OpenAI Going Public, the China–US AI Race, and How AI Is Reshaping the S&P 500 and Jobs w/ Salim Ismail, Dave Blundin & Alex Wissner-Gross | EP #205

27:10It peaked at 12 ,000 views. And then there was a fake live stream that peaked at 95 ,000 views. Let's take a look at the fake live stream. Cutting edge hardware with decentralized finance. It's about proving that crypto works reliably, globally, and for everyone. A couple of things to keep in mind. only use the QR code you see right here on the GTC broadcast. Don't trust any links floating around online. They're not us. I love that. The fake broadcasting, don't trust anything else. That's hilarious. The numbers are pretty staggering. It's$1.5 billion lost globally for deepfake-related fraud since 2019.

27:56And only 24.5%, according to the numbers here, people can actually spot deepfake quality. And AI detectors fail up to 50 % of the time. So this is going to be a thing. This is going to be a thing. I think reality may have just lost the algorithm more. Yeah, I think in this case, if you look closely at the video, the lip syncing was poor. I think in the short term, detecting counterfeit - All right, you're part of the 24 % that can notice this. Okay, Alex, you win on that. Definitely notice the poor lip syncing. I think in the short term, detecting counterfeit live streams in real time doesn't seem like a terribly deep technical challenge.

28:39In the long term, we're going to have more solutions like ubiquitous watermarking, perhaps, cryptographic guarantees of reality, perhaps. I don't think in the long term, this is a deal killer for us drowning in AI-generated slop counterfeit live streams. I think this is very tractable. Yeah, within the US, I agree. I think if you look globally, it's a little more of a concern. There's a real possibility that regimes lock themselves in. Control of media content is going to be so easy with AI assistance. And then convincing your population of virtually anything gets trivially cheap and easy. So I would be more concerned about, you know, in some nation where they're not as aware of AI watermarking or, you know, whatever.

29:25They're seeing things and just believing them. Well, I would say our nation isn't aware of that. And Alex, you made a point, you know, when we were discussing a year ago that AI-generated speech is far more compelling than human speech. All right. Let's move on to the AI wars. This is XAI versus OpenAI versus Google. What happened to Anthropoc? We're just writing off Anthropic just like that? Well, no. I mean, they're in there, but these are our major players today. We'll talk a little bit about Anthropic. Oh, Dario would really object to that. Well, listen, and I love Anthropic. Don't expect to get him on the pod now.

29:59No, I love Anthropic, and I want him on the pod for sure. All right, but let's jump in here. So I'm going to rant on this one. So XAI launches Grokipedia. I had a friend of mine, remember Justine from Singular University? Celine? Sure. Yeah. So she sends me this text. She goes, because she heard our pod, I was arguing or lamenting Wikipedia's inability to correct all the wrong things in my... And I actually hired consultants to fix Wikipedia for me because I'd make the changes and they'd be changed back. It's like, this is ridiculous. So she says, hey, Grokipedia is out with your Grokipedia entry.

30:41What do you think about it? And I look at it and it's amazing, right? It covers everything in detail, super well referenced. So Grokipedia is being written by Grok. It's writing, updating, fact-checking it on a real-time basis. And they have 900 ,000 articles compared to Wikipedia's 8 million articles. My particular entry here was 8 ,500 words on Grokipedia versus 4 ,800 on Wikipedia. But it was so well organized. And I just absolutely loved it. Any comments? I'll throw in a comment. I want to reason by analogy. There's a process that those not steeped perhaps in semiconductor manufacturing may not be familiar with.

31:29It's called zone melting. And it's a process for purifying, not knowledge in this case, but purifying semiconductors. And the idea is you take a rod and you pass it through a heater. And because there are more ways for impurities to exist in the solid state, in the melted state, rather than in the solid state, the impurities migrate out of the solid into the liquid state. And you do this over and over again. You get a pure and pure semiconductor. We don't have a science right now for knowledge purification. But one could imagine somehow near future, we have a science for it and we decide there are more ways for correct knowledge to be self-consistent than incorrect knowledge.

32:10And I think we're starting to see the beginnings of almost a knowledge equivalent of zone melting, where you take the raw slop, human slop of the Internet. You pass it through multiple passes of AI gen synthesis, creating what aspirationally would be more correct versions of the ground truth. Do this over and over again, and maybe aspirationally, because there are more ways for the truth to be self-consistent than whatever the starting knowledge was. Maybe we arrive at some sort of ground truth through this result. That's really cool. But it would be interesting if that ended up being the case.

32:49That's Elon's objective, right? Basically trying to, from first principles, derive truth. I put a quote down here. He says, from Elon, says, a step towards XAI's goal of understanding the universe, which I love. I have two comments. Please. One is, you know, Peter, you know, I've talked for a long time and written in the book that staff on demand and a community doing work is essentially a proxy for AI. Yes. Right? Driving is a great example. But now we see it actually applied. If you take a Wikipedia article, for a human editor to go through and track down all the links and ratify everything, it's just a pain in the ass, and it's not the strength of a human being.

33:30Whereas an AI can do this without even blinking. I think that now propagates to a level where Wikipedia, with an AI interpretation, per Alex's metaphor, which I think is absolutely fantastic, now gives us the ability to have closer and closer to pure truth. I never quite understood what Elon was talking about when he says maximizing or truth-seeking maximizer. But now I'm starting to get a sense of it, and it's absolutely brilliant. It's fantastic if we can get it there because it can cross-reference all the stories and cross-check things in a way that no human being will take the time to do, and it'll do it much more accurately.

34:05What Alex just described is really, really similar to the original Google PageRank algorithm, where it's just starting from nothing, iterating. That's a great point, too. a reference link and a site and assigning credibility back and forth in this self-annealing process, a simulated annealing process. And it worked. I mean, they don't need it anymore because they have so much data flowing in. But when they were just a little startup bootstrapping, it worked really well. So Alex's idea actually might work. This is also my comment. I use PageRank as an example when people ask what is AI, etc. And I say, look at PageRank.

34:36It's evolving a completely separate type of intelligence for crawling billions of pages making sense of it. That's very orthogonal the way human intelligence works, not replicative. So I think AI tends to have this totally different type of intelligence of mass crunching data and finding a signal from noise in a way that we're not designed to do at all. Well, just thank you to Justina Zander for pointing this out to me. I'll also mention real quick, I checked and it turns out Wikipedia has a budget of$170 million a year. About$100 million of that is labor, paying everybody to do this work. Some of it's voluntary.

35:13A lot of it is not. All right, let's move on. Just also just a shout out to Jimmy Wales for creating Wikipedia and managing for all these years. I mean, what an unbelievable gift to humanity. I mean, we were watching the transition from the Encyclopedia Brantonica to Wikipedia to Grokipedia. But Grokipedia was low-hanging fruit. Any of the AI companies could have taken this on. And I think it's going to become pervasive. I know I'm standing up a new website for diamandis.com. And the very first thing I'm putting at the top is my Grokipedia link, right? You want to go deep? There it is. So true to AWG's vision, this is an important conversation.

35:56And Alex, I actually read this paper. So this is a new AGI benchmark, which gives chat GPT a 57%. But I would prefer if you explained it because it's pretty amazing. And Salim, this one's for you, buddy. Finally, we have a definition of what AGI is and how to measure it for the first time ever. This was a paper that was co-authored by Eric Schmidt as well. And it's a pretty powerful concept. Alex, would you take us through it? Sure. So there's an enormous cottage industry of AI researchers trying to define what intelligence even is. I've been guilty of that in past years as well. My bias has always been to look for a universal, elegant definition of what intelligence is that isn't necessarily grounded in human behavior or human psychology.

36:46This paper, and as Peter, you mentioned, we know a number of the co-authors on this paper, that the basic idea behind this paper is to do the exact opposite. Instead of trying to look for some human agnostic definition of intelligence so that we can build more of it, The idea is instead to look at human psychology. So there is a theory that's popular in human psychology. It's called the Cattell-Horn-Carroll, or CHC, theory that decomposes human intelligence into 10 different factors, like the ability to reason quantitatively or to do visual processing. So the idea behind this paper is to define a benchmark that's directly inspired by the CHC theory to decompose intelligence of frontier models into 10 different categories with various subtasks associated with each category.

