It sort of feels like we’re in the midst of continuous event horizons. >> We all agree on this podcast that we’re right in the middle of the singularity at this moment. >> I think it’s increasingly likely that the singularity is an optical illusion. It’s an optical illusion that appears at a distance. It it looks like a vertical asmtote, but when you’re in the middle of it, as I I increasingly suspect we are, feels quite continuous. If you froze technology today and just assimilated what we invented in the last two years, it would take decades to to realize all the implications. QPT5 Pro sets record at frontier math. We now have clear line of sight to to solving all of math or substantially all of math as we understand it in 2025. Now with AI, >> how do we navigate this this future? Because we can see it’s coming now. So what does that future look like? And let’s start painting that picture. What does it mean when math is soft? What’s the implications for, you know, all of our subscribers here?
[00:01:02] >> Now, that’s a moonshot, ladies and gentlemen. >> Everybody, welcome to Moonshots. Another episode of WTF Just Happen in Technology with my incredible moonshot mates, Salem, Ismael, Alex Wezner, Gross, and Dave Blondon. Gentlemen, good morning. Exciting day here. >> Morning. >> Good morning. Oh, yeah. >> You know, we’re we’re recording this at 6:30 a.m. at least Pacific time. And I remember a couple minutes ago, we’re talking and and uh Salem, you were saying, “Oh my god, you’re getting up so early.” And Alex, what was your response? >> This is the slowest it’ll ever be for a while. >> Yeah. And there’s no >> You said don’t sleep through the singularity. >> I think that was it. >> Honestly, it’s like every day waking up more excited than the than the last. Uh and it’s just fun. So to our subscribers and listeners, uh we’ve spent the last uh four or five days gathering articles that are in our mind the most significant things going on. Um the
[00:02:01] speed is accelerating. We’re going to actually close this podcast conversation with a discussion about what is Ray Kros singularity that we’re going to hit by 20 245. What does it actually mean? Because it feels like we’re hitting it a lot earlier. Uh Dave, what’s been on your mind this week? Actually Ray was at MIT making that presentation that we’ll get to at the end of the pod. So I’ve been thinking about uh the the specifics of the timeline from here to there and then this incredible compute shortage. You know we’ll be in Riad next week. Uh and you know that’s data center central right now. >> So thinking about that a lot too. >> Yeah. This week is X-P prize visionering. Uh we’re going to have the moonshot mates uh in Malibu uh LA. If you’re interested in joining us at the X-P Prize Visionering, which is where we debate and we discuss all of the competitions, what we should be launching next, uh we’ll drop the uh uh the link in here. You can join us there
[00:03:02] uh and just go to >> x.org. Arguably the most important conference of the year globally. >> Yeah. >> Just because trying to solve which problems do we all want to go about solving is such a critical thing today. >> Yeah, I agree. And I think the order the order of operations matters a lot especially for people investing or career planning in this area and you know that the timelines are becoming more concrete now and so I think we’ll we’ll really cement our ideas a lot next week and then Alex will refine it into the perfect message for the audience. >> Yeah, a agreed. So if you want to join us uh again we’ll drop uh the link in the chat notes below. Uh Sem, what’s on your mind this week? you’re going to be with me in uh in Malibu at Visionering and then Dave, you and I are off to Riad for FII9, the future investment initiative. A lot of AI conversations happening there. >> We’re pretty much all there, right? I mean, uh huge conversations going on. the the the it looks like the big
[00:04:01] dominant conversation will be how do we use AI to solve everything to Alex’s points that he repeatedly makes on this pod cuz now we can apply AI as a tool to any of these domains. It’s huge. >> Yeah, Alex, how about your week? I’m sorry you’re not going to be with us. Uh but hey, I’m sure you’re busy. >> Yeah, it’s been an exciting week. I I think arguably and we’ll we’ll get to it. One of the most exciting developments over the past weekish was the solution of math. I I would argue that we now have clear line of sight to to solving all of math or substantially all of math as we understand it in 2025 now with AI which then topples physics and chemistry and biology and what what did you say to me? We’re going to have we’re going to have an accelerated uh accelerated play of Star Trek. It’s like it’s all going to happen. We’re we’re speedrunning Star Trek over the next 10 years. It’s not the 24th century. It’s more like 2035.
[00:05:00] >> Crazy. Crazy. All right. So, hold on your seats, everybody. >> So, the future is collapsing into the present basically. >> Linear versus exponential. See? >> Yeah. Yeah. Seriously. All right. I I added some slides here at the beginning to talk about the speed of change because I want everyone to understand this. Um and we’ll begin with this image here uh which is the AI the adoption of AI is now eight times faster than we saw uh with the internet years right so we went from zero to 200 million users uh in AI in you know in 1/8 the time it took for us to get there in the internet years. Um any comments on this? Well, this slide is understated, too. That’s showing chat GPT alone against the entire internet. And so, if you include Gemini and the other engines, it’s well over a billion on that left chart. Uh, so yeah, it’s it’s even more acute than this chart makes it look.
[00:06:00] >> I think this is likely to be in keeping with the notion that this is the slowest that things are likely to be for for some time to come. This is actually still pretty slow. I it maybe superficially one can look at the AI curve and say okay well we deployed super intelligence upgrades and reasoning models to a chunk of humanity over a few years it’s actually still pretty slow we don’t have yet a conduit for deploying physical world upgrades to to most of the world chat GPT obviously rode on top of prior platforms like the internet and personal computers and and smartphones but we don’t yet have a conduit for deploying material or physical changes to the world I think it’s going to look robotics, nanotechnology, a few other key technologies. I think things will actually be moving pretty quickly once we have those conduits that we don’t really have yet. >> Like you said, >> your point your point being that once we have 5 million robots out there, then you get an instant upgrade to everybody. We don’t have that for humanity >> or 5 billion. >> Let’s let’s make sure we come back to that in a future pod because I think um the the constraints to manufacturing are
[00:07:01] going to hold back robotics and there’s a separate curve for that. But then you have the nanobots which don’t require a lot of material and the nanobots are you know from a medical point of view are going to be massively impactful and I think those are actually going to come sooner than people are predicting uh because they don’t have the the you know the component supply chain bottlenecks that the you know the robots >> we don’t hear a lot about nanotechnology in the classical Eric Drexler means we hear about wet nanotechnology in terms of uh DNA origami and so forth but the idea that you can build an assembler uh a m a you know nanocale robot subcellular robot that’s able to pluck atoms of different types and build materials out of pure carbon sort of diamondoid materials is a is still a few you know few years out but I think it’s going to be arriving on the back of AGI and what follows um I’ve seen a team that seems to have a viable credible line of sight path to molecular
[00:08:00] manufacturing Peter Yeah, I I think that’s likely. You know, you’re going to see later in this pod, you know, Greg Brockman designing chips using AI and like what the heck does Greg Brockman know about designing chips? But with the help of AI, you know, anything is is possible. Very similar to Demos, you know, solving protein folding, but like how how does Demos know anything about protein folding? Well, with the help of AI, anything becomes possible. So I think you know these areas like molecular manufacturing and nanobots are going to come very soon from unexpected places from early adopters of the tools that are that are tuned to the problem. Every week my team and I study the top 10 technology meta trends that will transform industries over the decade ahead. I cover trends ranging from humanoid robotics, AGI and quantum computing to transport energy longevity and more. There’s no fluff, only the most important stuff that matters that impacts our lives, our companies and our careers. If you want me to share these metat trends with you, I write a newsletter twice a week, sending it out as a short two-minute read via email.
[00:09:00] And if you want to discover the most important meta trends 10 years before anyone else, this report’s for you. Readers include founders and CEOs from the world’s most disruptive companies and entrepreneurs building the world’s most disruptive tech. It’s not for you if you don’t want to be informed about what’s coming, why it matters, and how you can benefit from it. to subscribe for free. Go to dmmandis.com/metatrends to gain access to the trends 10 years before anyone else. All right, now back to this episode. You know, one of the comments we get sometimes is, “Okay, you guys are wealthy.” Well, hey, you know, none of us started that way. At end of the day, we created some wealth. But here’s the point I want to make for everybody who’s listening who’s struggling. And there are people who listening are struggling. These technologies are massively demonetizing at, you know, we going to have autonomous cars that are four times cheaper than owning a car. We’re going to end up with nanotechnology where if, you know, and this sounds insane until it all materializes. If I have a nanobot, um, I can literally throw that
[00:10:02] nanobot in the ground and say, “Muf anything out of raw materials.” you know, the information set is free, the energy is around, and you’re basically plucking atoms of whatever you need. I mean, that’s how an oak, you know, an an oak tree starts from a an oak seed and grows over time and just at a very slow pace. >> Can I mention a couple of things about this? >> Yeah, please. >> You know, you you write one of the things you pointed out in abundance, right? Is that is that if you went back a few generations ago, the richest people in the world exclusively had inherited their wealth. And today you look at the richest people in the world and exclusively they’ve earned their wealth. They went from zero to to everything. Uh this is not a this is a mindset problem more than anything else. You take Vitalik Bhutarn, 18-year-old kid out of Toronto, ignores his professors, gets together with a few friends, and boom, you have a $600 billion ecosystem that you know nobody understands. Uh so there’s this unbelievable potential to go from zero
[00:11:01] to everything. And this has never been true before in the history of humanity. >> So I’ve had so many friends using >> that’s completely the mindset that gets you there >> using AI to say okay just going deep and saying this is what I love doing. How can I start a business here? Right? How can I, you know, uh what do I do first? How can I learn about this? And it really is making a commitment to yourself to use these tools to educate yourself and then to build on top of those tools. >> Yeah. Can I give a crazy example? Sure. >> Um, uh, a friend of the probably we all know, uh, I was talking to him. I I won’t use his name just I’m not sure he’s comfortable with it, but he he told me that last month he launched 47 startups with using AI. >> Okay. Just in a month, he and his team just coded and pushed on. That’s just insane. That’s unbelievable. >> Yeah, it is. Uh, all right. So, speed is going fast. Here’s another article showing sort of the speed of change. AI
[00:12:00] content overtakes human uh content online. So this is fascinating. This is just over the past five years, you know, it used to be that 100% or near 100%. uh we had AI uh AI’s writing articles for certain uh for certain magazines as early as the you know late ‘9s I mean so late 19 uh 2018 or thereabouts but we dropped from 100% human content to below 50% and AI written content is exploding uh Dave or or Alex what do you guys think about that >> huge deal I mean there’s enormous opportunity to what you were saying a second ago Peter uh the ability to use AI to create new content is a business opportunity and a life-changing opportunity for everybody. The tools are incredibly democratized, easy to come up the curve. The AI will explain to you exactly how to use Sora 2 or or whatever, you know, V3 and uh just a huge amount that you can do with this. You know, uh, if you think about all the friction that people have in life, if
[00:13:01] you if you’re operating locally, you know, you’re a local government or you’re a local, you know, business owner or whatever, the interface, putting an AI interface on your business makes it dramatically more usable for all of your customers. That alone, it can be a life-changing business opportunity. So, you know, anyone who’s a reporter, who’s a an editor, a video creator, whatever, this is this is such lowhanging fruit all of a sudden. And and it’s not all slop, you know, this is highquality capability. It’s used for slop a lot. I’m not going to deny that. >> But but the ability to to create genuine highquality content with these tools that’s much more compelling than just writing an article would have been it’s right in front of you. >> Alex, what are your thoughts? >> Maybe add there is this cliche out there that we’re just going to drown in AI slop. And I I don’t buy that for for one second. If you rewind 20 years or so, there was a cliche that we were going to drown in email spam and that also did not happen. >> Agree. >> Better filters get brought into
[00:14:00] existence and ultimately the the same tools that would empower uh spammer also empower the ability for small individuals to have basically their own electronic printing press and to disseminate their ideas to millions of people via email campaigns that people subscribe to. So I I I don’t subscribe to the notion that when we look at a chart like this that it just means humanity is drowning in online slop. I I think if anything the there are sufficient economic motivations for the quality of slop to ultimately disrupt from below innovators dilemma style the quality of human writing and if anything I think we end up merging with the slop. >> Well I mean listen one should not just take whatever their AI writes and publish it. You mean you should read it, make sure it represents you. I mean, you should be the originator of the basic idea, and use AI to help level up the content you’re producing. Uh, here’s, >> you know, we’ve had synthetic data,
[00:15:01] aren’t we now entering the world of synthetic culture? Don’t doesn’t this become like a hall of mirrors where the everything is reflecting on what is what’s happened before and just amplifying itself in a totally weird way? This is going to become completely unpredictable, is it not? I I think that principles is that culture was always natural. What is how do we distinguish between natural versus synthetic culture to begin with? >> Dave, >> I think Alex’s point on uh on spam filters is really important here, too, because the ability to have your AI friend agent filter the content and reduce it to the subset that you care about. It’s so easy compared to, you know, the spam or so much more powerful than the spam filter version of that and it works perfectly fine. I think when Tyler Cowan came out with his new book, one of the things he said when he launched it was 99% of the readers of this book are going to be AIS, not people. So I designed it to maximize the impact on the AIS and the AIs are going to summarize, translate and feed it to the humans. >> It’s the new >> So that’s kind of the future of writing.
