool. I mean, do you think there's any truth and value to the claim, OpenAI's claim that there's evidence that China's deep seek uses model to train? I think everyone has benefited regardless because the data's on the internet.

3:40:16And therefore, it's in your portraying, right? There are like subreddits where people share the best chat GPT outputs. And those are in your... I think that they're trying to ship the narrative. Like, they're trying to protect themselves. And we saw this years ago, and bite dance was actually banned from some opening AI APIs for training on outputs. There's other AI startups that most people, if you're in the AI culture, or like, they just told us they trained on opening eye outputs and they never got banned. Like that's how they bootstrapped their early models. So it's much easier to get off the ground using this than to set up human pipelines and build a strong model.

3:40:50So there's long history here and a lot of the communications are seem like narrative. Actually, like over the last couple days, we've seen a lot of people distilled deep -seek's model into Lama model because the deep -seek models are kind of complicated to run inference on because they're mixture of experts and they're 600 plus billion parameters and all this and people distilled them into the Lama models. And then because the Lama models are so easy to serve and everyone's built the pipelines and tooling for inference with the Lama models, right? Because it's the open standard. So, you know, we've seen it, we've seen it sort of round about, right?

3:41:20Like, is it bad? Is it illegal? Maybe it's illegal, whatever, I don't know about that. But like, it could break contracts. I don't think it's illegal. Like, in any illegal, like, no one's going to jail for this ever. I think, like, fundamentally, I think it's ethical or I hope it's ethical because like, Like the moment becomes, we ban that kind of thing, it's going to make everybody much worse off. And I also actually, this is difficult, but I think you should be a lot to train on the internet. I know a lot of authors and creators are very sensitive about it. That's a difficult question. But the moment you're not allowed to train on the internet.

3:41:57I agree. I just get so take on how you can solve this because it already works. I have a reasonable take on it. All right, all right. So, so, you know, Japan has a law which you're allowed to train on any training data and copyrights don't apply if you want to train a model. A, B, Japan has nine gigawatts of curtailed nuclear power. C, Japan is allowed under the AI diffusion rule to import as many GPUs as they'd like. So all we have to do, we have a market here to make. We build massive data centers, we rent them to the labs, and then we train models in a legally permissible way, and there's no if -ands or buts.

3:42:33And now, the models have no potential copyright lawsuit from New York Times or anything like that. No, no, it's just completely legal. No, so genius. The early copyright lawsuits have fallen in the favor of AI training. I would say that the long tail of use is gonna go inside of AI, which is if you do, if you scrape trillions of data, you're not looking at trillions of tokens of data, you're not looking and saying this one New York Times article is so important to me. But if you're doing a audio generation for music or image generation and you say make it in the styled x -person That's a reasonable case where you could figure out what is their profit margin on inference I don't know if it's gonna be the 50 50 of you to create or program or something But I would opt into that program as a writer like please like that It's just it's gonna be a rough journey But there will be some solutions like that that makes sense but there's a long tail where it's just on the internet.

3:43:30I think one of the other aspects of that financial times article implied, and so that leads to a more general question. Do you think there's how difficult is spying, espionage, and stealing of actual secret code and data from inside companies? How much of that is being attempted? Code and data is hard, but ideas is easy. Silicon Valley operates on the way that top employees get bought out by other companies for a pay raise. And a large reason why these companies do this is to bring ideas with them. And there are, there's no, I mean, in California, there's rules of like certain non -compete or whatever are illegal in California.

3:44:10And whether or not there's NDAs and things, that is how a lot of it happens. Recently, there was somebody from Gemini who helped make this 1 million context length and everyone is saying the next llama, who I mean, he went to the meta team, is gonna have 1 million context length. And that's kind of how the world works. As far as industrial espionage and things, that has been greatly successful in the past. The Americans do the Brits. The Chinese have done it to the Americans. And so on and so forth. It is a fact of life. And so to argue industrial espionage can be stopped is probably unlikely.

3:44:48You can make it difficult. But even then, there's all these stories about like, hey, F -35 and F -22 have already been given to China in terms of design plans and stuff. Code and stuff like between, I say companies not nation states is probably very difficult, but ideas are discussed a lot, right? Whether it be a part house party in San Francisco or a company changing employees, or the always the like mythical honey pot that always gets talked about, right? Like someone gets honey -potted, right? Because everyone working on AI is a single dude who's in their 20s and 30s. Not everyone, but like insane amount of, insane percentages.

3:45:25So there's always like all these like, and obviously. So honeypotted is like a spy, a female spy approaches you and like, yeah. Yeah, or or male, right? You know, it's San Francisco, right? But as a single dude, I will say, in his late 20s, right, is like we are very easily corrupted, right? Like, you know, like, not corrupted myself, but you know, like, we are, we are, right? Every day is not me. Yeah, exactly. I'm too oblivious and I am not single. So I'm saved from one espionage access. Yeah, you have to make sure to close all security vulnerabilities. So you don't collect a lot of information about each of the mega clusters for each of the major AI companies.

3:46:07Can you talk about the buildouts for each one that stand out? Yeah, so I think the thing that's really important about these mega cluster buildouts is they're completely unprecedented in scale. right? US, you know, it's sort of like data center power consumption has been slowly on the rise and it's gone up to 2, 3%, even through the cloud computing revolution, right? Data center consumption as a percentage of total US. And that's been over decades, right? Of data centers, et cetera. It's been climbing climbing slowly. But now, 2 to 3%, now, by the end of this decade, it's like even, even under like, you know, when I say like 10%, a lot of people that are traditionally by like 20, 28, 20, 30.

3:46:46People traditionally non a traditional data center people like that's nuts. But then like people who are in like AI who have like really looked at this at like the end -theropics and open AI's or like that's not enough from them. Okay. But like, you know, this is both through globally distributed or distributed throughout the US as well as like centralized clusters, right? The distributed throughout the US is exciting and it's the bulk of it, right? Like, hey, you know, opening eye or, you know, say, meta is adding a gigawatt, right? But most of it is distributed through the US for inference and all these other things, right?

3:47:21So maybe we should lay out what a what a cluster is. So You know, this is include AWS. Maybe it's it's good to talk about the different kinds of clusters and what you mean by mega clusters And what's the GPU and what's the computer and what it's getting not that far back? But yeah, so like what do we mean by the clusters? I thought I was about to do the Apple ad, right? What's the computer? So traditionally data centers and data center tax have been a distributed systems problem that is capable of being spread very far and widely, right? IE, I send a request to Google, it's rather to a data center somewhat close to me, it does whatever search, ranking, recommendations, sends a result back, right?

3:48:03Right. The nature of the task is changing rapidly in that the task, there's two tasks that people are really focused on now. Right. It's not database access. It's not sort of me the right page, sort of me the right ad. It's now a inference. An inference is dramatically different from traditional distributed systems, but it looks a lot more simple, similar, and then there's training. Right. The train inference side is still like, hey, I'm going to put thousands of GPUs in blocks all around these data centers. I'm going to run models on them, user submits a request, gets kicked off, or, hey, my service, they submit a request to my service, right?

3:48:38They're on word and they're like, oh, yeah, help me copilot and it kicks it off from on my windows, copilot, whatever, Apple intelligence, whatever it is, it gets kicked off to a data center, right? Now, that data center does some work and sends it back. That's inference. That is going to be the bulk of compute. But then, and that's like, there's thousands of data centers that we're tracking with satellites and all these other things. and those are the bulk of what's being built, but the scale of, and so that's like what's really reshaping and that's what's getting millions of GPUs, but the scale of the largest cluster is also really important, right?

3:49:11When we look back at history, right, like, you know, or through the age of AI, right, like it was a really big deal when they did Alex Net on I think two GPUs or four GPUs, you just gotta remember, it was a really big deal. It's a big deal because you use GPUs. It's a big deal that you use GPUs, and they used multiple, right? But then over time, its scale has just been compounding, right? And so when you skip forward to GPT -3, then GPT -4, GPT -4, 20 ,000, A100 GPUs, unprecedented run, right? In terms of the size and the cost, right? A couple hundred million dollars on a yellow, right? A yellow run for GPT -4.

3:49:46And it yielded, you know, this magical improvement that was like perfectly in line with what was experimented and just like a log scale, right? I have that plot from the paper. The scaling, the technical work part. The scaling laws were perfect, right? But that's not a crazy number, right? 20 ,000 A100s, roughly each GPU is consuming 400 watts. And then when you add in the whole server, everything, it's like 15 to 20 megawatts of power, right? You know, maybe you could look up what the power of consumption of a human person is because the numbers are gonna get silly. But like, 15 to 20 megawatts was standard data center size.

3:50:20It was just unprecedented. That was all GPUs running one task. That was also the toaster. Yeah, the toaster is like a good example similar no power consumption to an A 100 right H100 comes around They increase the power from like 400 to 700 watts and that's just per GPU and then there's all the associated stuff around it So it's so once you count all that it's roughly like 1200 to 1400 watts for everything networking CPUs memory Boba Blasso we should also say so what's required? You said power so a lot of powers required a lot of heat is generated as cooling is required and and because there's a lot of GPUs that have to be or CPUs or whatever, they have to be connected so there's a lot of networking.

3:51:01Yeah, so I think, yeah, sorry for skipping past that. And then the data center itself is like complicated, right? But these are still standard sized data centers for GPT -4 scale. Right, now we step forward to what is the scale of clusters that people have built last year? Right, and it ranges widely. It ranges from like, hey, these are standard data centers, and we're just using multiple of them and connecting them together really with a ton of fiber between them a lot of networking etc That's what open AI and Microsoft did in Arizona, right? And so they have a you know 100 ,000 GPUs right meta similar thing They took their standard existing data center design And it looks like an H and they connected multiple them together And you know they got to they first did 16 ,000 GPUs 24 ,000 TGPUs total only 16 of them Thousand of them were running on the training run because GPUs are very unreliable So they need to have spares to like swap in and out all the way to like now 100 ,000 GPUs that they're training on llama for on currently right like 128 ,000 or so right This is you know think about a hundred thousand GPUs With roughly 1400 watts a piece that's that's that's 140 megawatts 150 megawatts right for 128 out right so you're talking about you've jumped from 15 to 20 megawatts to 10x you know almost 10x that number 9x that number to 150 megawatts in in two years, right, from 2022 to 2024, right?

