Own or Be Owned: Why Every Company Needs Its Own AI Model (Yash Patil, Co-Founder & CEO of Applied Compute)

29:54For folks that I'm sure most of the audience or a lot of the audience will be really intimately acquainted with this, but for folks that maybe aren't, can you explain a little bit about why post-training matters and how it sort of led to this work you're doing with Applied at the moment? Totally, yeah. So kind of a high-level overview of training models in general is there's a couple phases of training. A lot of people talk about the big, expensive, high-capex phase, which is pre-training, which is let's go and take internet scale data, use this architecture called the transformer to sort of learn language patterns.

30:32And out of that falls intelligence, right? So you get this model that can do next token prediction. And that's kind of like how we got the first sort of GPT models where they just got really good at predicting the next token. And it felt like these models could actually reason and answer questions and things like that. After you do pre-training. You basically have a unaligned model that will just predict the next token until you tell it to stop. To make it useful, that's when you move to things in the post-training realm. So there's a couple of different types of post-training. When we were first training these chat models, where it's like a user sends a message and an assistant sends a message, there's a variety of techniques of supervised fine-tuning, as well as reinforcement learning with human feedback to sort of steer the models to give responses that are sort of coherent, you know, in the format and tone that a human would like and whatnot.

31:31But you're sort of just basically doing behavior cloning on the supervised fine-tuning side. So you're saying, here's a bunch of examples for good answer. If someone inputs this prompt, these are examples of what good completions look like. And then you just train the model to reduce the loss on those like target tokens. Now, what people have been talking about when they say post-training today is something quite different than supervised fine-tuning or reinforcement learning with human feedback. It's this idea of RLVR. So reinforcement learning with verifiable rewards. It turns out if you take a really small, like a relatively small amount of high quality data as compared to like when you were doing pre-training or SFD or something like that, and you actually have a model try to attempt the same problem a bunch of different times and you have a verifiable way of checking if the model gets the answer right, you can use this new reinforcement learning algorithm to sort of show the model, hey, these are examples of what good looks like.

32:36These are examples of what bad looks like. Yes. And sort of nudge the weights to do more of the good reasoning versus bad reasoning. This thing that people talk about when they talk about chain of thought, So all these tokens before the model outputs an answer is actually entirely an emergent property. Turns out like models learn to steer themselves, learn to think, and that's how they get more accurate and get, you know, basically like hill climb any sort of task that you give it. So what the first domain that this started with was math. part of it is because first of all it's like got a really a lot of really good properties which it's like easily verifiable you don't need that many complex tools or things like that to be able to do math and the second thing is just like everyone on the research team at open ai was like competitive math people or things like that so it was like a domain that they deeply understood and also were very excited about and there's also a lot of other physical philosophical reasons about like why math is kind of the root of a lot of other things.

33:41So basically, the first experiments were actually hill climbing on a variety of different math problems ranging from basic arithmetic to harder proofs, algebra, calculus proofs, whatnot. And it turns out that the model is able to learn on easier problems and internalize that and be able to unlock harder and harder problems. So that's this hill climbing that you've seen. And there was kind of like a couple of scaling laws that came out. So obviously you have the pre-training scaling laws. Now you have this like post-training scaling law. And now you have this test time compute scaling law, which is like if you actually allow the models to reason for longer when you're using them in production, they give you better answers.

34:26So that's this like reason when you pick a reasoning level in Chachipati or something like that. That's it. Just reasoning for longer, expending more test time compute before giving a final answer. This episode is brought to you by Persona, the B2B identity platform helping businesses verify users, fight fraud, and build trust. Fraudsters are already using AI to spoof faces, voices, and documents, so your defenses need to adapt just as fast. Persona helps secure some of the Internet's largest and most trusted platforms with identity verification. If you're building a product where trust matters, identity should be a priority.

