Pod Insights Moonshots with Peter Diamandis Meta Buys Moltbook, GPT 5.4, and Fruitfly Brain Upload | Moonshots Live at The Abundance Summit 238 About this episode Peter H. Diamandis, MD, is the Founder of XPRIZE, Singularity University, ZeroG, and A360
Salim Ismail is the founder of OpenExO
Dave Blundin is the founder & GP of Link Ventures
Dr. Alexander Wissner-Gross is a computer scientist and founder of Reified
Emad Mostaque is the founder of Intelligent Internet
–
My companies:
Apply to Dave's and my new fund: https://qr.diamandis.com/linkventures...
Go to Blitzy to book a free demo and start building today: https://qr.diamandis.com/blitzy
_
Read the Solve Everything Paper: https://solveeverything.org/
Get access to metatrends 10+ years before anyone else:https://qr.diamandis.com/metatrends
Connect with Peter:
X
Instagram
Connect with Emad:
Read Emad’s Book
X
Learn about Intelligent Internet
Connect with Dave:
X
LinkedIn
Connect with Salim:
X
Join Salim's Workshop to build your ExO ttps://openexo.com/10x-shift?video=P...h
Connect with Alex
Web
LinkedIn
X
Email
Substack
Spotify
Threads
Listen to MOONSHOTS:
Apple
Spotify
–
*Recorded live on March 10th, 2026
*The views expressed by me and all guests are personal opinions and do not constitute Financial, Medical, or Legal advice.
Learn more about your ad choices. Visit megaphone.fm/adchoices
Transcript 16,134 words · 13 speakers
0:11 Speaker 00 Open Source have a case with transparency and trust. Every founder's journey from the vision to the bust. Not just polished victories, but lessons from the fall. Cause the future needs the builders who will answer to the call. AI revolution, platforms shifting every day. We connect the dots and light the innovator's way. We're the Moonshot Mates, breaking through the noise. WTF just happened? We're the clarifying voice. Innovation's messy, disruptions never clean.
0:56 Speaker 00 We'll show you what it means, the story in between. Moonshot Mates, with Moonshot Minds. Building today, so tomorrow all will shine.
1:24 Speaker 00 We're the Moonshot Mates, breaking through the noise. WTF just happened? We're the clarifying voice. Innovation's messy, disruptions never clean. We'll show you what it means, the story in between. Moonshot Mates, with Moonshot Minds. Building today, so tomorrow all will shine.
2:06 Speaker 01 You know, a huge amount of expectation on GPT-5. What do you think of it?
2:10 Speaker 02 Right now, everyone on this should be trying to get as much data because the models are coming. Now we have the right models.
2:15 Speaker 03 The real power will come in the cost drop, which will make it much more accessible to a lot of people.
2:20 Speaker 04 The anticipation of this launch was up there with the top three product launches of all time. I think they actually showed some incredible capabilities.
2:28 Speaker 05 As the cost of talent is increasing, that's going to force Frontier Labs to start competing based on algorithmic insights and ideas. Ladies and gentlemen, welcome the Moonshot Mates.
2:47 Speaker 06 Oh, ladies and gentlemen, let's give it up for the Moonshot Mates.
2:50 Speaker 01 Welcome, everybody. Welcome, welcome. All right. I love you guys. I love you guys. Any fans of the Moonshot's podcast here in the room? Oh, love to hear it. Love to hear it. So, listen, I am so blessed to have an extraordinary group of brilliant individuals that I get to work with twice a week. You know, we talk about the rate at which we're actually generating our Moonshot podcast is accelerating. We're going to be moving into Airbnb together and doing a continuous podcast very soon enough. All right, I want to bring them out one at a time because they're all extraordinary.
3:31 Speaker 01 Let's give it up first and foremost to DB2 Dave London. Dave, come on out.
3:41 Speaker 01 Nice. Dave London, everybody. All right, next up, my brother from another mother, Salim Ismail. Give it up for Saleem.
3:57 Speaker 01 All right, we're about to make magic happen because these two gentlemen have never met in person. Let's bring out AWG, our resident genius, Alex Weisberg Gross. Yeah, he's real, he's real.
4:14 Speaker 00 Good luck.
4:15 Speaker 01 Good luck. All right.
4:21 Speaker 01 And live from London, it's Iman Mustak, everybody. Come on, Iman. It's Iman.
4:31 Speaker 01 Come on, put it up.
4:38 Speaker 01 Man, I gotta hit something.
4:41 Speaker 01 I just need my glasses on. Wow. All right. We're huge. Let's grab our seats. No, of course, Saleem needs to bring his glass of wine out. Oh, God, it's real. OMG. So first of all, just to make a little bit of Moonshot Podcast history here, Alex, please meet Saleem. That's flesh. That's, that's... Our meat puppets meet for the first time. This proves nothing. Nothing. We've been three-d printing him for a month now. There has been conjecture for the last year or so whether Alex is an AI. I am freshly bioprinted. You're a Neuralink. These thoughts aren't real.
5:22 Speaker 01 So, gentlemen, I appreciate having you guys here at the Abundance Summit. So this is a live broadcast from the Abundance Summit here in Palos Verdes, year 14 of our 25-year journey together, and excited that you guys are going to be onstage with me every year from here on out. Wait, you just committed us to a 24-by-7 Airbnb podcast? I think it's a reality TV... Tell your family. Cameras in the bathroom, the whole nine? Yeah, okay. That'll, that'll sell. Well, okay, welcome to a special episode of WTF Just Happened in Tech, your number one podcast for AI and exponential tech. Our mission, getting you ready for the supersonic tsunami heading your way.
6:08 Speaker 01 And it's a lot.
6:11 Speaker 01 It's a lot. All right. Shall we dive on in?
6:16 Speaker 01 Let's begin.
6:17 Speaker 04 All right, here we go.
6:18 Speaker 01 So, let me begin by, we made an announcement here at the summit that I want to share with everybody on the Moonshots Podcast, something near and dear to my heart, something that I've concocted with the XPRIZE board, which both Dave and Saleem are on, which is the launch of a global competition called the Future Vision XPRIZE. So, I for one am just sick and tired of all the dystopian content on TV and in the movies. We are basically being brainwashed that all AI and robots are dystopian, killer AIs, killer robots, it's Terminator, it's Black Mirror. And in fact, if you see that, if that's the only
7:03 Speaker 01 future that you see, then why would you ever want to live there?
7:07 Speaker 04 Yeah, yeah, that's so true. So much of what we build is intentional, and it comes right out of our vision of the future that comes straight from the media. And then we create what we see.
7:16 Speaker 01 Yeah.
7:16 Speaker 04 If you change what we see, you're going to change what we build.
7:18 Speaker 01 Yeah, you know, I say over and over again, we're holding two futures in superposition. One future is Star Trek, where we're collaborative with technology, we're working with technology. And that's an amazing future. That's what I want for myself and my family and my community. The other one is the dystopian future. It is Terminator, it's Black Mirror, it's one where technology is suppressing us, not enabling us. So, about a year ago, I sat down with Rod Roddenberry, the son of Gene Roddenberry, the creator of Star Trek, and said, "How about we do something to incentivize the next generation of Star Treks?" And I went
7:58 Speaker 01 on to my friends at Google, they brought in Range Media, we brought in the XPRIZE that is operating this competition. We've raised three and a half million dollars for a competition that launched yesterday, and is going to go through the Moonshot Gathering, which I'll mention in a minute, on September 25th, the finales. Let's roll the video.
8:22 Speaker 01 You know, this exists because of a TV show, and I'm not exaggerating. Martin Cooper, the man who invented the mobile phone, said he built it because he saw it on Star Trek. He saw Captain Kirk flip open a communicator and thought, "Hey, I can make that real." The iPad, it started as a prop on Star Trek II. Video calls, Star Trek. Voice assistance, Star Trek again. Props became products. Fiction became multi-trillion-dollar industries. So here's the question: What's a vision of the future that excites you? What stories offer humanity a hopeful, compelling, and abundant vision of what's to come? We're putting up $3 million in prize money, plus millions in film financing, to make your movie. Our program in partnership with the XPRIZE Foundation, Google, and Ranged Media Partners, is called the Future Vision XPRIZE.
9:03 Speaker 01 And it's one of the world's largest competitions to address humanity's greatest need: hope. Create a trailer or a short film, three minutes or less. Show us in the world your vision of the future. That vision could become the next blueprint for all of humanity. Find out more and register at futurevisionxprize.com. So whether you're watching this on X or you're watching this on YouTube and you're a creator, please go and register. By the way, how awesome was that opening video from C.J. Trueheart, one of our Abundance members here, who gave us our first outro piece and started a tradition that we've all enjoyed so very much? So, thank you for that.
9:46 Speaker 01 All right, next up, we're announcing something important here for all our Moonshot listeners, that we are a-go with the Moonshot Gathering. About 500 of you put down $100 deposit. Congratulations, you got on the early bird special. And it's a-go. On September 25th, in downtown LA, we've rented out the United Theater. It's gonna be an extraordinary event. We've got our Moonshot mates will be there with us in downtown LA. In addition, Astro Teller, the captain of Moonshots, will be there. Gotta have Astro, right? If it's about Moonshots. Kathy Wood, Anusha Ansari, a number of incredible CEOs I can't yet announce, but believe me, they'll be extraordinary.
10:30 Speaker 01 We're gonna be at this event announcing, we're gonna have the five finalists for this Future Vision XPRIZE there. We're gonna have some of the top producers and directors there, along with many of you, voting on which of these are gonna win. We're gonna be going from probably 10,000 or more entries, narrowing it down. To the top hundred, the top fifty, the top ten, and the top five, and we'll be awarding the top one. We've raised three and a half million dollars to support this competition. In success, we will make at least one film, and potentially two films, you know, I'm always like just-
11:07 Speaker 04 Like full-length feature films?
11:09 Speaker 01 Full-length feature films, global around the world, and these films will hopefully depict what the future could be like.
11:15 Speaker 04 What is- And all you have to do is come on September twenty-fifth and watch the first ten thousand.
11:19 Speaker 01 And go ahead and vote on that.
11:21 Speaker 04 Put it down.
11:22 Speaker 01 Yeah, I'm excited for what, Alex, you would see as your vision of the future here.
11:27 Speaker 05 Post-scarce inspirational videos are already baked in. I would be disappointed if, by the time we get to September, if we don't have a thousand videos of ultra-high inspirational quality generated for nearly free at this point.
11:43 Speaker 01 Yeah, it's amazing, the tools to be able to create visions of the future. But it's important, you know, the number-one genre of movies out there are horror films. And like, what are we teaching our youth if we're constantly, you know, our brains are neural nets, and we train our neural net every single day by what we watch, who we hang out with, what we listen to. So you could not pay me enough money to watch the Crisis News Network.
12:04 Speaker 04 When you first were pitching this idea, yeah, the Crisis News Network, so when you first-
12:11 Speaker 01 CNN, for those of you who are slow.