3:52:19And some people like Elon, he admittedly, right? And he says himself got into the game a little bit late for portraying large language models, right? XAI was started later, right? But then he bent heaven and hell to get his data center up and get the largest cluster in the world, right? Which is 200 ,000 GPUs. And he did that, he bought a factory in Memphis. He's upgrading the substation at the same time, he's got a bunch of mobile power generation, a bunch of singles, cycle combine. He tapped the natural gas line that's right next to the factory and he's just pulling a ton of gas burning gas He's generating all this power.

3:52:51He's in a factory and an old appliance factory that shut down and moved to China long ago Right like you know and and and he's got 200 ,000 GPUs in it and now what's the next scale right like all the hyperscales Have done this now the next scale is is is something that's even bigger right and so you know Elon just to stick on the topic He's he's building his own natural gas plant like a proper one right next door He's deploying tons of Tesla MegaPak batteries to make the power more smooth and all sorts of other things He's got like industrial chillers to cool the water down because he's water cooling the chips So all these crazy things to get the clusters bigger and bigger But when you look at like say what open AI did with Stargate That's that in Arizona and in Out of lean Texas right what they've announced at least right.

3:53:38It's not built right Elon says they don't have the money You know, there's some debates about this But at full scale, at least the first section is like definitely money's accounted for, but there's multiple sections. But full scale that data center is going to be 2 .2 gigawatts, right? 2 ,200 megawatts of power in and roughly like 1 .8 gigawatts or 1 ,800 megawatts of power delivered to chips, right? Now this is an absurd scale. 2 .2 gigawatts is like more than most cities, right? You know, to be clear. and it delivered to a single cluster that's connected to do training, right? To train these models to do both the pre -training, the post -training, all of this stuff, right?

3:54:16This is insane. This is a nuclear power plant again. Everyone is doing this, right? Everyone is doing this, right? Meta, meta in Louisiana, right? They're building two natural gas plants, massive ones, and then they're building this massive data center. Amazon has like plans for this scale. Google has plans for this scale. XAI has plans for these scale, right? Like all of these, the guys that are racing, the companies that are racing are racing hard and they're doing multi -gigawatt data centers, right? To build this out because they think that, yeah, if I now have, you know, obviously pre -training scaling is gonna continue, but to some extent, but then also all this post -training stuff where you have a RL sandbox for computer use or whatever, right?

3:54:58Like, you know, this is where they're gonna, and all these very full -byable domains where they just keep learning and learning and learning self -play, whatever it is. makes the AI so much more capable because the line does go up. As you throw more compute, you get more performance, the shared is about scaling laws. To some extent, it is diminishing returns, you 10x the compute, you don't get 10x better model, you get a diminishing returns, but also you get efficiency improvements so you bend the curve. And these scale of data centers are doing, reeking a lot of havoc on the network. Nathan was mentioning, there's Amazon has tried to buy this nuclear power plant, a talent.

3:55:35And if you look at the talent stock, it's just skyrocketing. And they're building a massive multi -gigawatt data center there. And you just go down the list. There's so many ramifications. Interesting thing is certain regions of the US transmitting power cost more than actually generating it. Because the grid is so slow to build and the demand for power and the ability to build power and re -ramping on a natural gas plant or even a coal plant is easy enough to do. But transmitting the power is really hard. So in some parts of the US, like in Virginia, it cost more to transmit power than it cost to generate it.

3:56:07Which is like, you know, there's all sorts of like second order effects that are insane here. Can the power grid support this kind of growth? You know, Trump's executive orders, there's a Biden executive order before the end of the year, but then Trump had some more executive orders, which hopefully reduced the regulations to where yes, things can be built. But yeah, this is a big, big challenge, right? Is building enough power fast enough? Are you going to basically have a nuclear power plant next to a data center for each one of these? So the fun thing here is this is too slow to build the power plant.

3:56:37To build a power plant or to reconfigure an existing power plant is too slow. And so therefore you must use data center power consumption as flat. It's like nuclear is also good for it. Like long term nuclear is a very natural fit. But you can't do solar or anything in the short term. I'll be back. Because data center powers like this, right? Like, you're telling me, you know, I'm gonna buy tens of billions of dollars of GPUs and idle them because the power's not being generated, like power is cheap, right? Like, if you look at the cost of a cluster, less than 20 % of it is power, right? Most of it is the capital cost and depreciation of the GPUs, right?

3:57:14And so it's like, well, screw it. I'll just like, you know, I'll just build natural gas plants. This is what Meta's doing in Louisiana. This is what OpenAI's doing in Texas and like all these different places. They may not be doing it directly, but they are partnered with someone. And so there is a couple hopes, right? Like one is, you know, an Elon, what he's doing in Memphis is like, you know, to the extreme. They're not just using dual combine cycle gas, which is like super efficient. He's also just using single cycle and like mobile generators and stuff, which is less efficient. But he's, you know, there's also like the flip side, which is like solar power generation is like this.

3:57:45And wind is another like like this, different correlate, you know, different. So if you stack both of those, plus you get a big chunk of batteries, plus you have a little bit of gas, it is possible to run it more green. It's just the timescales for that is slow, right? So people are trying, but you know, Meta basically said, whatever, don't care about my sustainability pledge, or they'll buy like a power, it's called a PPA power purchasing agreement, where there'll be a massive wind farm or solar farm like wherever, and then they'll just pretend like those electrons are being consumed by the data center, but in reality, they're paying for the power here and selling it to the grid and they're buying power here.

3:58:20And then another thing is like Microsoft quit on some of their sustainability pledges, right? Elon, what he did with Memphis is objectively somewhat dirty, but he's also doing it in an area where there's like a bigger natural gas plant right next door and like a sewer next or not a sewer but like a wastewater treatment and a garbage dump nearby, right? And he's obviously made the world a lot more clean than that one data center is going to do, right? So I think like it's fine to some extent and maybe AGI solves global warming and stuff, right? Whatever it is. This is sort of the attitude that people at the labs have, right?

3:58:52Which is like, yeah, it's great. We'll just use gas, right? Because the race is that important. And if we lose, that's way worse. I should say that I got a chance to visit the Memphis data center. It's kind of incredible. I mean, I visited with you on just the teams and the rate of innovation there's insane. because my sense is that nobody's ever done anything of this scale and nobody has certainly ever done anything of this scale at the rate that XIA is doing. So they're like figuring out, I mean it's all sitting in on all these meetings where their brains are me. It's like it's insane. It's exciting because they're like, they're trying to figure out what the bottlenecks are, how to remove the bottlenecks, how to make sure that you know there's just so many really cool things about putting together a data center because everything has to work.

3:59:46The people that do like the Sys admin, the machine learning, all that is the exciting things so on. But really the people that run everything are the folks that know the low level software and hardware that runs everything, the networking, all of that. So you have to make sure you have procedures that test everything. I think they're using Ethernet. I don't know how they're doing the networking, but... They're using Nvidia Spectrum X Ethernet. There's actually, I think, yeah, the unsung heroes are the cooling and electrical systems, which are just like glossed over. But I think one story that maybe exemplifies how insane this stuff is, is when you're training, right, you're always doing, you're running through the model a bunch right in the most simplistic terms, running through the model a bunch, and then you're gonna exchange everything and synchronize the weights, right?

4:00:37So you'll do a step. This is like a step in model training, right? And every step your loss goes down, hopefully, and it doesn't always. But, you know, the simplest terms, you'll be computing a lot and then you'll exchange, right? The interesting thing is GPU power is most of it. Not working power is some, but it's a lot less. But so while you're computing, your power for your GPUs is here. But then when you're exchanging weights, if you're not able to overlap communications and compute perfectly, there may be a time period where your GPUs are just idle and you're exchanging weights and you're like, hey, the model's updating.

4:01:03So you're exchanging the radiance, you do the model update and then you start training again. So the power goes, right? And it's super spiky. And so funnily enough, right? Like this, when you talk about the scale of data center power, right, you can blow stuff up so easily. And so Meta actually has accidentally opened up stream something to code in PyTorch where they added an operator. And I kid you not whoever made this, like I want to hug the guy because it says, says PyTorch, it's like PyTorch .PowerPlat that no blow up. Equals zero or equal one. And what it does is amazing, right? Either when you're exchanging the weights, the GP will just compute fake numbers.

4:01:42So the power doesn't spike too much. And so then the power plants don't blow up because the transient spikes like screw stuff up. Well, that makes sense. I mean, you have to do that kind of thing. You have to make sure they're not idle. Yeah. An Elon solution was like, let me throw a bunch of Tesla mega packs and a few other things, right? Like there's, everyone has different solutions, but like Metins at least was publicly and openly known, which is just like set this operator and what this operator does is it just makes the GPUs compute nothing so that the power does its bike. But it just tells you how much power you're working with.

4:02:11I mean, it's insane. It's insane. People should just go to Google like scale, like what does X Watts do and go through all the scales from one watt to a kilowatt to a megawatt and you look at stare at that and you're high on the list and gigawatt is and it's mind blowing. Can you say something about the cooling. So I know Elon's using liquid cooling, I believe in in all cases. That's a new thing, right? Most of them don't easily look at cooling. Is there something interesting to say about the cooling? Yeah, yeah. So air cooling has been the de facto standard, throw a bunch of metal heat pipes, etc.

4:02:46and fans, right? And like that's cold. That's been enough to cool it. People have been dabbling in water cooling. Google's TPUs are water cooled, right, so they've been doing that for a few years. But with GPUs, no one's ever done, and no one's ever done the scale of water cooling that Elon just did, right? Now next generation Nvidia is for the like highest NGPU, it is mandatory water cooling, you have to water cool it, but Elon did it on this current generation, and that required a lot of stuff, right? If you look at like some of the satellite photos and stuff of the Memphis facility, there's all these external water chillers that are sitting basically, it looks like a semi -truck pod thing, what's it called the container, but really those are water chillers and he has like 90 of those water chillers just sitting outside, 90 different containers, right, with water, you know, like chill the water, bring it back to the data center and then you distribute it to all the chips, pull all the heat out and then send it back, right, and this is both a way to cool the chips, but also as an efficiency thing, all right, and going back to that like sort of three vector thing, right, there is, there is, you know, memory bandwidth flops and interconnect, the closer Those are the chips are together, the easier it is to do high speed interconnects.