35:01You've probably already experienced Persona without realizing it. verifying your LinkedIn profile, signing up for Etsy, or renting a scooter with Lime. Trusted by leading companies like Square, Brex, and Twilio, Persona gives you the building blocks to create identity flows that adapt to your customers, risk tolerance, and locales you operate in, whether you're verifying age, onboarding businesses, or automating KYC. It's fully configurable, so you can launch in days, not quarters. Want to see for yourself? Generalist listeners get a free year of the starter plan. Head to withpersona.com slash generalist and check it out.

35:39You're doing this work at OpenAI. What was the moment that you realized there's a need for the sort of service that applied compute offers where you are sort of going into these companies and doing this more hands-on work? So I think Microsoft AI, when they put out their thinking models recently, you know, the title of their paper is like the hill climbing machine or something like that. I think the way of thinking about RL is it really is that it's like if you define the hill to climb well the algorithm works so the way we were training these models was actually just like on the post-training side was hey let's go and make a bunch of high quality data sets that target the capabilities that we want the models to get really good at like basically we'd start with an eval and be like okay if this is the what the eval looks like let's just like go make training data that looks like the eval.

36:29Yes. And that was the way that we were approving these models. So the natural conclusion was kind of like, hey, you know, everyone is going to have their own evals and the things that they want these models to be really good at. And evals extend beyond just, you know, the intelligence of the model. It's like, oh, it needs to, you know, be this intelligent in this cost and latency regime or this modality or something like that. So if everyone's going to have their own evals, everyone should also be able to train their own models that do well on those evals right so so our sort of natural conclusion was like hey maybe since we have this hill climbing machine the best way to sort of manifest it in the world obviously the frontier models are amazing you know these will be like workhorse models for a lot of things yes but if people want to be very specific about the task that they're trying to optimize for they should have the ability to go and push drain their own models on their own data.

37:26Yeah. Evals are, you know, the scorecard for all of these models. And that's so general. It's the new PRD. Yeah. Yeah, exactly. And the reality is that every organization has a different scorecard for what they consider valuable. And that's, you know, sort of what you're saying is like, you know, if you have a different set of things that you're optimizing for, it starts to make sense that you have a machine that is made to hit those optimizations. Exactly. And if we go back to the Goldman example, right? Like if a lot of those evals are going to say private, people are going to want to own their intelligence, own the post-training capability to go and create those models that are good at those evals.

38:02Right. Yeah. So I think that was the primary motivator. And then I think the main sort of driver for why people are thinking about this right now is cost. The token apocalypse has really shown, hey, you need to use the right model for the right task, even beyond just like optimizing for evals. rate. Yeah, that's I'm curious, like how you think about it today. Clearly, that's top of mind. I think in the past you've talked about how general models sort of set the floor and, you know, specialized models can sort of raise the ceiling in some sense. Yeah. In some ways, it feels like, you know, the distance between those two might be getting smaller, but the cost at which you are accessing these different offerings is like, has it become more of a cost story in your head?

38:46cost has become like a a the main driver i think right now but i think in principle the argument still stands which like take kirkland and alice right they caused waves in the twitter community and i'm sure broadly as well about how they're going to spend like half a billion dollars to i forgot what the exact quote was but the essentially the idea was to create ai tools that their competitors can't also use, right? So I think the idea is, and this extends beyond just the model, to be clear, right? The model is part of the moat, but also training that model inside of your harness with your context, that's actually how you're going to get differentiation.

39:29So I think a lot of people are thinking about, hey, if I'm using the same stuff as my competitors, how am I different? And that's also another driver for why they should train their own models on their own data. And I think it's been just almost a little over a year, right? And you've gone from starting the company to a last valuation, 1.3 billion. Been a good year. Been a good year, yeah. Can you tell me a little bit of how you've started to work with folks and what that process looks like with these companies? Totally, yeah. So we work with a variety of companies across AI natives, digital natives, and large enterprises, and hyperscalers.

40:08So basically, if you look at how money is spent on these models right there's some amount spent on like training and it's like a non-trivial amount but like most of the money and most of the usage is on on inference right so when we were thinking about what the future of open models looks like or models that people are going to inference a lot for different things our conclusion was kind of if you are going to train or if you are going to use open weight models and specialize that model you're going to do some sort of training on top of it. So the actual wedge into inference is getting really good at training.