[00:16:00] >> That’s so important to realize that when you’re writing, you know, the major impact you’re going to have on the world is through an AI interpretation of what you’ve written. That’s amazing. there people are using the LLMs to essentially run SEO. So they’re feeding these things with uh with great images of their company and and and stories about their companies etc. >> So this next article large language models are for you know improving their forecasting ability is fascinating. Uh there’s this thing called a super forecaster. Uh I remember reading about this years ago right? It’s a it’s a person who can demonstraably and consistently uh make accurate predictions about the future as compared to the general public. And there’s a terminology there. And so it looks like, you know, GPT 4.5 is now scoring uh very close to these human super forecasters and it’s likely to exceed the best human super forecasters by by late 2026. Alex,
[00:17:02] thoughts on this one? >> Yeah, so maybe first a bit of background. This is a benchmark called forecast bench by the forecasting research institute. Consists of 500 constantly updated binary questions, yes no questions that look something like, will the following happen by the following date? Yes, no. That can be automatically verified. And when I see an experience curve like this, my mind immediately goes to sci-fi writers like Ted Chang and Frank Herbert who who’ve written extensively about what happens when AI or superhuman intelligence can predict the future to ultra high accuracy. Like what does civilization look like when we can predict things that are right around the corner? I I think arguably if we can predict the future of civilization, we can also steer the future of civilization. Uh, and this isn’t just sort of a centralized steering mechanism since everyone has access to First Order to to GPT4.5. It’s not some sort of like centralized command economy type future.
[00:18:01] Imagine a future where everyone has the ability to predict the the future of markets of of social outcomes and then if you can predict it, you can steer it. You you can optimize outcomes. I I think that’s what we find ourselves in in a few years. >> Yeah. You know, Sealem, we you and I talk about linear to exponential. Uh, and you want to take a second and digress to what that means. >> Yeah. I mean, look, you talk about this a lot in all your presentations, Peter, right? Like if you went back 100 years ago, every anything important happened within a day’s walk. And today, something that happens around the world hits us in seconds. And it’s really hard to get our heads around this because 4 billion years of evolution has guided all of our intuition, training, education about the world to be linear. For the last few decades, if you were running a business, you took your past performance, you drew a line as to where it might be in order to predict the future. But we’re entering this exponential phase. And I love the example you use of of so if you take a piece of A4 paper or 8 1/2 x 11 or A4
[00:19:01] paper like this, it’s like.1 mm thick. If you fold it, it becomes 2.2. If you fold it again, it becomes point4. Here’s a thought experiment for everybody. How thick is it if you fold it 50 times? Okay. And this is a very very unintuitive question. >> And it turns out at like full 20 you’re the size of a football field. At full 38 uh you’re the around the earth. And at the 50th fold, you’ve reached the sun >> 93 million miles. >> Now granted, it’s hard to fold that 50th time. It’s it’s pretty small at that point. But the very very very few people maybe Alex would get to that answer right away. Everybody else is going well I think it’s about this big. I think it’s about this big. Maybe it’s the size of a room. Going to the sun is a very very big difference than going to the whatever idea. And yet the world is running on this dynamic in this heristic. >> Yeah. We’re running we’re running on linear mindsets in a world that is growing you know at a hyper exponential not just exponential these days. Well,
[00:20:00] just uh some very practical advice for all my nephews and family out there. Uh the data behind this comes from Metaculus, I believe. Is that right, Alex? >> This is for the this is for the forecasting. >> Half of the 500 binary questions come from markets including Metaculus, but there are other markets as well. And the other half come from Wikipedia and other time series sources. >> Oh, interesting. Okay. Well, everybody should check out Poly Market, Metaculus, Kelshi. These are these are the prediction markets where you can actually invest or bet on future events. And they’re growing like wild. They’re they’re becoming very valuable companies, but they’re part of this new refactoring of the economy where you have prediction markets, you have AI forecasters and benchmarks. And then later in the pod, we’ll talk about new exchanges. And you know, this this whole process of investing and creating has worked really well in America for 100 years or more, but it needs to accelerate like crazy. And this is part of part of that acceleration. So we’ll we’ll probably follow up on this in more detail, but Alex was a huge early
[00:21:00] adopter of originally Metaculus and Poly Market and and brought it to my attention, but now I’m trying to bring it to everybody out there. Just go check them out and see what’s happening there. >> Shout out here to uh Ralph Merkel who created Merkel Hash Trees, which is the basis of Bitcoin and the encryption there. Uh he’s also one of the world’s top nanotech experts. But a few years ago you wrote a paper where he suggested that the poly market prediction markets are going to be the future of democracy because you could do policy formulation using prediction markets. It was a really profound idea. Fascinating. And and Ralph’s been a member of the Singularity University faculty uh from the inception. All right. Uh let’s move on to the AI wars. Uh but before we do that um here’s a sound bite from Sam Alman. Uh and I found it fascinating. Uh it’s AGI won’t feel like the singularity. Let’s take a listen. >> We talked about the terran test. AGI will come. It will go whooshing by. >> The world will not change as much as the
[00:22:01] impossible amount that you would think. But one of the kind of like retrospective observations is people and societies are just so much more adaptable than we think that you know it was like a big update to think that AGR was going to come. You kind of go through that. you need something new to think about. You make peace with that. It turns out like it will be more continuous than we thought. >> So, what do you guys think about that? You know, we went whooshing through the touring test, didn’t notice it. Uh, Alex, you’ve you’ve sort of argued that we’re at AGI right now and didn’t notice that. But what are your thoughts? >> It’s m maybe 5 years in our past, uh, 2020 or so. I I I think it’s increasingly likely that the singularity is an optical illusion. It’s an optical illusion that appears at a distance, it it looks like a vertical asmtote, but when you’re in the middle of it, as I I increasingly suspect we are, feels quite continuous. Uh that rapid change, if you follow it closely enough, actually just
[00:23:00] feels completely smooth. And I I almost it it’s sort of ironic that the notion of singularities in in math and physics evoke black holes uh and and relativity. And there there’s I I almost want to draw relativistic metaphor that the singularity perhaps only appears from an outside observer’s reference frame. Maybe from the reference frame of 1900 or so, it looks like a singularity. But if you’re right in the middle of it, spacetime is perfectly smooth. >> Fascinating. So you agree with Sam Foley? >> Yeah. >> Yeah. I >> I have no I have no indication that that this is not the case. >> I have three quick comments, >> please. Number one, I have my normal rant on what the hell do we mean by AGI because as last count at last count there were 14 different definitions. So, leave that to the leave that to the side. Um, I really do agree that we’re in the middle of the singularity and it it looks like normal spaceime. We really are in the middle of it. It’s been something and I think let’s talk about it at the end when we get more into what we mean by this. >> We’re going to debate what the singularity means at the end of this
[00:24:00] episode. Yes. What does what does Ry mean by we’re going to reach it in 2045? All right, Dave, what are your thoughts on this? >> Well, I love the fact that we all agree that we’re right in the middle of the singularity right now. That’s not common. You know, I’m positive it’s right, but it’s not common knowledge. And >> it’s so cool to have us all say, “Yeah, this is this is actually this incredibly magical moment in human history.” And exactly like Alex and See said, when you zoom out and look at the long term of human history, it looks like a step function. But because we’re right in the middle of it, we’re experiencing all the week- toeek changes right here on this podcast. all these week- toeek changes >> and and as Sam is pointing out >> humans are shockingly adaptable and as Peter always says they go back to sleep in a hurry >> so they see a new capability and the impl like hey look we’re launching private rockets into space you know we can like like the cost per kilogram plummeted the implications of that and gluing it all into all the different things we can suddenly do the backlog is now decades deep of of you know if you
[00:25:02] froze technology today and just assimilated what we invented in the last two years. >> Oh, decades. >> It would take decades to to realize all the implications. And people go, “Yeah, okay. I saw that. I’m going back to work.” >> Yeah. Um, full disclaimer here. I’ve been resisting this idea that we’re in the middle of a singularity, but I’ve now fully entered Alex’s reality distortion. So, >> oh my god. Incredible. All right. Well, stay tuned for some more conversation on this uh on this subject here. All right. The AI wars here. Uh GPT5 Pro sets ARC AGI record. Uh Alex, our resident expert on the ARC AGI. Tell me. >> Yeah, so as a reminder, ARC AGI is a benchmark that measures the ability of AI to synthesize new computer programs in response to uh challenges that can be interpreted almost as like 2D flat puzzle games. uh the the ability to extrapolate sequences of images and and
[00:26:02] patterns without any natural language help. And I I think it it’s a beautiful sequence. Now there’s more than one ARC AGI benchmark for the ability to do this sort of visual reasoning. It’s a beautiful benchmark, but but also one of the things I I love about the sequence of benchmarks and I’ve donated to to ArcGI in the past is that they pay close attention to cost per task, not just raw capabilities. So, so we can see a price performance frontier. And to the extent that that the goal of many in the AI community is to drive the cost of intelligence down to zero, we can watch in real time the cost of solving hard, arguably in in some cases superhuman challenges be driven to zero. ARGI specifically is focused on problems that are easy for humans to solve, hard for current AIs to solve, but as the cost plummets, we’re going to see superhuman performance. And and and to your to your point, Peter, GPT5 Pro is demonstrating
[00:27:00] exceptional score. So, so exceptionally high uh performance, but still at a relatively high cost. And over time, I would predict over the next year or so, we’re going to watch all of these curves on on the scatter plot that you’re showing shift to the uh shift to uh upward and shift to the left, at which point cost of intelligence too cheap to meter. >> Let’s put a few numbers on this. So, GPT5 Pro hit uh 70.2% on ARC AGI1. uh that compares to 65% for sonnet 4.5 and 66.7% for Gro 4. And what we’re seeing is just this constant leaprogging where everybody’s just incrementally, you know, moving towards towards 100%. And to hit your number on price, uh it’s uh for GPT5 Pro, it’s uh when I looked it up, it said it’s $4.78 per task, right? And all this stuff demonetizes rapidly. I I love that the horizontal axis here is is is in logarithmic terms. E every
[00:28:02] chart of of every good and service in our economy should be on logarithmic terms so we can watch the hyperdelation >> Dave. >> Uh yeah, no zoom in on that. Uh when you’re looking at the if you’re not driving right now, zoom in, look at the x-axis, you know, on on the right side you’ve got 10 bucks and then you in the middle you’ve got below a dollar. So it’s it’s a huge range of price points for very similar performance. You know, if you look at the peaks. Uh so what you’re seeing mostly on this chart is massive cost reduction, which you know makes it more accessible. Uh we’ll we’ll talk later about some of the other innovations that are driving down that cost, but that’s that’s perpetual. All right, Alex, this one is for you. You know, uh you and I have been going back and forth on text. Oh my god, we’re sol we we’re solving math. We’ve solved math. So this article in particular comes out on the heels of this. GPT5 Pro sets record at Frontier Math. What does it mean and what are the implications here? >> Yeah, I think this is arguably the most exciting development over the past week
[00:29:00] and a half. So, as a reminder, Frontier Math Tier 4 consists of math problems that professional teams of of mathematicians would take few weeks to solve. The these are very hard math problems. And over the past week and a half or so, we’ve seen first Gemini 2.5 Deep Think and then GPT5 Pro demonstrate breakthrough performance. GPT5 Pro at 13% on Frontier Math Tier 4. And I’ve Dave insisted that I make an internal recorded prediction for for just to get on the record what would it mean for math to be solved in quantitative terms. And and this is months ago. uh and and I put on the record as as Dave will I I think attest we can reasonably declare that math has been solved when Frontier Math tier 4 passes 10% uh scoring. So more than 10% of the problems can be sort uh can be solved by uh a bleeding edge model and and the reason why I picked 10% is because at at some point
[00:30:00] uh in the logistic regression of like predicting you know you’ve solved 10% at some point you just pour compute on and you get more results. I I think we’ve seen this over and over again. We saw this infamously with uh Ray Kerszers pointing out that you’re you’re halfway complete with sequencing the human genome once you’ve passed 1%. 10% is is sort of my my arbitrary benchmark. I predicted that we would be past this by the end of this calendar year. Christmas arrived early. Uh we math is on now on a trajectory if you just pour more compute on arguably with no new innovations, math will be solved. At least math as we currently know it. Okay. So, when math is solved, I’ve asked you this before, but I just want to hit it home because it’s a kind of an esoteric subject for most most people. What does it mean when math is solved? What’s the implications for, you know, all of our subscribers here? >> It’s the ultimate canary in the coal mine, as it were, for solving physical sciences, solving engineering, solving medicine. If if we can have machines