37:36And the main upshot of benchmarking GPT-4 and GPT-5 auto, critically not GPT-5 pro, according to my reading of the paper, the main upshot is surprise, intelligence is jagged. The frontier models are stronger at some skills, weaker at other skills, whereas the archetypical human would perhaps have a more uniform distribution of their skills across these 10 categories. But I would add the important caveat, again, just based on my reading of the paper, they didn't actually benchmark the bleeding-edge frontier models like GPT-5 Pro. For those looking at this on YouTube, you'll see these 10 different categories.

38:17These are human-like skills, knowledge, reading and writing, math, reasoning, working memory, memory storage, memory retrieval, visual, auditory speed, and they're benchmarking GPT-5 and GPT-4 against those. But it's a measurable benchmark, right? I mean, the other option is the pornography definition. We'll know AGI when we see it. Dave, what do you want to add on this? I'd love to get Alex's take. I assume a 10 on each axis on this radar chart is human, so you're trying to match the outer ring here. But it's really a... It's all humans. I mean, no human is going to match 10 on this. A quote-unquote well-educated adult.

39:00Okay. Yeah, okay. Well, you know, it's going to be, you know, the difference between the best human and an average human is a rounding error in the grand scheme of AI, you know. It's almost identical, actually. But, you know, it's so asymmetrical, and I don't understand the memory and storage axis and the speed axis. That one is surprising, right? How can you say the best AI is miles behind humans in speed? I don't quite get that. That was the part that struck me also, speed tuned off. So when you make a query, you know, to do something, your AI goes off and thinks about it for a while before it comes back with an answer.

39:35That is speed. A human, if I ask a question, Alex will typically not go away for five minutes and think about it. He'll give me at least his version of an answer right off the bat. But memory I found incredibly perplexing because I thought these AIs have incredible memory. Alex, what's your take on speed and memory? Based on my read, the memory storage access corresponded, or the deficiencies thereof, corresponded roughly to the fact that off-the-shelf vanilla language models and foundation models have a finite input context window. So if you ask them questions that reference older information by default, unless there's some sort of compression or memory compactification or rag type mechanism, they don't have the ability to remember things that you told them a long time ago.

40:26But again, I want to caveat this benchmark. I love benchmarks in general, as I mentioned previously on the pod. But these are off the shelf models without any agent bureaucracies on top of them, without prompt optimization, without even access to bleeding edge reasoning efforts. This is just, if you read the paper, it's just GPT-5 auto. So I'm wary to put my finger on certain deficiencies as being in any way indicative or instructive of the limitations of AI. Meaning there are other models that would perform much better on these 10 parameters. Or light modifications of existing models that, as with RAG, retrieval, augmented generation, make them superb at certain skills.

41:12So I think where this is useful in my mind is just having yet another benchmark as a proxy. It's a start, an important start for measuring human capabilities against AI capabilities. But when I see 57%, we've talked in the past on the pod about in the style of Ray Kurzweil, the moment you've passed 10 % or maybe even less, you're basically halfway there. In getting 57 % on a general human psychological benchmark, that indicates to me that probably with a little bit of reinforcement learning, a little bit of bureaucracy, agents, framework scaffolding, you probably get to 90 % today. Amazing. Well, by the way, everybody, just some forward-looking news.

41:57Ray Kurzweil is going to be joining us on the pod next month. Talk about his predictions for 2026. Yeah, it's going to be a lot of fun. I have my standard responses to this, which is I think this is really great for approximating or getting to kind of the frontal cortex and neocortex activities. But it doesn't deal with emotional intelligence or spiritual intelligence or any of the other dimensions of intelligence that we typically attribute to human beings. But I thought about you, Salim. I thought about you specifically on this one because it's going to be a measurable benchmark that we can at least point at.

42:33And we're going to discuss whether we're going to hit AGI according to OpenAI in late 26, 27, or 30, 32. But what I'm saying is I disagree with the premise because AGI for me would incorporate these other things. Okay. So if you're measuring pure IQ test type stuff, fine. This is a great benchmark and we can kind of go on. I wonder if we're going to have the first AI spiritual leader who proclaims a religion. Oh, I think that's very doable. You know, I remember one spending time in the Himalayas with some of these gurus, right? And I sat with these gurus with the orange robes and the long beards.

43:08And I came out with the conclusion, there's about 10 or 15 questions, like what is the meaning of life type of thing. If you have a pretty good answer for those 10 or 15 questions, you can become a guru. And that's kind of an LLM. That's a neural network. All right. I'm predicting here. I don't think that's going to be hard. I think that's very doable. By this time next year, there's going to be an AI-based religion that is going to scale at a hyper-exponential. It's going to be amazing. All right, big news this week. OpenAI restructures to become a public benefit for-profit corporation and a nonprofit.

43:41So OpenAI will hold – the OpenAI Foundation will hold$130 billion stake, 26 % of the new company. and the OpenAI group is now what's called a PBE, a public benefit corporation. Salim, you and I did that move with Singularity University, converting it from a non-profit to a for-profit and spinning out a non-profit. Exactly this was a benefit corporation with a non-profit alongside. And so here's the ownership. That's an important point, actually. So a B Corp can do anything a C Corp can do. So go public, raise money, be profitable, all those things. This is a great point that you're making, Dave.

44:17for the viewers, from a taxation and legal perspective, a B Corp is exactly the same as a C Corp. Except for every other public company. The detachment is that in a C Corp, if the board is obligated towards financial optimization and can be sued if they're seen not to be doing that, whereas in a B Corp, the board is obligated towards whatever the mandate is of the B Corp and can be sued for that in theory. I love the percentage ownership here. So here we go. Microsoft owns 27 percent. The nonprofit owns 26 percent. And the remaining is owned by OpenAI PBC at 47 percent. And this restructuring is going to allow OpenAI to go out and raise money.

45:01But here's the rub. Here's the rub. Elon's lawsuit against OpenAI remains active. And his bid to try and block the restructuring was denied in court. But the case will proceed to trial in the spring of 2026 is what I read about. And the implications are interesting, right? So number one, the court could order a rescission that unwinds the open AI for-profit PBC structure and restore the nonprofit control. Number two, the key deals such as revenue sharing with Microsoft could be voided or renegotiated. negotiated. And number three, there are potential damage fees and reduced fundraising flexibility for the PBC that could result.

45:49So that's going to be interesting drama in a year from now. Yeah. If you look at the valuation of the company, the market does not believe any of those problems will actually be material. For sure. So it seems unlikely. But I think it is, you know, Elon's got a very valid point in that that whole time you're a nonprofit, you're not paying any tax. And if you're secretly building a massively profitable trillion-dollar company while avoiding taxes, that's a terrible precedent. You can't do that. And Elon even said it. Like, if that were legal, everybody would do that and start your company as a charity.

46:25So I think the courts will have to say, yeah, you can't do that. And the penalty would be like a dollar or something, just like they did with Microsoft antitrust. You're like, okay, you're guilty. You're totally guilty. You're fined a dollar. They would find a dollar? One dollar, yeah. That was the whole Microsoft, you know, the whole kill Netscape, destroy the entire, Marc Andreessen, you're out of a job. Like, what's the cost? A dollar. Wow. The same thing happened, by the way, in the 50s when Goodyear and GM banded together and bought all their train tracks in LA and ripped them out and privatized them and just ripped them out.

46:57And the court, there was an antitrust and they got fined a dollar. Okay. I'd like to point out maybe quickly two possible societal goods here. One is that this results in one of the world's largest nonprofits being created that now has the backing of a frontier lab. And the stated goal of the new OpenAI nonprofit, one of their first goals is to spend$25 billion using AI to solve disease. And I think that's a tremendous societal good. We've spoken here in the past about how AI has the potential to solve disease and biology in the next five years. I think this is another arrow in the quiver of making that happen.