4:03:56So this is also like a reason why you're going to go water cooling is because you can just put the chips right next to each other and therefore get higher speed connectivity. I got to ask you. So in one of your recent posts, there's a section called Cluster Measuring Contest. There's another word there, but I won't say it. You know, what, who's, who's, who's got the biggest now, who's going to have the big today? Individual largest is Elon, right? What? Elon's cluster. Elon's cluster in Memphis, 200 ,000 GPUs, right? Meta has like 128 ,000, opening has 100 ,000 now. Now to be clear, other companies have more GPUs than Elon.

4:04:40They just don't have them in one place, right? And for training, you want them tightly connected. There's some techniques that people are researching and working on that let you train across multiple regions, but for the most part you want them all in like one area, right? So you can connect them highly with high speed networking. And so, you know, Elon today has 200 ,000, H -100 ,000, H -100 ,000, H -100 ,000, H -100 ,000, H -200, right? Meta, OpenAI, you know, and Amazon all have on the scale of 100 ,000 a little bit less. But next this year, right, this year people are building much more, right?

4:05:14andthropic and Amazon are building a cluster of 400 ,000 Trainium 2, which is Amazon -specific chip, trying to get away from Nvidia, right? You know, meta and open AI have scales for hundreds of thousands, but by next year, you'll have like 500 ,000 to 700 ,000 GPU clusters and those GPUs are much higher power consumption than existing ones, right? Hopper 700 watts, blackwell goes to 1200 watts, right? So the power per chip is growing, and the number of chips is growing, right? No, it's, yeah, you think you think you know, you'll get to a million. You think that's actually feasible. I mean, I don't doubt Elon, right?

4:05:53The filings that he has for like, you know, the power plan and the Tesla battery packs, it's clear he has some crazy plans for Memphis. Like permits and stuff is open record, right? But it's not quite clear that, you know, what and what the time scales are, I just never doubt Elon, right? You know, that's he's gonna surprise us. So what's the idea with these clusters? If you have a million GPUs, What percentage in, let's say two, three years is used for training and what percent, pre -training, what percent is used for like, for the actual competition? So these mega clusters make no sense for inference, right?

4:06:28You could route inference there and just not train. But most of the inference capacity is being, you know, hey, I've got a 30 megawatt data center here. I've got 50 megawatts here. I've got 100 here, whatever. I'll just throw inference in all of those because the mega clusters, right, multi -gigawatt data centers, I want to train there because that's where all of my GPUs are collocated where I can put them at a super high networking speed connected together, right? Because that's what you need for training. Now with pre -training, this is the old scale, right? You could increase parameters, you do increase data, model gets better.

4:06:59That doesn't apply anymore because there's not much more data in the pre -training side, right? Yes, there's video and audio and image that has not been fully taken advantage of. So there's a lot more scaling, but a lot of people have transcripts, took and transcripts of YouTube videos, and that gets you a lot of the data. It doesn't get you all of the learning value out of the video and image data, but there's still scaling to be done on pre -training. This post -training world is where all the flops are going to be spent, right? The model is going to play with itself. It's going to self -play.

4:07:26It's going to do verifiable tasks. It's going to do computer -yay. Use in sandboxes. It might even do simulated robotics things, right? All of these things are going to be environments where computers spent in quote -unquote post -training. But I think it's gonna be good. We're gonna drop the post from post -training. It's gonna be great training and it's gonna be training, I think. We're trying to be king. At some point. Because for like bulk of like the last few years, pre -training has dwarfed post -training. But with these verifiable methods, especially ones that scale really, you know, potentially infinitely, like computer use and robotics, not just math and coding, right, where you can verify what's happening.

4:08:03Those infinitely verifiable tasks, it seems you can spend as much compute as you want on them. especially at the context length increase. Because the end of pre -training is when you increase the context length for these models. And we've talked earlier in the conversation about how the context length, when you have a long input, is much easier to manage than output. And a lot of these post -training and reasoning techniques rely on a ton of sampling and it's becoming increasingly long context. So just like effectively, your compute efficiency goes down. I don't think Flops is the standard for how you measure it, with RL and you have to do all these things where you move your weights around in a different way than a pre -training and just generation.

4:08:43It's going to become less efficient and Flops is going to be less of a useful term and then as the infrastructure gets better, it's probably going to go back to Flops. So all of the things we've been talking about is most likely going to be in video, right? Is there any competitors? Google, Google, I kind of bored them. Yeah, I was kidding. I was like, what's the story Why would TPU? Like what's the... TPU is awesome, right? It's great. Google is... There are a bit more tepid on building data centers for some reason. They're building big data centers. Don't be giving me wrong. And they actually have the biggest cluster.

4:09:15Let me... I was talking about Nvidia clusters. They actually have the biggest cluster period. But the way they do it is very interesting, right? They have two sort of data center super regions, right? In that, the data center isn't physically... All of the GPUs aren't physically on one site, but they're 30 miles from each other. or not GPS, right? They have like in Iowa and Nebraska, they have four data centers that are just like right next to each other. Why doesn't Google flex its cluster size? Go to multi data center training. There's good images in there. So I'll show you what I mean. It's just semi -analysis multi data center.

4:09:49So this is like, you know, so this is an image of like what a standard Google data center looks like. By the way, their data centers look very different than anyone else's data centers. What are we looking at here? So these are, yeah, so if you see this image, right? In the center, there are these big rectangular boxes. Those are where the actual chips are kept. And then if you scroll down a little bit further, you can see there's like these water pipes, there's these chiller cooling towers in the top, and a bunch of like diesel generators. The diesel generators are backup power. The data center itself is like look physically smaller than the water chillers, right?

4:10:21So the chips are actually easier to like keep together, but then like cooling all the water for the water cooling is very difficult, right? So Google has like a very advanced infrastructure that no one else has for the TPU. And what they do is they've like stamped these data center, they've stamped a bunch of these data centers out in a few regions, right? So if you go a little bit further down, this is a Microsoft. This is an Arizona. This is where GPT -5 quote -unquote will be trained. If it doesn't exist already. Yeah, if it doesn't exist already. But each of these data centers, I've shown a couple images of them.

4:10:53They're like really closely collocated in the same region, right? Nebraska, Iowa. And then they also have a similar one in Ohio complex, right? And so these data centers are really close to each other. And what they've done is they've connected them super high bandwidth with fiber. And so these are just a bunch of data centers. And the point here is that Google has a very advanced infrastructure, very tightly connected in a small region. So Elon will always have the biggest cluster fully connected, right? Because it's all in one building, right? And he's completely right on that, right? Google has the biggest cluster, but you have to spread over three sites and buy it by a significant margin But you have to go across multiple sites.

4:11:29Why doesn't Google Compute with Nvidia Why don't they sell TPUs? I think I think there's a couple problems with it. It's like one TPU has been a Form of allowing search to be really freaking cheap and build models for that, right? And so like a big chunk of the search or TPU purchases or big chunk of Google purchases, and usage, all of it is for internal workloads, right? Whether it be search, now Gemini, right? YouTube, all these different applications that they have, you know, ads, these are where all their TPUs are being spent and that's what they're hyper focused on, right? And so there's certain like aspects of the architecture that are optimized for their use case that are not optimized elsewhere, right?

4:12:15One simple one is like they've open sourced the Gemma model and they called it Gemma 7B, right? But then it's actually 8 billion parameters because the vocabulary is so large. And the reason they made the vocabulary so large is because TPUs, like Matrix Multiply Unit, is massive. Because that's what they've like sort of optimized for. And so they decided, oh, I'll just make the vocabulary large too, even though it makes no sense to do so in such a small model, because that fits on their hardware. So Gemma doesn't run as efficiently on a GPU as a Lama does, right? But by subversa, Lama doesn't run as efficiently on a TPU as a Gemma does, right?

4:12:46And it's so like there's like certain like aspects of like hardware software co -design so all their search models are they're ranking and recommendation models all these different models that are AI but not like Gen AI right have been I've been I've been optimized with GPUs forever the software stack is super optimized But all of this software stack has not been released publicly at all right Very small portions of it jacks and xla have been but like The experience when you're inside of Google and you're training on TPUs as a researcher you don't need to know anything about the hardware in many cases, right?

4:13:15Like it's like pretty beautiful, but as soon as you step outside, they'll go, a lot of them go back. They leave Google and then they go back. Yeah. Yeah, they're like, they leave and they start a company because they have all these amazing research ideas, and they're like, wait, infrastructure is hard, software is hard, and this is on GPUs. Or if they try to use GPUs, same thing because they don't have access to all this code. And so it's like, how do you convince a company whose Golden Goose is search where they're making hundreds of billions of dollars from to start selling GPUs, or TPUs, which they used to only buy a couple billion of, I think in 2023, they bought like a couple billion, and now they're buying like 10 billion to 15 billion dollars worth.

4:13:52But how do you convince them that they should just buy like twice as many and figure out how to sell them and make 30 billion dollars? Like who cares about making 30 billion dollars? Won't that 30 billion exceed actually the search profit eventually? I mean like you're always gonna make more money on services than the number. Always hard. I mean, yeah, like, to be clear, like today, people are spending a lot more on hardware than they are the services, right? Because the hardware front runs the service spend. But like, if there's no revenue for AI stuff or not enough revenue, then obviously, like it's gonna blow up, right?