[00:31:01] that solve arguably humanity’s most uh rigorous intellectual endeavor, which I would suggest is math, then everything else I would expect over the next few years, call it 5 to 10 years, I expect to come. >> As a tangible example, encryption is all math-based. So when you can have an AI deliver that, you can kind of get super encryption at whatever level you want instantly. >> And physics and material sciences. Yes. Yeah. >> And all of the physical sciences and and conversely, as I’ve articulated in the past, any any discipline that relies on math, at least the current math being hard, is also in danger. >> So, it’s not just learning math, it’s inventing it at some point pretty quickly. >> Absolutely. >> Okay. >> Yeah. Wow. >> Yeah. I think I think just the the implications you know one thing people misestimate continually is if if some brilliant mathematician human solves a hard problem that makes the news but it doesn’t imply that all other problems
[00:32:01] will be solved the next day the AI version of the same achievement >> great point >> like in biotech in math in physics in all these other areas because it has near infinite scale instantaneously if it can do one frontier math problem. It’s very close to being able to do many and then then, you know, billions, you know, just right after that. And so people need to factor that into their rate of change thinking and as as it cracks these different areas, >> we better find some more problems. >> This is dominoes falling and heading towards again Alex’s words, solve everything. >> It’s bulk discovery and we’ve only seen this in in narrow areas. We saw this with alpha fold 3 and protein folding where almost overnight we had arguably highquality protein structures for most known proteins. We’re going to see bulk discovery across a number of disciplines. The the other story here is um we don’t have a slide for this but uh the Erdish problems. This is a a set of order of magnitude a thousand problems
[00:33:00] that were identified by famous mathematician Paul Erdish. AI and GPT5 are being used to bulk solve those and the solutions are starting to pour out. If you go to the Erdish Problems website, you’ll see just in the past few days now there there are folks who are just bulk applying GPD5 to all of these open problems and they’re getting switched from open to solved. >> Wow. >> Don’t sleep. Don’t blink. It’s happening. All right, let’s move on to this one. Open AI uh on building chips with AI. Here’s a quote from Greg Brockman. We’ve been able to apply our own models to designing this chip. We’ve been able to get massive area reduction. uh you take components that humans have already optimized and just pour compute into it and the model comes up with its own optimization. Dave, talk to me. >> Yeah, I love this story because this is this is it ties together so many things we’ve been talking about. You know, one of them is the the short timeline to about 100 to 10,000x improvement in AI
[00:34:00] performance because of the the AI self-improving. And when we say AI self-improving, a lot of people in academia are like, “Well, it’s not that smart yet.” But it is because AI self-improving is nothing more than math, algorithms, and chip designs. And it can do those point tasks. And then a lot of the academics then say, “Yeah, but that’s not true reasoning. That’s not true genius. That’s not true whatever. It doesn’t matter because that’s all you need to do to self-improve.” And so I I just love the story. I also love the fact that, you know, Greg and Sam out of the original cast at OpenAI are the two guys that dropped out of college, didn’t finish undergrad, and they’re the two survivors. And so, you take the, you know, take Greg, the guy that dropped out of MIT. He’s designing chips now because he’s a master of AI. And I don’t know if you guys did any VSSI design or chip design at MIT. I I I did a fair amount of it actually doing neural net designs. Um, and it’s absolutely laborious. And the amount of improvement is incredible if you just had the time and you know 10,000 people to work on
[00:35:00] it. >> And so you AI is just going to rewire and re redesign and relay out the thing. And the simulators are near perfect. >> It’s an acceleration of the acceleration. >> It’s so cool. It is to me. >> I can attest to this. One of my student jobs at Waterlue was to run field tests for VLSI boards and it was like freaking linear painstaking hell. >> Yeah. And so the other thing this ties in is Leopold Dashen Brunner buying Broadcom stock >> and you’re like, okay, I didn’t see that one coming. Why is he? And of course, you can see it in his 13F filing. But now it’s obvious, right? Brockman is designing chips. The chips need to go to Broadcom and then they get manufactured on TSMC. Therefore, Broadcom stock is the one that Leopold buys. And of course, Leopold was at OpenAI and knows Brockman. So it all ties together. >> You you mentioned something that I think is important again. And we hit on this sort of week on week that Sam and Greg dropped out of college to go and pursue this. And you know, one of the things you and I have discussed, Dave, is uh
[00:36:00] majority of the entrepreneurs who are succeeding today aren’t the ones who’ve gone on to get PhDs or gone on to do graduate work, right? They’re the ones that have launched either just after college or have dropped out of college to go pursue it. And it’s almost as if going to get your graduate degree and going very narrow into a deep uh you know it’s like I remember I was if I was going to explain to my grandfather what I did when I was at uh at MIT he says okay you’re an expert in uh in I would say to him like look in the dirt over there there’s this thing called this bacterium and he goes oh you’re an expert in that. No no no I’m not an expert in that. In the bacterium there’s this thing called DNA. You’re an expert in that. No, in the DNA there’s this thing called a gene. You’re an expert in that. No, no. In the in that gene there’s a promoter sequence and I’m an expert in that, right? And so our graduate work right now is this hyper narrow focused effort instead of being able to step back and look at, you know, reinventing an entire field. >> Yeah. I I’ll put a slightly different
[00:37:01] spin on it too, which is the the people Greg’s age and Sam’s age that dropped out had very high situational awareness around the urgency of what’s happening right now. If it were 2001, you know, and 9/11 had just happened, uh going and getting a PhD would make a ton of sense because the world isn’t moving at warp speed during that time frame. But the reason the undergrad dropouts and the other people just graduate and start a company right away are way overperforming is because they recognize the urgency of the moment and they know that four years from now is is just not a good choice in this moment. So it’s that >> also their brains haven’t been calcified by studying that one thing that Peter talked about. I don’t know how Alex broke through all of this. Somehow he’s >> program calls it the hardest lesson to unlearn. It takes a lot of unlearning. >> Yeah. >> Crazy crazy. All right. Um, let’s let’s move on here. Sora hits 1 million app downloads in less than 5 days faster than Chat GPT. I remember when ChatGpt
[00:38:00] came out. Uh, I was doing a podcast. I was like, “Oh my god, a million downloads in 5 days.” You know, new world’s record, but something is going to beat that. Well, here it is. It’s Sor1 uh beating that in the app store. And of course, something will beat this as well uh probably within the next year. Uh Dave or Salem, what’s your thought here? >> Well, one thing we’re seeing here, I I was talking to the CEO of Pacfi uh two days ago, really thriving marketplace company, and they were asking for my advice on how to implement AI within their product. And I said, well, just ask the AI. And so what you’re seeing here is more and more of what you do next is what was recommended by the last thing you had on. So if you’re a Gemini user, you go with Gemini. If you’re a chat GPT user, go with Chpt PT. But you say, “Hey, I want to create an amazing video. What should I do?” It tells you what to do and then nine times out of 10 you just do what it said. And so the distribution of these new capabilities is actually within the AI installed base. And that’s why it’s got this self-reinforcing loop.
[00:39:00] >> I mean I I would have expected nothing less just because the next thing will be two days and the next thing will be one day. And as Alex points out in our earlier conversation, at some point we’ll get the ability to upgrade everybody at the same time and then we’ll be that’s what Ray is talking about. >> Yeah, for sure. All right, I’m going to turn to you on this one, Alex. Samsung’s tiny recursive model redefineses AI efficiency. So researchers have built a mini AI model at Samsung with 7 million parameters to test reasoning ability. And this compares to models with uh billions or trillions uh of parameters. So talk talk me through this, Alex. Well, I would say at a high level, remember first that this revolution, this soft or gentle singularity, if you will, that we’re living through was arguably the result of just compressing information. That one of the biggest lessons I take away from h how intelligence was arguably solved is just like you could take a lot
[00:40:02] of matter and if you compress a lot of matter in a tiny volume, you get some phase transitions and ultimately you can recover free energy for example via fusion. Similarly, if you take a lot of information and you compress it down into a model that has a relatively small number of parameters, at some point you get intelligence out almost for free. That’s like a big meta lesson in my mind from from the past 10 years. So there is an enormous amount of of opportunity for taking large hard problems, large information spaces and compressing them down not to models with billions or hundreds of billions of parameters, but to models with millions of parameters. I I’ve also spoken about my expectation. This is one person’s speculation that we’re we’re marching toward an ultimate end state where we’ll achieve like a perfect model that’s like a micro kernel. It’s like a diamond of of a model that may not even be it may not consist of differentiable endtoend parameters. Maybe it’ll only be a million parameter equivalents or a
[00:41:00] million bytes or something like that. And I when I see progress like Samsung’s TRM which which looks I I it’s it’s a it’s a lovely paper if you look at the architecture sort of resembles a diffusion model in some sense the the premise is it’s it’s domain specific in in this case in the case of the result you’re showing it was trained on a bunch of examples for the ARGI1 challenge that we were talking about earlier has a notion of a scratch pad but it’s a relatively tiny model and it rewrites the scratch pad several times and then goes back to work. it has some latent expectation of what it’s trying to do and then rewrites some more times. It looks like a diffusion model if you squint hard enough at it. But I I think that the the big takeaway is such a tiny model achieving breakthrough performance. You know, if if folks who can see the chart here, it’s competing with Oclass models in terms of price performance, but it only has a few million parameters. So I I think >> so the implications are what? You’ve got this on your phone. You’ve got this on every device. Even if it’s if it’s not
[00:42:01] connected to the internet, it’s able to deliver you intelligence. >> Oh, the implications are so much bigger than just that. Sorry. Sorry, Alex. Go ahead. We’re going to say the same thing. I’m sure. >> Well, I mean like look, if you can if you can take, you know, normally to achieve that same level of performance, you need about about, you know, on the order of a trillion parameters, say 700 billion just to keep it simple. So, you go from 700 billion down to 7 million. then you can use all the rest of that compute capacity to expand it again and make it better at that domain. So if you can take a general massive model that cost you know a billion dollars to build and compress it down to a specific use case like protein folding or you know just longevity related use case or just a chip design related use case. Get all that original capability but now you’ve got it in a very tiny package. Then you can start training it out, reuse those billions of parameters to make it much better at that domain. And so you have this combination of many many effective
[00:43:00] workers, you know, many agents working on the problem plus the ability to build back out the intelligence level with more and more data purely through this process of distillation and expansion. Distillation and expansion. It’s the most powerful thing in the world. And it’s also really really empowering for a whole new wave of startups. And I don’t want to get too deep into it, but we could talk about this for hour, but I think this this slide encapsulates the most profound idea in all of intelligence, which is so it was something that Elio Sudskver said at MIT with Lex Freriedman about God, it must have been like seven or eight years ago at the dawn of this wave of AI that compression and intelligence are actually the same thing. And everyone goes, what that that makes no sense. >> This is the point Alex has been making, right? That’s right. >> Yes. Yes. And this is why Alex says intelligence is going to turn out to be everywhere. There there’s going to be DNA and how much is packed into a few DNA, a few genes, right? This is the same basic principle. >> I’m going to name this episode the singularity is now. That’s how I’m feeling. >> Uh, >> okay. But here here’s two questions
[00:44:00] about this. >> One, does this obiate the need for all the radical energy expansion or do we just have so much compute to do that we need it all any? >> No, no, no. We we were going to chew this up so fast >> because theility the second thing that occurs to me is right now when we run these big LLMs it’s kind of a mainframe model. This is kind of tilting towards more client server type architectures. >> Yeah. Distributed like yeah the foundation models are going to be everywhere in your pocket in your body and out in the cloud. >> Amazing. >> Yeah. Let’s come back to that. Let’s pile that away because that’s a hu Sem what you just said is also hugely important and profound this distributed uh training concept. >> So let’s come back to that. >> All right. Our next article uh diving into anthropic introducing Claude Haiku 4.5 uh which delivers near frontier coding and reasoning performance at onethird the cost of Sonnet 4. Let’s play a quick video here uh while it’s playing without sound and it’s just uh showing its capabilities
[00:45:02] here. Uh, Alex, what do you think of Haiku 4.5? >> So, I I ran it through one of my I I have a a suite of evals, a suite of tests or benchmarks that I I throw at codegen models and and classes of other uh classes of other models. So, so my my favorite eval for for new codegen models and arguably Haiku 4.5 can do more than codegen, but I I ask it to generate a uh in one shot a visually stunning cyberpunk firsterson shooter and Haiku 4.5 was able to do that to near perfection, but critically was fast. It’s very very fast model. So I I I maybe wall clock time 30 to 45 seconds before I had a playable cyberpunk FPS. That was lovely. >> That’s stunning. >> That’s nuts. >> Stunning. >> All right. Congratulations Anthropic on that one. >> Uh moving on. Here we have Google. uh