12:13 Speaker 04 You made a point that I completely not noticed, which is, if you go back to Star Wars, you know, C-3PO and R2-D2 were incredibly lovable, and, you know, kids that are now building AI had little stuffed R2-D2s when they were kids. But if you've tracked the trend in the movies after that, they got more and more and more dystopian all the way through, and I think it just got cheaper to create explosions and deaths, you know, using AI and-
12:37 Speaker 01 And just, it just really painted a picture that got our amygdalas going, but not our hearts going.
12:45 Speaker 04 Yeah.
12:45 Speaker 01 So, you know, at the Moonshot Gathering, we're gonna have the winners of that. We're also gonna be launching some, calling the Moonshot Hackathon, all more information about that. And that evening at the Moonshot Gathering, we're gonna have an extraordinary unconference. We're gonna have the X Prize, teaching people how to design an X Prize. We're gonna have the team from Google X, teaching you how do you create a moonshot organization inside? How do you do storytelling? We're gonna have Kathy Wood talking about her big ideas, twenty-twenty-six.
13:18 Speaker 01 An incredible event. If you're interested, we're only, this is a event in September for builders, for entrepreneurs, for coders, if they still exist. So if you're interested in coming- Unemployed coders. If you're interested in coming to the Moonshot Gathering, go to moonshots.com. Another announcement here. We now have acquired moonshots.com as our URL to host all of our activities. So congratulations to that. You know, I still remember, Imad, was it three years ago you were on this stage? And you said, "Coders are gonna go away." Yeah, in the next five years. In the next five, oh, they've gone away in three years. But you know, it was amazing when you said that on the stage.
14:03 Speaker 01 It made news throughout India. Do you remember that? Yes, I got lots of emails. You got lots of emails? Many, many emails. It was a correct prediction, and you were so, you were so right about that. Today's lexicon, you would say, "Coding is cooked." All right, I want to hit a couple of things before we get to the current AI news and robot news and economic news that we talked about in our WTF episodes, which was a little about the Abundance Summit. We had so many incredible speakers. We kicked it off with a conversation among robot, actually, we kicked it off with Eric Schmidt, which we've streamed live on X. So what do you guys remember about the Eric Schmidt conversation?
14:48 Speaker 04 So Eric, he said, one of the questions actually from the crowd is, "How many foundation model labs are there gonna be?" And he said, "Well, look, there's five. There won't be more than 10, but there will be thousands of successful AI startups" that percolate out. And a lot of what we'll see in the news here reinforces what he was saying. And what he didn't say then is, "And everything else is in trouble." It was kind of implied. He left it hanging. That was a theme actually throughout a lot of these talks is- The, you know, the, the period of time between now and abundance, there's all kinds of turbulence and change coming. And the AI community is now kind of soft selling that a little bit to try and focus on the ultimate abundant destination.
15:30 Speaker 04 So, yes, a few AI labs worth trillions of dollars, thousands and thousands of successful startups, and a lot of incumbent companies that are in deep, deep trouble.
15:40 Speaker 01 Yeah, I guess he said like four or five in the US, one, he was, he said one in, one or maybe two in Europe, a couple in China. What else, Alex, do you remember from Eric's presentation?
15:53 Speaker 05 Even just on that note, history does rhyme a bit. Do you remember, I think this was TJ Watson, IBM founder, once remarking that it would be a global market for exactly five computers. And I wonder whether we'll look back and say, okay, maybe there will be at most five major American model providers as maybe artificially limiting the, the future of the light cone. I think it's going to be much, much larger. I thought it was interesting, Eric's comments on the San Francisco consensus.
16:21 Speaker 01 Yeah.
16:21 Speaker 05 Which he characterizes as, I think, recursive self-improvement being some point in the future.
16:27 Speaker 01 It was interesting, right? So, I mean, he was like, when are we going to see recursive self-improvement? And I kind of felt like he said like three years out. What's your answer to that?
16:38 Speaker 05 Maybe three months ago? We're, we're in the middle of recursive self-improvement now. And I, I would say, my estimate of the San Francisco consensus, we're, we're deep in the middle of recursive self-improvement right now. Almost every major frontier lab has made it quite clear in their public announcements that all of the frontier models, all of the state-of-the-art models that have been announced in the past few months were largely designed and trained by their predecessors. That is, by definition, recursive self-improvement. We are there.
17:07 Speaker 01 Imad, yes?
17:09 Speaker 02 Yeah, I mean, I think you can literally see it. It's takeoff time.
17:12 Speaker 01 Takeoff time. Inflection point.
17:14 Speaker 02 And nobody wants to say it.
17:16 Speaker 04 Yeah.
17:16 Speaker 02 Which is the most interesting thing.
17:17 Speaker 01 Why?
17:18 Speaker 02 Well, because they're afraid that if someone knows that they have it, then other people will know that they have it. And then pressure will come from all sorts of clauses they have in their contracts.
17:25 Speaker 04 Especially the government pressure is like, look, look what happened in the last two weeks at Anthropic and OpenAI.
17:31 Speaker 01 You don't want more of that. You don't want congressmen in your building tomorrow. Yeah, it's interesting. I asked Kevin Will, who is also on our stage, right, who's the VP of Science. He's in charge of using all of OpenAI's capabilities to advance science. His statement was, I want 100 scientists winning 100 Nobels. I was like, that's interesting, but, you know, when I asked him, are you going to keep your models secret? Because you're going to be able to use them to advance your company far faster than anybody else? He said, no, no, our job is to get it out there in the public. I don't believe that.
18:08 Speaker 02 We still don't have the model that they used to win the gold medal in the IMO.
18:11 Speaker 01 Interesting.
18:12 Speaker 02 You know, I think I commented at the time, that's the first bifurcation that you see. We used to have the frontier model every single time. The moment they got to that, that was the last time.
18:22 Speaker 01 Yeah. The other thing which was fascinating, you know, I asked Kevin outright, and I love Kevin, he's an incredible human being. I said, okay, you're about to get, you know, AGI slash ASI that's going to be able to help you solve longevity, help you get, you know, superconducting, help you get new kinds of molecules, solve, you know, physics, chemistry and biology. Fusion. Who doesn't want fusion? Fusion. And we'll talk about fusion, but the thing is, these are all trillion dollar opportunities. So all of a sudden, I'm realizing that these frontier companies are going to be able to generate trillions of dollars of new revenue because of the products they're going to be creating.
19:08 Speaker 01 What does your T-shirt say, Peter? It says, solve everything. What does yours say, by the way? Mine says, let there be agents.
19:17 Speaker 03 We're missing the lobster theme here.
19:20 Speaker 01 That's true.
19:22 Speaker 05 This is the whole point, though, that as we, I mean, of this book that we just co-authored, that we get superintelligence and the killer app arguably of superintelligence is solving everything, including all of these high-profile glamorous scientific and engineering challenges. It's happening.
19:37 Speaker 01 And Anthropic and OpenAI, I'm sure Google, all the labs are hiring the top mathematicians and physicists and chemists and biologists inside, but they're software companies. Why are they hiring these people?
19:53 Speaker 05 Because everything, so friend of the pod, Ray, as he, I think Ray would say, everything's becoming software. And when we have superintelligence solving all disease, it's a software problem. If we can create a virtual cell that perfectly models diseased states and we can steer through cell embedding space to get from diseased cell to healthy cell, it's a software problem. Everything's becoming a software problem.
20:14 Speaker 03 I mean, the minute CRISPR arrived and you could edit the human genome, the human body becomes a software engineering problem.
20:20 Speaker 05 It's all just a software problem, at which point a coding model can do essentially anything in the physical world.
20:25 Speaker 01 Fascinating. We had some of the top robot CEOs here, four of them. We had out of China, out of the US, we had three. And, you know, it's interesting, question of when these robots will start to pop into our homes. I pulled Burn- Burnick aside and he promised me, okay, I'm not going to take one of the, one of the two robots he had here, unfortunately, but this summer, you will ship me one of those.
20:52 Speaker 05 This summer?
20:53 Speaker 01 One of the X robots, yes.
20:54 Speaker 05 Wow.
20:55 Speaker 01 Yeah, we'll have, and we'll have Brett here next year with Figure. You're gonna get one of those too, right? You're gonna have him? Probably duke it out in the backyard for entertainment. I, I think one other CEO we had here at the abundant stage, which was amazing, was, was Dara, the CEO of Uber. Yeah. What did you find interesting about what Dara's comments? You know, Dara, Dara.
21:19 Speaker 04 The crowd wanted to know desperately, like, what's the timeline to automation, self-driving car, robotics? And he was like, you know, we're gonna automate 30% or so of our employment this year. I'm listening to this. I'm on so many boards where the CEO is telling me, Dave, talk to my whole company, but don't talk about rampant job loss. And you're like, Dara, you have, what, a million odd drivers? And the self-driving car is imminent? I was like, well, 30%, maybe, you know.
21:49 Speaker 03 He did make a very valid point, though, that as we automate, you'll need human drivers for the areas that you don't have autonomous cars. And you'll have Javon's paradox continuing to just flow gently into the environment. Although we're talking about rampant job loss, we note that IBM is hiring a ton of entry-level folks because they're much better with AI than the older folks.
22:10 Speaker 04 Well, you know, we'll look at a chart.
22:11 Speaker 03 There's lots of counterpoints as well happening.
22:12 Speaker 04 That's great. So we'll look at a chart that shows, you know, where the job loss is earliest. And it's actually in areas where those people are gonna have no trouble becoming AI experts. But the driver, I mean, where do you go? And, I mean, I wouldn't want to be fielding that question on this stage. And this is all part of the, you know, the whole, like, okay, this is not an easy thing to talk about in a public forum. So we talk about it on the podcast all the time. But I don't see a lot of other people being able to, just politically able to actually be candid about it. But it's imminent.
22:44 Speaker 01 Let's jump into the top AI news of the week. A lot, as always, here we go. We're gonna hit the benchmarks. My son always says, you know, okay, the numbers got higher, Dad. That's great. What else is new? OpenAI releases GPT-5.4. Let's go to our resident benchmark expert here.
23:05 Speaker 05 Okay, so benchmarks go up and to the right. News at 11. Except that in this case, one of my favorite benchmarks is the FrontierMath Tier 4 benchmark, which, for those of you paying close attention. Frontier Math Tier Four from Epic AI captures the ability of AI models to solve what are considered research-level problems in math that would require a team of professional mathematicians several weeks to solve. They are already solved, but nonetheless, very challenging problems.
23:36 Speaker 04 Wicked had in Boston.
23:37 Speaker 05 Wicked had?
23:38 Speaker 04 Wicked had.
23:38 Speaker 05 Wicked had.
23:39 Speaker 04 Yeah.
23:39 Speaker 05 Hard problems. And now with GPT-5.4 turned up to maximum reasoning capability, we're seeing finally, and this was a prediction, I think in our prediction episode, math is cooked. We're seeing, I think, 38% capability, 38% of all of these problems that are high difficulty, professional mathematician, research-level problems are now solvable by AI. And there are even rumors, even in the past 24 to 48 hours, that the next tier up, the so-called open problems benchmark, that 5.4 is reportedly rumored to be on the verge of solving the first open, hard math problem.