4:14:26People won't continue to spend on GPUs forever. And Nvidia is trying to move up the stack with software that they're trying to sell on license and stuff, right? But Google has never had that DNA of like, this is a product we should sell, right? The Google Cloud is a separate organization from the TPU team, which is a separate organization from the DeepMind team. It shows a separate organization from the search team. There's a lot of bureaucracy. Like Google Cloud is a separate team within the TPU team. Technically, TPU sits under infrastructure with sits under Google Cloud, but like Google Cloud for like renting stuff and TPU architecture are very different goals, right, in hardware and software, like all of this, right?

4:15:04Like the Jax XLA teams do not serve Google's customers external where as Nvidia's various CUDA teams for like things like nickel serve external customers, right? Internal teams like Jackson, XLA and stuff, they more so serve deep -minded search, right? And so their customers different, they're not building a product for them. Do you understand why AWS keeps winning versus Azure for Cloud versus Google Cloud? Yeah, Google Cloud is tiny, isn't it relative to the other? Google Cloud is third. Microsoft is the second biggest, but Amazon is the biggest, right? And Microsoft deceptively sort of includes like Microsoft Office 365 and things like that like some of these enterprise wide licenses So in reality the Gulf is even larger Microsoft is still second though, right?

4:15:48Amazon is way bigger why because using AWS is better and easier and in many cases it was a cheaper and it's first it was first Yeah, but there's a lot of things that are first that well, it's easier. It's harder to switch than it is Yeah, okay, because there's big fees for switching to AWS generates over 80 % of Amazon's profit. I think over 90 % right insane. The distribution centers are just like one day will decide to make money from this, but they haven't yet, right? Like they make tiny little profit from one day of Amazon Prime will triple in price. You would think they would improve AWS interface because it's like horrible.

4:16:22It's like clunky, but everybody is. I allow. Yeah. You want to thank. I think actually Google's interface is sometimes nice, but it's also like they don't care about anyone besides their top customers. And like their customer service sucks, and like they have a lot less like. I mean all these companies they optimize for the big customers, yeah. It's supposed to be for business. Well, Amazon has always optimized for the small customer too though, right? Like obviously they optimize a lot for the big customer, but like when they started, they just would go to like random Bay Area things and give out credits, right?

4:16:51And then they like, or just put in your credit card and use us, right? Like it's went back in the early days. So they've always, the business has grown with them, right? in Burgeon. So why does Amazon, why Snowflake all over Amazon? Because Snowflake in the beginning, when Amazon didn't care about them, was still using Amazon, right? And then of course, one day, Snowflake and Amazon has a super huge partnership. But this is the case. Amazon's user experience and quality is better. Also, a lot of the silicon they've engineered makes them have a lower cost structure and traditional cloud storage CPU networking, that kind of stuff.

4:17:18Then in databases, basis, right? Like, you know, I think like four of Amazon's top five revenue products, margin products are like gross profit products are all database -related products like Red Shift and like all these things, right? Like, so, so Amazon has a very like good Silicon 2 user experience like entire pipeline with AWS. I think Google, they're, they're silicon teams. Yeah, they have awesome Silicon internally, TPU, the YouTube chip, you know, some of is they're not serving external customers or serving internal customers, right? I mean, Nvidia's entire culture is designed from the bottom up to do this.

4:17:56There's this recent book, The Nvidia Way, by take him that details this and they're how they look for future opportunities and ready their CUDA software libraries to make it. So that new applications of high performance computing can very rapidly be evolved on CUDA and Nvidia chips. And that is entirely different than Google as a services business. Yeah, I mean, Nvidia, it should be said as a truly special company. Like, I mean, they're the whole, the culture of everything. They're really optimized for that kind of thing. Speaking of which, is there somebody that can even challenge Nvidia hardware -wise?

4:18:32Intel AMD. I really don't think so. We went through a very long process of working with AMD on training on their GPUs and for instance, stuff. And they're decent. Their hardware is better in many ways than in Nvidia's. the problem is their software is really bad. And I think they're getting better, right? They're getting better faster, but they're just the Gulf is so large. And like they don't spend enough resources on it or haven't historically, right? Maybe they're changing their tune now, but you know, for multiple months we were submitting the most bugs, right? Like, oh, semi -analysis, right?

4:19:05Like what the fuck? Like why are we submitting the most bugs? Right? Cause they only cared about their like biggest customers. And so they'd ship them a private image, bubble blood. It's like, okay, but like I am just using PyTorch and I want to use the publicly available libraries. You don't care about that, right? So they're getting better. But like I think AMD is not possible. Intel's obviously in dire straits right now and needs to be saved somehow. Very important for national security for American. Can you explain the obviously so why are they in dire straits? Going back to earlier, only three companies can R &D, right?

4:19:39Taiwan, since you, to Samsung Pyongyang and then Intel Hillsboro. Samsung's doing horribly, Intel's doing horribly. We could be in a world where there's only one company that can do R &D. And that one company already manufactures most of chips. They've been gaining market share anyways, but that's a critical thing, right? So what happens to Taiwan means the rest of the world's semiconductor industry and therefore tech relies on Taiwan, right? And that's obviously precarious. As far as Intel, they've been slowly steadily declining. They were on top of servers and PCs, but now Apple's done the M1 and Nvidia's releasing a PC chip and Qualcomm's releasing a PC chip and in servers Hyperscalers are all making their own arm -based server chips and until has no AI silicon like wins right they have very small wins and They never got into mobile because they said no to the iPhone and like all these things have compounded and they've lost their process technology Leadership right they were ahead for 20 years and now they're behind by at least couple years right and they're trying to catch back up and we'll see if like their 18A, 14A strategy works out where they try and leapfrog to SMC.

4:20:42But like, and Intel is just like losing tons of money anyways, right? And they just fired their CEO, even though their CEO was the only person who understood the company well, right? We'll see. He was not the best, but he was pretty good, relatively, technical guy. Where does Intel make most of its money to CPUs, though? PCs and data center CPUs, yeah. But data center CPUs are all going cloud. and Amazon, Microsoft, Google, are making our own based CPUs. And then PC side, AMD's gained market share, Nvidia's launching a chip. That's not going to be success, right? MediaTek call -com every launch chips.

4:21:13Apple's doing well, right? Like they could get squeezed a little bit in PC, although PC generally I imagine will just stick intel mostly for Windows side. Let's talk about the broad AI race. Who do you think wins? We talked about Google, the leader. The default leader has been Google will be comes up their infrastructure advantage. Well, like in the news, open AI is the leader. They're the leading in the narrative. They have the best model. They have the best model that people can use and they're experts. And they have the most AI revenue. Yeah, open AI is winning. So who's making money on AI right now?

4:21:48Is anyone making money? So accounting profit wise, Microsoft is making money, but they're spending a lot of catbacks, right? You know, and that gets depreciated over years. Meta is making tons of money with recommendation systems, which is AI, but not with Alma. Alma's losing money for sure. I think anthropic and open AI are obviously not making money because otherwise they wouldn't be raising money. They have to raise money to build more. Although theoretically they are making money. You spent a few hundred million dollars on GPT -4 and it's doing billions in revenue. Obviously it's making money.

4:22:20Although they had to continue to research to get the compute efficiency wins and move down the curve to get that 1200X that has been achieved for GPT -3. Maybe we're only at a couple hundredX now, but with GPT -4 Turbo and 40 and there will be another one probably cheaper than GPT -4 -0 even that comes out at some point. And the research costs a lot of money. Yep, exactly. That's the thing that I guess is not talked about with the cost that when you're referring to the cost of the model. It's not just the training or the test runs. It's the actual research, the manpower. Yeah, to do things like reasoning right now that that exists.

4:22:58They're going to scale it. They're going to do a lot of research. I think people focus on the payback question, but it's really easy to just be like, well, GDP is humans and industrial capital. If you can make intelligence cheap, you can grow a lot. That's the dumb way to explain it, but that's sort of what basically the investment thesis is. I think only Nvidia is actually making tons of money and other hardware vendors. The hyperscalers are all on paper making money, but in reality, they're like spending a lot more on purchasing the GPUs, which you don't know if they're still going to make this much money on each GPU in two years.

4:23:34You don't know if all of a sudden OpenAI goes kapoof, and now Microsoft has hundreds of thousands of GPUs they were renting to OpenAI that they paid for themselves, but they're investment in them. That no longer have a customer, right? Like this is always a possibility. I don't believe that, right? I think opening I will keep raising money. I think others will keep raising money because the investments, the returns from it are gonna be eventually huge once we have AGI. So do you think multiple companies will get, let's just say? I don't think it's winner tickle. Okay. So it's not, let's not call it AGI, whatever.

4:24:11It's like a single day. It's a gradual thing. Super powerful AI. But it's a gradually increasing set of features that are useful and rapidly increasing rapidly increasing set of features So you're saying a lot of companies will be It just seems absurd That all of these companies are building gigantic There were companies that will benefit from AI But not because they trained the best model like meta has so many avenues to benefit from AI and all of their services people are there, people spend time on Metas platforms and it's a way to make more money per user per hour. Yeah, it seems like Google X, slash XAI, slash Tesla, important to say, and the meta will benefit not directly from the AI, like the LLMs, but from the intelligence, like the additional boost of intelligence to the products they already sell.

4:25:07So whether that's the recommendation system, or for Elon who's been talking about Optimus, the robot potentially the intelligence of the robot. And then you have personalized robots in the home, that kind of thing. He thinks it's a 10 plus trillion dollar business, which at some point maybe, I don't not soon, but who knows what robots are. Let's do a tam analysis, right? Eight billion humans and let's get eight billion robots, right? And let's pay them the average salary. And there we go, 10 trillion, more than 10 trillions. Yeah, I mean, you know, if there's robots everywhere, why does it have to be just eight billion robots?

4:25:47Yeah, of course, of course. I'm gonna have like one robot, you're gonna have like 20. Yeah, I mean, I see a use case for that. So yeah, so I guess the benefit would be in the products as well, which is why opening eyes in a trickier position because they, all of the value of open AI right now as a brand is in Chatchy PT. And there is actually not that, for most users, there's not that much of a reason that they need open AI to be spending billions and billions of dollars on the next best model when they can just license llama five and for be way cheaper. So that's kind of like chat. that GVT is an extremely valuable entity to them.