[00:46:00] kind of a pre-announcement, but the drums are are beating in the off in the distance that Google is getting ready for uh releasing Gemini 3 likely this December. It turns out that Google has released its Gemini models in December over the last two years. And of course, once Gemini 3 drops, the moonshot mates are going to be there to explain what happened, what does it mean, what are the most exciting attributes in it. Uh any other thoughts here, Alex? It’ll be interesting to see with this uh annual cadence how quickly all of the the Frontier Labs leapfrog each other. I I do think there have been uh reports of an experimental checkpoint floating around the internet uh preliminary release of Gemini 3 that is really impressive if if the reports are accurate uh fluent music graphics 3D generation. So I’m very excited to play with Gemini 3. >> We’re going to see that in our next slide here or listen to our next slide
[00:47:01] here. So purportedly this is uh from Gemini 3 that it can now compose original music. Let’s take a listen to this as you’re sipping your coffee or driving your car [Music] uh you know it’s it’s gorgeous. What are the implications of this? Um what do you guys think will come out? >> So hold on. Let me ask a question because that just sounded like Shopan to me. And um we’ve had for a while the ability to say uh hey play me something like Shopan. It’ll play you something. There’s something that’s happened here. What am I missing? >> What I think is is interesting is this is MIDI generation. This isn’t generating raw waveforms or or wavelets or raw audio as we’ve discussed in the pod in the past like 5. This is treating
[00:48:01] music as a as an almost first class modality. There are a variety of of languages like ABC notation or MIDI notation. And what I take away from demonstrations like this and and others that are uh allegedly floating around from this experimental checkpoint is Gemini despite you know whatever limitations in in in its omnimodal training data set has the ability to understand music notation uh and music languages as a first class modality. that that sort of transfer is highly difficult and non-obvious if you’ve ever tried to ask like uh an oak class model say in years past to generate compelling music. It was very challenging. >> Well, Andre Carpathy just kind of emerged and started podcasting again out of nowhere and he made the point that these capabilities are using the same generic neural net design transformer design that the language models use and that the self-driving car uses. And so the implication of that is that look, it’s it’s exploding into these areas
[00:49:00] purely with more data. There’s no, you know, hard heavy lift for humanity to design something new to make this work. You just put in more data and now suddenly it can do MIDI, you know, actual music creation. But what you know, that implies that there’ll be something next week and next week and next week and next week. It just as quickly as you can gather the data, it’s it’s uh developing these new abilities. I want to double click on that because I don’t think majority of people understand that these models when we’re hitting these scaling laws of more data, more parameters, more compute are evolving capabilities that were never predicted. It wasn’t like we tried to build these capabilities. It’s like they’re emerging properties as the systems are becoming more intelligent and and that’s fascinating. Scary in some ways, but fascinating. >> Yeah. And hugely important. hugely important because because when people see a new capability like this, they assume that some team was grinding away on it for 20 years in some basement and it just got launched. That’s not the case. This this is just a purely
[00:50:00] emergent capability on top of the core platform. It’s like, oh wow, look what we can do now. >> And so ironic, Peter, I mean when when I was in undergrad at MIT, one of my first research adviserss was Marvin Minsky, and he would slap my hand every time I use the word emergent. He would say, “No, that that’s preposterous. emergent. Using that word just means you don’t know what you’re talking about. Uh and and yet, ironically, decades later, in fact, we we see all of these emergent capabilities. >> Amazing. We should do a whole episode on things they told us that that turned out to be wrong. >> There’s a bunch. >> That would just take too long. All right, let’s go on to another uh another property that’s been announced for VO3.1. So, Deep Mind releases the next uh next image video model. Let’s take a look at a quick video about VO 3.1. >> New enhanced capabilities give you control like never before. Let’s take a look. You can use a reference image. VO puts them together into a fully formed
[00:51:00] scene complete with sound. >> Hello, is anybody here? >> You can extend your clips and transform any shot into a full scene. Reimagine any shot by adding or removing elements from subtle details to impossible objects. VO matches scale, lighting, and shadow for real world physics and cinematic outputs. Life with audio using sound effects and dialogue. >> Just got to listen. >> I mean, incredible, right? So there are three critical differentiators between VO3 which was amazing in itself and in V3.1. The first that they point out is uh superior audio quality and synchronization and you could hear that when it was cooking the food in the frying pan. It’s much richer, more natural. The second which I find fascinating is you can give it three ingredients like you know here’s an image of a person and here’s an image of an object here’s an image of a room and
[00:52:00] it will take those three effectively and tell a story around those three combining them and you can give it a opening scene like here’s the starter image I want you to start the video with and here’s the ending image I want you to end it with and it’ll create a logical construct that goes between uh those two points. I mean, pretty amazing. Other comments on this one, please? >> Yeah. Well, so a couple of our uh friends have made, you know, fulllength movies now and and to do that, it only gives you a minute or two back at a time and so you have to take the ending image and use it to stitch together the starting image, which is frustrating because, you know, they could do that automatically. They just don’t have enough compute to keep up with all the people that want to use this. Uh so, but you can do it. you can hack your way around it and make incredibly seamless fulllength movies all of a sudden. And if you go back just six months ago, actually remember 18 months ago, Peter, you had um Aristotle and Plato on stage at Abundance 360 debating with each other. >> And if you look at all the comments from
[00:53:01] the movie crowd, you know, the producers, the Paramount Studios crowd, they’re like, “Yeah, but.” And if you look at the list of yeah, buts, almost every one of them has been solved in less than 18 months. You know, yeah, but the characters aren’t consistent from scene to scene. Yeah, but it it’s obviously digital and I can see the seams between the scenes. Yeah, but yeah, but yeah, the physics aren’t quite right. The arm detached from the body. Every one of those things banged out in this, you know, less than 18month timeline. >> Extraordinary. And we’re going to be at the Abundance Summit this year. We’re going to be diving into this again. What’s the implications for Hollywood and in particular the creator economy, right? Um, I’ve got a project, a secret project I can’t talk about it yet, but working on with Google, uh, around production of content like this and, uh, we’ll be releasing that hopefully in the next couple of months. Uh, See, do you want to comment on this? >> Um, I don’t fully understand the
[00:54:00] implications of this except if you can go end to end up per scene like this, then stitching together a whole movie becomes every amateur can now go full on, right? And that completely changes Hollywood radically. I guess we’ll see an explosion of consumer generated movies now uh with all sorts of weird plots etc. >> Oh, I I I suspect I don’t know but I suspect that there are many many people all over the world that have something to say that could never get the attention of a studio or an actor before and now they don’t need it. Now they can individually put something professional studio caliber together to illustrate the point they were trying to make. I think that’s going to be a crazy explosion. >> And what a win for YouTube, right? I mean, Google is sitting pretty because the only distribution engine uh for this explosion of content, >> that’s right, >> is going to be on on YouTube. >> I I I also think with some of these models, VO 3.1 and Sora 2 Pro, for example, these are just training wheels
[00:55:00] for videobased reasoning models. I think that will generate an enormous amount of additional economic value. So, so generating consumer grade video, this is wonderful. This is radically democratizing Hollywood. Excellent. But then imagine now we’re about to have reasoning models that can think with images, think with videos. That’s going to solve a whole bunch of transformative problems that I think >> with the associated downsize. My son just served me a surro clip of me at 800 lb playing a video game. And I’m like, if you if you put that out there and people think that’s me, it’s going to be terrible. It’s going to be the hut. >> Oh my god. Uh our next article uh again following on Alex’s sort of predictions here, Gemini 2.5 and GPT5 win gold at the International Olympiad on astronomy and astrophysics. I didn’t realize that there’s an International Olympiad on astronomy and astrophysics. Alex, over to you. >> There there is and it’s quite competitive. So this is for high school students. There are international olympiads for math and physics and computer science. I I was on the
[00:56:01] computer science Olympic team for for the United States in high school. This is a very competitive Olympiad. Uh and I I think so this is work from Ohio State University, another lovely paper. And I the authors make the point in in achieving this this benchmark. We’re drowning in pabytes of data from astronomical sources from automated sky surveys and there aren’t enough human waking hours in the world. >> Yes. To analyze it. >> Yeah. to analyze it. So, this is more than just uh beating high school students at astronomy and astrophysics. This is about AI solving astronomy and astrophysics and enabling us to understand our universe. There aren’t enough human waking hours to look at all the skies. >> I remember changes the game for SETI, doesn’t it? I remember Gerald Soffen who ran the Viking program back in the mid70s in the dark ages and he said you know we have we’ve looked at less than a fraction of 1% of all the data that came back from Viking because there isn’t enough human
[00:57:02] capability to do that and now I mean all the data generated especially we’re about to have the interplanetary internet too right laser links between orbiting satellites around earth around the moon around Mars uh with high bandwidth capability And we’re going to be able to constantly be sorting through the data and making discoveries sort of go to sleep at night, wake up in the morning, you know, early because you can’t you don’t want to find and critically again these are generalist models that are accessible to anyone. Almost anyone can access Gemini 2.5 or GPT5. This isn’t some sort of like siloed data center that only folks who have the time and ability to to purchase observatory time from the the great telescopes can get. Anyone can do this. So I think it’s going to radically democratize situational awareness into our universe. >> Well, it goes, you know, using this as a uh an analogy, anybody out there who is a large amount of data in your business, in your industry, right? It’s this gold
[00:58:02] mine that hasn’t been tapped yet. And so if you have that data, uh, being able to actually, you know, feed it into one of these models to start to extract value should be your very first stop. >> Wait, Peter, can I just get back to something? Alex, you just said this democratizes situational awareness across many industries. Can you unpack that >> into our universe? I I I said so. So there are public publicly available pabytes of data gathered by different observatories that Historically, if we’d had this conversation 10, 20 years ago, we’d be talking about some sort of citizen science initiative, maybe people spotting, looking for interesting looking things in the sky at home. We don’t need to have that conversation anymore. Now, anyone can turn loose Gemini 2.5 or or GBD5 or some other frontier model with a decent inference time compute budget and say, “Go look for interesting things in the sky. Make a discovery. It would have been 10, 20 years ago. It’s remarkable when a high
[00:59:01] school student discovers a new asteroid uh or discover. >> How does this deliver us situational awareness into the universe? >> We live in a very dynamic universe. Uh there are all sorts of interesting things going on all the time and we’re drowning in pabytes of new data all the time. >> This episode is brought to you by Blitzy, autonomous software development with infinite code context. Blitzy uses thousands of specialized AI agents that think for hours to understand enterprise scale code bases with millions of lines of code. Engineers start every development sprint with the Blitzy platform, bringing in their development requirements. The Blitzy platform provides a plan, then generates and pre-ompiles code for each task. Blitzy delivers 80% or more of the development work autonomously while providing a guide for the final 20% of human development work required to complete the sprint. Enterprises are achieving a 5x engineering velocity increase when
[01:00:00] incorporating Blitzy as their preIDE development tool, pairing it with their coding co-pilot of choice to bring an AI native SDLC into their org. Ready to 5x your engineering velocity? Visit blitzy.com to schedule a demo and start building with Blitzy today. >> Now for our book corner. Uh favorite science fiction books. I’ve added one and of course AWG has. And Dave and Seem, I need your book entries for next week. >> Mine was Fahrenheit 451. It’s going back. >> It’s going back. I’ll mention mine first. It’s a book called After On by Rob Reid. Uh I’ve read it twice. My son Jet and I have read it together once and it’s a it’s a fascinating story. It’s a very fun read. Takes place in Silicon Valley. And it’s really the story of the emergence of the first uh conscious AGI or or ASI. And
[01:01:02] you know what is it like when something comes online and it looks around and it sees fear in this case I won’t you know ruin too much of it but the AGI is called flutter and it hides for the first period of time and it starts to then talk to certain people to understand how society and military and culture is going to you know you know feel about a super intelligence out there on the internet. It’s a great story a lot of fun. And Dave, you’re reading this now, you said, right? >> Yeah. Yeah, I’m about a quarter of the way through it. It’s, you know, a lot of a lot of books that deal with AI are esoteric. This is just a great fun awesome story that also deals with AI. So, it’s it’s thick. There’s a lot of words. It’s a it’s a long book, but it’s really good. >> It’s really It’s really fun to read. So, that’s Afteron by Rob Reed, and it gets you thinking about what it might really be like, what could be going on right now that we just don’t know about. Uh, Alex, tell us about Diaspora.