40:47So what we do with our customers is we have a post-training platform where they can integrate all of their data. And we have a team of applied researchers that actually deeply embed with the customer, work with their ML and research teams to train these custom models. And we really want to make sure that the economics are in their favor. So we try to have most of the stuff that they pay for come on the inference side, right? So it's like, once we have a really good model and you guys want to use it, then we can go and host it for you. Yes. The sort of forward deployed engineer model is something that I think people obviously associate with Palantir as a successful example of it.

41:29With Palantir, I think the way they've made it work is that clearly they found a way to get good leverage on it such that the 50th version of it of them taking on a project is very different than the fifth. What's the version of that for you that stops it from being sort of just pure consulting leverage? There is always some component of this that is going to be unique per company. That's kind of the whole thesis, right? It's like each company is going to have different data and want a different model, But really, it's the infrastructure that is repeatable across all of these companies, right?

42:05It's like the scaled infrastructure for training these models, the scaled infrastructure for serving them, and then the sort of techniques for capturing production usage, turning that into more data that we can go and continually train. So our equivalent of Foundry, which is Palantir's platform that they use to give to all their FDs to go and move faster. Yes. We have our own post-training platform where all of our applied researchers and researchers from the companies we work with can both collaborate to train different models. And really, I think the way of thinking about it is it's the infrastructure to make training performant and easy.

42:45Mm hmm. Yeah. Does it end up becoming more of an ongoing engagement than, you know, people might realize from the outside? I could imagine that, you know, one version of it would people would think, oh, you go in, you're sort of doing an AI transformation and then, you know, you're out the door. But I would imagine like the demand for, you know, improving this, deepening it is like leads to quite long engagements. Totally. Yeah. I mean, so I think there's always a desire to continually improve these things, which is why, you know, when we deploy a model in production, a lot of our product is built around, hey, how do we take the usage of this model and turn it into more data that we can continue to continue to train on or build context around the model so that, you know, it's learning actually the more you use it.

43:34So, yeah, I think the idea is that the engagements don't end. Obviously, like the way we generate revenues by serving the models and we need to make sure that those models are always performant. And maybe even to get even more sort of tangible, like can you maybe talk us through an example or two of like what taking this more specialized contextual approach unlocks for a customer, for instance? Yeah. So I think one of the case studies we have on our website is this engagement we did with DoorDash, which is, you know, they have more than 100 ,000 merchants that come to their platform every year and they want to create a new DoorDash storefront.

44:16Right. So when they when they do this, they come with a bunch of unstructured information about their business, you know, what what food they serve, things like that. And one of the things is like a picture of their their menu. So it turns out this was actually an example of where we were able to exceed the frontier in terms of quality because we trained a small specialized model that was able to take pictures of menus and turn them into high-fidelity DoorDash storefronts. Basically, the process was create an eval for what good looks like, create the training data for us to go and do RL training on top of, train the model, and then deploy that and host that.

44:58The trend feels like, you know, if applied compute continues to grow and this is part of sort of a broader movement is like sort of trending towards more of these small models where you do sort of say, here's a set of tasks that like are maybe a bit more narrow, a bit more deterministic that like we can actually very inexpensively do this very effectively. Is that roughly how you see things going? Yeah. And so I think the motto is better, cheaper, faster. what people are starting to realize is that these open weight models provide a ton of flexibility around ownership and they're actually good.

45:37Like you can actually go and use them to perform optimally on your task. So I think you're going to see just a lot of people for, you know, reasons related to cost. I think cost is the big driver right now, but there's just so many other downstream benefits you get by owning your own model. When you sort of move towards the world of you know, a company having multiple of its own models for different tasks, does that create like some sort of maintenance debt for them that they have to sort of stay on top of? And yeah, I mean, there's probably still cases where that's very much worth it for them.