4:26:23But it like, they could make more money just about that. The chat application is clearly like, does not have tons of room to continue, right? Like the standard chat, right? Where you're just using it for a random question and stuff, right? The cost continues to collapse V3 is the later one. It'll add ads. Biggest, but it's gonna get supported by ads, right? Like as, you know, a lot of metal already serves 405B, probably loses the money, but at some point, you know, they're going to get, the models are gonna get so cheap that they can just serve them for free with ad supported, right? And that's what Google is going to be able to do.

4:26:52And that's obviously they've got a bigger reach, right? So chat is not going to be the only use case. It's like these reasoning, code, agents, computer use, all this stuff is where OpenAI has to actually go to make money in the future. Otherwise, they're complex. But X, Google, and Meta have these other products. So this isn't likely at OpenAI and then and then the topic disappear eventually. Unless they're so good at models, they are. But it's such a cutting, I mean, it depends on where you think AI capabilities are going. You have to keep winning. Yes. You have to keep winning. As you climb, even if the AI capabilities are going super rapidly, awesome into the direction of AGI, like there's still a boost for X in terms of data, Google in terms of data, meta in terms of data, in terms of other products, and the money, and like there's just huge amounts of money.

4:27:44The whole idea is human data is kind of tapped out. We don't care. We all care about self play verifiable. Yeah, self play. You think about AWS, which is an RNG problem. AWS does not make a lot of money on each individual machine. And the same can be said for the most powerful AI platform, which is even though the calls to the API are so cheap, there's still a lot of money to be made by owning that platform. And there's a lot of discussions as it's the next compute layer. You have to believe that, and yeah, there's a lot of discussions that tokens and tokenomics and LLM APIs are the next compute layer or the next paradigm for the economy, kind of like energy and oil was.

4:28:20But there's also like, you have to sort of believe that APIs and chat are not where AI is stuck, right? It is actually just tasks and agents and robotics and computer use. And those are the areas where all the dev -value will be delivered, not API, not chat application. Is it possible you have, I mean, it all just becomes a commodity and you have the very thin wrapper like perplexity just joking. There are a lot of wrappers making a lot of money. Yeah, so but do you think it's possible that people would just even forget what open AI and the thought it is and just because there would be wrappers around the API and it just dynamically.

4:29:00If model progress is not rapid, yeah, it's becoming a commodity, right? right deep seek V3 shows this, but also the GPT -3 chart earlier, Kurt Chart showed this, right? Lama 3B is 1200X cheaper than GPT -3. Any GPT -3, like anyone whose business model was GPT -3 level capabilities is dead. Anyone whose business model's GPT -4 level capabilities is dead. It is a common thing that the best businesses being made. Now are ones that are predicated on models getting better. Right. Which would be like rappers, thing that is riding the wave of the models. The short term of the company that could make the most money is the one that figures out what advertising targeting method works for language model generations.

4:29:39We have the meta ads which are hyper targeted in feed, not within specific pieces of content. And we have search ads that are used by Google, and Amazon has been rising a lot on search. But within a piece, within a return from Chat GBT, it is not clear how you get a high quality placed ad within now put. And if you can do that with model costs coming down, you can just get super high revenue. That revenue is totally untapped and it's not clear technically how it is done. Yeah, that is sort of the add sense innovation that Google did. The one day you'll have in GPT output an ad and that's going to make billions of not a lot of money.

4:30:20And it could be in conversation. We have voice mode now. It could be some way of making it so the voice introduces certain things. It's much harder to measure and it takes the imagination, but yeah, and it wouldn't be so shate. It wouldn't come off shady so that you would receive public blowback. That kind of thing. So you have to do it loud enough to where it's clear it's an ad and balance all of that. So that's the open question that's trying to solve and thropic and open AI. They need to they might not say that they're trying about that at all. They don't care about it right now. I think it's I think they're perfectly are experimenting.

4:30:52Yeah. Yeah, I'm so excited about that. Oh, interesting. Yeah, for sure. Like, for complexity, Google Meta care about this. I think open -eye and enthropic are purely laser focused on AGI. Yeah. Like, agents and AGI, and if I build AGI, I can make tons of money, right? Or I can pay for everything, right? And this is, this is, you, it's just predicated, like, back on the, like, export -controlled thing, right? If you think AGI is five, ten years away or less, right? These labs think it's two, three years away. Obviously, your actions are, you know, if you assume they're rational actors, which they are mostly, what you do in a two year AGI versus five year versus ten years, very, very, very different, right?

4:31:34Do you think agents are promising? We have to talk about this. This is like the excitement of the year that agents are going to rubber. This is the generic hype term that a lot of business folks are using. AI agents are going to revolutionize everything. Okay, so mostly the term agent is obviously overblown. We've talked a lot about reinforcement learning as a way to train for verifiable outcomes. Agents should mean something that is open -ended and is solving a task independently on its own and able to adapt to uncertainty. There is a lot of the term agent applied to things like Apple Intelligence, which we still don't have after the last WWDC, which is orchestrating between apps.

4:32:16And that tool use thing is something that language models can do really well. Apple intelligence, I suspect, will come eventually. It's a closed domain. It's your messages app integrating with your photos with AI in the background. That will work. That has been described as an agent by a lot of software companies to get into the narrative. The question is, what ways can we get language models to generalize to new domains and solve their own problems in real time? Maybe some tiny amount of training when they are doing this with fine -tuning themselves or in context learning, which is the idea of storing information in a prompt and you can use learning algorithms to update that and whether or not you believe that that is going to actually generalize to things like me saying book my trip to go to Austin in two days.

4:33:06I have XYZ constraints and actually trusting it. I think there's a HCI problem coming back for information. Well, what's your prediction there? Because my gut says we're very far away from that. I think opening eyes statement, you've heard it, if you've seen the five levels, right, or it's Chat is level one, reasoning is level two, and then agents is level three, and I think there's a couple more levels, but it's important to note, right? We were in chat for a couple of years, right? We just theoretically got to reasoning. We'll be here for a year or two, right? And then agents, but at the same time, like people can try and approximate capabilities of the next level, but agents are doing things autonomously, doing things for minutes at a time, hours at a time, et cetera, right?

4:33:52Reasoning is doing things for tens of seconds at a time, right? And then coming back with an output that I still need to verify and use and try, check out, right? And the biggest problem is, of course, like, it's the same thing with manufacturing, right? Like there's the whole six sigma thing, right? Like, you know, how many nines do you get and then you compound the nines onto each other? And it's like, if you multiply, you know, by the number of steps that are six sigma, you get to, you know, a yield or something, right? So like in semi -gunner manufacturing, tens of thousands of steps, 99999 is not enough, right?

4:34:24Because you multiply that by that many times, you actually end up with like 6 % yield. Right? Yeah, or low yield. Yeah, or zero. And this is the same thing with agents, right? Like, chaining tasks together each time, LLMs, even the best LLMs in particularly pretty good benchmarks, don't get 100%. Right? They get a little bit below that because there's a lot of noise. And so how do you get to enough Nines, right? This is the same thing with self -driving. We can't have self -driving because without it being like super geo -fenced like Google, like Google's, right? And even then they have a bunch of tele operators to make sure it doesn't get stuck, right?

4:34:58But you can't do that because it doesn't have enough Nines. And self -driving has quite a lot of structure because roads have rules. It's well defined, there's regulation. When you're talking about computer use for the open web, for example, or the open operating system. Like, there's no, it's a mess. So like, the possibility, I'm always skeptical of any system that is tasked with interacting with the human world, with the open, messy human. That's the same thing. If we can't get intelligence that's enough to solve the human world on its own, we can create infrastructure, like the human operators for Waymo, over many years that enable certain workplaces.

4:35:42There is a company, I don't remember it, but it is, but that's literally their pitches. Yeah, we're just gonna be the human operator when agents fail. And you just call us and we fix it. It's like an API call and it's hilarious. There's gonna be teleoperation markets when we get human robots, which is, there's gonna be somebody around the world that's happy to fix the fact that it can't finish loading my dishwasher when I'm unhappy with it, but that's just gonna be part of the Tesla service package. I'm just imagining like an AI agent talking to another AI agent. One company has an AI agent that specializes in helping other AI agents.

4:36:15But if you can make things that are good at one step, you can stack them together. So that's why I'm really, if it takes a long time, we're gonna build infrastructure that enables it. You see the operator launch, they have partnerships with certain websites, with DoorDash, with OpenTable, with things like this. Those partnerships are gonna let them climb really fast. Their model's gonna get really good at those things. It's going to prove a concept that might be a network effect where more companies want to make it easier for AI. Some companies will be like, no, let's put blockers in place. And this is the story of the internet we've seen.

4:36:48We see it now with training data for language models where companies are like, no, you have to pay. And business is working it out. That said, I think airlines have a very, and hotels have high incentive to make their site work really well and they usually don't. Like if you look at how many clicks it takes to order airplane ticket, it's insane. I don't... You actually can't call an American Airlines agent anymore. They don't know about number. I mean, it's horrible on many on the interface front. And to imagine that agents will be able to deal with that website when I as a human struggle, like I have an existential crisis every time I try book an airplane ticket that I don't...

4:37:30I think it's going to be extremely difficult to build an AI agent that's robust in that way. But think about it, like United has accepted the Starlink term, which is they have to provide Starlink for free and the users are going to love it. What if one airline is like, we're going to take a year and we're going to make our website have white text that works perfectly for the AI's? Every time anyone asks about an AI flight, they buy whatever airline it is. Or like, they just like, here's an API in, it's only exposed to AI agents and if anyone queries it, the price is 10 % higher. And for any flight, but we'll let you see any of our flights and you can just book any of them.

4:38:05Here you go, aging man. That's like, oh, and I made 10 % higher price. Awesome. And like, am I willing to say that for like, hey, book me a flight to C -Lex, right? And it's like, yeah, whatever. I think computers and real world and the open world are really, really messy. But if you start defining the problem in narrow regions, people are gonna be able to create very, very productive things. and ratchet down cost massively, right? Like now crazy things like, you know, robotics in the home, you know, those are gonna be a lot harder to do, just like self -driving, right? Because there's just a billion different failure amounts, right?