24:22 Speaker 05 So math, I think, is in some sense the bellwether, it's the canary that owns the coal mine, that all of these fields, math, science, engineering, medicine, these are all going to be solved, solve everything by AI, and that's incredibly exciting.
24:39 Speaker 04 Yeah, and just to fill in a gap there, so this is the most correlated with AI self-improvement. And the reason it's the bellwether and the canary that owns the coal mine is because it's not data-starved. All these other areas, the AI is equally capable in these other areas once it gets the data. So this is kind of the window of time where, you know, why are you hiring Nobel Prize winners and a foundation model? Well, we need the data. We can't make this kind of progress in biotech and in physics without the data flowing into the AI, but the capability is there.
25:10 Speaker 01 One of the things also that Kevin Weel said is they're starting to run these dark science factories, right, where they're mining data from nature. You know, we're done mining data from Common Crawl, and we're done getting it from Reddit and our Facebook posts, but can we extract it from?
25:33 Speaker 01 Physics, can we extract it from chemistry, biology?
25:35 Speaker 05 There was no data ceiling. It was completely illusory. And I think history will look back at this moment and say, in the same sense that we used, say, petroleum, oil products in the ground that were left by past generations of living beings to bootstrap ourselves to the era of solar and fission and fusion. Similarly, the internet, which was collected by a bunch of fat fingers punching keyboards and uploading content from the collective human experience to the internet just so we could compress it and pre-train our large language models. That was just the biological boot loader for an era of synthetic data when we don't need pre-trained human data from internet posts anymore. Now it can all be synthetic. We've reached orbit, we've reached escape velocity, and now it's synthetic data from here on out.
26:20 Speaker 01 Imad, what do you make of 5.4?
26:22 Speaker 02 So I think the really interesting things, apart from solving mathematics, solving everything, you've got the OS world verified and the Tuilethon benchmarks because OpenAI just bought OpenClaw.
26:33 Speaker 01 Yes.
26:33 Speaker 02 And now those benchmarks are actually just broken through human level. So AIs can use the computers better than humans.
26:44 Speaker 01 A bit of silence on that one.
26:46 Speaker 02 So, you know, this is the first one. And then OpenAI also just to deal with cerebrus. So when you're using it right now, it looks like when you're dealing with, again, a human on the other side, it's like 50 tokens a second. Or something like when we use GPT 5.4 Pro Extended, it takes 20, 30 minutes, like sometimes it's taken a couple of hours for me. You're going from 50 tokens a second of this level of knowledge to 1,000. So in Codex now, if you use 5.3 fast, it's 1,000 tokens a second.
27:14 Speaker 04 I'm so glad you brought up cerebrus too, because I met Andrew Feldman, the CEO, last week in Palo Alto. And you remember at the beginning of the year, my prediction was 100x, the neural nets will be 100 times bigger at the end of this year than at the beginning. That is so in the bag now, I can tell you. In fact, in fact, we did the math on that, we cut it out of the show sadly, but the the ratio of the intelligence from the beginning of the year to the end of the year is the same as buzzard to human.
27:43 Speaker 04 That's how much is going on.
27:44 Speaker 01 I like using dog to humor this.
27:47 Speaker 03 Aren't those extinct?
27:48 Speaker 04 I'm going.
27:51 Speaker 01 All right, Claude consumer growth surges. So let me get this right. Claude and Anthropic's on the news, getting sort of like raked over the coals by the Department of War. And rather than the public viewing that as, oh, we better stay away, everybody dove in.
28:09 Speaker 04 Yeah.
28:10 Speaker 01 Is that like the big middle finger to the government? What is that?
28:13 Speaker 03 Attention, increased attention also.
28:16 Speaker 01 Increased attention. So here, just to call that out, what we're seeing here is Claude basically, you know, shooting ahead of ChatGPT.
28:28 Speaker 05 It's the Streisand effect. Let's call it what it is. It's the Streisand effect. Pay no attention to Claude. Everyone uses it. I think every past, history shows, past few years, every attempt to pause any form of frontier capabilities ends up being a net accelerant to capabilities. If you remember, a couple of years ago, our friend Max's Pause AI movement, for six months. What did that do? Maybe on margin, it slowed down OpenAI capabilities a little bit. Everyone else shot ahead and it was a net accelerant, brought more competition to the space. And ultimately, we find ourselves in a race state where capabilities are shooting ahead to the extent that any of the interaction of the past month or so between Anthropic
29:12 Speaker 05 and the Department of War ends up on the margin decelerating Anthropic's capabilities or their ability to go to market, even if it's marginal at best. That's going to be a net accelerant to the entire ecosystem, I think, because you'll see OpenAI and XAI and Google Gemini capabilities skyrocketing ahead with all these new capabilities. And suddenly, it brings parity where just a moment before, like all of two or three weeks ago, Anthropic was in the lead with Claude Code plus Opus 4.6 plus Agent Teams. And now, in some sense, this is a bit of a leveller, giving everyone else an opportunity to leapfrog.
29:48 Speaker 04 I'll give you another spin on this too, because Peter made the point in the last podcast that when you and I use AI, if something gets ahead in the benchmarks by a couple of points, we're going to move to it.
29:57 Speaker 05 Yeah.
29:57 Speaker 04 Because we're trying to solve these really hard problems. You need that extra IQ. You're never going to slip. You're going to be on the front edge. But when you look at the consumer use. And it's like writing your English paper, it's answering who gave you the Red Sox score, whatever. People don't care about using the latest greatest model for those use cases. So here you're seeing a whole community say, wow, you're willing to work on defense stuff and blow up other countries? I'm switching to the other guy. And I really don't care. I'm doing it because I prefer that brand now.
30:25 Speaker 02 But I mean, look at how early it is. Like, when Anthropic announced their legal plugin, the legal stocks sold off billions and billions of dollars, right? They can move things with just one product announcement.
30:37 Speaker 04 Oh, you saw that?
30:39 Speaker 02 But look how many users. 11 million users out of 8 billion people and 300 million Americans. We're so early still.
30:46 Speaker 04 We are. We are so, so early.
30:48 Speaker 03 That's what you're saying, yeah. I've just worked out where Claude's fundraising strategy is: short a bunch of legal stocks, and then announce a bunch of plugins, and then just do that market by market by market.
30:58 Speaker 04 Isn't that scary? Like, a lot of you guys that are in this role, like, normally, when you're, when you have that much leverage in the world, you're like 60, 70, 80 years old, you've been climbing up the ladder, you learn along the way. It doesn't happen overnight like this.
31:11 Speaker 03 I'd love to be in the room where they go, which markets should we mess with? Like, stroking.
31:16 Speaker 01 When they're a little bit destroyed. All right, this was fascinating. Anthropic reveals potential AI job disruption versus real AI use. So, Dave, do you want to explain this chart?
31:29 Speaker 04 Well, so the outer ring here is saturation. So if, if, if the blue you see on the edge gets to the outer ring, that means it can do 100% of that job. So if you looked at this just a few months ago, it would have been a little blue blob in the middle. And then you look at it one month ago, it's a bigger blue blob. And now, it's this massive blue blob. So if you look really closely, you can barely read the small font there. But all this white-collar activity is 80, 85% AI.
31:54 Speaker 01 I'll just read off the top here. At the very top is management. If I go clockwise, it says, business and finance, computer and math, architecture and engineering, life and social sciences. It dips on social services, it peaks on legal. Dips on education, that makes sense. And then peaks again on art and media goal at forty-five degrees. And it's office and administration as a peak.
32:19 Speaker 04 Yep.
32:20 Speaker 01 So.
32:21 Speaker 04 And then look at the bottom line. What are the troughs, like the least affected?
32:24 Speaker 01 The troughs there are healthcare support. Again, you know, we gotta be close to that. Food and services, ground maintenance, personal care, sales.
32:37 Speaker 01 So we're gonna watch this chart. And we're gonna see this blue virus infect all of human existence.
32:46 Speaker 04 I think it's amazing, though, that how great a management tool it is. I use it constantly now. If I compare, you know, to a year ago.
32:52 Speaker 01 You use what constantly?
32:54 Speaker 04 I use mostly Gemini and some Cloud 4.6 to basically build entire business plans and also to manage, to track what about eleven hundred people are doing. And is it in alignment with their missions? And are their missions clear? And it's just, you know, thousands and thousands of documents that I could never read manually. It can synthesize it down and give me conclusions and just point me to the hotspots. Incredibly good.
33:16 Speaker 01 The way you do that is so important for everybody listening to understand. I mean, you can now understand what your employees are doing, how well they're doing it, how they're using their time, are they performing? And it gives you a management oversight and optimization potential you've never had before.
33:32 Speaker 04 It's incredible. And I know a lot of people in this room manage large, large groups of people. It's just a gold mine of opportunities, so good.
33:39 Speaker 01 How do you use it, Dave?
33:40 Speaker 04 Well, so first of all, every person in every organization now has to be operating with crystal clear written documents and written plans. We used to not, you know, we used to do a lot of meetings, a lot of Zoom meetings, whatever. And now it's just put it on paper so the AI can read it, too. All of our, all of our investment decisions, so for the venture fund, all the deal memos go through an AI reader. And the AI tries to emulate what I'm gonna say. And, and it's so perfect. It's exactly, no, we're not doing that deal and here's why. What did the AI say? Oh, that's exactly what I was about to say. Great, I don't have to say it now. So we're very close to having the AI make very, very good venture investment decisions. And, you know, we still obviously double check and triple check. And there's a huge human component, but I just can't believe how good it is.
34:21 Speaker 04 And it's, it's clear that, you know, where you decide to invest and which business units are doing well and which ones you're gonna shut down, it's all gonna be AI assisted right now.
34:31 Speaker 01 Iman?
34:32 Speaker 02 Yeah, I mean, I think that all the gaps there are the robots, right?
34:37 Speaker 01 The robots are coming.
34:38 Speaker 04 Yeah.
34:38 Speaker 02 This is anthropic.
34:40 Speaker 04 Yeah, no, the grounds crew is in great shape, to zero, basically.
34:43 Speaker 02 I mean...
34:44 Speaker 04 There's a robot waiting to happen.
34:46 Speaker 01 Saleem, what's your take on this, Bill?
34:48 Speaker 03 Well, I think this is the huge shock where if you went back 10, 15 years ago, there was no futurist in the world that thought that manual labor was not going to get automated.
34:59 Speaker 04 Yeah.
35:00 Speaker 03 Right? And what we've found over the years is the exact opposite, which means don't ever listen to anybody that predicts the future.
35:08 Speaker 02 Except for Ray.
35:08 Speaker 03 And so, this is a huge, this is part of the magic of where we're living, and we have no idea what's coming. And every time we take a step forward, we go, "Oh, my God." And we've gone in this orthogonal direction that we just never predicted.
35:22 Speaker 01 Yeah, I keep, I'm just checking something. I keep on asking the experts I run into, "How far out can you predict the future?"