47:36Second, societal good. One of the things I worry about is what happens if a private frontier lab develops incredible superintelligence and decouples from the human economy? I think putting OpenAI on a trajectory where it can reasonably be expected to go public sometime in the next two to three years, I think an IPO by OpenAI and other frontier labs and putting the equity in the hands of retail investors and index funds is almost certainly a net societal good because it keeps the economic interests of large chunks of humanity aligned with these frontier labs and vice versa. Well, corollary to all that too, Alex, I think that everyone's like, hey, Brendan Foody, Mercore, he's a billionaire at age 23.

48:20He spent an immense amount of time inside OpenAI's building. And we saw him there in the lobby last time we were there. If you think, what's my life mission? Am I starting a company? Am I changing the world? Am I solving all disease? Regardless of what your life mission is, think about the impact of$25 billion of charitable money just to solve disease. What about the other$100 billion? Where's that going to go? So if you're involved in this in any way and you don't have a strategy for how you interact with open AI, how am I in that building? How am I relevant? How am I going to be – when they start turning to commercialization and goods through the AI engine, how do I interconnect with that?

48:58You know, I ask all these entrepreneurs, what's your open AI strategy? And a lot of them have no answer. But you think about the scale just of what Alex just said. You have to have an answer to that. This is the largest nonprofit in terms of capital base, and it will be even bigger. It will reach a half a trillion dollars. And Dave, you remember you and I met with – I'm not going to say who it is because I don't think it's been officially released – the individual who is a co-founder there that will likely run this OpenAI Foundation. We're talking about potentially spinning up some X-Prizes as a means.

49:32He was very knowledgeable about these ideas of incentive competitions to sort of leverage capital 10x. And we just learned this year that the numbers from XPRIZE is we leverage every dollar in the purse by 60-fold. So imagine if, you know,$100 billion becomes$6 trillion of leverage. What a fun time ahead. Yeah, well, think about the scale, too. It's just exactly a great point that, you know, a normal big, big X prize is a$100 million prize. You know, and here you're talking about$130 billion, which, you know, if the stock goes up post-IPO, could be$260 billion. So all they have to do is, you know, sell some shares and fund a$100 million prize.

50:15Could do that every day of the week. I dream about having 10$1 billion prizes for the 10 biggest problems. It would steer where students spend their time, where founding partners focus on building companies. I mean, it would be sort of a flame to the entrepreneurial moths out there. Can I take the other side of this just for a second? Sure. Of course you should. Signaling, greenwashing, they're putting all this money over there and then going full speed towards the IPO and hoping that the good that they can do will kind of balance the crazy path to greed. Do not notice the man behind the screen.

50:59And speaking about greenwashing, here's our next story with Sam Altman. I turned green here, a little bit of Shrek in his DNA. And the title here is OpenAI Plans a$1 Trillion IPO and to spend a trillion dollars a year in AI infrastructure. I love this. We've said this before, a trillion dollars here, a trillion dollars there. It's becoming a word far too popular these days. I mean, the speed is incredible. So OpenAI is - Use your word, Peter. It's incredible. It is incredible. Take a drink. You also got to hand it to Sam for the unbelievable sheer gumption for just going for it. Yes. Wow. No, amazing.

51:42Just incredible. Well, hey, let me put some numbers behind that. So here he is. I'm going to spend a trillion dollars a year. I'm going to do a trillion dollar IPO, and then I'm going to spend a trillion dollars a year on data centers. Your actual revenue today, dude, is$13 billion. Now, you're saying you're going to get to$100 billion. But the equivalent would be is if you have a household with$100 ,000 of income, and your husband or wife comes home and says, honey, we should spend$10 million a year on houses and stuff. That's the equivalent metric, okay? Okay, so just to put it in context, that's the gumption, like you said, behind this claim.

52:16But hey, he's done everything he's said so far, so it's plausible. Yeah, so their IPO's planned in— I'll try that with Lilly and see how far I get. Well, you did that when you mortgaged your house to buy Bitcoin, which in retrospect turned out to be a good idea. We didn't put all of it in, unfortunately, but still. Well, you still get it. Mortgage it again, Alex. I think it's actually like a pretty tiny number. Global GDP is upwards of$100 trillion. So just saying we're going to spend 1 % of global GDP on AI infrastructure. Yeah, but this is one guy, Alex. This is one company. One company, right?

52:52Yeah. So if you have like five frontier labs each doing that, that's still like 5 % of global GDP. I think this is a drop in the bucket. And that's before AI starts to radically grow the global economy. This feels on the low end to me. And we saw this last time compared to the railroads or telecom infrastructure. The AI build is still as a percentage of the US GDP on the low side. But here are the numbers. OpenAI is working on the largest IPO in history with targets to do this in 26, 27. And the other point made here is that they're planning to build one gigawatt capacity per week at 20 billion per gigawatt.

53:37There are 52 weeks in the year. That's a trillion dollars a year, which is pretty extraordinary. Just to give people a context, gigawatt is enough to power the whole of Dallas-Fort Worth. It's a truckload of energy.

53:55Incredible. There we go again. Sorry. Okay, time to drink. This episode is brought to you by Blitzy, autonomous software development with infinite code context. Blitzy uses thousands of specialized AI agents that think for hours to understand enterprise scale code bases with millions of lines of code. Engineers start every development sprint with the Blitzy platform, bringing in their development requirements. The Blitzy platform provides a plan, then generates and pre-compiles code for each task, Blitzy delivers 80 % or more of the development work autonomously, while providing a guide for the final 20 % of human development work required to complete the sprint.

54:42Enterprises are achieving a 5x engineering velocity increase when incorporating Blitzy as their pre-IDE development tool, pairing it with their coding co-pilot of choice to bring an AI native SDLC into their org. Ready to 5x your engineering velocity? Visit blitzy.com to schedule a demo and start building with Blitzy today.

55:06Okay, let's go. Let's visit our friends at Claude. And I love this article, Alex, that you found. Claude shows signs of introspection, a model that is partially self-aware. Claude 4.1, take it away. Alex? It's a really interesting paper. You'll recall that historically the best suggestion that was floating around the AI research community for diagnosing self-awareness was maybe you train a model on the internet, excluding any notion or any mention of self-awareness and see whether the model is then able to articulate something about self-awareness. I think that original proposal was probably somewhat impractical.

55:49This is a far more practical diagnostic for self-awareness. And the idea is basically take the internal hidden activations of a model and graft on an external thought, sort of incepting an externally imposed thought onto a model and detecting whether a model is able to recognize that it's having external thoughts intrude upon its internal activations. Isn't that called psychosis? I think that might be slightly different. I think this is closer to some sort of maybe telepathic forcing. You're taking an external activation vector and forcing it upon the internal hidden activations of the model and then checking whether the model realizes that it's being externally influenced.

56:36And pretty remarkably, some of their stronger models, the Opus 4, 4.1, were about 20 % of the time able to articulate not only that they were being externally influenced through this sort of vector activation injection, but were also able to reasonably well articulate precisely the nature of the external thought that was being forced into their internal streams. So the question is, what does self-aware mean? It understands it's an AI model, it understands what, Alex? The proposal in this paper is that self-awareness means that the model is able to think about its own thought. It's able to understand what its own inner thoughts, if you will, its own inner activations are and able to reason based on that.

57:26Well, just a high level point for the neural net geeks out there too. This research can only be done if you have access to the internal weights and activations of the neural net. So it's done inside anthropic. But now that meta is not going open source, you have to actually use a Chinese model to do this kind of research or you're screwed, which is really very sad because I think, you know, before I switched to computer science at MIT, it was cognitive psychology. And I think that experimenting with the parameters and activations of a neural net will tell you far, far more about how a human brain works than the normal stick a little probe into a rat.

58:01And so it's an incredible research playground. And these ideas of what's the definition of self-awareness and can I inject a thought? So what you do is you say, here's the neural net thinking about a very specific topic. I'll grab the actual activations from part of the neural net. And then while it's thinking about something else, I'll inject those and see if it somehow is complementary. And then, of course, the result is the neural net is like, where did that thought come from? So that's the introspection and self-awareness. But you can only do that if you can splice thoughts, which is an incredibly powerful, cool tool.