4:38:42But like agents that can like navigate a certain set of websites and do certain sets of tasks, or like look at, you know, look at your, you know, take a photo of your grocery, your fridge and or like upload your recipes and then like it figures out what to order from, you know, Amazon, slash Whole Foods, food delivery, that's gonna be pretty quick and easy to do, I think. So it's gonna be a whole range of business outcomes and it's gonna be tons of optimism around people can just figure out ways to make money. To be clear, these sandboxes already exist in research. There are people who have built clones of all the most popular websites of Google, Amazon, blah, blah, blah, to make it so that there's, I mean, opening, I probably has them internally to train these things.

4:39:21It's the same as DeepMind's robotics team for years has had clusters for robotics where you interact with robots fully remotely. They just have a lab in London and you send tasks to it, it arranged the blocks and you do this research. Obviously, there's text there that thick stuff, but we've turned these cranks of automation before. You go from sandbox to progress and then you add one more domain at a time and generalize. In the history of NLP and language processing, instruction tuning in tasks per language model all used to be like one language model did one task. And then in the instruction tuning literature, there's this point where you start adding more and more tasks together, where it just starts to generalize to every task.

4:40:01And we don't know where on this curve we are. I think for reasoning with this RL and verifiable domains, we're very early, but we don't know where the point is where you just start training on enough domains and poof, like more domains just start working and you've crossed the generalization barrier. What do you think about the programming context? So software engineering. That's where I personally, I know a lot of people interact with AI the most. There's a lot of fear and angst too from current CS students, but there's also that's where that is the area where probably the most AI revenue and productivity gains have come.

4:40:38Whether it be co -pilots or cursor or what have you, right? This is or just standard chat GPT right like a lot of I know very few programmers We don't have chat GPT and actually many of them have the $200 tier because that's what it's so good for right I think that in that world We already see it like sweet bench and if you've looked at the benchmark Made by some Stanford students. I wouldn't say it's like really hard But I wouldn't say it's easy either. I think like it takes someone who's been through at least You know a few years of CS or a few a couple years of programming to do sweet bench well And the models went from 4 % to 60 % in like a year, right?

4:41:16And where are they gonna go to next year? You know, it's gonna be higher. It probably won't be 100 % because again, that 9's is like really hard to do. But we're gonna get to some point where that's, and then we're gonna need harder software engineering benchmarks and so on and so forth. But the way that people think of it now is it can do code completion easy. It can do some function generation and have to review it. Great. But really the software engineering agents, I think can be done faster sooner than any other agent because it is a verifiable domain. You can always like unit test or compile.

4:41:47And there's many different regions of like, you can inspect the whole codebase at once, which no engineer really can. Only the architects can really think about this stuff, the really senior guys, and they can define stuff. And then the agent can execute on it. So I think software engineering costs are going to plummet like crazy. And one interesting aspect of that is when software engineering costs are really low, you get very different markets, right? So in the US, you have all these platforms asked companies, right? Sales force and so on and so forth, right? In China, no one uses platforms asks.

4:42:20Everyone just builds their own stack because software engineering is much cheaper in China, partially because like people, a lot of STEM graduates, et cetera. So it's generally just cheaper to do. And so at the same time, code for, like code LLMs have been adopted much less in China because the cost of an engineer there is much lower. But what happens when every company can just invent their own business logic really cheaply and quickly? You stop using platform SaaS, you start building custom tailored solutions, you change them really quickly. Now all of a sudden, your business is a little bit more efficient, too potentially, because you're not dealing with the hell that is some random platform SaaS company stuff not working perfectly and having to adjust workflows or random business automation cases that aren't necessarily, I require it's just logic that needs to be built that no one has built.

4:43:04All of these things can go happen faster. So I think software and then the other domain is like industrial chemical mechanical engineers second coding right just generally and like There are tools like semiconductor engineers their tools are 20 years old all the tools run on XP including a sml lithography tools run on windows XP right it's like you know and like A lot of the analysis happens in excel, right? Like it's just like guys like you guys can move 20 years forward with all the data You have and gathered and like do a lot better. It's just you need the engineering skills for software engineering to be delivered to the actual domain expert engineers.

4:43:36So I think that's the area where I'm like super duper bullish of generally AI creating value. The big picture is that I don't think it's going to be a clip. We talked to anything, a really good example of how growth changes is when meta added stories. So Snapchat was on an exponential. They added stories at flatline. Software engineers, then up until the right. AI is going to come in. It's probably going to be flat. It's like, it's a lot like Everyone's gonna lose their job. It's hard because the supply corrects more slowly, so the amount of students is still growing, and that'll correct on a multi year delay, but the amount of jobs will just turn, and then maybe in 20, 40 years, it'll be well down.

4:44:20But in the few years, there'll never be the snap moment where it's like software engineers aren't useful. I think also the nature of what it means to be a programmer and what kind of jobs programmers do changes. because I think there needs to be a human in the loop of everything you've talked about. There's a really important human in that picture of correcting the code. Like thinking smarter than the context lens. And debugging also. Like debugging by sort of reading the code, understanding the steering the system, like, no, no, you missed the point, adding more to the prompt. kind of like, yes, adding the human.

4:44:59Designing the perfect Google button. Google's famous for having people design buttons that are so perfect. And it's like, how is AI gonna do that? Like, they could give you all the ideas. Perfect. I mean, that's the thing. You can call it taste. Humans have, one thing humans can do is figure out what other humans enjoy better than AI systems. That's where the preference you loading that in, but ultimately humans are the greatest preference that's where the preference comes from. And humans are actually very good at reading, or like judging between two things versus, this goes back to the core of what Arley Jeff in preference tuning is, is that it's hard to generate a good answer for a lot of problems, but it's easy to see which one is better.

4:45:39And that's how we're using humans for AI now, is judging which one is better, and that's what's off -engineering could look like. The PR of you, here's a few options. What are the like, here's some potential pros and cons, and they're gonna be judges. judges? I think the thing I would very much recommend is people start programmers start using AI and embracing that role of the supervisor of the AI system and like partner of the AI system versus writing from scratch or not learning coding at all and just generating stuff because I think there actually has to be a pretty high level of expertise as a programmer to be able to manage increasingly intelligent systems.

4:46:18I think it's that and then becoming a domain expert in something. Yeah, because you like seriously if you go look at aerospace or semiconductors or chemical engineering everyone is using really crappy platforms really old software like the job of a data sciences is like is like a joke right in many cases. In many cases is very real but it's like bring what the forefront of human capabilities are to your domain and like even if the forefront is like from the AI your domain you're like at the forefront. So it's like you have to be at the forefront of something and then leverage the like rising tide that is AI for everything else Oh, yeah, there's so many low hanging fruit Everywhere in terms of wear soft work and like help automate a thing or digitize a thing in In the legal system.

4:47:03I mean that's why doge is exciting You ever I mean I got the Hang out with a bunch of the doge folks and they I mean government is like so old school It's like begging for the modernization of software of organizing the data, all of this kind of stuff. I mean, in that case, it's by design because bureaucracy create protects centers of power and so on. But software breaks down those barriers. So it hurts those that are holding on to power but ultimately benefits humanity. So there's a bunch of domains of that kind. And one thing we didn't fully finish talking about is open source. So first of all, congrats.

4:47:51You released a new model. Yeah, this is the... To Lou. I'll explain what a To Lou is. A To Lou is a hybrid camel when you breed a Drowmorderry with a backbuckery in camel. Back in the early days after ChatGPT, there was a big wave of models coming out like Alpaca and Vikuna, etc. That were all named after various mammalian species. So, Tulu is the brand is multiple years old, which comes from that. And we've been playing at the frontiers of post training with Open Source code. And this first part of this release was in the fall where we built on Lama's Open Models, Open Weight Models, and then we add in our fully open code or fully open data.

4:48:32There's a popular benchmark that is Chatbot Arena, and that's generally the metric by which how these chat models are evaluated and its humans compare random models from different organizations. And if you looked at the leaderboard in November or December among the top 60 models from 10s to 20s of organizations, none of them had open code or data for just post training. Among that, even fewer or none have pre -training data and code available, but post -training is much more accessible at this time. It's still pretty cheap and you can do it. And the thing is like how high can we push this number where people have accessed all the code and data.

4:49:05So that's kind of the motivation of the project we draw on lessons from Lama in video had a Nemotron Model where the recipe for their post -training was fairly open with some data and a paper and it's putting all these together to try to create a recipe that people can Find two models like GPT -4 to their domain. So to be clear in the case of Tulu Maybe you can talk about Alma too, but in the case of Tulu you're taking Lama 345b Too, Lu has been a series of recipes for post -training. So we've done multiple models over years. And so you're open sourcing everything. Yeah. If you start with an open weight based model, the whole model technically is an open source and you don't know what Lama put into it, which is why we have a separate thing that we'll get to.

4:49:50But it's just getting parts of the pipeline where people consume in and customize. I know I hear from startups and businesses that are like, okay, I can take this post -training and try to apply it to my domain. We talk about verifiers a lot. we've used this idea which is reinforced learning with verifiable rewards, RLVR, kind of similar to RLHF, and we've applied it to map. And the model today, which is like we applied it to the Lama 405B base model from last year, and we have our other stuff, we have our instruction tuning and our preference tuning, but the math thing is interesting, which is like it's easier to improve this math benchmark.

4:50:28There's a benchmark MATH math, all capitals, tough name. When the benchmark is name is the area that you're evaluating, we're researchers, we're not brands, brands, strategists. And this is something that the deep seek paper talked about as well as like at this bigger model, it's easier to elicit powerful capabilities with this RL training. And then they distill it down from that big model to the small model. And this model we released today, we saw the same thing as we're AI2, we don't have a ton of compute, we can't train 405B models all the time. So we just did a few runs and they tend to work and it's like, it just shows that there's a lot of room for people to play in these things and they crushed Lama's actual release, right?