[01:02:00] >> Yeah, Diaspora by Greg Egan, I would say, is my second favorite book after Accelerando by Charlie Stross. So, Diaspora is a book about, and I’d argue it’s utopian. Some might disagree, but it’s a book about what life after the singularity looks like. And intriguingly, it depicts a future where we don’t take apart our solar system to build the Dyson swarm, where this this tiling the earth motif of of turning everything into data centers actually peaks uh and then declines and most humans in the future of diaspora are living as uploads or emulations living in data centers underneath mountains on Earth. But it’s a heterogeneous mix of biological humans, some who chose to remain purely biological in uploads and cyborgs. And I I think it it’s one of the most vivid posts singular depictions uh that I’ I’ve ever read. Big fan book. >> There you go. This is all our subscribers. This is your this is your reading assignment uh for the week ahead or listening assignment as my kids say.
[01:03:01] Dad, you don’t read books. You listen to them. Uh, by the way, after on is incredible. >> Time to read anymore. We’re just trying to keep pace with all the stuff that’s happening. How how are we going to navigate this? >> Yeah. Insane. All right. Uh, let’s move on to robotics. Uh, a few things coming out in the robotics world this week. Uh, the first one, super fun. This is Tesla’s FSD uh, version 14.1.2 brings on Mad Max mode. All right, let’s take a quick look at this video, which is insane. So, if you’re listening, uh, what we’re seeing here is someone in their Tesla Model Y or Model S, whatever it might be, and they’re in MadMax mode. They’re going 83 mph, but their car autonomously is weaving in and out of traffic like in the middle of a car chase for your favorite uh adventure movie. Um, and I just imagine in the day, you know,
[01:04:00] people are in Mad Max mode. They’re pulled over by the cop and they say, “No, no, no. It wasn’t me. It was Mad Max who was driving this. Please let me go.” Uh, thoughts on this. >> This is nuts. How How is this legal is my one number one question? >> Um, well, it’s not all with the driver, but this is going to cause a lot of issues. I I think you’re going to see some spectacular car crashes around some of this. Well, I mean, listen, the the point I think is I would rather have uh my FSD v14.1.2 doing this than me doing it. Uh it’s able to see, you know, with millimeter precision how far you are from the car you’re cutting off next to you. >> I I think it’s it’s also ironic if you remember the original MadMax movie was was supposedly set in a civilizational collapse type scenario where energy had collapsed in particular. there were worldwide energy shortage shortages and and sort of perverse in in some sense that now we’re on the verge of energy abundance uh at a global scale thanks to
[01:05:01] the AI data center boom uh and and yet we’re we’re sort of looking back and trying to build uh MadMax almost nostalgic retrospective while at the same time building minority report level rapid transit. >> I I would call this Gran Turismo. >> Yeah, let me let me read read the description uh from Tesla. When you select MadMax mode for your Tesla vehicle, the FSD system, which by the way still requires driver supervision, right? Uh will attempt to drive more assertively. Faster acceleration, more frequent overtakes, more aggressive lane changes, especially under traffic conditions. >> I mean, it’s how I drive normally anyway, but but that would be unable. >> All right, the news this week. Uh figure three, Brett Atcock’s company made the cover of Time magazine. I love this image. Best innovations of 2025. Robots are coming to your home. And here we see uh figure three folding laundry. Uh got to love it. Uh let’s take a quick look at the video of figure three uh which
[01:06:03] has been reimagined for the future of labor. Let’s go play this video then we’ll chat about it. So uh figure 3 is here. It’s got a new brand new look. Here we see it uh washing dishes, serving food. Here’s your key. You’ll be in room 23. The elevators are past the door on the right >> and playing at the front desk of a hotel and then delivering packages. Interesting. >> Very slowly. >> Very slowly. But uh uh fascinating uh figure 3’s changes here, right? It’s taken on a new home ready design. Uh we you know we were at 1x. Uh I done a podcast. I’m a full disclosure. I’m an investor in in Figure uh a couple of times through my venture fund. >> Congratulations. >> Yeah. No, it’s it’s done amazing. They did a we reported last week or two weeks ago that they did a a billion dollar financing at a $39 billion valuation. Of
[01:07:01] course, this is a multi-t trillion dollar marketplace for sure. Um but if we look at figure three versus figure two, there are five outstanding things. Number one, uh, their new Helix AI system has a suite of sensors that make the vision, language, action much more capable than Figure 2. It’s got redesigned hands with cameras in the hands, upgraded audio for voice reasoning. Uh, it’s got a home ready design, you know, it’s looks uh, more soft and easy. It’s taken on the same sort of look and feel as as Neo Gamma from 1X. And then it’s a lighter chassis at 61 kg versus 70 kg and still able to lift 20 kg. Thoughts, Dave? >> 20 kg. I you know it goes funny because the 1X can lift its own body weight. >> Yeah. >> Yeah. Yeah. That’s interesting. But the form factor seems to be settling in on this human but slightly smaller and soft
[01:08:00] exoskeleton and not going to hurt you. Uh so moving a little slow, but it doesn’t need to move slowly. it it just does because it’s less likely to slap you in the face that way. >> Um, so they’re all settling on this very similar design. So, I think that’s going to be the final answer. >> I I think the most interesting design change that I saw in Figure 3 is the Palm cameras that that you mentioned, Peter. So, that that immediately in my mind rhymes with Tesla vision and and Tesla issuing LAR in in favor of pure vision. If if you imagine the the difficulty, sure it it has tactile sensors on the fingertips, but it it’s really difficult to achieve, at least at this point in time, human quality, tactile sensation throughout the body. We we have human biological meat bodies, we’re covered in tactile sensors. Humanoid robots, not so much. So to the extent that figure is able to move the future to the left by using cameras, using vision to substitute for tactile sensation all around the arms, I think
[01:09:01] it’s a very clever move. And I I think that in combination with vision language action VLA models. I I think this is just going to be like a homework assignment for for a K12 student in a few years. Just implement a a working humanoid robot using an off-the-shelf VA model from a Frontier Lab. Embody it. this will all be viewed as pretty easy. >> Crazy. And then the other thing announced by Brett is uh a price point. Uh he wants to hit a $20,000 price point on this. You know, we talked in the past, we heard from Elon that, you know, cost of goods sold sold would be about $20,000 and it would be priced depending on volume. But, you know, Brett’s getting very aggressive here on pricing. I think he wants to take out the competition. Brett is a brilliant engineer and entrepreneur and he’s very competitive. So he’s going up against Tesla and 1X and is playing to win. >> Selena, >> part of the new American just quickly part of the new American dream, right? So it used to be a house in the suburbs with a car
[01:10:01] >> and we need to add humanoid robot at least one to that given that humanoid robot order of magnitude price is a little bit like a cheap car. >> Yeah, it’s you know 300 bucks a month, 40 cents an hour. Uh it’s how many would you own? I mean probably a couple at at that >> sky’s the limit. >> Yeah. >> See, any rants? Any rants on this, Sem? >> I have a r you’re pre you’re pre suggesting things. Okay. A couple of thoughts. One is um I I go back to Immod’s comment about cap capital does not need labor anymore. And this is kind of another indicator along that spectrum, which is pretty a very big uh outcome. The second is you know we’ve we took us like so long to get um FSD working um almost 15 years now almost 20 years now uh um with very bounded edge the edge cases here are infinite when you have a humanoid in
[01:11:00] the home like what the hell what happens when it thinks the baby is a doll and and put or thinks it’s a it’s it’s a teddy puts it through the washing machine I don’t understand all those nuances I don’t think it’s a fair anal technology. I mean, it took FSD so long to get here because the AI models had to evolve to get to the point where they can work with the data in flow, the VA models. >> We have those. >> Okay, I I understand the speed of it and the fact that it may work, but there’s so many things that could happen, right? I used the example before of your neighbor calling up going, “Hey, your damn robot’s charging itself on my Tesla charger. Freaking get it off.” you know, like h I don’t know how we’re going to navigate all of those boundaries, which we naturally do. And maybe the uh the the u a broad approach of like learning from each other is will will kind of solve a lot of this. But I’m >> these robots are going to be smarter than any human that you possibly know.