58:31And as soon as you go API only and you start – because on that radar chart we saw earlier, you're operating outside the neural net and trying to define AGI from outside the neural net. But it's so much more powerful to operate inside the neural net. But we may be in danger of losing that as a tool. Hopefully, the Chinese models will keep coming out. Alex is warning me against using them too much. Salim, a final word from you. This reminds me of Hod Lipson, who's a professor at Columbia, and he builds self-assembling robots and evolutionary robots that have a feedback loop to improve themselves.

59:03And he actually tried out his approach to self-awareness, which was ask the AI what it would look like in five years. And then by the feedback loop of constantly forcing itself to go, oh, well, who am I that I might look like something in five years? He thought that would generate self-awareness. he thought that's what happened in Facebook a few years ago when they shut it down. And that question, by the way, is blocked in all the major models. But somebody will do that to DeepSeek, and you'll get to that same point. And I remember Dan Barry talking about the frog. We may have talked about this on the podcast before.

59:36He's watched a ton of free-floating animals in labs at NASA. And his opinion of self-awareness was frog. And we're like, frog? He goes, well, a mosquito is an automaton, doesn't really know it's a mosquito. A dog definitely has self-awareness, knows it's a dog. For him, the boundary condition was a complexity of about a frog, where in his opinion, a frog kind of goes, oh, I'm a frog. Above that, more, below that, less. Okay. Let's move to our friends at Alphabet. An incredible quarter for them. They topped$100 billion in quarterly revenue for the first time ever. Google Cloud grew at 34%. So just good on on Alphabet and Google.

1:00:15They're rocking it. And a couple of other elements that Google and Alphabet have announced, a new marketing tool, which I love called Pomely. And this is Google's AI marketing tool. Let's take a look at this video and then we can discuss it. Because I think this is, you know, again, Google provides all of these incredibly useful end user tools that make them so powerful as a company.

1:00:50so pomaly will understand your business dna

1:00:57and prompt your own campaigns or get suggestions.

1:01:06All right, bottom line. Yeah, bottom line, this is an AI tool from, yes, go, yeah, from DeepMind that helps small businesses create on-brand marketing campaigns. So Pameli will analyze your business website, learn the tone, the color, the style, to create ads that match that brand, and generate ready-to-use posts that can be edited in the tool. So they're basically helping their customers who are advertisers do better advertising. I think super smart. Any comments on this, Dave? Well, before Google had AdSense, they thought they were going to hire 10 ,000 salespeople and be kind of like Lycos and AltaVista.

1:01:50Nobody remembers all this. But there was a very, very smart original employee there, an Iranian guy who said, hey, why don't we create an auction marketplace and people can just come and bid on Google and we'll open it up to the economy. We'll democratize it. We'll make every entrepreneur in the world able to thrive along with Google. And it worked incredibly well. And that created the Google we see today. They're going to do that again with all these schools. Gil Elbaz created that engine, right? And yeah, incredible. Oh, there I go. Yeah, incredible. So, hey, you just, as they roll these capabilities out.

1:02:25I just want to point out the latent trading ability of people in the Middle East is off the charts. And when you apply that to kind of deep internet paradigms, that's kind of incredible. What this struck me as was another example of, Peter, in our book, Exponential Organizations, we have the concept of interfaces, right? And Google AdSense succeeded because you automated the supply side and the demand side of the ad business. And this is now pushing the boundaries of that further and further into the creative process. Just to comment, Peter, on this as well. I think the elephant in the room here is that the visual ads that are being generated are not being generated pixel by pixel.

1:03:02I've spoken on the pod in the past about how in the future user interfaces, I think, are just going to be every pixel is purely generative. In this case, it's almost charmingly retro in the sense that it's not pixel by pixel generated. It's vector graphics. It's images and photos clipped from the original underlying website. And I think the elephant in the room that it's not purely generative means that it's going to be ultra low compute cost to generate. And that is suggestive that we may live in a very near-term future where display ads on the internet are generated on demand because it's relatively compute expensive at the moment to generate a custom image pixel by pixel for an ad.

1:03:40But it's relatively lightweight. I love that. Yeah, especially as agents are cruising all of my tabs on my search engines and listening to my conversations. They know exactly what I want in that moment and can generate an ad to influence me until such time that I just give my AI permission to do all the buying, in which case it's game over for advertising. Well, you know what else is charmingly retro in all of this. Hold on, Peter. Peter, you've hit on something unbelievably huge here. Okay, what's that? it on all of this is assuming a human consumer right yeah and very quickly which is short through that yeah i remember having this conversation with you know we were advising procter and gamble and we did a workshop with them and they're like spend a huge amount of r &d on what color should the pampers box be to attract some design at what level i'm like well my wife has an amazon subscription to diapers and doesn't care what the box says anymore and they're like huh and it's just the dissonance between the old way and the way that you're talking about once we have our own ai interfacing that changes everything i think that also has to be taken into account so maybe this is just a short-term thing and jarvis will just buy equivalents take over jarvis will buy everything i need because it knows when i'm running out and it knows what the best quality is and it doesn't really care what the ads say all right let's move on in the google verse here so uh We're seeing Google AI Studio introduce Vibe Coding.

1:05:07So Vibe Coding is now available on Google Studio. No coding or API needed. Dave, do you want to take this one or Alex? Well, it looks a lot like Replit and Lovable. So we'll see how that shakes out. This is the big guys stepping on the toes of the little guys. Yeah, I used it. It was a fantastic experience. And I think somewhat differentiated from the Vibe coding, the in-browser Vibe coding experiences from OpenAI or Anthropic. For one, it creates multiple files. If you ask it to create an app, it's not just sort of fixated on a single self-contained file. So it can create multiple files of code, which is very important for certain sorts of apps.

1:05:47So I ran it, of course, through my favorite eval for Vibe coding, which is create a visually stunning cyberpunk first-person shooter. and it created a visually stunning dashboard, sort of an intro lobby dashboard for the FPS, but I had to prompt it to create the rest of the game. But what it did create was visually stunning and I think it's a promising first step. I am curious what the interaction with Repl.it and Lovable will be. We spent a few days with Amjad Massad, the CEO of Repl.it, and I've been playing with Repl.it on my phone and computer, sort of vibe coding different apps, which is fun.

1:06:24And one of the things I had a long conversation with Jack Hittery while we were in Riyadh. And one of the things that Jack said, which I love, is every morning, instead of becoming a consumer, become a creator. So usually I get up and I'm reading all of Alex's texts, all the breakthroughs that he found last night. And I'm just constantly, as all of us are, consuming hundreds of articles over the course of the week, maybe 20 or 30 per day. and Jack was, no, no, no. Every morning, I'm going to vibe code something. Every morning, I'm going to create something. And I think that creator mindset is so critically important for us to be using.

1:07:08So a conversation with your AI and creating something every day would be super fun. Always be creating. I was creating an app on my phone last night to remind me to take my pill packs because I have five pill packs a day. and so it will now text me in certain windows and remind me, did you take your pill pack? And then I can dismiss it if it didn't. So it's sort of an aging adjunct to my health. Okay, moving on. Let's go on to the chips and data center wars. A lot going on here. Again, a trillion here, a trillion there. First story is Samsung is building a facility with 500 ,000 NVIDIA GPUs automating chip manufacturing.

1:07:53This is an AI mega factory that will combine NVIDIA's Omniverse with Samsung's chip making for up to 20 times faster performance. Blackwell chips have generated $500 billion in business so far. Again, a nice chunk of change. Alex, what do you make of this one? This is what recursive self-improvement looks like, Peter. This is GPUs, AI being used to optimize chips to make more AI. And there are so many applications ranging from computational lithography to fab optimization for this. I think when I've spoken in past of the innermost loop of civilization looking like some linear combination of chips, robots, data centers, power sources, all of these, this is what the innermost loop of civilization spinning faster and faster looks like.

1:08:45This is the economy, right? That innermost loop is the economy going forward. certainly the future of the economy. Yeah. I did some reference checking here. So 500 ,000 NVIDIA GPUs that will draw somewhere between a quarter to 0.4 gigawatts of energy. There's no single site in China or the West that compares. So the Chinese top centers topping out at 10 ,000 to 35 ,000 GPUs. And in the US, clusters like Azure and Meta are ranging from 30 ,000 to 55 ,000. And this is 500 ,000. Wow. Does this mean the machines are now basically manufacturing their own evolution? Is that basically where we're going?