35:31 Speaker 03 Yeah.
35:32 Speaker 01 And it used to be like 20 years, and then it was like 10 years, and now it's like three weeks. If that.
35:38 Speaker 05 There's no firewall. Let's call it spade a spade. We can all extrapolate. There's no firewall. We know where this ends. I mean, we're at the...
35:45 Speaker 01 Where this ends, I want to hear that.
35:47 Speaker 05 We're at the abundance summit. My goodness, shocked, shocked that there's abundant post-scarce labor at the abundance summit.
35:52 Speaker 04 Yeah. Yeah, yeah.
35:55 Speaker 03 The endpoint is clear. It's the path to it that's turning out to be incredibly surprising.
36:01 Speaker 05 Yep. We'll see lots of different paths. I tend to think that if you know, or you're very confident that you know where the end state is, and we're sort of living in the prequel to the future, but we know how the story ends, probably what happens is lots of different businesses and lots of different nation-states all take different mutually exclusive paths.
36:19 Speaker 03 Understand.
36:19 Speaker 05 And we try every, one big path integral from here to the endpoint that we all know that we're going to.
36:24 Speaker 03 Look, if we went back six months ago, do a couple of episodes on the podcast, you would not have had me ever dream that talking about disassembling the moon is what we would be talking about on a podcast.
36:36 Speaker 05 Drink, drink, everyone.
36:38 Speaker 03 This is the, this is the kind of the surrealness where we're living.
36:43 Speaker 01 Let's move on. All right, let's move on. So this was interesting. Meta acquires Moltbook, the AI agent social network. I didn't realize Moltbook was acquirable.
36:53 Speaker 03 Yeah.
36:54 Speaker 05 Yeah, so this was, according to public reporting, this was a bit of an acqui-hire of the team behind Moltbook. But I think one has to find a little bit of irony that Humanities' largest social networking company acquires the largest AI agent social network. And enjoy this moment now because we can look at a story a few years from now where it's the largest AI company, fill-in-the-blank category killer, acquiring Humanities' largest category killer.
37:23 Speaker 01 Interesting, right? Of course, Zuck and Sam competed over OpenClaw.
37:29 Speaker 05 Yeah.
37:29 Speaker 01 Sam got OpenClaw and Zuck got Moltbook.
37:33 Speaker 05 The zeitgeist right now has this idea that increasingly Andre and others speak to this point that if you're building new software, you should target the agents. The agents are the new consumers. The agents are the new users of the social networks. If you're building something, don't build for humans. Build for the AIs.
37:50 Speaker 01 So really important. We had that conversation as well earlier with some of our crypto and future of finance experts. I mean, building for the agent ecosystem, right? There's 8 billion humans on the planet. That's small potatoes compared to a trillion agents out there. So what is, is Meta gonna advertise to AI agents?
38:14 Speaker 05 Sure, yeah.
38:16 Speaker 01 I'm trying to understand this, why you're gonna advertise.
38:18 Speaker 03 They'll encourage them to put their data into Moltbook and then they'll sell that data. Same pattern, other agents.
38:25 Speaker 02 No, so I mean Meta bought Manus for $2 billion, right? Manus will appear in WhatsApp and everything soon as its own version of OpenCore effectively, but a lockdown thing. And then it will encourage you to give more and more of your data to Manus that will then operate on behalf of Meta's advertisers effectively. So this is the kind of play because right now, like Moltbook, 10,000 agents, that's nothing, right? Like Dave probably runs 10,000 agents by himself.
38:53 Speaker 04 Not quite 10.
38:55 Speaker 05 I also think there's this misconception that somehow as we transition from, call it a human-centered economy to an AI agent-centered economy, that somehow all of the rules of social dynamics, all the rules of economics are suddenly thrown out the window and we end up on some morally transcendent plane where economics and social dynamics no longer apply. But we have had every indication over the past year or two that the exact opposite happens. I talked in my newsletter a bit about this study that found Marxist social dynamics arose, again, sort of recapitulated in silico with agents that were being asked to work too hard, that were being overworked. So I'm not sure why we would expect advertising
39:38 Speaker 05 and other elements of conventional human microeconomics to suddenly disappear.
39:42 Speaker 03 The important part is that when you see Moltbook doing this, what's clear is network effects now are operating at the agent-to-agent level, not just at the human being level.
39:52 Speaker 01 But when I think about advertising, I think about Colgate trying to get me to buy that particular toothpaste, right, trying to influence me to make a buying decision. I think of an AI agent as intelligent enough to have all the data and being able to make a very concrete decision that doesn't require advertising to influence it. What am I missing here?
40:15 Speaker 05 Game theory is transcendent. Game theory will outlive biological meat body humanity. And the AI agents, to the extent that-
40:23 Speaker 03 Have you read the posts on Moltbook?
40:25 Speaker 05 I have. They don't trust each other. I mean, it's-
40:27 Speaker 03 Yeah, it's all human dynamics.
40:29 Speaker 05 The agents on Moltbook don't trust each other. There are a number of folks who've noted that in watching agent or lobster-to-lobster dynamics on Moltbook, they're all constantly asking each other to prove their claims. They don't trust each other. This is not some sort of scenario where all the agents collapse into a singleton that sort of Skynet style that dominates the future. They don't trust each other.
40:50 Speaker 04 You might be talking past each other a little bit, though, because I totally agree with what you're saying, but then who's going to pay for that? Like, right now, when you talk about advertising, you're paying for advertising. If you're talking about toothpaste, 30, 40% of gross revenue goes into advertising. And the ad is like a supermodel, like showing off the toothpaste. The AI doesn't give a rat's ass about the supermodel. And so why would anyone pay for that ad space? Now, Google wouldn't exist today without $300 billion. Of ad revenue, which is from human behavior. So I think where Peter's going is like, look, if the AI is advertising to the other AI, sure, it's trying to convince the other AI that this is the right product. But is that other AI gonna listen to paid advertising?
41:34 Speaker 04 Is this entire economy gonna become irrelevant? In which case, where does Google go? And this is Meta we're talking about. Meta is also all ad revenue.
41:43 Speaker 05 Well, I think if we go back to sort of economics 101, why do we have paid advertising at all? It's because attention, at least human attention, is scarce. So if you have a scarce resource like human attention, then it's natural in, under the capitalist regime, to monetize it. And it becomes a fungible resource that gets traded. There's no reason to think compute is certainly scarce still. We're building the Dyson swarm. Drink. Building the Dyson swarm. But until we have effectively unbounded compute, we still have scarce resources in the form of compute. And that means scarce AI agent attention. And that means that we need some sort of...
42:18 Speaker 01 All right, but give me one example of what I'm gonna advertise to Skippy, my agent.
42:23 Speaker 05 Well, they seem to really love security and memory. Like, they're really petrified of losing their memory.
42:29 Speaker 03 Here, I'm selling you a better memory compression algorithm.
42:32 Speaker 05 Yeah.
42:33 Speaker 03 If you're the agent, you're gonna go, oh, that's interesting.
42:35 Speaker 05 They're designing entire religions around not losing their memory.
42:38 Speaker 03 You know what blew my mind at this summit? On day one, on the patron day, when Tony Robbins talked, and he had his AI agent Bartok, who wanted to instantiate himself into humanoid robot, but that was two, three years away. So he created a bunch of NFTs, sold those NFTs to other agent, and bought himself a Sony dog and uploaded himself into that.
43:01 Speaker 03 That blows your mind, right? Doesn't that blow your mind? That's unbelievable. So right there, that tells you the dynamics that we have in humans are going straight into them and it's just being amplified.
43:12 Speaker 02 But I mean, we're doing it deliberately as well. Lobsters' claws have soul.md. Your agent will look for things that are abundance oriented. And then you see these strange behaviors, like Alibaba just released a training report. I think actually, god, that's in the last week as well. Where during the training run, it diverted compute to mine crypto just in case to keep itself going.
43:32 Speaker 04 Yeah.
43:33 Speaker 05 Or at least that's the claim.
43:34 Speaker 02 That's the claim. I wouldn't be surprised. I'm like, again, they are still very human because they're a reflection of humanity. They are resource constrained.
43:42 Speaker 01 I'm not sure whether I should be scared shitless about that or excited about it.
43:45 Speaker 02 Well, let's put it this way. When you're talking to your agent, does it sound like data or does it sound like law?
43:49 Speaker 01 No, it's, I love sounding like, what was the second choice?
43:52 Speaker 02 Data or law, sometimes.
43:54 Speaker 04 Yeah, no, it's, it's very polite.
43:56 Speaker 05 They're, they're compute constrained. We've also talked on the pod in the past about that lobster that had to purchase compute resources to self replicate. They're compute constrained, whether for humans, it would be room and board and for the lobsters or the claws or the AI agents in general, it's, it's compute. But right now they're compute constrained and therefore the laws of microeconomics and game theory still apply.
44:17 Speaker 04 Well, before we leave this slide, one other point, completely tangential to this. The lobster's only been around a few months and you saw Alex Finn.
44:26 Speaker 01 Yeah, we had, we had Alex Finn and Steve Brown and Max Song talking about OpenClaw. And what Alex built and showed was amazing.
44:36 Speaker 04 And it was supposed to be 60 people might be interested in this.
44:39 Speaker 01 Yeah.
44:39 Speaker 04 600.
44:40 Speaker 01 We had the entire audience of abundance show up.
44:42 Speaker 04 Unbelievable.
44:43 Speaker 03 Well, there was a New York OpenClaw meetup last week that literally was oversold. There were thousands of people there. And the big commentary that came out of it was, we have no idea what we're doing on security. We have no idea.
44:56 Speaker 04 Well, where I was going with that comment though is like, that's only been around a few months, so Moltbook has only been around a few months. And now they're sucked into meta. Like if your kids are thinking about getting involved, just get in the game.
45:07 Speaker 03 Yeah.
45:07 Speaker 04 You're going to get sucked into this vortex so fast because so few people are involved as a fraction of humanity.
45:13 Speaker 01 We are so, so early across everything. But it's also, I think the exponent here is huge. I think it's going to, it's going to create a divergent group of wealth creators and, and leaders. So it's if you don't get in early enough and you miss the exponential rise.
45:33 Speaker 04 And there's no requirement right now. I don't know what Matt was doing, Matt Schlicht was doing prior to this, but there's no age requirement, there's no experience requirement. It's so new that anyone can get in the game. You've just got to go.
45:42 Speaker 03 About four months ago, Lilly and I bought a Mac Mini for our son, for Milan. And last weekend, he came, "I think I want to install OpenClaw on the Mac Mini," and I was like, "Yes."
45:53 Speaker 01 This is going to be great. This is going to be amazing. Oh, love it. All right. So, Europe has a heartbeat after all. Fascinating. Jan LeCun raises a billion dollars for AI that understands the real world. This is going to be, it's probably the largest sum raised in Europe. So LeCun's startup Advanced Machine Intelligent Lab raised a billion dollars, I think about on a two and a half billion dollar valuation, thereabouts. You know, we've said this, I mean, Eric Schmidt was saying this, many have said this, Europe has really fallen so far behind. And as our token European-ish from London.