46:13But is that something you've noticed come up? Yeah, I mean, so this is part of like the continuous improvement part is there's obviously like model drift and you know, the distribution of tasks might be changing slightly, but over a long period of time, that really, really matters. Our view is just like training and inference loops. They're like this right now, but they're starting to move closer and closer together. Right now, the form is like people are just getting a lot faster at making new data and training the model again. But eventually, these things are going to be quite intertwined. I think there are some cool examples of this that are kind of like maybe the first sort of beginnings of it.

46:55Like if you saw Cursors, Composers, Real Time RL work, I mean, that was them doing like online training by like taking a bunch of user trajectories, you know, getting, extracting the implicit reward, taking a training step and then deploying a new model in production. I think like that is just kind of an indication of where things are going to go. It's like you're actually going to be doing more and more continuous training because as your model is being used. It's kind of like an exploration technique, right? Yeah. And so what you're describing there is almost like a continuous learning model where it's like every time I happen to do this task, it's actually training this model to be better the next time around.

47:38Totally. And the name of the game is learning from sparse rewards, right? So if you think about the different phases of training, pre-training is like super data inefficient, right? You need to like learn the whole, or you need to train on like the whole internet in order to like learn the language. Then we had this SFT stuff, which was kind of behavior cloning. So you say, hey, here. I'm not sure about that. Yeah, yeah. So SFT is described as behavior cloning because you're giving it examples of what the correct output should look like. And so you give it a bunch of examples. And then for a new task or a new prompt, the model is sort of interpolating, like, oh, I should say these things.

48:16And then we kind of move to RL, which is a lot more data efficient, which is I'm going going to take one data point, one math problem, generate a thousand trajectories, all with different thinking tokens or whatnot, all with different rewards. And then I'm going to do backprop on those. And there the model will kind of learn what good reasoning looks like and bad reasoning looks like. The ultimate thing, right, is being able to do super, super sparse learning, which is like you go and roll out a trajectory, you have some indication of whether it was good or bad or something, and you're able to directly learn from that.

48:53That is difficult, and that is the problem to solve in continual learning. On the open weights models, you were mentioning earlier that you think maybe the gap between open weight and closed models has sort of narrowed over the past year. In some ways, if I look at the hardest benchmarks, I almost feel like it's widened, but actually what we're learning is that just beyond a certain threshold, it almost doesn't matter. Yeah. Like, I don't know. Do you think that's wrong? Yeah, yeah. So I think what we were talking about before is Frontier models are amazing and you should still use them for your hardest tasks.

49:29But if you have like$100 of spend, maybe 20 of those dollars go to the Frontier models and the$80 can be actually$20 that go to better, cheaper, faster models that you train on this task, right? So I think like like the frontier models are obviously very, very, very good. But, you know, you don't need to use them for everything. There's all these memes on Twitter, right? Like where like people are using some big sword to cut like a blowtorch. Exactly. So I think like, you know, there's a lot of truth to that. And that's what you know, you're talking about model routing and things like that. I think there's going to be pretty interesting innovations there.

50:07We're working on some of that as well. The other thing I think about with open wage models is like China has really dominated that so far. Yeah. You know, it feels like America is is clearly leading with some of the frontier models. But yeah, it's like how do you think about the fact that an implication of this could be that a lot of American companies become like actually quite dependent on Chinese models that the rules could get changed up on them? of those could go closed at some point. Is there a robust enough open weight system right now? So I think that you're right. It all boils down to incentives.

50:46So I'm extremely excited that companies like NVIDIA are investing a ton in open source models. And NVIDIA has a real reason to do it, right? The dream for NVIDIA is that every company has their own model and is running it on Nvidia chips. So I think that there are real, there are companies like Nvidia, Nvidia being the big one, that have true incentives to keep models open. And I think there is enough of investment and a lot of smart people working on open models that I think it's going to be a pretty safe and big bet over a period of time. One of the things that I think is really interesting about applied compute sort of outside of the technology piece is that you seem to have managed to attract a huge amount of ex-founders, which is always a very bullish sign, I think, for an investor, you know, to see that happen.