1:09:29Absolutely. Yeah, that's the right way to think about it for sure. And also, you know, the AI fabs themselves have always been automated. They're all roboticized on day one. But it's the periphery around that, turning it into a data center or feeding the front end of the fab, that huge amount of investment opportunity there. to close the inner loop. I think Alex really should write a book if books still exist with that title. But it's such a brilliant insight. But there's so much leverage in the inner, inner, inner loop. And if you focus on, okay, where are the bottlenecks on the innermost inner loop?

1:10:05And you're going to find that it's the chip getting out of the fab and into a data center and actually doing something useful. That's where most of the bottlenecks are now. Huge robotics. And energy. And energy. Yeah, and that's in the feeding on the front end of it. All right. Here is a fun conversation, an article, Extropic Creates Thermodynamic AI Chips Combating Industry's Energy Crisis. This comes from a friend of ours, Gil Verdon, who's been on my stage at the Abundant Summit. He's been on our podcast here. And he's talking about a breakthrough in hardware, a technology called thermodynamic sampling units, TSUs, that are using 10 ,000 times less energy than GPU-based systems.

1:10:48using probabilistic bits. Alex, so what does that all mean? Well, so I read Gil's paper, and I'm a huge fan in general of trying to get closer and closer to the physical limits of computing. Seth Lloyd famously 20 plus years ago discovered slash reported that the ultimate physical computer would probably, at least for serial computing, look like a black hole, black hole as the ultimate supercomputer. So I'm a big fan of approaching the physical limits of computing. In this case, though, my worry is there's such a sort of a sordid history of probabilistic computing approaches being attempted and failing to keep up with Moore's law and algorithmic improvements.

1:11:32This is my worry. I want to believe I want something like this to succeed, but I'm not super optimistic that that this isn't just going to get steamrolled by algorithmic advances and advances in good old fashion CMOS digital logic. It looks too much like analog computing, probabilistic computing. And remember, 10 ,000 times, even generously, 10 ,000 times energy improvement at the rate that models and algorithms are advancing and the rate that good old fashioned digital CMOS is improving, That may only be a few years of headroom, which a new architecture would need anyway to get off the ground.

1:12:12But can't we view it from the energy perspective? Because reducing the energy requirements on Earth by 10 ,000-fold seems staggeringly beneficial. In principle, yes. But in practice, the workloads that the economy demands have to be able to run on these computers in order for you to realize this hypothetical energy advantage. And I think for better or for worse, the burden of proof is on Xtropic and Gil to demonstrate that his hardware can host workloads that are as commercially valuable as, say, the NVIDIA workloads. I saw Elon and Gil going back and forth on X. And Elon saying, so do you have something I should be looking at?

1:12:55And Gil said, yes, let me show you. So we'll see if the Muskverse gets behind this technology. I have a quick comment. Yeah. Yeah. Alex, can you go back to the black hole being the ultimate superhuman world? Just because you lost me right there and my head's like stuck now. Can you just go and turn into that just for a second? It turns out black holes are wonderful computers. They'd be a little bit difficult on the input-output side, especially the output side. A little. But in principle, so you can, and Seth and others have refined this notion, you can define a generalized notion of computation and then in pure physics.

1:13:32So it deals with terms of how quickly can internal state changes happen inside a physical system. And turns out black holes are absolutely the physical limit based on the physics that we have today for the fastest serial computer because state changes in terms of their quantum state evolve at the physical limit. So programming them may be a little bit challenging. Maybe you'd have to fire in like an X-ray laser or gamma ray laser. And maybe you have to parse the Hawking radiation. But if you can solve input-output, black hole supercomputer is the way to go. I can see the title of this episode this week is Black Hole Supercomputer is the Ultimate.

1:14:08Black hole supercomputer on your desktop. Okay, let's move on. This is a fun article. So Elon Musk on data centers in orbit. SpaceX will be doing this. So here we see in the image here V3 of Starlink. So Starlink version 3 will be coming out. It will be delivering 10 ,000 times or 10 ,000, 10x more capacity, one terabit per second and enabling large-scale off-world processing. And I love this quote that you shared with me last night, Alex. I added it. 100 terawatts per year is possible from a lunar-based producing solar-powered AI satellite locally and accelerating them to escape velocity with mass drivers.

1:14:53So basically turning lunar material into compute and then accelerating them off the moon with a mass driver, again, the work of Gerard K. O 'Neill, and into Earth orbit. So this is the beginning of a lot of things, Alex, we've been talking about. Dyson swarms, matriarchical brains. Let's talk about what we're talking about. We're talking about disassembling the moon to build more computers, to build Computronium and the Dyson swarm. And what's more remarkable, I mean, just in the past few episodes of the pod, I've been beating the drum for how, you know, mark your calendar now. We're at the very beginning of the construction of the Dyson Swarm.

1:15:33Maybe I was overly pessimistic. Maybe we're actually going to see multiple competing Dyson Swarms and SpaceX is going to launch one. You'll see other companies, maybe other frontier labs launch competing Dyson Swarms. At this point, in the style of worrying about overpopulation on Mars, I'm starting to wonder whether I should be instead banging the drum for ensuring good interoperability between all of the Dyson swarms. Let's take a moment. Freeman Dyson, a brilliant individual who said at some point, you're going to disassemble all the planets in the solar system and create a sphere around the sun that captures all of its energy.

1:16:11And that's going to be the hallmark of an advanced civilization that's called a Dyson sphere. If it's not one sphere, but a bunch of different satellites and computes that can be viewed as a Dyson swarm. But the real fun concept is a Matryoshka brain. So, Alex, over to you. Yeah, so these are three overlapping concepts. So, Dyson sphere, Dyson swarm, Matryoshka brain. Dyson sphere was this notion of having basically and there was a even a Star Trek Next Generation episode that I saw that one yes a solid sphere at roughly Earth radius from from the Sun that lots of people perhaps could live on the interior of uh and and enjoy nice environments and things probably not practical from a material science perspective the stresses would be enormous us.

1:17:02That's Dyson Sphere. Dyson Swarm says, let's rather than having this be a solid enclosure that's rigid, let's instead have this be lots of orbiting satellites that are nonetheless collecting the energy from the sun. Matryoshka Brain says, let's take multiple Dyson spheres at different radii from the sun and have the innermost spheres consume the light, the solar insulation at certain frequencies, and then radiate waste heat outward to the outermost spheres, which then will consume progressively more and more infrared shifted radiation and use that to power their compute. So these three concepts, Dyson sphere, Dyson swarm, and Matryoshka brain, and also Jupiter brain is another popular depiction.

1:17:52These are all interrelated concepts. I think if we go the trajectory of taking apart our solar system, whether we brand it as one or the other, they're pretty similar. Maybe a black hole is a Matryoshka brain circling a star and we can't see the light. You know, I wonder about this. If this is the fate of intelligent civilizations, I would expect to see more infrared shifted solar systems elsewhere in the galaxy. To my knowledge, we haven't observed this. That makes me suspicious that even though I bang the drum for Dyson swarms, maybe there's something out there lurking in our technological future that will cause us to not actually need to take apart our solar system.

1:18:39Two points. One, for those of you interested, Matryoshka brain comes from the Matryoshka dolls, which are the nested Russian dolls. So you can imagine nested spheres around the sun, each one absorbing energy, utilizing energy, and then radiating waste heat that becomes the input for the next sphere that it radiates to and the next sphere and so forth. Second point is we've got to get Elon back on the pod here to talk about this. I think it would be a lot of fun. All right. Let's talk about energy and robotics, our final topic for today. This is a big deal. So California invests big in battery energy storage and leaves blackouts behind.

1:19:24So it used to be pretty awful. And I remember this, we had rolling blackouts in California, but the state has done something amazing, which they've increased battery storage by 3000%, going from 500 megawatts in 2020 to 15.7 gigawatts this year. and the batteries store solar for evening demand, replacing underperforming gas plants. And, you know, I'm glad to see this is happening. And Salim, thoughts? I think this is awesome. I mean, what a testament to 15 gigawatts is an incredible number. No, it's not. No, it's not. No, it's not. Stay away from the word incredible. Yeah. It's a stupid, trivial rounding error, guys.