4:51:09Like they're way better than it. Yeah, so our Val numbers, I mean, we have extra months in this, but our Val numbers are much better than the Lama Instruct model that they released. And then you also said better than deep seek V3. Yeah, on our Val benchmark. The most deep seek V3 is really similar. We have a safety benchmark to understand if it will say harmful things and things like that. And that's what draws down no silhouette. It's still like, it's like an amalgamation of multiple benchmarks. Or what do you mean? Yeah. So we have a 10 evaluate. This is like, this is standard practice in post training is you choose your evaluations you care about in academics and smaller labs.

4:51:41You'll have fewer evaluations in companies. You'll have a really one domain that you really care about in frontier labs. You'll have 10s to 20s to maybe even like 100 evaluations of specific things. So we choose a representative suite of things that look like chat, precise instruction following, which is like respond only in emojis. Like does model follow weird things like that? Yeah. Math could. And you create a suite like this. So safety would be one of 10 in that type of suite, where you have like, what is the broader community of AI care about? And for example, in comparison to deep seek, it would be something like our average of our model would be 80, including safety and similar without.

4:52:17And deep seek would be like 79 % average score or without safety and their safety score would bring it down to like some time. So you beat them even ignoring safety? Yeah, so this is something that internally, it's like I don't want to win only by like how you shape the Val benchmark. So if there's something that's like people may or may not care about safety in their model, safety can come downstream, safety can be when you host the model for an API. Like safety is addressed in a spectrum of locations and they have applications. So it's like if you want to say that you have the best recipe, you can't just gated on these things that some people might not want.

4:52:50And this is like the time of progress. We can release a model later. We have more time to learn new techniques, like this RL technique. We had started this in the fall. It's now really popular as reasoning models. The next thing to do for open source post training is to scale up verifiers, to scale up data to replicate some of deep -seek results. And it's awesome though we have a paper to draw and then it makes it a lot easier. And that's the type of things that is going on among academic and closed frontier research in AI. Since you're pushing open source, what do you think is the future of it?

4:53:25You think deep seek actually changes things since it's open source or open weight or is pushing the open source movement into the open direction? This goes very back to license discussion. So deep seek R1 with a friendly license is a major reset. So it's like the first time that we've had a really clear frontier model that is open weights and with a commercially friendly license with no restrictions on downstream use cases since that data distillation, whatever. This has never been the case at all in the history of AI in the last few years since Chat GBT. There have been models that are off the frontier or models with weird licenses that you can't really use them.

4:53:58So is it meta's license pretty much permissible except for five companies? And so this goes to what open source AI is, which is there's also use case restrictions in the Lama license, which says you can't use it for specific things. So if you come from an open source software background, you would say that that is not an open source license. What kind of things are those though? Like are they like, at this point, I can't pull them off. I have my stuff like competitor or starboard. It used to be military use was one, and they removed that for scale. It'll be like, like CSAM, like trial abuse material.

4:54:31Or like, that's the type of thing that is forbidden there, but that's enough from an open source background to say it's not an over source license. And also the Lama license has this horrible thing where you have to name your model Lama, if you touch it to the Lama model. So it's like the branding thing. So if a company uses Lama technically the license says that they should say built with Lama at the bottom of their application And from like a marketing perspective that just that just hurts like I could suck it up as a researcher I'm like oh this line like it says Lama dash on all of our on all of our materials for this release But this is why we need truly open models, which is We don't know deep seek our ones data.

4:55:06So you're saying I can't make a you know cheap copy of Lama and pretend it's mine But I can do this with the Chinese model. Yeah Yeah. That's what I was saying. And that's why we want this whole open language models thing, the all -mote thing, is to try to keep the model where everything is open with the data as close to the frontier as possible. So we're compute constrained, we're personnel constrained, we rely on getting insights from people like John Schoenman tells us to do our own outputs. We can make these big jumps, but it just takes a long time to push the frontier of open source and fundamentally, I would say that that's because open source AI does not have the same feedback loops as open source software.

4:55:46We talked about open source software for security. Also, it's just because you build something once and you can reuse it. If you go into a new company, there's so many benefits. But if you open source a language model, you have this data sitting around, you have this training code. It's not that easy for someone to come and build on improve, because you need to spend a lot on compute, you need to have expertise. So until there are feedback that loops of open source AI, it seems mostly an ideological mission. People like Mark Zuckerberg, which is like America needs this. And I agree with him, but in the time where the motivation ideologically is high, we need to capitalize and build this ecosystem around what benefits do you get from seeing the language model data.

4:56:27And there's not a lot about that. We're going to try to launch a demo soon where you can look at an Olimo model and a query and see what pre -training data is similar to it. which was like legally risky and complicated, but it's like, what does it mean to see the data that the AI was trained on? It's hard to parse. It's terabytes of files. It's like, I don't know what I'm going to find in there. But that's what we need to do as an ecosystem if people want open source AI to be financially useful. We didn't really talk about Stargate. I would love to get your opinion on like, what the new administration, the Trump administration, everything that's doing, that's being done from the America side and supporting AI infrastructure and the efforts of the different AI companies.

4:57:10What do you think about Stargate? What are we supposed to think about Stargate and the SAM have the money? Yeah, so I think Stargate is a opaque thing. It definitely doesn't have $500 billion. Doesn't even have $100 billion, right? So what they announced is this $500 billion number, Larry Ellison, Sam Altman and Trump said it. They thanked Trump and it's used, the Trump did do some executive actions that do significantly improve the ability for this to be built faster. One of the executive actions he did is on federal land, you can just basically build data centers in power, pretty much like that.

4:57:49The permitting process is basically gone or you file after the fact. One of the again, like I said, I had a skitzo take earlier, another skitzo take if you've ever been to the Presidio in San Francisco, beautiful area. You could build power plant in a data center there if you wanted to because of this federal land used to be a military base. But you know, obviously this was like piss people off. You know, it's a good bit. Anyways, Trump has made it much easier to do this, right? Generally, Texas has the only unregulated grid in the nation as well. Let's go Texas. And so, you know, therefore like our cot enables people to build faster as well.

4:58:24In addition, the federal regulations are coming down. And so Stargate is predicated. This is why that whole show happened. Now, how they came up with a $500 billion number is beyond me. How they came up with a $100 billion number makes sense to some extent, right? And there's actually a good table in here that I would like to show in that Stargate piece that I had.

4:58:49It's the most recent one, yeah. So anyways, Stargate, You know, it's basically right like there is it's a table about cost There you passed it already it's that one So this table is kind of explaining what happens right so Stargate is in Abilene, Texas the first hundred billion dollars of it That site is 2 .2 gigawatts of power in about 1 .8 gigawatts of power consumed. Per GPU, they have roughly, Oracle is already building the first part of this before Stargate came about. To clear they've been building it for a year, they tried to rent it to Elon, in fact. But Elon was like, it's too slow, I need it faster, so then he went into his Memphis thing.

4:59:34And so opening, I was able to get it with this weird joint venture called Stargate. They initially signed a deal with just Oracle for the first section of this cluster. right? This first section of this cluster right is roughly $5 billion to $6 billion of server spend right and then there's another billion or so of data center spend but the and then likewise like if you fill out that entire 1 .8 gigawatts with the next two generations of NVIDIA's chips, GB200, GB300, VR200 and you fill it out completely that ends up being roughly $50 billion of server cost, right? Plus there's data center cost, plus maintenance cost, plus operation cost, plus all these things.

5:00:15And that's where OpenAI gets to their $100 billion announcement that they had, right? Because they talked about $100 billion as phase one. That's this Abilene Texas data center, right? $100 billion of total cost of ownership, quote, unquote, right? So it's not CAPEX, it's not investment, it's $100 billion of total cost of ownership. And then there will be future phases they're looking at other sites that are even bigger than this 2 .2 gigawatts, by the way, in Texas and elsewhere. And so they're not completely ignoring that, but there is the number of $100 billion that they say is for phase one, which I do think will happen.

5:00:49They don't even have the money for that. Furthermore, it's not $100 billion. It's $50 billion of spend, and then like $50 billion of operational cost, power, et cetera, rental pricing, et cetera, because they're renting it open a eyes renting the GPUs from the Stargate joint venture. What money do they actually have? Softbank is going to invest oracle is going to invest. Open a eyes on the line for $19 billion. Everyone knows that they've only got $6 billion in their last round and $4 billion in debt. But there is news of softbank investing $25 billion into open a eye. That's part of it. 19 billion can come from there.

5:01:26Open a eye does not have the money at all, to be clear. Ink is not dried on anything, opening has zero dollars for this 50 billion, right? In which they're legally obligated to put 19 billion of CapEx are into the joint venture and then the rest they're going to pay via renting the GPUs from the joint venture. And then there's, then there's Oracle. Oracle has a lot of money. They're building the first section completely. They were spending for themselves, right? This is $6 billion of CapEx, $10 billion of TCO. But they and they were going to do that first section. They're paying for that, right?

5:01:56As far as the rest of the section, I don't know how much Larry wants to spend, right? At any point, he can pull out, right? Like this is again, it's like completely voluntary. So at any point, there's no signed ink on this, right? But he potentially could contribute tens of billions of dollars, right? To be clear, he's got the money, or it was got the money. And then there's like MGX, which is the UAE fund, which technically has $1 .5 trillion for investing in AI. But again, like, I don't know how real that money is. And like, whereas there is no ink signed for this, Softbank does not have $25 billion at cash.

5:02:27they have to sell down their stake in ARM, which is, you know, the leader and CPUs and they, they IPO'd it. This is obviously what they've always wanted to do. They just didn't know where they'd redeploy the capital. Selling down the stake in ARM makes a ton of sense. So they can sell that down and invest in this if they want to and invest in OpenA, if they want to. As far as like money secured, the first 100 ,000 GB 200 cluster is like, can be funded. Everything else after that is up in the air, money's coming. I believe the money will come. I personally do. It's a belief. It's a belief that they are going to release better models and be able to raise more.

5:03:01But the actual reality is that Elon's right, the money does not exist. What is the US government have to do with it? What does Trump have to do with everything? He's just a hype man. Trump is reducing the regulation so they can build it faster. And he's allowing them to do it. Any investment of this side is going to involve anti -trust stuff. So obviously he's going to allow them to do it. He's going to enable the regulations to actually allow it to be built. I don't believe there's any US government dollars being spent on this though. Yeah, so I think he's also just creating a general vibe that this is regulation will go down and this is the era of building.