51:43How have you managed to do that? It's something like two thirds, right? Yeah. So it's almost part of like the business model is ever because we're doing this like sort of forward deployed approach where we're embedding deeply with with customers. it requires people who are extremely high agency and like ambitious and move fast. So I think it just naturally like the kind of work just naturally attracts these these types of folks. And then, yeah, I think I think like what we we in our culture, we sort of decided, hey, we want to hire ex-founders who are who fit this archetype. And they just themselves attract other like minded folks.

52:26so yeah I think it hasn't been anything really special like we're not trying like recruiting from some very specific pipeline or something like that but it's just this kind of natural snowball effect where you know the type of work that we do plus the initial team that we got together just attracts more people of the same mentality at a higher level like what are the things that you think are maybe particularly distinctive about applied computes culture Sure. Like, what do you care about that, you know, other other startups, you know, or other places you've worked maybe wouldn't care about as much?

52:59So there's a lot of this like nine, nine, six stuff. Yes. We're not a shop like that. Like we and I deeply respect people who who choose to operate like that. I think the type of culture that we want to embody is people are just very excited about their work. and we have no requirements on how long people work or things like that, but it's a very bullish sign when people are so excited about what they're doing that they want to come in on weekends and sort of spend a bunch of time in the office. So yeah, I think the things that we prioritize or what we look for when people are joining the team is like being deeply mission aligned.

53:40So, you know, we want people who want to come and work on hard problems, but also really believe that, you know, something like this should exist in the world and that this is actually very important for basically every company out there. And then take a lot of pride in working on hard problems that don't have like the thing that's very unique about our company. Right. And it's actually somewhat related to a Palantir. I remember having this unique moment where I was like, Palantir is one of the few companies that signs massive contracts without knowing if they can solve the problem. Right. So interesting.

54:17That's true. Right. Like, you know, they'll sign like huge, you know, a hundred million dollar deals where this is like before they've started working out the problem. Right. So. So I think there are there there's like a unique set of companies that that do that. And and obviously, like, I think we fit into that category where there's open research questions about like what we're trying to go and get really good at and solve. And people who like to work on things that are uncertain and there's a lot of greenfield research to be done are just naturally attracted to companies like AC. When you talk about, you know, wanting people who want to do the hard thing, like what's been the hardest, thorniest technical piece of this so far for you?

54:56Yeah. So I think like one of the really cool places of work in the company right now is basically building the best ML infrastructure for training open width models. So, you know, We're obviously in this compute crunch. The more performantly you can train models and the more accurate they can be, that's true differentiation. There's actual work. It's not like RL is completely figured out and it's kind of like a commodity thing. There's real research happening at the ML infralayer. So we've been hiring a bunch of these really, really smart researchers that are actually able to go deep into the stack, work very close to the hardware to actually make training more performant.

55:39And that's one of the ways we're differentiating amongst like other things that are out there. Can you, you know, maybe part of this is too secret sauce, but like, can you talk a little bit about what that training stack is, like what that looks like? Yeah. So, so basically the, you know, the, we, we procure our own GPUs and then we, when you're actually going and doing training, right, the way training works is you're alternating between like doing inference. So, So you have some checkpoint, some weight checkpoint that's like policy step zero or something. You go and try a bunch of attempts at a problem, right?

56:13And that problem is actually in some sort of replayable environment, right? So say we're training a coding model. You're like spinning up a sandbox that has a code base on it and potentially even more complex than that. Like you have, you know, services that are running that the model need to interact with, yada, yada, yada. All needs to be replayable though, right? because you're repeating this like a thousand times. Yes. You go and do the inference on that, get a bunch of model attempts, assign reward to them based off of like some verifier and then use some sort of objective to be able to go and take the next step on the model and get your policy step one.