1:20:15How about stonking? I don't know. We're going to get our subscribers drunk. So interestingly, the price - Batteries don't store gigawatts. Batteries store gigawatt hours. Yes, gigawatt hours. And if you read the story underneath this, it's only got three hours at that power level. This is a joke. Okay. This is just like, this is like all of Alex and my interactions with government trying to do things that sound important that are these stupid little edge rounding cases. This, at peak load for California, this is one hour of storage. And on a typical day, they can store up to about a day of an average.

1:20:49All right. But let's take a look at the numbers, right? So blackouts have been cut by 90%, from 15 a year to two a year, which is great. And here's the problem. We're going to see the CPI of electricity just skyrocketing, right? So the price for electricity was 22.5 cents per kilowatt hour in 2020. It's increased now to 32.4 cents per kilowatt hour, a 44 % increase. And if we continue to make the demands that we have on data centers, there's going to be a problem. The proverbial shites can hit the fan sometime soon. I agree with that. But I mean, California has done everything humanly possible to self-destruct at the government level, despite having the greatest tailwinds, like the most incredible state, massive state, all the innovation in the world, every advantage in the world.

1:21:45And the government claiming to do something good by piling up a bunch of batteries. It's like, what, you're down to two blackouts a year? I mean, seriously, that's our expectation of what we get for 13 % tax rate? At least they're down a lot. That's ridiculous. There's something there. I'll point out also just for California specifically, if folks are familiar with the infamous so-called duck curve of California, where the demand for electricity peaks in the evening, like right after sunset, and also in the morning, there is a mismatch between California, which is rich in insulation, in solar energy on the one hand, and the need for early evening power.

1:22:25I think even just a few hours of battery storage can help to smooth out the duck curve. And that's transformative for California in a way that we here, perhaps in New England, don't have quite the same problem. In our energy story, Google to buy power from NextEra, a nuclear plant being revived. So Google signs a 25-year deal with NextEra to buy power from a revived Dwayne Arnold nuclear plant in Iowa. It's reopening in 2029, provide 615 megawatts round-the-clock carbon-free. It's a$1.6 billion project. And it's interesting that we've got these, again, these hyperscaler companies that are buying energy.

1:23:09They're basically – it used to be that this was something the government did. The government provided a distribution network for power and you would buy it off the grid. That is no longer the case. Companies need to provide their own energy. So going all in on fission plants, SMRs, soon fusion plants, hopefully solar plants. Salim, do you want to jump in? And, ooh, look, it'll create 400 jobs. Such negativity. This is huge. This is great. Yeah, this is an important bridge, I think, to the near future. I think what this shows is that we're basically dissociating the energy sources from the grid.

1:23:56And now we can have energy wherever it is happening and plunk a data center next to it and then leverage it. So the marginal energy usage around the world will totally explode. I can't wait for geothermal to really kick in here, right? There's so many places that geothermal is easy. Why? But wait, Alex, you're putting way more emphasis than I would have thought. Tell us why. Well, I think bridges are important. And right now, the limiting factor for tiling the world with compute is, as a number of executives have recently pointed out, we have the GPUs. The problem is having warm racks to put them in, as Satya Nadella said in the past few days.

1:24:33I saw that. Microsoft has, you know, what, hundreds of thousands of GPUs they can't turn on because they don't have the energy for it? Yeah. So I think like having bridges like reactivating otherwise disused nuclear plants, I think this is an incredibly important bridge to the future until we get SMRs and fusion and maybe solar and maybe new forms of nat gas. Totally, totally right. And, you know, what Chase Lockmiller is doing also the same bridge kind of structure where you start with your regular fuels, use natural gas or whatever. but it's steam turbine generation right into the grid right into the data center you can reuse all that when you move it to small nuclear you move it to smr and then a fusion comes online in 2030 2032 maybe replace the boiler you replace the boiler it's just like by far the most efficient way to get to the ultimate end state the dyson swarm or whatever but but between here and there That's the right stepping stone.

1:25:33It's much harder to do with solar because solar doesn't feed into a generator. It feeds into a battery pack. And so you're not reusing any of that when you move it to fusion in 2032. I wonder what takes four years, right? This is four years away. What takes four years to get an existing nuclear plant up and going? Are they retrofitting it? Are they updating it? Or is this all paperwork? It's all of the above, Peter, because we're advising Fermi America on this stuff. They're planning to do six gigawatts of gas turbine and six gigawatts of nuclear. It is very, very complicated to spin up a nuclear power plant.

1:26:12And that's the inner loop. That is the Alex Wiesner-Grosse inner loop of the inner loop of the inner loop. Like, focus on that. Like, Peter, why is it four years? Could we make it three years? You can build a starship and go to Mars in four years. I don't understand why you can't get a nuclear plant up and going in four years. To their credit, they've had an AI generate an S11 in record amount of time and got it out the door. So it's starting to happen. It's just that this is why that SAGE project is so important, Peter, to rewrite policy as we need it. Listen, when the Trump administration finally says we're going to accelerate this tenfold, that's when they'll get serious about energy production.

1:26:54All right. These are some fun articles coming out. we're about to see the Robotaxi wars coming online. So NVIDIA is planning a Robotaxi project to challenge Waymo and Tesla. So here are the numbers. NVIDIA is launching a$3 billion Robotaxi project in the self-driving car race. This is a partnership between NVIDIA, Uber, and Stellantis. For those of you who don't know Stellantis, they are behind brands like Chrysler, Jeep, Peugeot, and Fiat, One of the largest automakers in terms of not the brand, but building the parts. And it will use an end-to-end AI system called Cosmos that NVIDIA has built to handle driving simulation.

1:27:42Interesting. They want 100 ,000 robotaxis launched by 2027. It's coming, right? We saw, God, 100 years ago, there was a 10-year period where we went from 99 % horse and buggy and 1 % cars, flipped it all over to 99 % cars and 1 % horse and buggies. And the question is, is that this decade between all these players and all the capital going in? Thoughts, gentlemen? I made a prediction 10 years ago that all driving would be automated. My son, who's now 14, would never get a driver's license. So we've got two years to satisfy that. Yeah, my boys too. So here's some of the numbers. Tesla cybercabs right now, they have 200 vehicles operating, I think, in Austin.

1:28:36And their plan is to scale up to 10 ,000 this coming year. Can't wait for them to be in Santa Monica, where I live. Waymo has 700 vehicles. And by the way, I see them all the time as I'm driving around. I must see 20, 25 of those a day. So 700 vehicles is a pretty small number. So they must have concentrations here in LA and up in San Francisco. And there's 500 ,000 miles between collisions with Waymo. It is the safest player out there. And NVIDIA is partnering now to go live with this. I love that$3 billion is a stonking amount of money. And yet for NVIDIA, it's like a drop of the bucket given their market.

1:29:18It is a drop of the bucket. It's like they don't even notice there's a little side project. But this is where their GPUs are going next. They're going into humanoid robots and autonomous cars, right? It's automating the entire world around us. That's right. Alex, what's your take? They have spoken in the past about how this innermost loop is not going to remain contained inside data centers for very long. I think, as I've noted in past, the computer is literally going to walk out the door of the data centers. In this case, it's going to drive out the door. But I think for many people, these driverless cars, I have one, and many people I know surprisingly haven't even driven in one or haven't had the experience of driving in one.

1:29:57For many people, this is going to be their first encounter with a generalist robot. It's going to be either seeing or driving in or owning a driverless car. And it's not going to stop there. I think the same stack that we're seeing NVIDIA with their autonomous vehicles pushing, it's going to generalize to humanoid robots on the timescale of one to three years. So this is, again, the beginning of the expanded innermost loop of civilization that we're seeing. On the flip side, here's the article from Uber's perspective. And just to remind folks, Dara, who's the CEO of Uber, will be on stage with us at the Abundance Summit in March.

1:30:36It's going to be my best ever. We have only 30 seats left for the Abundance Summit. It's selling out faster than any time ever in history. If you're interested in grabbing one of those 30 seats, you can apply at Abundance360.com. But this is not a commercial the program will fill, but it's going to be incredible. Can't wait. We're going to be talking with Dara about not only autonomous vehicles, but flying cars. and Uber is going in so many new directions, just driving revenue. So their goal is 100 ,000 NVIDIA-based robotaxis beginning in 2027. And this puts them in direct competition with Waymo and Tesla.