[01:12:01] They’re all going to be running the most advanced models out there. >> All right. They can’t tell the difference between a doll and a and a doll and a baby or your Tesla charger and your neighbors. >> I still don’t know why you can’t have two extra hands on it. This makes no sense to me. >> Okay, we’ll get to that finally. >> How? >> All right, Dave, you were going to say >> uh well, two things. First, Brett Brett Adcock. I spent a bunch of time with him backstage at your last event, Peter. He is just awesome at the abundant summit. >> Yeah. Yeah. Salt of the earth. >> He’s a guy that everyone will be cheering for to win. That really really makes you feel good about you know everything Sem is worried about Brett is the guy you want I hope I’m wrong right but I I’m I’m struggling to make the leap on how we navigate all those edge cases that’s all >> with intelligence the other thing is you know you know to to what Alex was saying a second ago a lot of people are saying well look it doesn’t have the same sensory feedback that a human hand has you know I’ve got a you know millions and millions of neurons touching and feeling it yeah no doubt that’s true but it has visual acuity and dexterity
[01:13:02] that’s crazy superhuman. And so it it frustrates the heck out of me when people are saying, “Well, I’m going to work on the the the the smell sensor for this. You know, I’m going to do fundamental research on it.” When all you have to do to succeed right now is glue together the obvious. Well, for the first time in human history, we have perfect visual pattern recognition. It can easily tell your baby from a doll easily. Uh and and it can do all these fundamental mundane tasks trivially easily. and it has a brilliant AI voice to program it. You don’t have to get into coding some arcane language. You just tell it what to do. Those capabilities alone should explode technology. >> So, I’m I’m suffering from a gross lack of imagination here. >> It’s possible. It’s possible. >> Just ask your ask your favorite. >> I’ve been wrong before you know rarely, but I >> Let’s move on. We got a lot to cover. Let’s get to chips, energy, and data centers. Uh the first the first data point here is uh something that’s going
[01:14:00] to be a real challenge. So US electricity prices reach all-time high. Uh this is a spike in dollars per kilowatt. Uh we’ve seen it, you know, it was pretty level between 2014 and 2022 and we just see this asmtoic uh rise uh in energy and this is going to hit everybody in their pocket at home, right? And I can imagine a situation where uh communities will start to say no no no you’re not allowed to build a data center in on our grid because we don’t want to be you know uh subsidizing your your AI system. So there’s going to have to be some policy changes here. either differential pricing where consumers pay a pretty flat price and the uh the data centers are are paying uh for the additional required or something has to happen here thoughts Dave >> or Alex go ahead please two two two
[01:15:00] important caveats here one this is nominal dollars per kilowatt hour so you have to subtract off inflation so you subtract off inflation which obviously spiked in in the years following 2020 you you still get a material increase in electricity price, but this once you’ve subtracted off inflation, we’re looking at real dollars per kilowatt hour. This is the market doing what the market does. It’s signaling via prices that there’s strong demand for utility electricity. And if utility electricity can’t supply the the thirst for for energy for data centers, then we see what we’re already seeing, which is data centers will will have their own colloccated off the grid NAT gas and uh SMR nuclear facilities and maybe eventually reconnect to the grid, which is obviously more ergonomic once the grid is available. But that this I I think again subtracting off inflation, this is a reflection of the the thirst by intelligence for the foreseeable future for energy. >> Yeah. And >> you can see the markets you you can see the markets reacting to this in the other way also. Fermy energy were re
[01:16:02] advisers to that project just went public and it went from 0 to 17 billion valuation in 8 months. >> Insane. Insane. Okay, I’m going to come to you on this one uh for fun and giggles. It’s insane. The US cancels the nation’s largest solar project uh which was planned to generate 6.2 gawatts of power. The US government cancelled as Moralda 7, Nevada’s massive solar project set to become the loyal the US’s largest solar project in history, capable of powering 2 million homes across 118,000 acres of desert land. >> What are they doing, Seem? >> Okay, so I did a little research on this. It’s not as bad as it seems, right? On the surface, you read this and you go, “Oh my god, they’re cancelling solar. What a bunch of idiots. What kind of lit revolt are we dealing with here?” or da da da da and I would go on a full kind of madness rant like we did last week. I think what’s actually seems to be actually happening here is that the
[01:17:00] regulatory headwinds on such a huge project are kind of slowing it down. What they’ve decided to do is break it up into they haven’t canceled it. What they’ve done is said this big thing can’t go forward. Uh break it into smaller projects and you can reapply smaller projects to do this. Now this will still add a huge amount of timeline to it because now everybody has to go back to square one. uh rebuild this as individual uh applications, seven I think in total and uh reapply which will add a lot to the thing. So there it’s it’s not all um uh it’s not all the the headline says it is not as bad as the headline says it is, but it’s pretty bad. And and why can’t you just rush this through and get this out there? I don’t understand. >> And we’re accelerating nuclear. We’re accelerating. >> Well, the administration is fond of breaking all the regulatory to do what it wants to do. Why isn’t it doing this? We need the energy as you we saw in the previous slide. This makes no sense at that level. But there’s there’s a bunch of nuances in the thing. So, I don’t
[01:18:01] want to go full crazy. >> All right. Um, also this week, the US Army announced the Janis program for next generation nuclear energy. So, this is deploying commercial micro reactors to secure power for US defense. Uh let’s see, Alex, what are your thoughts on this? >> I I think this is a very exciting development. Um there are many US bases that are almost tyrannized by their their need to to ship oil. If you remember how in part the the Pacific theater of World War II started, there there was an element to which it was a a blockade of of oil. uh and I I think to the extent that uh the US army can serve as an additional demand function for pushing forward micro reactors, SMR, nuclear reactors in in general. This is going to be a net boon for artificial intelligence. >> So these micro reactors are like what like one think
[01:19:01] >> these are these are >> these are smaller. >> Yeah. >> Small vision reactors. >> Yeah. So it’s like like a fision reactor in a uh how big a dimension it’s like in a shipping container. >> I I I don’t know the the the dimensions. Not sure if those have been made publicly available. >> But I find is these are commercially owned and operated, right? And these are sort of like being rented to military operations. >> Yes, that’s right. >> I think that’s super smart because it creates demand for all of that stuff. But you know, we’ve been running nuclear submarines perfectly well for 50 years without any issues. Why is this such a big deal? >> I don’t know. >> Why can’t we just go without any issues? >> Why haven’t we jumped to this point 20 years ago? >> I I think that that that is the elephant in the room. Uh so and the the premise also for uh for all mankind, one of my favorite television series from Apple TV Plus, an alternative universe where nuclear energy didn’t get kneecapped in
[01:20:00] the 1970s, but instead had continued to advance. Yeah, I’m really glad you brought that up because both both of these slides Yeah. Uh the answer to both questions is more politics than than anything else. >> You know, it’s you know, whose idea was it originally? Okay, we don’t like that person anymore, so cancel their idea and replace it with this idea. Um but yeah, there’s the the answer to like why didn’t we do this 20 years ago is purely because of public backlash, you know, uh to to you know, the perceived risk and some movies. Actually, the movies are incredibly damaging to some of these. You know, if you >> found the movie was a killer. >> Yeah, exactly. It’s like the Jaws equivalent for nuclear reactors. You know, nobody goes in the ocean anymore. >> Type movies. Yeah. >> Let’s move on. So, next article here is US chip plant investment to outpace China, Taiwan, and South Korea uh by 2027. Um, Alex, do you want to hit this one? Yeah, I I think we’re we’re seeing the the innermost loop of civilization
[01:21:01] finally recursively accelerating. So between chips >> Whoa, whoa, whoa, hold on a second. Can you just repeat that word for word? The innermost loop, >> the innermost loop of civilization is recursively accelerating. So, so if you look at, as I’ve tried to articulate previously, sort of a technology tree or maybe a supply chain of technologies, we have chips, we have energy that we were just talking about, we have robotics, we have data centers. All of these arguably form a sort of a recursive feedback loop. We’re going to be using robots to build fabs that that produce chips that go into the data centers that are powered by the energy that build better robots and so on in a >> Why is it the innermost? Cuz it’s going straight down to the energy equation. It >> I would argue it’s the innermost because it’s the most recursive and it’s also the fastest improving. >> Yeah. The the Yeah. The f the fastest loop of reinforced learning and improvement. All right. But that flywheel, that this innermost flywheel
[01:22:00] of civilization, I expect to just fly out into the rest of the economy in the next few years. It’s not going to stay contained to just those four or five technologies. Contrary to to to those who are worried about some sort of like circular Nvidiaesque economy that that’s uh just one big wash sale, I I think it’s going to spread pretty quickly. And well, listen, the we’re seeing this constantly with billions of dollars being deployed uh hundreds of billions of dollars being deployed uh by all the all the frontier labs. Um it wants to be sovereign. I mean that that’s the the other story here with uh one might project reasonably into the future that just as we’re seeing in interest from from different uh sovereign powers interest in inference compute and and making sure inference compute you know you get a Stargate and you get a Stargate uh that that inference compute uh is sovereign we may reasonably see all of these other sort of core elements of this arguably
[01:23:01] innermost loop also become geographically distributed. Silicon is the new steel >> possibly. >> Cool. >> Our next article here is Nvidia to sell a $3,999 G uh DGX Spark mini PC. Uh and this is pretty epic. Anybody who is uh playing with a PC on their desk, uh the Spark computer, the the NV the DGV DGX Spark Mini uh can support models with about 200 billion parameters. It’s got one pedlop or a thousand terra operations per second. A thousand trillion operations per second capabilities on them. Uh god, this is going to blow away the MacBook as your preferred preferred computer on your desktop. Uh Dave, what do you think about this? >> Yeah, I’m going to get one for sure right away. Um the question I had was, you know, I needed dramatically more
[01:24:01] compute. Uh, it’s obvious to me and everybody will soon. I can’t get it right now through cursor. I can’t get it through the APIs. I’ll give this a shot. It’s not quite clear how I’m going to get it integrated into my agent world and my coding world because I don’t think that’s going to be trivial. But I’ll get to work on it and see if we can make it work. But but uh yeah, you know, the device wars are just beginning now because there’s a view of the world where you have a lightweight, you know, Johnny IV device from OpenAI and it’s connected to a massive amount of comput on the cloud. Then you’ve got the Nvidia vision here where no, bring it into your home, bring it into your office because then at least you know you’ve got it. No one else is going to take it from you right as you’re producing something. It’s not going to disappear. So you feel you have that feeling of like it’s mine, it’s under our control. So, we’ll try both, you know, both versions of the future. Uh, but it’s very much in flux right now, but I I can’t wait to try it. >> Yeah. All right, let’s jump into some of the uh interesting news in the economy here. Uh, this article, data centers and
[01:25:00] AI account for 92% of the GDP growth in the first half of 2025. So, uh, our GDP grew by 1.6% in the first half of 2025. And of that 1.6%, 1.5% um, of the 1.6 Wow. >> Uh was data centers and AI right now. >> So now Alex’s inner loop comment is making sense. >> Yeah, >> that’s a brilliant. I love that comment. By the way, >> this See, this is just the opening act. So So the opening act is we spend a bunch of our our GDP on building out, you know, tiling the earth as it were. The the next act I I would predict is transformative applications that pour out of all of these data centers that we’re building that that solve math, science, engineering, medicine, the the works to to justify all this capex. But this is the economy right now. >> This is divide by zero. It goes to infinity pretty fast. Um I I think we’re just beginning. I mean, we’ve talked about this, right, of of the explosion of the GDP in the United States and to a
[01:26:02] to a large degree other parts of the world. Um, and we’re beginning. >> So, the economy run now runs on math and we’ve just solved math. >> Well, we better find harder math or move on to to non-math problems. >> But the question becomes if we’re if we’re exploding the GDP, right, there’s wealth being created and you know, one of the big challenges is how do we redistribute that wealth, right? How do we have it not be concentrated? Part of it’s going to be that the cost of living, which is of great concern to a lot of people, right? Cost of living has arguably increased because of inflation, uh, but it hasn’t yet come down. Uh, are we going to see it drop? >> Here’s something to Here’s something to watch out for in this. If you follow Alex’s kind of train of thought here, the three areas where we have increased costs are healthcare, education, financial services because those are highly regulated. When we see major breakthroughs in demonetization in those
[01:27:01] areas, then we’ll know we’re winning. How’s that for a thought? >> I I think that’s true. We’re going to reinvent healthcare. I mean, the best healthcare in the world should be free and fully democratized. The best education in the world uh will be free and fully democratized. It just hasn’t happened yet. Two points on this story real quick. Uh one of them is that the spin on the story was well without without AI data centers, the economy would be in terrible shape. Not true. All that happened here is all the capital and effort went into this instead of something else because this is much more urgent. So the economy would have been in fine shape either way. But we just chose to use all of our time, money, real estate, investment, everything on this problem because it’s such an important and >> we’re planting seeds. We’re planting seeds for future growth here. >> Exactly. The other thing is that this is where AI benefits everybody. You know, to the point you made at the front of the podcast, Peter, uh, anyone who who wanted to help Elon Musk build his data center in Tennessee and was willing to
[01:28:00] go there and help him do it could instantly double their salary because he was willing to in the urgent race to build all this, more than willing to pay whatever it takes to get everyone to come and work on that project. >> Electricians, air conditioning, plumbers. >> Yeah. So, this is really really democratizing the AI revolution in a big way. And uh and it benefits it benefits any state actually that wants to jump on start building data centers is going to create jobs. >> This is like the second industrial revolution. It’s like electrification or like intercontinental railroad. The the fun stuff happens once all the infrastructure is already in place and you >> Exactly. It’s we’re planting the seeds that will blossom into a whole, you know, almost unimaginable. And speaking of that, here’s our next article. I found this one fascinating. the Dallas Federal Reserve is preparing for AGI. And so, uh, they’re we see here a chart in which they’re making some predictions. So, the Dallas Fed seriously considering a benign singularity. By the way, we should tell them we’re living through it right now.