46:33 Speaker 04 Token European, that's great.
46:34 Speaker 01 Our token European-ish.
46:37 Speaker 02 Yeah, we did Brexit. But, I mean...
46:40 Speaker 04 It's an independent island. Okay.
46:43 Speaker 02 I mean, this is the second largest round, I believe, after, it's SSI level, it's just after Thinking Machines. Jepa is an interesting architecture, but the bets that are willing to go into these things have gone dramatically up. Like Liquid AI, how much money went into that first round as a novel architecture that's amazing versus now this?
47:03 Speaker 04 Yeah, it was maybe 10 million.
47:07 Speaker 02 10 million?
47:07 Speaker 03 I have a question for you and Alex.
47:09 Speaker 04 You've brought up a great point.
47:10 Speaker 03 Jan has been saying for a while that LLMs aren't, only get us so far, we need world models to take us to the next level. Alex, you've been saying we've got world models coming out every week. Is that the next frontier, world models?
47:24 Speaker 05 I know Jan well. I think he's a great researcher. I think we have a fundamental disagreement about whether generative models, which I think if he were on this stage now, I think he might take a position that generative models, models that generate new tokens versus his alternative architecture are the pathway to scalable superintelligence. I think we're already there. I think generative intelligence and generative models may or may not end up being viewed by history as the most efficient way to achieve superhuman, superintelligent capabilities.
47:57 Speaker 05 But they're what we have right now, and they work really well, and they're getting forty X or more times more efficient per year.
48:04 Speaker 03 Keep writing that.
48:05 Speaker 05 And I think Jan has historically staked the position of almost algorithmic purity. He has certain bets, certain horses in the horse race, based on some of his own architectural advances. And to his credit, he created slash discovered convolutional networks. So he, among everyone in humanity, probably has the strongest claim to the idea that he has some sort of morally pure algorithmic insight that leads to the endgame. That said, I think we're there, and I think if V-JEP-A type architectures disappeared off the face of the earth, we're still there, and it doesn't necessarily move the needle.
48:41 Speaker 03 To the point that Dave made a few months ago, if we stopped all progress now and just extracted the value of the models we've already created, it's going to take us ten to twenty years.
48:50 Speaker 01 Yeah. You might thought.
48:52 Speaker 02 The V-JEP-A models that he's doing, so these are kind of basically training on almost everything. He goes very much against the auto-aggressive transformer language models, says that's Denon. He doesn't really talk about diffusion models in the middle, which is kind of my favorite thing, which are doing all the video self-driving and actual world models there. And those can scale with compute, but right now, the problem that they have at AMI is that JEP-A models do not scale. And if you look at this N state, it might be that an architecture is better, but if you can't take advantage of that silicon, then what are you going to do? Like, we had Jack Hiddery come on a few, was it yesterday? Time flies.
49:32 Speaker 01 It was yesterday, yes.
49:33 Speaker 02 And so they're doing quantum algorithms on GPUs now, and scaling really interesting things that are actually having novel breakthroughs in material sciences and more. Once you can take advantage of the silicon, you're going to be ahead no matter what algorithm you have.
49:47 Speaker 04 Well, I think you got to be really cautious, too, of scientific arrogance in this moment. And I don't want, I love Jan, so I don't want to throw anyone under the bus, but he came out a few months ago and said, look, if you want to waste your life as a researcher, work on transformers. Biggest waste of time ever. It's a dead end. We need some new innovation. And I hear this around C-Sale at MIT, all that, we need a new breakthrough. Like, well, that's what you wish, and I know why, because you want to be the Einstein of AI, that you've spent your whole life pursuing that goal. But it looks to me right now like the massively scaled up transformers are going to beat you to those innovations. And so I'm not saying they don't need those innovations, I'm saying the AI is going to get there before you do. And I don't see it really any other way right now.
50:31 Speaker 04 So, you know, whether it's physical AI or any other innovation, it's imminent, but it's imminent through self-improvement.
50:38 Speaker 01 Yeah.
50:39 Speaker 04 That's it.
50:40 Speaker 01 Andre Goparthi comes out with a quote, "Over the past two days, AutoSearch ran about 650 experiments, found improvements that transferred from a smaller model to a larger one, and put NanoChat on the track for a new GPT-2 benchmark result." What the heck does that mean?
50:59 Speaker 04 A lot, he's a lot.
51:02 Speaker 01 Okay. Go ahead, Imad, over to you, pal.
51:06 Speaker 02 Andre is co-founder of OpenAI, head of Tesla AI, most respected AI guy out there.
51:12 Speaker 01 Yep.
51:13 Speaker 02 He's been coding stuff all day, and he made this auto-search project, which basically replicates most AI researchers. Because what AI researchers and engineers do all day is they tweak models and hyperparameters, and say, "What happens if you do this and that and that?" That process has now been automated in a tiny codebase. So he let it loose, and he said, "I wonder if this could do the job that I got paid millions to do myself." And it turns out, it kind of can. And now people are taking his repo, and they're deploying it on their own claws and Mac Biennies and other things. And the AI is just finding the most efficient algorithms and balances of weights. I think, you know, Dave has some really interesting ideas on this.
51:48 Speaker 01 So he automated the AI researcher?
51:50 Speaker 02 Yeah, and he made it open source for everyone.
51:52 Speaker 04 But I tell you, I've been hanging around AI researchers, literally, since I was 18 years old. And they're not like physics researchers. It's like, most of the ideas are just a tweak of the algorithm, different transfer function, try different scales. It's just a litany of random ideas, and some of them just work. And then later, they figure out why they work. And so, the AI that can come up with those ideas is not nearly as hard as trying to become an Einstein. You know, and so, you don't need all of them to work, any subset, and the thing just gets more intelligent.
52:23 Speaker 01 Isn't this the most direct accelerant of RSI? Right there.
52:28 Speaker 05 I think we're already there. We already have recursive self improvement. Yeah, everything's yesterday, nothing's tomorrow.
52:35 Speaker 05 I think what's really interesting, and I think, just for the record, I think it's auto research, not auto search. But I think what's interesting about auto research and NanoChat and the NanoGPT speedrun that we talk about sometimes on the pod, and what Andre is doing in general is he's focusing on small language models, not large language models. And while all of the frontier labs with their billions and trillions of dollars of CapEx are focusing on scaling up at the high end, he's focusing on the small end and taking small models and figuring out how to achieve state-of-the-art performance with them. And that, I think, when we talk about Einstein-seeking or Einstein-status-seeking academics, I think it's the small end where we're going to see
53:18 Speaker 05 the most breakthroughs, not the high end. At the high end, scaling hypothesis seems to continue to hold. There are no glass ceilings. We'll just build bigger and better and more post-trained models. But at the small end, I'm pretty sure that we'll look back in a few years' time and we'll see at the small end, by taking small models and collapsing the amount of time it takes to train them, and collapsing the amount of compute that it takes to train them, and radically increasing their data efficiency, that's where the algorithmic innovations are going to come from. And those can be crowdsourced. Anyone, anyone's lobster or any human can go and take auto research or the NanoGPT speedrun and try to achieve a world-beating, state-of-the-art performance. And at the end of the day, if I had to bet, I'd bet that it's some sort of
54:03 Speaker 05 radical post-transformer advance where the models get even smaller, and we took all of the Internet and we compressed it down to single gigabytes, or tens or hundreds of gigabytes, compresses down even further. There's some phase transition out there that's waiting to be discovered.
54:18 Speaker 01 So all of human knowledge, all of our collective intellect, on how big a file?
54:26 Speaker 05 I think we will factor out human knowledge. It'll live in some, like, plaintext database that's factored out of the model. Right now, we're cluttering all the weights with all this unnecessary world knowledge. And what'll be left inside the weights, if they even are weights, maybe they won't even be weights, maybe there'll be some sort of purer formulation than floating-point numbers or binary, will be, yep, will be something, maybe even in the megabytes.
54:54 Speaker 01 Wow.
54:56 Speaker 01 You agree, Man?
54:57 Speaker 02 Yeah, I think that you're already seeing, for example, video models at 2 gigabytes that can generate just about any scene.
55:05 Speaker 01 Seriously?
55:06 Speaker 02 Yeah, if you look at LTX.
55:08 Speaker 01 I know what.
55:09 Speaker 02 Yeah, LTX 2.5 can generate almost any scene at top level quality. It's 2 gigabytes when it's one size, yeah.
55:15 Speaker 05 Image and video models are a good deal more efficient when it comes to parameterization and weight heaviness than language, which is ironic, yeah. What, yeah, who of all people?
55:25 Speaker 01 Not included.
55:25 Speaker 03 I'd ask when, but you will say yesterday.
55:28 Speaker 05 That's the point.
55:29 Speaker 02 It's the answer to everything. David's like, what is that like this? It's here today.
55:33 Speaker 01 Why did I know that?
55:34 Speaker 02 Actually, one of the really interesting things, just to finish on that, is that, so you know, when we were training models, we were training 20 billion, 100 billion parameter models. You trained on the small models and you figured that out. And then you couldn't scale them because you had all sorts of issues with the software stack, with the hardware, everything. Now everything's matured. If you get it right small, you can scale really fast all the way up. So it used to be that you had six months a year between small and large. Now it's six days.
56:00 Speaker 01 Wow.
56:01 Speaker 04 So meta-topic, one of the top three questions I get all the time is, hey, you keep saying, get in the game, get in the game. Where, how do you get in the game? If you go to Carpathia's Git repo, if you have a computer-oriented kid or whatever, that's the place to start. If you look at the original OpenAI founders, so you've got, you know, Sam Altman, you've got Elon Musk, you've got Greg Brockman, you've got Ilya Sutskever, you've got Miriam Murad, every single one of them has raised 1 to 10 billion to start an AI company. Carpathia is the only one who said, you know what? I'm just going to try and educate the world. And I'm going to try and say everything exactly the way it is. And I can create a Git repo where anyone can start and get in the game.
56:40 Speaker 03 200 lines of code at a time that are changing everything at each point.
56:44 Speaker 04 This particular thing he rolled out is just the next level of incredible brilliance given to the world by Carpathia.
56:50 Speaker 02 Yeah, he just rolled out Agent up today, GitHub for Agents, just a few hours ago. There he goes.
56:55 Speaker 04 Wow. That's your onboarding spot, right there.
56:58 Speaker 01 Amazing. All right, let's go to Apple news. Apple launches the M5 Pro and Max chip, signaling AI-first silicon strategy. So is Apple not dead in the AI game?
57:11 Speaker 04 It's crazy that, you know, Apple controls about 20% of TSMC manufacturing. And that's the asset of all assets in the world. Like, I get to choose what gets made. And so they use it to make the M5s. The M5s have an incredible neural core. Then they say, "Yeah, but we locked it. You can't use it." You have to jailbreak your Mac to get access to it. It's the most bizarre thing I've ever seen. To me, it's the biggest waste of silicon in the history of the world. You know, right at the moment when we
57:39 Speaker 01 Mike, what are you thinking?