56:50And so one of the really interesting problems is like, okay, you're doing this sort of alternating of sample, train, sample, train, sample, train, sample, train. Yes. it's like always slower to do things in sequentially right so what do you actually want to do is you want to do things asynchronously where you can actually be sampling on your next set of problems while training is happening oh wow and you can actually there's like really cool science questions around this of like how far can you push that how asynchronous can you be how off policy can you be can i be sampling problems using policy you know step one or something or sorry policy some some earlier policy while i'm training like the last policy so this is where a bunch of the ml dynamics come in or wait how do you make these off policy corrections where you're trading off performance and speed and throughput versus like ml dynamics and and stuff like that so these are the kinds of like infra problems that a lot of our team is working on it's like how do we make this stuff really performant and and it really matters for longer and longer horizon tasks, right?

57:57Like, you know, if you're running your model for like an hour, your total training runs can be super, super long if you're doing things synchronously, right? Train inference, train inference. You really do want to do things super asynchronous in order to like cut down on your total GPU time. You're sort of optimizing around your layering GPU time because that's like the most constrained. And so this is the other cool thing, right? It's like, so there's that side of the house, but then there's also environment creation, right? Why do people care so much about performance sandboxes and things like that?

58:29The reason is because if you're waiting on CPU workloads, if you're idling your GPUs for CPU workloads, you're probably wasting a lot of money. So there's like a ton of performance engineering on the environment side of the house to make training faster as well. You know, as you were talking about the sort of imperative for every company to own their own intelligence. It sort of occurred to me, maybe this is a very naive thought, like, isn't the sort of logical extension of that, that at some point it'll also trickle down to the human level and all of us will need to have our version of that. Like this, you know, does AC in five years, you know, start to become, you don't need a forward deploy engineer, but you have a version of it that's like more self-serve and, you know, each of us has our own model, our own set of models.

59:17Yeah, I mean, so far we've been pretty enterprise facing, but maybe this is market number two. Excellent. No, I think you're right. I think a lot of people have been very excited about this idea of personal AI. And I think maybe a bit of a jump to a broader point, but if you think about personal AI, that's something that deeply understands you, how you work, and maybe running all the time. right so i think um maybe a broader point is just like as these things start to trickle down the demand for compute and inference is gonna like balloon like crazy you know that that that's something that's like kind of like a broader maybe i actually don't even know if the market is fully pricing in how much compute that we'll need yeah all of this stuff like if everyone has their own personal models and that are constantly being trained and inference and things like that like we just need a ton of compute which i don't know it seems likely that will happen to me maybe that's yeah maybe that's too maybe i'm no no i think i think people totally want it to happen the problem is i mean i think they're just like right they're like physical limits right like uh to how you know how much the supply chain can accommodate and like not even like on the chip like you know there There are ASML machines, like lithography machines that take a really long time to build.

1:00:37So that's why I think we're in for... This kind of ties back to tokenpocalypse. We're going to be in a compute crunch and inference demand is going to crank like crazy. The ability to train Pareto optimal cost performance models is actually going to matter a lot. And I think we're seeing the early beginnings of that. And it's just going to get worse until people really start to invest in this. what's your sort of uh you know not ai 2027 but your view on the next couple years like what do you think yeah the world is looking like from your perch you know i know there's a lot of folks who are like oh 50 of jobs are going to go away in the next two years and things like i think like personally i i think that is is not going to happen um and it it probably is inspiring a lot more fear that maybe doesn't is maybe misplaced yeah i and i and i think it's coming from a good place, which is just like we need to be prepared for this sort of stuff.

1:01:34But kind of being on the front lines, like doing this for deployed work and whatnot, I think just the diffusion of AI into the real world take time. There's a lot of change management that needs to happen. Companies are like large institutions that take time to adopt new things. And we were talking about this before, right? Like OpenAI is this, you know, it's OpenAI, the research and deployment company. And, you know, deployment is so hard that they created a company called Deployco, the OpenAI Deployco, right? And same with Anthropic and stuff like that. So I think like people know that it is going to take some time to go and deploy this stuff.