[01:29:01] Uh, where the economy productively explodes between now and 2035, right? Everybody, the next 10 years, the next five years is the game. um you’re alive during the most extraordinary time ever to be alive and you’re witnessing it uh which is incredible. So they actually uh in this in this graph uh they look at a couple of things. So their scenario planning uh the they have a baseline versus AI boosted uh singularity and they have two curves sort of one in which it’s a benign singularity and things literally uh go into hyperexponential growth. They have another one which is a singularity extinction path which which I don’t want to be thinking about but if hits the fan and uh and we’ve got we’ve got real problems. Uh so I find it fascinating the Federal Reserve is thinking about this. Yes. >> Two two thoughts here. So this might have a second or third level impact from us because we’ve actually had some folks
[01:30:01] talking to these people for a while. Um I’m very impressed that they’re doing this. This is great. This is essentially um pricing exponential growth into the economy, which is amazing. >> Well, I I think it’s utterly I I posted this on our chat, but utterly frustrating and idiotic that they settled. You know, they looked at these scenarios where the red line goes through the roof, the purple line, we all die next year. What we’ve decided is that the final impact, our best prediction is.3%. It’s the difference. You see those two lines that you can’t even tell the difference between them. Yes, >> that’s what we think the our best guess that the final answer is. Utterly utterly idiotic. >> I mean, but this is what’s happening in all all of our interactions with government agencies that we’re having. It’s there. It’s the same like absent any idea, I’m going to forecast something incredibly timid. I can’t >> This is your rant from last week, Seem. >> It is. It is. It is. It is. And what they’ve done is going, let’s put this in this in to cover our bets so we don’t
[01:31:01] get flamed later. >> Yeah. But otherwise, let’s predict a marginally small increase over the next decade. >> But I I’d like to take the positive out of this that they’re at least they’re bringing it into the models and then we can have the basis for fixing it later. >> I don’t blame the people. I don’t I don’t blame the people. They’re in a system where they they cannot they they get punished for anything outside of.3%. So it’s not >> I think it’s very encouraging that the Fed is considering encouraging that the the Fed is considering the possibility. And but I would I would also just question is GDP per capita necessarily the right measure for productivity? Does GDP properly capture productivity? >> Yeah, we need a new metric there for sure. >> This is a huge huge conversation because when you have demonetization, GDP collapses, but everything is 10x better. So what the hell? >> Yeah. So Texas is really playing at a big game here, right? with Starbase there, with uh you know Tesla there, companies moving there. And so the SEC
[01:32:01] approves that the Texas stock exchange uh can ease the rules for a public listing. You know, one of the things that powers our economy is the whole venture model, which invests in companies that then build their valuation and then build to an IPO and liquidity. and it’s been slow and painful over the last five years. Dave, uh, your thoughts on this one? >> Oh, huge amount of thoughts. I’ll keep it short because I could talk for hours on this topic, but you know, Peter, you and I have taken companies public before. Uh, the process needs to be simplified and AI, >> it’s arcane. And also what gets reported, you know, with gap accounting and and you know, all your 10 cues and all that, what gets reported is very it’s just a huge amount of garbage compared to just the simple accurate truth. And so the AI version of this is going to be phenomenal where you know everything is reported. It’s perfectly accurate. It’s perfectly accountable, but it’s much simpler. The AI can
[01:33:00] interpret it. You know exactly what you’re investing in. Uh but the beautiful thing about this is that we’ll now have another choice and then hopefully a bunch of other choices other than just a New York Stock Exchange and and NASDAQ and choice is the answer. You know, as long as there are competing exchanges, this will all get solved. And but we desperately need it because the pathway to liquidity is the pathway to investment. The pathway to investment determines whether the US wins or loses the AI race against China. And so this is critically important to >> so important. The whole idea of of all the accountants and lawyers you have to hire to take a public a company public is to assure that a grandma who buys your stock isn’t being ripped off. Right? That’s basically it’s cover your ass across the entire uh entire process. And it should be possible for an AI system to evaluate a company and its stock and make sure that in fact everything they they say is correct and just accelerate this by a factor of a
[01:34:00] 100fold, not just twofold. >> Yeah. Yeah. And for a while there, you know, we thought the blockchain by itself would solve this problem and that that you could raise capital through a coin offering, an ICO, and that would replace the IPO. But you do need it just doesn’t work without some regulatory guarantee. Otherwise, it can get corrupt way too quickly. So, this is the perfect hybrid. I think I like that. I like the general trend that we’re moving from centralized systems to decentralized systems. And this this will bring in the abundance economy in a way because it’ll allow um much better choice and markets will thrive. >> All right, we’re going to close on health and uh and a discussion about the singularity. You’re not going to want to miss. All right, so uh in our health segment here, I’m going to play a video. This is an ALS patient with a neural link feeding themselves. Um here we can see individual, I think it’s their their third patient, they’re controlling a robotic arm. uh and being able to with their thoughts uh pick up food and feed
[01:35:00] it to themselves. I mean, this is the beginning of the merger of of humans and machines. Um it’s crude. Uh we’re going to see, by the way, this year at the Abundance Summit, I’m going to be bringing two of the top uh BCI companies, a company that’s got, you know, an order of magnitude or couple orders of magnitude more uh bandwidth connectivity between the Neoortex than Neuralink has and another one that’s being backed by Sam Alman that is a brand new approach to BCI. So, uh, we are 90% full on the Abundance Summit this year. We were selling out way before. If you’re interested, you can go to, uh, abundance 360.com to learn more about it. But super excited about this uh, the acceleration of BCI. Alex, you want to add something on this one? >> Yeah, I mean, I first obviously this is transformative for for people living with ALS and and tremendous progress. I I also think more more broadly this is
[01:36:02] the beginnings of democratizing access to our visual excuse me to our motor cortex. Imagine anyone being able to to augment themselves in a variety of ways to control an exoskeleton. And also in some sense every sci-fi author I’ve mentioned in past I’m a sci-fi snob. Many many sci-fi authors just can’t resist the tendency to just extrapolate a single dimension like oh we get AI or oh you know the apocalypse happens in a very narrow way but but actually I think the future we find ourselves in is one where every single sci-fi scenario happens all at once. So, we’re getting AI and we’re getting cyborgs and we’re getting this and that and that that’s a much more exciting future and present to be living in >> in we often talk about it as you’re going to get Star Trek or MadMax and we thought it was one or the other and it’s clearly happening both like Ukraine >> same world same universe all happening it’s it’s it’s the ultimate crossover >> everything all the time everywhere all
[01:37:00] at once for me for me that robot ALS thing was a little bit in the hoham category in with in the sense that I would expect to see that happen like we expect we should be expecting these innovations at this point. >> How quickly our expectations get reormalized. >> Yes. >> Yeah. How quickly the miraculous becomes boring. Yeah. Well, honestly, you know, we’ll talk about this more in the next section, but I remember when Singularity came out for we’ve launched Singularity University, an article came out on CNET saying it was being led by Ray Krosaw, Peter Diamandis, and the noted transhumanist Selma. And I had to go look up look the term what is a what is a transhumanist? And it turns out in the definition, transhumanist is a human being who’s augmented themselves with technology. And it makes no sense to me that whole framing. So, we can talk about this more, but you know, Dave, you wear glasses. Are you a transhumanist? Like, where do the spectrum is ridiculous on this? >> All right, let’s move forward. I want to I want to hit our last two articles and
[01:38:00] get into our singularity debate here. So, uh this article is Google’s AI cracks a new cancer code. So, Google deep mind developed the AI model called cell to sentence-cale generating a new cancer treatment that’s never been seen before. analyzed tumors and tested 4,000 drug candidates virtually. Uh, incredible. Uh, Alex, what are your thoughts? >> Well, my immediate thought now is that Google cracks cancer with a new modality for a sentence model and Selma is going to say, “Ho, uh, what what what’s next?” No, this is this is I think a very important development. So, so the the first I think big thing to understand is the cellto-s sentence model. This is a a new modality in in some sense. So we we know models speak text, they speak video now very popular, they speak audio. Speaking cell uh and speaking cell proteomic expression is a whole new
[01:39:00] modality. So a Google coined this notion of a cell sentence. A cell sentence is a a sentence of genes. So it’s literally like text with a sequence of gene names ordered in descending order by how much the gene has been expressed in the cell. That that’s a cell sentence. So treating cell sentences as first class citizens alongside English sentences enables you to have conversations with virtual cells. We we’ve spoken on the pod previously about how in principle medicine can be solved by simply having the world’s best virtual cell simulator and virtual organ simulator and virtual organism simulator and just having questions with these these simulations. This is I think a very very important step being able to have a conversation with the cell and asking it like how do we how do we solve cancer? >> I I’m not pulling the whole hum card. This is amazing. This is this is really huge. I totally get where you’re going. This is monstrous. Yeah, it’s it’s amazing in the narrow sense and in the broad sense. In the broad sense, it’s a
[01:40:01] great example of a domain where AI can think unlike a human >> uh you know about you know things where we just don’t have have any intuitive intelligence because we don’t live at the cellular level. The AI doesn’t care. It’s just data from the AI’s point of view. So it gets very very good intuition far better than any human being you know. So you know in parallel with that you’ve got things like magnetic containment of a fusion reaction very hard to visualize. AI just cuts right through it and like all these other areas because we we tend to continually show those charts that benchmark here’s Gemini, you know, 2.5. Uh here is a human doing this exact same task. But what about all these domains where people just don’t operate naturally? This is a great case study to to track closely. >> Yeah. You know, we heard this tickling AI. >> Would you would you agree with that? >> No. How would you how would you define AI? >> I don’t know. It makes sense. >> You know, we we heard uh Demesis Habus
[01:41:02] say curing all disease within a decade. We heard Daria Amade talk about doubling the human lifespan in the next 5 to 10 years. I mean this is you know sort of the bent twig that shows us where we’re heading. And I just you know for me one of the most important mindsets someone can have is a longevity mindset. And it’s the belief that in fact we’re going to be heading towards longevity, escape velocity, which is uh our next article here. We’re going to be wrapping on this and in discussion of the singularity. So let me just read this out loud. Ray Kerszswwell reinforces his optimism on longevity escape velocity lev by 2032. So he’s predicting that by uh the early 2030s nanobots could connect with human brains directly to the cloud. By 2045, humans will reach the singularity. So, hitting on the first topic of reaching longevity, escape velocity, what is lev? Uh, so for the last century, most of 1900 through 2000, we were adding about
[01:42:02] uh 3 months per year that you were alive. So, you for every year that you were living, you’re extending your life for a quarter of a year. The idea of longevity escape velocity is that there’s going to be a point that for every year that you’re alive, science is extending your life for more than a year, right? And at that point, it’s a choice of how long you might want to live. >> I have I have a great example. I have a friend who has a kidney problem and they’re giving him the drugs to stabilize his kidney for the next few months because then the drugs will be available to solve it for another few another five years. So you don’t have to solve the whole thing. You just have to solve to the next hop. >> Right? And with all the stem cell therapies, gene therapies, we’re going for 3 months to 6 months to 9 months. Then we cross that threshold where we’re adding more than a year to your life per calendar year that goes by. And at that point, you can live for a theoretically arbitrarily long period of time. >> Yeah. And we just I just finished my uh abundance longevity trip and we had 50 companies that were each contributing
[01:43:00] towards this direction. Uh and some amazing amazing tech. It’s it’s coming. It’s coming so fast and it’s not because we’ve gotten smarter or have uh you know done anything uh linear. It’s the impact of AI. Every single biologist who’s driving breakthroughs is driving it on the back of of AI. Um >> so the two two truisms used to be death and taxes. So we’re going to solve death now and taxes may be solved by crypto. So there’s like we’re kind of there. There’s nothing more to do. Uh, I want to I want to hit on the conversation of Ry stating that we’re going to have the singularity by 2045. Uh, and I’d love to get some thoughts on here. So, you know, what is the singularity by Ray’s definition? I asked my I asked Gemini 2.5 since since Ry was the futurist and residence at Google and uh I said, “What are the three things that cannote the singularity as Ry defines it?” And number one, it’s non-biological intelligence exceeds biological
[01:44:00] intelligence. Well, I kind of feel like it’s there now. Number two, human machine merger. Humans become hybrids or non-biological again moving there very quickly. And then three, radical transformation of the world. Biology, physics, and society beyond what we can easily predict. And this is, you know, Alex, what you speak about solving everything. So, what do you guys think is uh you know the singularity? >> Well, I’ll give you my two cents on this this comment and because Rey to me is just brilliant and yeah, he he predicted the singularity and he named it back in the 1990s and his timelines and his curves and everything. He got crapped on so hard by so many people for so many years and he’s going to land it like right on the nut. Um, and I hope that history documents it that way. I think what he’s wrestling with right now is we all agree on this podcast that we’re right in the middle of the singularity at this moment. And so I think he’s trying to push off the date to a point in the future where it doesn’t people