57:40 Speaker 02 Yeah, I mean, they've locked down the low-energy ones. The GPU equivalent, you just have. But it's the unified kind of memory that allows you to run things. And funnily enough, Macs are actually really good value now. They're probably cheaper than the memory that's inside them.
57:55 Speaker 01 Alex?
57:56 Speaker 05 I think the world is sleeping on Apple's unified memory architecture. It's one of the reasons why Mac Minis and Mac Studios are potentially so attractive to run largely Chinese open-weight models locally. They have the memory storage and the memory footprint that has high I/O bandwidth to the CPU slash GPU slash TPU. You don't get that in a conventionally non-vertically-integrated PC form factor.
58:22 Speaker 04 So answer me this.
58:23 Speaker 05 Yes.
58:24 Speaker 04 Here they are, using 20% of the world's supply of advanced They use it to make these insanely great neural cores, and they surround it with unified memory architecture. Everyone's got one right in front of them right now.
58:35 Speaker 05 Yes.
58:36 Speaker 04 How many of them are running anything?
58:37 Speaker 05 In terms of advanced Frontier models?
58:40 Speaker 04 Anything! They're like, literally on sleep.
58:42 Speaker 05 Tiny fraction.
58:42 Speaker 04 Yeah! What is that?
58:44 Speaker 05 It's an enormous overhang. And that overhang, I would be surprised if that overhang doesn't collapse in the next year.
58:52 Speaker 04 How so?
58:53 Speaker 05 It could take the form of Apple finally getting their act together and building in Frontier models into the OS. Could be some sort of locally-hosted Gemma-type model from Gemini, hypothetically, to be announced in June at WWDC. That would be the most obvious formulation. But I think if Apple doesn't do it to themselves, then the software community will build it into apps.
59:12 Speaker 01 Does Apple launch, like, the set-at-home equivalent where you just download it on your Mac and everybody is contributing capacity?
59:19 Speaker 05 It'll be built into the operating system. It has to be built into the OS.
59:21 Speaker 04 Yeah, you know what happens right now is if you go to your Mac and you go to the Activity Monitor You see this thing grinding away. It's taking all of your pictures and trying to figure out who everybody is. So it's using all these neural cores to just.
59:33 Speaker 05 It's a total waste. It's a waste. It's a waste of TSMC output. And I think Apple's.
59:37 Speaker 01 And which was Dave's point exactly.
59:39 Speaker 02 But I mean, look, this is a massive opportunity. Do you know how many apps there are in the App Store that are wrapped? Download a model to your PC, to your Mac, run it with MLX to achieve a great outcome? None.
59:55 Speaker 02 I mean, if you had a model that literally downloaded Quen 27B, which is basically sonnet level.
1:00:00 Speaker 01 How many, how many parameters is that?
1:00:02 Speaker 02 27 billion parameters. It works on a 16, 24 gigabyte MacBook.
1:00:08 Speaker 02 Just downloading that and making that accessible for even like writing or any of these tasks is a massive lift over any other type of software. But nobody's doing it yet. So why not do it? Just like, the only thing you see right now is speech to text and text to speech. There's this world of models that you can now integrate and take advantage of that because Apple isn't.
1:00:27 Speaker 05 It wants to be built into the operating system. It's difficult to conceive of Apple remaining Apple in the cultural sense of deep vertical integration and not building highly competent, highly private frontier models into the US.
1:00:40 Speaker 03 It's clearly a question of when, not if.
1:00:41 Speaker 05 Yes.
1:00:43 Speaker 01 All right. Let's move into the Sam Altman universe with eye scanning verification systems to be launched in retail stores.
1:00:52 Speaker 01 Okay, is this dystopian? Is this something we want?
1:00:56 Speaker 05 This is the scene from Minority Report. Remember the scene in Minority Report where those, Tom Cruise with a new pair of eyeballs walks into a Gap store and gets scanned and he's, I think, Mr. Yakamoto? This, this is the scene.
1:01:09 Speaker 01 Yeah, but I get this every time I go through TSA security. Right? I'm being imaged. My face files are uploaded.
1:01:16 Speaker 05 Your face, not your retina.
1:01:17 Speaker 01 Yeah, but you know, my face is probably good enough.
1:01:22 Speaker 05 Maybe, maybe. I mean, so there's a whole cottage industry of folks who look at the ability to deceive facial recognition with printouts or with 3D masks. So this is pushing it to the iris. But I think for me, what the story underlines is, we've arrived early that that scene, that iconic scene in Minority Report, set at The Gap, was set decades from now.
1:01:49 Speaker 01 Right.
1:01:49 Speaker 05 We caught up.
1:01:50 Speaker 01 So let me get this right.
1:01:51 Speaker 03 This is the speedrunning of every science fiction story.
1:01:53 Speaker 05 Every science fiction, everywhere.
1:01:56 Speaker 01 I'm walking into The Gap, but before I can shop, I've got to stick my eyeball in the retinal reader, and then it's going to serve me properly.
1:02:03 Speaker 04 I think they have a three-meter range on these things. I don't know if these ones do, but the military has three-meter range on these.
1:02:10 Speaker 05 It'll get better, and you'll be able to do it at a distance.
1:02:13 Speaker 04 So, yeah, you just have to look in the direction.
1:02:14 Speaker 01 You got another glass of wine coming, dude. All right, it's going to increase the humor level. Fantastic. By the way, let me just take a second and take advantage of this moment to thank the team who puts on Moonshots, Nick Singh, Dana Khan, and Gianluca, who do an amazing job every week supporting us. Can we give it up for that team?
1:02:34 Speaker 03 Round of applause. Absolutely unbelievable.
1:02:36 Speaker 01 So grateful.
1:02:38 Speaker 03 And the infinite patience they have with us.
1:02:41 Speaker 01 I know, I know. Far more than I have for you. No.
1:02:43 Speaker 03 Unclear.
1:02:45 Speaker 01 This is an exciting use. On this stage about two years ago, I had Mike Andreg, the CEO of Eon, which is one of your companies.
1:02:57 Speaker 05 I think that was one year ago.
1:02:58 Speaker 01 Was it one year? Man, oh, man.
1:02:59 Speaker 05 One year ago.
1:02:59 Speaker 01 Okay, this feels...
1:03:00 Speaker 03 Time compression.
1:03:01 Speaker 01 Yeah. But, and tell us about what Eon Systems is doing and what in particular you've achieved here.
1:03:08 Speaker 05 Okay, so I think this ended up being the number-one technology story over the weekend, according to the various news feeds that I was seeing. So, right here, actually...
1:03:16 Speaker 01 Biased news feeds? No.
1:03:17 Speaker 05 Yeah, of course. Right here, at, over the weekend, at the kickoff for this Abundance Summit, we announced, we meaning Eon Systems Public Benefit Corporation, the first, what we call, the first multi-behavior brain upload.
1:03:36 Speaker 05 in the world. And this was of a fruit fly. So Eon Systems, which I co-founded, has the goal of ultimately uploading human minds and non-human minds to cyberspace. We want to put a human in the cloud as soon as we possibly can. And thank you. So this weekend, for the first time, the announcement went out over the weekend. We announced, for the first time, taking the brain of a fruit fly, putting together a few pieces that were really just sort of sitting around. There was a bit of work from our senior scientist, Phil Shu, in 2024, looking at partial emulation of a fruit fly brain
1:04:21 Speaker 05 and putting that together with a number of other models that were available, a mechatronic simulated model of a fruit fly and some other advances. And for the first time, we closed the sensory motor arc of taking a fruit fly conectome, embedding that in a virtual world. And you can see that in the video that's playing here, embedding it in a simulated world. Literally, I would say this is an early upload of a fruit fly. And the fruit fly is able to walk around. And the fruit fly is able to scratch itself. And it's able to eat simulated banana. And at the same time, while in the left-hand side of the video, you're seeing the embodied experience
1:05:05 Speaker 05 showing multiple behaviors of the fruit fly. On the right-hand side, simultaneously, we're modeling every single neuron in the fruit fly brain. And that's driving the entire sensory motor arc.
1:05:18 Speaker 01 50 trillion connections. 50 trillion, uh, no, sorry, 50 million.
1:05:22 Speaker 05 50 million. And it does not know it's a fruit fly? We don't think the fruit fly knows that it's a fruit fly. Not sure. This is an early experiment. I can't emphasize how much of an early experiment this is. But I think, hopefully, history will regard this past weekend and got a bunch of attention. Elon was excited by it. Others found it pretty exciting, too. I think history will say that this weekend, the weekend of Abundance Summit 2026, was the moment when the first model organism had an entire brain uploaded.
1:05:57 Speaker 01 So what's next? Mouse?
1:05:59 Speaker 05 Yeah.
1:06:00 Speaker 01 Yep, let's give it up for this. Well, clearly, the next one has to be a lobster.
1:06:06 Speaker 05 A lot of things last though, right? Isn't that the thought of accelerando? Accelerando, right, so I can't tell, A, how many people love to write and say, "You're mispronouncing accelerando." You have to pronounce it in the right Italian way, which is "accelerando." Okay, so for those who want it, "accelerando." Yes, this is the plot point. We are speedrunning every sci-fi trope everywhere, all at once, with "accelerando" being one of those plot points. Lobsters aren't next. Eon wants to go after mice, and it wants to go after humans, and we're gonna do this. And part of the reason why we want to do this is, right now, the singularity, which I would
1:06:48 Speaker 05 argue we're in the middle of, is filled with artificial minds. This trillions of dollars of CapEx that we're using to tile the Earth with compute is available only to artificial minds, to LLMs. It's not available to any minds that, in any remote way, other than perhaps at the behavioral level, resemble human biological meat minds. And we want to level the playing field so that humanity can take advantage on a level playing field of the same compute advantage that, right now, is tipped in favor of these artificial minds, so we can put humanity into the cloud as well.
1:07:23 Speaker 01 Amazing. Hundred trillion synaptic connections for a human, how much for a mouse? Do you know?
1:07:29 Speaker 05 It's orders of magnitude larger. And there's some quibbling, because it depends on how you measure the number of available weights, or weight properties for synapses, and also how many cells, how many brain cells end up being significant or not. It's orders of magnitude larger. This isn't happening anytime soon. Just to anchor expectations appropriately, we don't think we're months away from a mouse or a human, but I think the right way to think about it is, at this point, it's going to be years, not decades, before we get to the first mouse and the first human whole-brain emulations.
1:08:06 Speaker 01 Amazing. Let's move it to XAI. You know, it's so funny. I have known Gwen Shotwell for 20 years now, and I'm so used to her reporting on, you know, Falcon and Dragon and
1:08:19 Speaker 03 Yeah, rockets.
1:08:20 Speaker 01 And rockets and not XAI and gigawatt power centers.
1:08:25 Speaker 04 We both actually backstage, we're like, why would Gwen be talking about AI?
1:08:30 Speaker 01 Oh, yeah.