1:02:16So I don't think overnight everyone's going to lose their jobs. There will be like areas of the economy that I think will experience more change than others just due to the nature of the job. But to me, this is like a multi-decade long rollout and diffusion of AI into the real world. Wow. What has surprised you most in terms of that change management, maybe the slowness of some of these enterprises? What has maybe you couldn't have counted on? I mean, in some ways we knew this would happen, but just seeing it up close makes a lot of sense. There are just established procedures inside of companies, and one of the difficult things is like you know to change those procedures you often have to touch multiple layers of management and and the actual people who are doing the work and it is just a difficult coordination problem right and and i think the other thing that is also obvious but we've just you know michael chan on our on our go-to-market team put out a really good like field notes guide to what forward deployed engineering has been like is like no one is data ready.

1:03:19Absolutely no one is data ready. Things are like, obviously, always very messy and whatnot. But if data is kind of everything, even just the state of where it is, and even the best companies is it's not where it needs to be to actually go and make any sort of deployment extremely smooth right now. So I think there are just these natural sort of roadblocks that will slow things down when you're trying to roll out AI and like big organizations. Well, I always like to end with a couple sort of thought experiment questions. If you were given the chance to conduct any experiment with no operational constraints and no budget constraints, what's an experiment you'd like to run?

1:04:01Oh, this is a fun one we talked about in the office, which was if we all reverted to cavemen and like lost all modern technology, how long would it take us to reproduce the iPhone? So knowing what we know now, how long would it take us to get all the infrastructure in place? Like no fabs, none of this stuff. Man, that's really - Just sand. I used to think about that for like a day at least. Yeah. I don't know. What did you come up with? Well, so, and I guess the darker side of this is you have to factor in everything. So like people going to war against each other and things like that. Yeah, yeah, true.

1:04:46If we were just lost all modern technology. 100%. I conclude that we just never would. You think we would never have anything? Yeah. Like, first of all, most of the population, this is getting a little weird, but I think most of the population would not just like know how to survive the first like 60 to 90 days. And then, you know, we'd have to, there's the ability to go and build all this infrastructure takes a ton of time. If you sort of are thrown into chaos and dystopia, people are obviously going to get mad at each other and create factions and fight. So I think we'd spend most of our time fighting them than actually building a thing.

1:05:23Wow, I'm going to be thinking about that for a while. That's a good one. Okay, sort of last question then, I suppose. If you could assign a book to everyone on Earth and you knew they would read it and they'd understand it, what would you want to give to people? I really like this book called The Design of Everyday Things. It's actually one of the first tech books that my brother gave me. And I think it really set a good framework for how to think about solving problems in non-complex ways. And here's like a crazy segue for you, right? I think in today's age, when you have AI tools, particularly coding, you can actually, by going faster, you can actually be going a lot slower.

1:06:07So like taking the time to, like the one thing I'm worried about people losing is this ability to actually deeply think about how things should be architected and designed. Yes. because if we rely too much on AI tools, and I think every company has seen this in some capacity with this AI vibe coding or whatnot. We're all for AI coding. My main thing that I worked on at OpenAI was codecs, but I think there is truly some atrophying happening by relying too much on AI tools. So back to the basics, back to the roots. Yeah, I mean, things like judgment, taste, all of that stuff is very easy to stop actually building when you have these tools in front of you, right?

1:06:49Totally. I think the best engineers right now are the ones who are people who learned how to code before AI because they actually deeply understand systems thinking and things like that. And that would be my worry as a new grad engineer or something like that is getting too reliant on AI tools. Because at the end of the day, I think humans will be the overlords over all of these AI tools and you will still need to do, you know, critical thinking to know what you want to build. They might help you build it, but you need to know what you want to build. Well, I think that's a that's a nice optimistic place to end.

1:07:26Thank you so much. I really enjoyed it. Oh, thank you. This was awesome. Yeah. Thanks for having me. That's it. Thank you for listening to this episode of The Generalist Podcast. Please subscribe on Apple Podcasts, Spotify, or your preferred podcast app. Ratings and reviews help others discover these discussions, so if you enjoyed the conversation, I'd be grateful if you could take a moment to leave one. For all past episodes and more, visit us at thegeneralist.substack.com. See you next time as we continue to explore the future.