[01:45:02] don’t bother him about it anymore. I’m just guessing. Uh and he’ll just let it play out from here because you know one one thing about futurists Peter I know you you deal with this all the time. when you’re right 99 times out of a 100red, you know, Peter, you remember he he said, “Yeah, everyone’s going to have a a computer that’s more powerful than the biggest supercomputers in the world in their pocket and they’ll be wearing it around.” And now everyone’s like got an iPhone and they’re like, “Oh, is that what he meant?” Huh. Well, who cares? You know, everyone knew that was coming. >> But they remember the one they remember the one time that you were wrong. >> Yeah, exactly. He thought we’d have full self-driving and we’d all be in our self-driving cars today. He got hit so hard last year and the year before because we’re not in our self-driving cars. and that’s because of regulatory slowdowns, but we’ll be there imminently. And and it’s just so frustrating to see him get picked on that way. So my guess is he’s saying, “Look, don’t bother me about singularity definitions till 2045 because it’ll it’ll be history by then anyway.” But we all know right here, right now, the AI is clearly
[01:46:01] self-improving. It’s doing its own chip design. Uh we saw that Brockman article earlier. We’re right in the middle of the singularity as he originally defined it. >> It is now. The singularity is now. Alex, >> I I think I agree with most of the prediction except for the discontinuity. That was that that third item. So So there’s an intellectual strain that that starts with Good uh coming up with this notion of an intelligence explosion. Then good passes the the torch to Verer Vji who popularizes the notion of technological singularity. Then passing the the torch again to to Rey. I I think this notion that we’re going to have an intelligence explosion that somehow leads to a point of discontinuity where we can’t predict what’s on the other side. That’s the part that I I I struggle with. >> That’s a definition of a singularity. I mean, the idea that there’s event horizon and beyond which you can’t see what’s happening next. And it sort of feels like we’re in the midst of continuous event horizons, right? We’re
[01:47:01] >> but I also think like I I I feel may maybe this is just one person’s perspective but I I kind of feel like we have line of sight as to what happens next. Like we’re we’re we’re solving intelligence. Uh it’s well on its way to having been solved using super intelligence to solve math, science, engineering, medicine, bunch of other things. Solve everything. Okay. And then presumably we’ll discover the nature of our universe. um, you know, nature of of the the trajectory of intelligent civilizations. I I would expect to gain deep insight into that over the next call it 10 plus years. And that that to me doesn’t feel like a discontinuity where you can’t see what’s next. I I think you have pretty good line of sight. >> And I think it’s such a good outcome. You know, if you if you go back to the original book, that’s mind-blowing original book from 1990s, the the step function version of this is kind of weird in that the AI is in a box somewhere and it’s self-improving itself and it finds some way to create its own nano nano tube based compute. So, it doesn’t need GPUs from Nvidia. It’s
[01:48:00] building its own compute inside its own kind of box world. And you you go to bed one night, you wake up the next day and it’s like it’s taken over everything. >> And that’s not a good outcome in any way, shape, or form. The way it’s evolving looks like a step function on any reasonable time scale, but as you’re living it, it’s actually manageable week to week. You if you really focus, you can actually see next week’s innovations coming and you can actually benefit from it and also guide it. So, it’s actually working out better than than you ever could have predicted. U you know, plenty of risk, plenty of things we need to get on top of, but it’s it’s a great version of the singularity and just embrace it. Um, can I say a couple of comments here? >> Of course, Liam. I would expect nothing less. >> You know, back when we had the founding conference of Singularity University, I’d never heard of you. I’d never heard of Ry. I’d never even heard of Singularity, right? I walked in totally blank. Um, cuz the NASA people, >> how did you get invited to that? >> What happened was I when I was at Yahoo running Brick House and heading up
[01:49:00] innovation there. Um, I got uh I got I set up a relationship with NASA to do some interesting projects together, right? In back in the old days they had millions of satellite images. We had millions of flicker users. Could we help tag those? My dream was can we have a satellite hanging in our office to show that’s what real innovation looks like. And can I tell a little little snippet of an anecdote? So we used to have speakers from NASA come and speak at at Brick House which was and you have to we had an event once where you have to imagine this event in San Francisco with 300 uh software developers all with tight jeans, slick back hair, MacBooks on their laps, white socks, etc. And we had this uh 75year-old guy from NASA come and speak. He was he’d worked on the lunar program on the Apollo program. Okay. And so he did this talk and everybody’s kind of mostly people are looking and Q&A comes along and I said what’s the biggest difference now between the space industry now and when you were you know working on the Apollo program and he goes huh good question he goes maybe it was computers and I’m like
[01:50:01] what do you mean he goes well we had no computers so all information was transmitted via carbon copy paper the pink sheet went there the green sheet went there the yellow sheet went there and all of a sudden you I was standing at the back of the room and you saw these 300 developers all look up and their brains all were exploding at the same time going, how did they how did they do what they did with carbon copy paper sheets being passed around, right? It’s like kind of an unbelievable thing. So that’s how was it got through the NASA discussion. The NASA people one day called me and said, “Hey, we’re helping host this founding conference for Singularity University. We’re bringing 70 thought leaders together. Come along.” So I was actually supposed to take Lily away that weekend, but it was so weird this thing. I was like, “You know what? Let’s cancel the weekend. Let’s let’s go to this thing.” And that’s where Peter, you and I met. Um, and this concept comes along called the singularity. A few weeks later, you said, “Hey, come along and help us run it.” And I remember getting a call from Brad Templeton uh that later that day. And he goes, “Hey, I didn’t know you were a singularitarian.”
[01:51:00] And I’m like, “Wait, what? Wait, what’s that?” And and so then I had to look that up. And what I what attracted me was the original thesis was we can now use technology to address grand challenges because these technologies scale naturally and that was the most profound and interesting thing which was the whole compelling thing about 10 to the 9th and Peter your uh vision of using these technologies to address global problems the secret thing that you don’t tell people is that you’re trying to find more teams to work on enterprises that’s the part that you don’t talk about public anyway so singularity gets bill and we were talking about the singularity which was at that time it was defined as the point where machine intelligence overtakes human intelligence that was the common framing >> which we’re not >> and I disagreed with it publicly I disagreed with because a go back to my intelligence rant we don’t know what intelligence is there was like a dozen facets of it and b my second part was what constitutes overtaking the minute I can prescriptively describe a task an AI robot’s going to do it much better than
[01:52:00] me anyway so it’s a bit of a nonsequittor so the kind of that tension comes along and then Ry writes the singularities nearer and becomes more of a process and I really like Alex Alex’s framing of it that it’s a if for us living in it it’s it seems normal. Um but when you step back in history it’s going to be this unbelievable inflection point. Um and this is I think the key part of it where we just kind of live through it and we get there. outcomes of it of us being able to solve all major problems with AI now with and because AI will then solve material science etc is the most profoundly amazing part which is why we get so kind enthusiastic and bubbly excited about this and I apologize for the whole hum cars now and then it’s just that sometimes you know you get to a point where you’re living this conversation >> you’re spoiled you’re spoiled >> you you get spoiled you’re kind of like oh yeah we yeah we we would expect to see that etc and then you go nuts at the lites who are going well this is a bad idea and and cancel the most important projects in the world. I mean, anyway,
[01:53:01] >> yeah. Well, that that last point I think is if anything’s changed in the last 3 weeks that’s really palpable to me, it’s that the you know, there’s always going to be doubters and haters and you know, they’re all over the place. And um but the counterveiling voice to that has been Dennis Habis and Sam Oldman and Elon Musk. But very recently, Sam has toned it down. Dennis has toned it down. And it’s not because they don’t believe that it’s happening right now. >> It’s because it’s happening anyway and they don’t need to promote it. They’re doing it inside their building at warp speed and they don’t need another picketer outside the door tomorrow. And so, so what you’ll see now is kind of this kind of, hey, did things get quiet? Did things slow down as they explode inside, you know, various rooms and labs? The the framing I like the best is that all our previous models for how the world operated break down and we need totally new models like Alex talked about needing new benchmarks, right? We need to formulate totally new models for
[01:54:01] where the world goes. GDP for example is not a workable model >> and it’s the new it’s going to be the new social contract that needs to be reformed as well. How do we how do people >> how do people uh you know use this? How do they get their dividend from AI? Whether it’s reduction in the cost of all the things that they need >> uh increased, you know, increased agility. >> Full disclosure here with my secret plan has been that I’ve been building this exo community around the book, which is now 40,000 people in 150 countries speaking 47 languages. What we’re actually doing is building kind of like a peace corps to help this transformation because we’re going to need an army of people that are practiced and versed in this in this model in the new models that are coming to be able to ease that transition. Otherwise, we’ll end up in several hundred years of the dark ages. >> Yeah. You know, See, you Dave and I are going to be in Saudi in about a a week, 10 days time. And one of the projects we’ve been working on with EMOD is how
[01:55:00] do you use AI to provide every sovereign nation the ability to govern better to establish policies cuz we’re going to have a lot of disruptive change coming. You know, all of a sudden when people are living 30 healthy years longer or humanoid robots are in the hundreds of millions and at 40 cents an hour. um these things are going to change nation states fundamentally and their ability to rapidly adopt new policies uh educate their their populace spread the wealth if you would um is is super important. So we’ll be doing that and then uh this coming week uh I’ll be with you and Immod uh and Eric Pulier recording a WTF episode live at uh uh at X-Prise Visionering in Malibu. Again, if you’re interested in joining us for that, we have a few tickets left for X-Prize Visionering. You can go to x-priseze.org uh to learn more and we’ll put uh the we’ll put the link as well for visioning
[01:56:00] down in the chat. >> I I would like to just again point out what Dave suggested at the beginning that Ry will go down as I I kind of think of him not as a real person. He’s in like an avatar from the future. I think he proves that time travel does exist cuz how the hell did he come up with this stuff decades ago and it’s proving he must be just coming from the past from the future into the into the present. And I’ll tell one quick anecdote. We once uh you know Ray would come and speak at Singularity. I’ve heard him speak maybe 60 times. I’ve never not learned something which is really really really frustrating because you have to listen for that little nugget of gold. We once were able to um we once got him two glasses of wine before he did his talk and and nobody has ever forgotten that to our session where he just kind of riff on and it was utterly brilliant. It was it was there he said that language is a really thin pipe to discuss topics as complex as some of the ones we’re discussing, right? And and you have this unbelievable wisdom coming from and this ability to perceive decades in the
[01:57:01] future with unbelievable accuracy and accurate framing etc. I think this is going to go the the accuracy of what he has uh kind of put down and he’s willing to put it down and be kind of gauged by it is going to go down in history. >> Amazing. That’s a good note to close on. Uh Moonshot Mates, love you all. Thank you for your intelligence, your predictions, your humor. >> We need to get Ray to write a book. The singularity is now or it’s come and gone. Uh either way, and maybe we should get Ray on this podcast with us. We should and I think a really important comment is what Alex pointed out is like how do we navigate this this future because we can see it’s coming now. So what does that future look like and and let’s start painting that piter. >> All right have an amazing week guys. Uh be seeing you and talking to you very soon. >> I have reading to do. >> Yes you do. Thanks Peter. Bye guys. >> Thanks guys. >> Every week my team and I study the top 10 technology meta trends that will
[01:58:01] transform industries over the decade ahead. I cover trends ranging from humanoid robotics, AGI and quantum computing to transport, energy, longevity, and more. There’s no fluff, only the most important stuff that matters that impacts our lives, our companies, and our careers. If you want me to share these metat trends with you, I write a newsletter twice a week, sending it out as a short two-minute read via email. And if you want to discover the most important meta trends 10 years before anyone else, this report’s for you. Readers include founders and CEOs from the world’s most disruptive companies and entrepreneurs building the world’s most disruptive tech. It’s not for you if you don’t want to be informed about what’s coming, why it matters, and how you can benefit from it. To subscribe for free, go to dmmandis.com/tatrends to gain access to the trends 10 years before anyone else. All right, now back to this episode.
[01:59:00] [Music]