1:31:17Today, in a number of cities, you can order up a Waymo on your Uber app, which is fantastic. Any final thoughts on this story? I think this is too slow for my taste because there's like such a huge demand for this. Even Waymo only has, I think, 2 ,000 is what I looked up. There's 800 in the Bay Area, 500 in LA. We need like tens of thousands. The good news is each Waymo car replaces dozens of cars that are sitting around 94 % of the time empty. Yeah, most definitely. All right, our next story here. I love this one. Foxconn did deploy humanoid robots at its Houston AI server plant. So check this out.

1:32:00Foxconn will expand AI server production in Texas to half a million square feet, producing GB300s and Blackwell series AI servers. It's got a partnership with Digit, which is Agility Robotics. We're going to have the CEO of Digit on our, or the CEO of Agility Robotics on our stage this year, Salim. At this stage, abundance needs to be like eight days long, Peter. No, it's tough. I mean, I'm trying to make sure that we have enough time for all of the community members to meet each other, hang out, have conversations, and have it fun. But yeah, it's four and a half days. But we'll have four, currently four robot companies there.

1:32:45I'm trying to get a fifth one from China. We'll see if we can get the new version of one of the top Chinese robots there as well. So check this out. These humanoid robots are going to be driven by NVIDIA's Isaac Groot-N model. You know, this is the innermost loop, isn't it, Alex? It is. This is robots operating factories that make servers that go in data centers that power the robots. That's the loop. Yeah. I won't say incredible. I'm not going to say it. I'm not going to say it, but wow. I will say there are nearly enough companies working on all the different form factors. There's room for a thousand more startups doing different variants of this.

1:33:31For all the mechies out there that are wondering what they should do. I'm going to build a robot company that has an eight-armed octopus-type robot and beat everybody else. I also think this is a preview of how we get... You were talking earlier, Peter, about Sam's forecast of a gigabot per week. I think this is a plausible technical trajectory for how we get there. We're going to have robots building the fabs and the factories producing the servers and the data centers. It's going to be one massive flywheel. All right. Our next story in energy is Blue Energy and Caruso partnered to develop an advanced nuclear-powered AI data center.

1:34:11But I think the more interesting story here is that they plan to stand it up with natural gas plants and then convert it to nuclear about three years after that, right? So don't wait for nuclear. Get it operating with what you can right now and then retrofit nuclear when you can. So the wait time for a gas turbine engine today for the gas turbine stuff is about four and a half years. It's crazy. It's insane. We advise Siemens Energy and they're sold out forever. And it's incredible. Everybody's trying to go to dump heaps and get spare parts for gas turbines out of their garbage dumps. It's like really, really in the recycling plants.

1:34:58It's really crazy right now. Wow. I want to talk a little bit about the China-U.S. battle and hit a couple of different points here. So here are some numbers, and they're important to note that China is dominating production in a few different areas. 66 % of electric vehicles are being built in China. 80 % of solar panels and batteries are in China. 60 % of wind turbines. Those are staggering numbers. Add to that that on the innovation metrics, 70 % of all global AI patents are coming out of China, and 75 % of clean energy filings are coming out of China. You know, this was part of the debate that on our evening in Riyadh that we hosted this dinner, and Kathy and Balaji were going back and forth on topics like this.

1:35:53Dave or Salim, you want to weigh in? We really ought to have just a focused session with Antonio Gracias and Chase Lockmiller on how to deal with this in the U.S. Because our whole investment cycle isn't geared up for this type of investing. And China is. It's just very manufacturing heavy. And, you know, hey, we need more power. We need more melters. We need more whatever. We don't really do that well in the U.S. venture economy. But that's all getting rethought right now. And Antonio and Chase are the guys right in the middle of it. So we should just get them on the pod and brainstorm our way through.

1:36:24How are we going to restructure? Because these are very expensive projects. They're not venture projects. They seem to work every time. The playbook is right on that prior slide. It's not a mystery. You have to be involved with the government. You need zoning. You need a location. You need permits and all that. So it's just a new format for American innovation. But it's going to last for 10, 20, 30 years. So you might as well get ahead of it. I mean, imagine if Elon weren't doing what he does right now. you know these numbers would be far worse so uh this i want to make a comment yeah there's something this is the one of the flaws of our democracy we're in four-year high metabolism election cycles nobody's thinking 20 years down the line yep and china may be authoritarian but they can look out 20 years and say we need that much energy that much water and do things to make that happen.

1:37:19I think I actually had a hack for this. I kind of did some brainstorming. Somebody asked me at a conference, what would you do? And I said, every four presidential terms, I would appoint a government, give them 10 % of GDP and say, your only job is to fix all the stuff that's 20-year-old range projects. And then you're out. One term only, full authoritarian, go, which is essentially some of what Trump is doing in this case. So you just want to rewrite the constitution. Okay. It's a great thought, Salim, though, because, you know, flying back from Saudi Arabia and just looking down and, you know, there's nothing out in the desert for hundreds and hundreds of miles.

1:37:57And then you get to Europe and it's just the most blessed Mediterranean greenfield. Like everything should be perfect in Europe. But the government dysfunction is preventing them from any kind of involvement in what's going on right now. So it's a good case study. Like you can mess it up in a real hurry. And so if we don't have, but we're not geared up to compete with China right now on this particular front. And it does need to get rethought. But if you don't rethink it, yeah, things can go really bad. Don't take it for granted. So guys, on one of our next WTF episodes, I want to bring some of the data that we found at FII9, the Future Investment Initiative event that we were just at.

1:38:37Some of the data is staggering about how the rest of the world looks at this. And I want to share that on an episode because it's really important because it's going to drive the near-term future. All right. This particular chart comes from a tweet that Balji put out, and he labeled it, it's happening. The AI flippening is here. And so this is a look at who's making the open-source models. And what we saw this summer was open-source models were being dominated by China versus the U.S. and most definitely not Europe. And I think what's most important here is that as governments start adopting different AI systems, their ability to get access to free open models versus paying for the models from the hyperscalers in the US, it's kind of a land grab going on.

1:39:33I don't have much more to say other than watching this. You know where this is really, really going to collide? I was talking to Brian Elliott over at Blitzy about exactly this topic. Like, you know, with Meta doing open source, we had a huge open source option in the U.S. And then Meta fell off the grid, you know, and last models were Terrellama 4, I guess, was terrible. And now they're trying to rebuild it, but they're rebuilding it closed source. So the U.S. doesn't really have an open source option. But then when you look at the, there's a bunch of projects for the military that, you know, the Blitzy will ultimately end up working on that need to be air-gapped.

1:40:05And you've got to use an open source model in an air-gapped environment. You can't just go to the OpenAI API with super proprietary government data. But right now, my only choice is Kimmy. Kimmy 2 running on Grok with the Q-chips and the Grok cloud, which I think is a phenomenally good option. But it's all Chinese code. God knows what's inside there. So we spent a bunch of time with Eric Schmidt. I want to play a short video of Eric from the FI9 event we were at last week talking about U.S. versus China, just to sort of provide the U.S. perspective. Who's winning this race? At the moment, the United States, without question.

1:40:48The U.S. has a deep financial market that allows you to raise literally a trillion dollars on a thesis and an idea, which is incredible. You have this massive build-out going on, and you have a real potential of solving hard problems. Tell me how close China is to overtaking us. It's not as close as I thought. I went to visit. China does not have the depth of the capital markets. They do have lots of energy, which we don't. They have lots of energy, but they don't have the depth of the capital markets, and they don't have the chips. The capital markets, they haven't figured out a way to make all that money the way the U.S.

1:41:24does. and the chips that haven't been able to make the chips that the United States and others won't give them. That keeps them behind by a good chunk. China is, however, focusing on exploiting AI in every aspect of its business much better than the United States. So I think the U.S. will win on the intelligence race, but China is likely to win on the deployment race. And that's a problem for America and Europe. All right. I want to jump into robots as our last topic here. So 1X, we've had a great pod with Bernd Bornik, the CEO of 1X. Look it up if you haven't seen it. Dave and I went and visited his factory.