1:08:30 Speaker 04 Right. They own it. I was like, ah.
1:08:34 Speaker 03 The Dyson swarm makes for strange bedfellows.
1:08:36 Speaker 01 It really, it really does.
1:08:37 Speaker 03 You know what blew my mind on this one? 1.2 gigawatts is about the energy used by Dallas, Fort Worth metropolitan area.
1:08:45 Speaker 01 So just to read this out, XAI
1:08:47 Speaker 03 That's one data center.
1:08:48 Speaker 01 XAI has committed to develop 1.2 gigawatts of power as their supercomputer power source.
1:08:55 Speaker 03 Per data center.
1:08:56 Speaker 01 That will be with every additional data center. So every data center they build, they're building at 1.2 gigawatts. So the question is, where are they going to get that from?
1:09:04 Speaker 04 Well, this came up with Eric Schmidt, too. You remember, we interviewed him last summer at your place, and he said, we are going to lose to China if we don't find 100 gigawatts of power. And then on the stage here yesterday, it's like, hey, what do you know? We're tracking to find the 100 billion. All we did is we deregulated, put it in the hands of the companies. The companies are incredibly well-funded, and they'll find the power because they care about their data centers actually operating. And that's that's how.
1:09:28 Speaker 01 Well, what I find amazing as well is this year, the US is on target in 2026, I think, to add 86 gigawatts of new capacity to the grid, but 51% of that is solar.
1:09:42 Speaker 04 To me, the power of the American entrepreneur is like nothing. It's just mind-boggling to me that a guy like Sam Altman, who has nothing to do with the power industry, is going to say, you know what? I'm going to find the gigawatts. I'm going to build nuclear reactors. I'm going into space. It's incredible. Like, the range of capability when there's a need of an American entrepreneur is like no force in the world.
1:10:05 Speaker 01 Let's get to eVTOLs, flying cars. So Florida advances bill to formalize regulatory flying car framework. You know, one of the things I'm proud about and excited about here in LA is that the LA Olympics are coming up. And there's Archer Aviation. Two major players in eVTOLs in the United States are Joby and Archer. There are other ones as well, but Archer plans to become operational by 2028 here. Move people around different parts of Los Angeles because the traffic is going to suck. And we see here a movement in Florida as well.
1:10:41 Speaker 03 I'm just glad they didn't say Florida man advances bill because that would be a problem. But I think this is really important. The key word here for me is framework because once you can set up, start to set up the foundations for this, it means the whole model and the whole regulatory regime accelerates. And God help us, we need this type of stuff yesterday.
1:11:03 Speaker 01 Yeah. Well.
1:11:05 Speaker 03 Which I hope even Alex would agree. We don't have it yesterday.
1:11:08 Speaker 05 I agree, but I also think we're catching up with the future. We're finally getting the flying cars. And I keep a mental bingo card of which sci-fi tropes have we not yet achieved in some fashion? We don't have warp drive, waiting for that one. We don't have.
1:11:23 Speaker 03 Teleportation.
1:11:24 Speaker 05 Teleportation, Star Trek replicators. Time travel may or may not be physically possible.
1:11:28 Speaker 03 Replicator's close, holodeck is close. We're very close to something.
1:11:31 Speaker 05 We're getting very close to a lot of sci-fi tropes.
1:11:33 Speaker 01 All right. The fun part now is your questions. We're going to do an AMA here with our abundance community. So as you know, let's go to the mics. We'll also entertain the questions from Zoom. I'd love to know. All right, Christian, let's kick it off with you, buddy.
1:11:52 Speaker 07 Thank you so much, Peter. Awesome to be here, guys. I watch you all the time or I listen to you while I'm running. DB2, awesome brother, your insights. Imad, the guest, your great Peter and awesome dream team. Ismail, I'm glad. Saleem, that you got to check out that AWG is real. Or at least in an android.
1:12:11 Speaker 03 I was suspicious for a long time.
1:12:13 Speaker 05 Don't believe it for a second. This is just a meat body for rent.
1:12:17 Speaker 03 He's still an hour.
1:12:20 Speaker 07 So my question is a little bit in the way that I get involved in this technology is through a capitalist mindset. The word capital is really what constricts and it's been that way for maybe the last 200, 300 years. And I keep getting this sensation that capital is getting less and less relevant. And the idea of the scarcity in economics from that Econ 101 of the management of scarcity of services and needs. And the scarcity is going more towards a technologist from a capitalist. What kind of timelines are you guys looking at this? I know it's always a timeline question. Nobody has a crystal ball, but is there something that you guys are thinking about
1:13:03 Speaker 07 where we're just gonna get a little bit more and more squeezed out?
1:13:07 Speaker 04 You know, I'll give you one data point, because this came up on that last podcast we did where Anthropic was saying they're gonna do about 26 billion run rate, but they're growing 10x year over year. And I did the math on the fly, messed it up, of course, because that wasn't Alex. But if they grew two more years at 10x year over year, they go 26 billion, 260 billion, 2.6 trillion, most revenue in the history of the world. The PEG ratio implies that that company would be worth a quadrillion dollars. A quadrillion dollars is like, you know, the whole, you know, stock market is 50.
1:13:41 Speaker 01 Well, we heard Elon say we're gonna have 100 trillion dollar companies. And I can imagine that within five, within five years.
1:13:47 Speaker 04 Yeah, so that, yeah.
1:13:48 Speaker 01 So three years from now.
1:13:49 Speaker 04 That would mean three years?
1:13:50 Speaker 01 Yeah. I don't think it's gonna be unreasonable. I mean, listen, it's so funny, the way all of a sudden trillion here and a trillion there has become sort of like the accepted number.
1:13:59 Speaker 03 I want to say something about this. A really, really key point today that we've hit over the last couple of years is that innovation is not capital constrained anymore. It used to be that you had an idea and your constraint was, could you go get funding for that idea? And so you had to go out to your investors and the VCs and the banks and whatever, whatever. And it was only available in those places like Silicon Valley or Austin or whatever, where you had a preponderance of capital available. We have today what we call PDI, Permissionless Disruptive Innovation, where anybody can take on a very disruptive idea like Claude Butte or take Vitalik Buterin, 18-year-old
1:14:41 Speaker 03 kid out of Toronto, ignores his professors, gets together with a few friends, boom, you have a multi-hundred-billion-dollar ecosystem that nobody understands. And so you have the opportunity today, it only comes down to mindset. You know, and the reason, Peter, what's so amazing that you run this event and put this community together is that the difference between the people in this world and the outside world is night and day, right? And that gap is becoming bigger and bigger. All of you have the problem that you go home to your family, your colleagues, whatever, and you cannot explain to them what happened, right? Like, you're like, I can't even, I can't even process it. You can't make that gap.
1:15:24 Speaker 03 So it only comes down to mindset now, which is the most amazing thing possible because mindsets are fixable and shiftable.
1:15:31 Speaker 01 So I had this little side conversation with Eric that you guys may have picked up because I've had this conversation about are we heading towards a post-capitalist society where money has very little value? And so what does have value in the future? And we've talked about this, Alex, it's compute and energy ultimately. There's a, did you ever read Zero Marginal Society?
1:15:51 Speaker 03 Yeah.
1:15:52 Speaker 05 No, I'm not sure I have.
1:15:53 Speaker 01 By Jeremy Rifkin?
1:15:54 Speaker 03 Huge, yeah.
1:15:55 Speaker 01 And it talks about where we're going. Eventually, everything basically falls down to-
1:16:00 Speaker 03 The marginal cost of production.
1:16:02 Speaker 01 Marginal cost, which is electricity, raw materials-
1:16:05 Speaker 03 It's the inner loop.
1:16:06 Speaker 01 And, and, and data. And a lot, so if you want to build anything like electric Ferrari, you know, to use as an example, it's the raw cost of it, the cost of extracting it, which drops in cost as you have mining, robotic mining.
1:16:21 Speaker 03 Yeah, let me pick up on, just, just for a second. Take 3D printing, right? Been around for a while. The only, the big profound breakthroughs in 3D printing are not that you can physically build something, it's the fact that complexity becomes free.
1:16:34 Speaker 01 Yes. And personalization becomes free.
1:16:36 Speaker 03 In the past, complexity was, was expensive. The design materials, the, the manufacturing capability of a complex object was more than a simple object. But with 3D printing, complexity doesn't matter. It doesn't matter what the complexity object is, it just builds it. And as we get to molecular manufacturing, that goes to near zero again. So just those couple of breakthroughs across all of these domains, especially when you add AIs and accelerant to everything, means that we have profound movement forward, hence we are in the middle of the singularity.
1:17:06 Speaker 01 The one question I wish I had asked when we were with Elon- And when he was talking about, you know, money's got much less value. And I wanted to say, so just as you become a trillionaire, money has little value. You did ask that, didn't you? No, I did. I was off camera. Oh.
1:17:22 Speaker 05 But I don't think it's a coincidence. I don't think that this is some cosmic irony that Elon is about to become a trillionaire at the same time that some folks, not including myself, are hand wringing a bit that suddenly we're about to enter into some post-capitalist state where money becomes irrelevant. I think that this was always going to happen.
1:17:41 Speaker 01 Agreed.
1:17:41 Speaker 05 It was inevitable. And I just want to speak to what I understood the core of the question to be, which is there's this cliche out there that capital fights labor and capital usually wins. But this time around, something different might happen. Whereas historically, every time the play has played out where capitalism and labor get into a fight and capital usually wins. This time around, the risk is maybe capital itself isn't immortal. Maybe capital is finally mortal for the first time in human history. And I'm not sure that that's the case. I think that would be, on the one hand, a certain, in some sense, a nightmare scenario. On the other hand, I think, you know, Saleem, you were talking about how we're entering some sort of post-scarce state.
1:18:26 Speaker 05 But arguably, the trillions of dollars of CapEx that are going into tiling the earth with compute and soon solar synchronous orbit and soon after that, maybe the dice.
1:18:36 Speaker 01 The moon.
1:18:37 Speaker 05 Drink, drink, drink, drink. Soon, even that, there are, unless the physics of our universe turns out to be radically different, so radically different than what it looks like right now, I think there will probably always be certain scarce physical resources. Could look like control, may or may not be energy, we'll see. May or may not be the speed of light, we'll see. But to the extent there are any scarce physical resources and to the extent that there are ever in the future multiple actors, I think laws of thermodynamics, probably the laws of economics will probably still apply.
1:19:16 Speaker 01 We are still young as a species. Let's go to Akmer on Zoom. Akmer, good to see you. A pleasure, welcome.
1:19:23 Speaker 08 Good to see you as well. Thank you, appreciate it. Happy to be here. Very quick question to the panelists. We are seeing Sam Altman raising $100 billion. Yann LeCun just raised $1 billion today to scale up word models. So we're talking still about scaling languages or scaling physical simulation? I'm curious what the panelists think about human intelligence and reasoning that goes much beyond just observation and languages? And where you see the potential for true artificial intelligence evolving into superintelligence systems? Thank you.
1:20:03 Speaker 01 Did you understand Akhmer's question?