5:03:40So if you're a builder, you want to create stuff, you want to launch stuff, this is the time to do it. And so look, we've had this 1 .8 gigawatt data center in our data for over a year now. And we've been like sort of sending it to all of our clients, including many of these companies that are building the multi -gigawatts. But that is like at a level that's not quite maybe executives like seeing $500, $100 billion and then everyone's asking them like, so it could spur like another, like an even faster arms race, right? Cause there's already an arms race, but like this, this like $100 ,500 ,000 ,000 number Trump talking about it on TV, like it could spur the arm race to be even faster and more investors to flood in and et cetera, et cetera.

5:04:15So I think, I think your right is that in that sense that Trump is sort of like championing people are going to build more and his actions are going to let people build more. What are you excited about about these several years that are upcoming in terms of cluster buildouts, in terms of breakthroughs in AI, like the best possible future you can imagine in the next couple of years, two, three, four years, what does that look like? Just it could be a very specific technical things like breakthroughs on post training or it could be just size big. Yeah, I mean, it's impressive clusters. I really, I really enjoyed tracking supply chain and like who's called and what?

5:04:59Yeah, I really do. It's really fun to see like the numbers, the cost, who's building what capacity, helping them figure out how much capacity should build, winning deals, strategic stuff. That's really cool. I think technologically, there's a lot around the networking side that that really excites me with Optics and Electronics, right? Like kind of getting closer and closer whether it be co -package optics or some sort of like forms of new forms of switching. This is internal to a cluster. Yeah. Also multi -data center training, right? Like there's people are putting so much fiber between these data centers and lighting it up with so much bandwidth that there's a lot of interesting stuff happening on that end, right?

5:05:36Telecom has been really boring since 5G and now it's like really exciting again. Can you educate me a little bit about the speed of things? So the speed of memory versus the speed of interconnect versus the speed of fiber between data centers? Are these like orders of magnitude different? Can we at some point converge towards a place where it all just feels like one computer? Oh, no, I don't think that's possible. It's only going to get harder to program, not easier. It's only going to get more difficult and complicated and more layers, right? The general image that people like to have is this hierarchy of memory.

5:06:11So on chip is really close, localized within the chip, right? You have registers, right? Those are shared between some compute elements. And then you'll have caches, which are shared between more compute elements. Then you have like memory, right? Like HBM or DRAM, like DDR memory or whatever it is. And that's shared between the whole chip. And then you can have pools of memory that are shared between many chips, right? And then storage and it keeps zoning out, right? The access latency across data centers, within the data center, within a chip is different. So like you're obviously always, you're always gonna have different programming paradigms for this.

5:06:44It's not gonna be easy programming this stuff is gonna be hard, maybe I can help, right? You know, with programming this. But the way to think about it is that like there is,

5:06:58there's sort of like the more elements you add to a task, you don't gain, you don't get strong scaling, right? If I double the number of chips, I don't get two exit performance. This is just like a reality of computing, because there's inefficiencies. And there's a lot of interesting work being done to make it more linear, whether it's making the chips more networked together, more tightly, or cool programming models, or cool algorithmic things that you can do on the model side. DeepSeek did some of these really cool innovations, because they were limited on Interconnect, but they still needed a parallelize.

5:07:29All sorts of, everyone's always doing stuff. Google's got a bunch of work. and everyone's got a bunch of work about this. That stuff is super exciting on the model and workload and innovation side, right? Hardware, solid state transformers are interesting, right for the power side. There's all sorts of stuff on batteries and there's all sorts of stuff on, you know, I think when you look at, if you look at every layer of the compute stack, right? Whether it goes from lithography and etch all the way to like fabrication, to like optics, to networking, to power, to transformers, to cooling, to, you know, networking and you just go on, up and up and up and up the stack, you know, Even air conditioners for data centers are like innovating, right?

5:08:04Like it's like there's like copper cables are innovating, right? Like you wouldn't think it but copper cables like are there some innovations happening there with like the density of how you can pack them And like it's like all of these layers of the stack all the way up to the models human progress is at a pace that's never been seen before I mean, just imagine you sitting back in a layer somewhere with screens everywhere Just monitoring the supply chain where all these clusters like all the information you're gathering I mean, there's a big team. There's a big team You do quite incredible work with semi -analysis.

5:08:35I mean, just keeping your finger on the pulse of human civilization in the digital world. It's pretty cool, just to watch, feel that. Yeah, thank you. I guess. I feel all of us doing shit. Epic shit. Feel the AGI. Feel the AGI. I mean, from meme to reality. What is there like breakthroughs that you're looking forward to potentially? I had a wild think about this while listening to Delan's beautiful shot. He did listen to me. No, I knew this was coming. And it's like realistically, training models is very fun because there's so much low hanging fruit. And the thing that makes my job entertaining, I train models, I write analysis about what's happening with models.

5:09:20And it's fun because there is obviously so much more progress to be had. And the real motivation why I do this for someone where I can share things is that there's just, I don't trust people that are like, trust me, bro, we're going to make AI good. It's like, we're the ones that it's like, we're going to do it and you can trust us and we're just going to have all the AI. And it's just like, I would like a future where more people have a say in what AI is and can understand it. And that's a little bit less fun than it's not a positive thing. I feel like this is just all really fun. Like training models is fun and bringing people in is fun.

5:09:53but it's really AI, if it is going to be the most powerful technology of my lifetime, it's like, we need to have a lot of people involved in making that and making it open helps with that. As successful as possible, is it open as possible? And my read of the last few years is that more openness would help the AI ecosystem in terms of having more people understand what's going on, rather that's researchers from non -AI fields, to governments, to everything. It doesn't mean that openness will always be the answer. I think that it will reassess of like what is the biggest problem facing AI and tack on a different angle to the wild ride that we're on and For me just from even the user experience Anytime you have the like a potty said the the aha moments like the magic like seeing the reasoning the chain of thought It's like there's something really just fundamentally beautiful about that It's putting a mirror to ourselves and seeing like oh shit It is solving intelligence as the cliche like goal of these companies is and you get to understand To what why we humans are special the intelligence within us is special and For now also why we're special in terms of we seem to be conscious and the add systems for now Aren't we get to solve we get to explore that mystery?

5:11:14So that's, it's just really cool to get to explore these questions that I don't think I Would have never imagined would be even possible Back when so just watching with excitement deep blue be cusp off Like I wouldn't have ever thought this kind of AI would be possible in my lifetime It's like this is really feels like AI. Yeah, it's incredible. I started with AI of learning to fly a Silio quadruder. It's like learn to fly and it just like it learned to fly up It would hit the ceiling and stop and catch it. It's like okay That's like really stupid. You can hear it. What's going on now? And now you could probably with natural language tell it to learn to fly and it's going to generate the control Algorithm and it requires into that All the way there's low level blockers like we had to do some weird stuff for that But you can do that to our robotics conversation Yeah, when you have to interact an actual physical world is hard what gives you hope about the future a human civilization.

5:12:12Looking into the next 10 years, 100 years, 1000 years, how long do you think we'll make it? You think we got 1000 years? Humans will definitely be around in 1000 years. I think there's ways that very bad things could happen. There'll be way too many humans, but humans are very good at surviving. There's been a lot of things that that is true. I don't think they're necessarily we're good at long -term credit assignment of risk. But when the risk becomes immediate, we tend to figure things out. And for that reason, there's physical constraints to things like AGI, hyper recursive improvement to kill us all type stuff.

5:12:54I'm for the physical reasons and for how humans have figured things out before, I'm not too worried about it. AI takeover. There are other international things that are worrying, but there's just fundamental human goodness and trying to amplify that. I think we're on a tenuous time. And I mean, if you look at humanity as a whole, there's been times where things go backwards. There's times when things don't happen at all. And we're on a, what should be very positive trajectory right now. Yeah, there seems to be progress, but just like with power, there's like spikes of human suffering. and we want to try to minimize the amount of spikes.

5:13:32Generally, humanity is going to suffer a lot less, right? I'm very optimistic about that. I do worry of like techno fascism type stuff arising as AI becomes more and more prevalent and powerful and those who control it can do more and more. Maybe it doesn't kill us all, but at some point every very powerful human is going to want to brain computer interface so that they can interact with the AGI, and all of its advantages in many more way and merge its mind with, you know, sort of like, and its capabilities, or that person's capabilities can leverage those much better than anyone else and therefore be, you know, won't be one person rule them all, but it will be, you know, the thing I worry about is it'll be like few people, you know, hundreds, thousands, tens of thousands, maybe millions of people rule whoever's laughed, right?

5:14:20And the economy around it, right? And I think it'll, that's like the thing that's probably more worrisome is like, human machine amalgamations, this enables an individual human to have more impact on the world and that impact can be both positive and negative, right? Generally humans have positive impacts on the world at least societally, but it's possible for individual humans to have such negative impacts and AGI at least as I think the labs define it which is not a runaway sentient thing but rather just something that can do a lot of tasks really efficiently, um, amplifies the capabilities of someone causing extreme damage.

5:14:55Uh, but for the most part, I think it'll be used for, you know, profit seeking motives, which will then reduce, which will increase the abundance and supply of things in there for reduced suffering, right? Yeah. That's the goal. Scrolling and a, a timeline. Just throwing the stasis. That is holding scrolling all the status quo of the world. That is a positive Outcomes like if I have food tubes and lunged out and I'm happy that's a positive outcome Well expanding out into the cosmos Well, this is a fun time to be alive and Thank you for pushing the forefront always possible in humans and thank you for talking to this is fun Thanks for having thanks for having us Thanks for listening to this conversation with Dylan Patel and Nathan Lambert support this podcast.

5:15:44Please check out our sponsors in the description. And now, let me leave you some words from Richard Feynman. For a successful technology, reality must take precedence over public relations. For nature cannot be fooled. Thank you for listening. And hope to see you next time.