1:42:03Let's take a look. They have a release of their Neo Gamma, and here's their promotion. My name is Bernd, and today we're launching Neo, our humanoid for the home. Ain't no sunshine when she's gone. Neo is a humanoid companion designed to transform your life at home. It combines AI and advanced hardware to help with daily chores and bring intelligence into your everyday life.

1:42:36As someone who lives with Neo every day, there is no experience quite as much. All right, so here are their commercials. They just went out. You can put a$200 deposit down on a Neo Gamma robot. And they announced their pricing,$20 ,000 to buy it in their early access. And I love this. Or$4.99 per month. And you can buy it in three skin tones. I find that fascinating. But check this out. On the right-hand side, I was walking to my workout gym, which is just outside my studio. and there was this giant sticker on the ground. So this is incredible marketing. I have to hand it to them. They're doing really a super job on direct-to-consumer marketing on this.

1:43:29Yeah, I pre-ordered mine. Can't wait to have the experience. I do think it's interesting that in many of the scenarios for the Neo, at least in the early days, according to 1X, they're going to be tele-operated. And that may turn some people off. Having someone tele-remote into their home doesn't bother me at all. I'm very, very excited to try this out. I think I got my order in when I was there at their facility. I'm hoping, you know, they're going to be, so Berndt will be at the Abundance Summit, and he's going to be bringing a number of the Neo Gamma robots. I'm just going to put one in my car and drive away at the end of the summit.

1:44:07So I'm getting mine in March. Peter, when we were out there, he was saying $100 ,000,$140 ,000 price point. I can't fathom how he's coming in at$20 ,000. It's competition. They would have got his film back in the factory, but there's so much going on inside this robot. I cannot believe that they can get it out the door at$20 ,000. If I were running this company, I'd be subsidizing. I mean, this is a data collection plate to get a huge VLA trade in.

1:44:35They're skating to where the puck is going to be. So Elon said Optimus is going to be at this$20 ,000, not price point, but cost of goods in when they get to millions of robots being built and eventually robots building robots. Again, the innermost loop over here. And this is competition against the prices out of China, the prices from Elon and from Brett Adcock. So you've got to be competitive. And it's training people to expect price points that resemble, if you're purchasing it outright, a cheap car. And if you're leasing it, it's like leasing a car. This will be for, you know, I've mentioned in past the American dream, you know, so-called of having a house in the suburbs, a car, and now a humanoid robot doesn't necessarily generalize that well to the rest of the world.

1:45:26But I think having at least one humanoid robot in your home becomes part of the new economy. And when the price gets down to$300 a month to lease, right? Again, I've made these numbers, I say them every time,$10 a day, 40 cents an hour. Everybody can afford that because your robot now becomes part of your earning potential. Your robot can go and do stuff for other people or for you, right? As Elon has said, this is all about creating the world of abundance. I love this story. I've got a couple of quick thoughts. Yeah, go ahead. One is I have a dog that literally looks like a teddy bear. And I'm wondering what would happen if the robot mistook the two.

1:46:11That would be one. And I know all the models out there. At least in this case with all the visuals and so on, it's not a kickboxing robot, which I thought was not a great marketing thing to say from the last couple of episodes. Yeah, Unitree. I'm excited to see what happens here. Unitree has definitely taken a different approach. and I have to say that Optimus has still got a hard metal exterior. Neo Gamma from 1X came out with this soft, cuddly, warm sweater-like look. Very important. And then figure copied it. The latest release of figure three has the same look. So anyway, I guess borrow from the best.

1:46:56this is a story, Alex, that you shared. Tornoyal, how do you pronounce it? I think it's Tornoyal. Tornoyal. It's an autonomous drone that ends mosquitoes. And I love this. Can you imagine you're a tech entrepreneur and you're someplace with a lot of mosquitoes and you're just being bothered and you go, how do we end these mosquitoes? And then your answer is intelligence drones. So here in this video, we see this drone that is very lightweight, and it is autonomously flying around. It's spotting the mosquitoes, and it's zapping them with an electric grid that the mosquito is flown through. It recharges and patrols 24-7 from its base station, uses ultrasonic sonar to detect the mosquitoes, beating of its wings.

1:47:49and a kinetic interception to eliminate them. This is like smart rocks in space or smart dust in space. We're going to get nanobots, I think, that will enable us to regulate ecosystems. And I think in the process, it's probably going to raise a number of bioethical questions. Peter, you and I and Dave have talked offline about near-term futures where, for bioethical reasons, or maybe even, dare I say, effective altruistic reasons. We're repairing butterfly wings on the one hand, but on the other hand, we have drones to obliterate mosquito populations. It's going to be a very interesting future.

1:48:29It will, and it is an interesting right now, being in the present. It's a super exciting time. I'm going to end this pod with a thanks to Ruken, one of our subscribers and one of our fans, who sent us over a musical piece called Don't Look Up, The Singularity Is Near. And I'm going to play it as our outro music. But before we do that, gentlemen, any closing thoughts? Awesome episode. I learned so much today. This is amazing. Yeah. Yeah, there's no doubt the pace of stories is really accelerating. You've got to anticipate it all 2x every month or two. So it's going to be interesting keeping up.

1:49:18I wake up in the morning at now 3.30, and I'm like, what happened while I was asleep? Well, discussing a flying mosquito killing drone was not part of my thinking for what I would be talking about today. Yeah, well, we're trying to— What about putting a black hole supercomputer on every desktop in every home? Yay, or Motryoshka brains, you know, and taking apart the moon to get us a dicey. Have to happen sometime, Peter. And butt breathing was our article last year, our last episode, our new episode, our new closing article. I think we should do that. We should have an unusual science closing piece or an unusual tech.

1:50:03and I'll take the mosquito killing drone as our one for this week. All right. We're ending the moon. The moon had it coming.

1:50:14Well, listen, I have to go there and start a city before we take it all apart. That's one of my goals. You know, anyway. Gentlemen, I love you much. And everybody, if you're an audio or video creator and you have an ending outro music piece you want to share with us, You can share it with us on the pod. We read all the comments and they are incredible comments. And thank you all for subscribing. Please share this with your friends, your family. One of the best parts of being in Riyadh, Dave and Salim, was all of our fans there. I kept on, everybody kept on coming up. I mean, this is a conference of like 5 ,000 people, sort of a world economic forum in the desert.

1:50:57And everybody's like, I love your podcast. So, Alex, you were sorely missed and a lot of fans. Yeah, people can tell me I'm going, where's Alex? Is he joining that episode you're doing? Yeah. Got to invite me next time. Okay. Well, we'll do FII in March in Miami altogether and try and line up a live podcast from FII Miami. All right. Here is our outro music. Enjoy this, everybody, and have a beautiful exponential day and don't sleep through the singularity. It's the most exciting time ever to be alive.

1:51:58Cattle lights, gossip, but who even cares? My cat just posted its own nightmares. Don't look up, the singularity's near. It's laughing in binary, loud and clear. We built a god from electric dust. Now it prays to us how to have it or trust. Don't look up, the code's gone divine. Heaven's a glitch in the command line.

1:52:33All right. Oh, the lyrics are amazing. Amazing. We built a god out of digital dust. Yeah. All right, guys. Awesome. Have an awesome day. Every week, my team and I study the top 10 technology metatrends that will transform industries over the decade ahead. I cover trends ranging from human-on-robotics, AGI, and quantum computing to transport, energy, longevity, and more. There's no fluff, only the most important stuff that matters, that impacts our lives, our companies, and our careers. If you want me to share these meta trends with you, I write a newsletter twice a week, sending it out as a short two-minute read via email.

1:53:08And if you want to discover the most important meta trends 10 years before anyone else, this reports for you. Readers include founders and CEOs from the world's most disruptive companies and entrepreneurs building the world's most disruptive tech. It's not for you if you don't want to be informed about what's coming, why it matters, and how you can benefit from it. To subscribe for free, go to dmandis.com slash metatrends to gain access to the trends 10 years before anyone else. All right, now back to this episode.