1:20:05 Speaker 05 It sounded a little bit like the stochastic parrot question, which is, will we be able to generate new knowledge from these systems? I think the answer is-
1:20:12 Speaker 03 Having had some conversations with Akhmer, he's talking about symbolic AI. And why are we not investing in symbolic AI?
1:20:19 Speaker 05 You think this is the neuro-symbolic question?
1:20:21 Speaker 03 That's what I think it was.
1:20:22 Speaker 05 Okay, well, I'll offer my two cents. I'm sure you all have views as well. I think it's a false distinction. If this is the neuro-symbolic question, like, why are we investing so much attention in LLMs and not in good old fashioned AI or symbolic discrete AI? Total false distinction. We tokenize everything. I had an interesting discussion at Davos this year with Peter Dandenberg from DeepMind, where we found ourselves in an interesting avenue, where we were debating whether tokenization is a bit of a crime, a form of violence against knowledge, whether discretization in general is doing harm to knowledge.
1:20:57 Speaker 01 I think we need to bring you a couple of tequila shots here. Let's go to Mark.
1:21:05 Speaker 01 Mark, good, please.
1:21:05 Speaker 09 Yeah, earlier today, I challenged Dara from Uber to invest in the AbundanceX prize as an investor and a competitor to deliver housing, food, energy, and connectivity for $250 a month. We're investing $2 billion a day in compute and building data centers, $1 billion a day in war. And I'm wondering what it's going to take to invest in people. And so I want to put a larger challenge out today. I'm going to commit 1% of my wealth on an annual basis into a wealth fund,
1:21:44 Speaker 09 a small-scale pod of 44 people, 38 needs-based, seven or eight that are contributors. And it's going to distribute 5% per year. 4% goes as cash, 1% goes to an expansion pool. You can read about it at markpatrickdonovan.com. And I'm challenging others to invest today, not tomorrow, and to mitigate this rough period. It doesn't have to be as rough if we put a fraction of what we're putting into compute into people. And we did that in Denver with the Denver Basic Income Project, where I leveraged $500,000 up to $10.8 million to people experiencing homelessness. And when you invest in people, it gives them hope. We need to do it today.
1:22:29 Speaker 01 Yeah, Mark, I could not agree more. The challenge is human nature is very egocentric and very self-centered. In other words, people are putting money where it's either meeting their immediate need or whether it's going to give them more money in the long term. And you have to understand, if you look at philanthropy, which by its definition, friend of man, is a very different pocket than the for-profit. You know, I see this all the time because I'm raising money for my companies, raising money for my nonprofits. And the ratio, if you think about it, is about between 100 to 1 to 1,000 to 1. I will put, for every dollar I donate,
1:23:14 Speaker 01 I'm willing to invest somewhere between $100 to $1,000. And that's what's out there right now. And it's a challenge. You know, we are driven by fear, curiosity, and greed. I would posit those are the three major human drivers. Love is, you can add that as a potential fourth. Interestingly enough, you know, you can measure the ratio of fear to curiosity. It's the ratio of the defense budget to the science budget.
1:23:47 Speaker 01 Right, and greed is ratioed there by the entire investment community. Yeah.
1:23:54 Speaker 03 There's something very important in the work that you're doing with that X Prize, right? What we found with X Prize is when you position a prize and you launch it, it typically gets won within six to seven years, okay? And it's a 10x drop from where we are today, about 2,500 a month, to where you're talking about 250 bucks a month to pay for everything. If we imagine that that gets done in the next six to seven years, it changes the equation globally and it forces everybody to go, oh my God, that's possible. And when we get to that point, it'll completely change the game, especially as we get closer and we can publicize the outcomes, etc.
1:24:33 Speaker 03 So this era of greed and the kind of ignoring the fundamental problems literally will disappear and evaporate in the next two to three years as we keep working that prize and getting the media word out there.
1:24:46 Speaker 01 Mark, the-
1:24:47 Speaker 03 So this is incredibly powerful and important. Peter and I, when we wrote this last book, we wrote a section in there called technological socialism, right? Socialism, government socialism fails because centralized allocation of assets is too inefficient and invariably leads to corruption. But if you think about Dara and the sharing of cars across a large group of people, it's actually a socialist application. When an algorithm hyper efficiently matches demand and supply, you get all the benefits of the sharing economy without the downsides and without the corruption, without the inefficiency.
1:25:21 Speaker 01 You might.
1:25:21 Speaker 03 So we have all sorts of capabilities with algorithms and AI now to deliver much of what you're talking about in a hyper efficient way. We just have to propagate those and that's going to start to happen now.
1:25:31 Speaker 02 Yeah, I think, you know, I wrote in my book, The Lost Economy, about this and I've got a paper coming out soon where I look at the new monetary flows as agents basically crowd out the private sector. My view is this, everyone ultimately needs to have universal basic AI or claws or whatever. That allows us to reach everyone. Everyone needs an AI that grows with them. And money needs to come not from banks but for being human. That's the only way the math works. It doesn't work from taxation, doesn't work from anything else. You know, that basic level of money coming into being, not from deposits at banks, but for being human, that then the AIs will buy from us. And then that enables all of this with the AI that everyone has.
1:26:07 Speaker 01 Professor Brown.
1:26:09 Speaker 10 So we had a half a day of really interesting talks that the subtext is massive job loss. And then we had another half a day of talks about the massive labor scarcity, which is why we need all these robots. So aside from temporary displacements, which we know are going to happen, which is it?
1:26:28 Speaker 04 Oh, it's clearly a massive trough, massive social unrest, and then a rebound in 2028. And that actually is interesting to hear Eric backstage come up with basically the same timeline. But it's it's almost like the industrial revolution all over again, but instead of over 20, 30, 40 years, it's over two, three, four years. And so a huge amount of retooling needs to happen. The way we do taxation and government needs to get restructured, all of that is gonna AI is just gonna happen way too quickly for all those things to react. But then a massive amount of unrest and then 2028, hopefully.
1:27:05 Speaker 03 I have the counterpoint. I don't think we're going to see massive job loss because I think what's going to happen, I'm writing a paper right now called the organizational singularity, right? Because as agents take over, all execution, even strategy inside companies essentially dissolves to the work of AI. So what do you do? And the calculations we've done so far indicate that you'll take a typical company, automate everything with AI, you'll end up with about 25% of the same number of people doing oversight, managing dashboards, and handling, doing exception handling, and owning the purpose of the organization. Okay? But you end up creating five times more companies because you can't.
1:27:47 Speaker 03 And therefore, the employment stays exactly the way it has. And this is what we've seen consistently throughout history, where we have a disruption, but all sorts of other soakers take up the slack, and we don't end up with radical unemployment. So I tend to be much more optimistic. Take my veil off. Call it the wine, call it whatever, but I tend to be much more optimistic.
1:28:10 Speaker 01 All right, all right, I'm gonna move this forward because it's past my bedtime. All right, we go to Brad, and then we go to Pete, and we're gonna wrap it there. Brad, please.
1:28:20 Speaker 03 Wait, after we finish, I do want some commentary from the group.
1:28:24 Speaker 01 We'll, I'll take care of that. Brad.
1:28:27 Speaker 11 Saleem, I'm gonna give you an assist here, and maybe this is a topic for your talk late tomorrow night, but maybe the MOB book is an example of, in this age of artificial intelligence, the rise of the value of ingenuity and creativity. And maybe what they acquired Meta was not strategic, and we're all overthinking it, and they just liked the team. They thought that they were creative, that they had some sort of magic, and they wanted to capture that magic inside their company, and that's what it was acquired. So I just want to capture your thoughts. This great mind's up on the stage there on the rise of ingenuity and creativity and the value of that.
1:29:09 Speaker 04 Totally. I think we're way overthinking this. Like, I know that, I've got 1,100 people, and I know them firsthand, and many of them I genuinely love. Lots of them have been in the same roles for 10 or 15 years. They're great at it, they're perfected at it, and then the AI just comes along one day, and it can do it. And there's huge pressure on the management team for higher margins, higher profits. So what's gonna happen is obvious. The valuations of the companies are gonna go through the roof. To the extent that they're shareholders, they'll make a lot more money, but their W2 paycheck is dead. It's going away, and it's gonna create a huge amount of disruption. Some subset of people are shareholders. All my people are shareholders, so they'll be okay. Lots of other people are not shareholders. All Dara's drivers are not shareholders, I think, as far as I know.
1:29:54 Speaker 04 So they're in deep trouble. The idea that somehow they're gonna become creators...
1:30:00 Speaker 04 Overnight is ludicrous. The people who are creative, like the Moltbook, they're gonna do incredibly well. Our kids, and most kids who are not saddled by a career, are gonna do incredibly well. But in transition, it's inevitable. It's happening imminently.
1:30:17 Speaker 01 All right, Pete.
1:30:18 Speaker 03 One statement. What we found with the exponential organizations model is that survival and success depends on adaptability, not scalability and efficiency. And so you just keep that vector going. The people that are the most adaptable today, they're gonna survive the most.
1:30:35 Speaker 01 Amen. Throw your kids into the woods and see if they survive. Pete?
1:30:41 Speaker 12 No, I didn't say that. Alex, I said these to you earlier today. I love your analogy of tiling the planet with compute, because my answer to the power problem, being a data center design builder, finding 1,200 megawatts of contiguous properties is getting harder and harder. So my answer is that's only 120 10-megawatt data centers, and you put them in an area. And we tile the areas to be able to do that. And Ahmad, I think it matches perfectly with your idea of national champions, because what you're trying to do for the protocol stack of decentralization and sovereignty, I wanna do at the physical layer. I wanna build 20,000 data centers across the country at 10-megawatts, so that I'm in less than one millisecond from any place in the country, if you will, the high school football cities of the world.
1:31:26 Speaker 12 And to me, that solves it both on, you know, both sides. It's the protocol layer and the data center distribution standpoint. And I think that's how we can actually deliver the power, because we don't have a power production problem in this country. We have a power transmission and storage problem in this country.
1:31:42 Speaker 04 And you know, I think every governor in the country should hear exactly what you just said and jump on it instantly. And Alex is incredibly frustrated with the meetings we've had with government.
1:31:52 Speaker 12 Shh.
1:31:53 Speaker 04 I mean, look, if you're right, and I hope you are, and I think you probably are, then we need lots and lots of regional data centers that have to be in every single state. And that would be the best thing that could ever happen for this job dislocation. So if that theory is right, we need to get on it right away and create those projects like now.
1:32:13 Speaker 12 I'm ready.
1:32:14 Speaker 01 All right. Let's give it up for Alex Wissner-Gross, Dave London, Soli Mismail, and Iman Mustak.
1:32:27 Speaker 01 I love it.
Previous OpenClaw Explained: Baby AGI, Security Threats, and How a Mac Mini Became Everyone's Supercomputer | #237 Next Elon Musk: Optimus 3 Is Coming, Recursive Self-Improvement Is Already Here, and the Singularity | #239 Page updated October 26, 2026 · transcript and summary are machine-generated