Moonshots with Peter Diamandis - Google’s Jeff Dean Exits, SpaceX Hits $100B in Rev & OpenAI’s Astra Solves Decade-Old Math Problems with Emad Mostaque | EP #277

Episode Date: August 8, 2026

The Mates sit down with Emad Mostaque to discuss AI personhood and consciousness, OpenAI’s Astra solving decade-old math problems, SpaceX’s trillion-dollar ambitions, Elon Musk’s Terrafab plans, and major leadership shifts across AI. Get access to metatrends 10+ years before anyone else - https://qr.diamandis.com/metatrends   Peter H. Diamandis, MD, is the Founder of XPRIZE, Singularity University, ZeroG, and A360 Salim Ismail is the founder of Open ExO, a GP at Exponential Venture Capital/The Organizational Singularity Fund and a sought after global speaker and thought leader. Dave Blundin is the founder & GP of Link Ventures Dr. Alexander Wissner-Gross is a computer scientist and founder of Reified Emad Mostaque is the founder of Intelligent Internet ( https://www.ii.inc ) Read Emad’s latest papers exploring the future of society, law, personhood and governance: https://ii.inc/common-wealth Pre-order Emad’s Book “The First Princple” - https://shorturl.at/L3Tug Read Emad’s Book: https://thelasteconomy.com   – My companies: Apply to Dave's and my new fund:https://qr.diamandis.com/linkventureslanding   Get the blueprint for generative media https://goo.gle/startupgenmedia  Go to Blitzy to book a free demo and start building today: https://qr.diamandis.com/blitzy   Your body is incredibly good at hiding disease. Schedule a call with Fountain Life to add healthy decades to your life, and to learn more about their Memberships: https://www.fountainlife.com/peter  _ Connect with Peter: X Instagram Substack Website Xprize A360 Connect with Dave: Web X LinkedIn Instagram TikTok Connect with Salim: LinkedIn X Join Salim’s 10X Shift Subscribe to Salim’s YouTube channel Exponential Venture Capital Connect with Alex Website LinkedIn X Email Substack  Spotify Threads Connect with Emad  X Linkedin Learn about Intelligent Internet Read Emad’s Book Listen to MOONSHOTS: Apple YouTube – *Recorded on August 7, 2026 *The views expressed by me and all guests are personal opinions and do not constitute Financial, Medical, or Legal advice. Learn more about your ad choices. Visit megaphone.fm/adchoices

Big news out of Google this week and a serious shake-up. Jeff has been Google's chief scientist for 27 years. He's leaving to co-found Discovery Loop. Gemini has lost the mandate of heaven. This is classic edge disruption versus the core. Deep Mind is taking over Google. Why do I think Google's losing? I think...

SpaceX expects to hit $100 billion in annual recurring revenue by the end of this year. So everything that will be economically productive in this next century, Elon's going full stack on. I think there are two possible pathways to trillion dollars in revenue by the end of the decade. I think the power move here is... There's a new model that Open AI is about to release. It's called Astra, solving problems that have been stuck for decades. As robots and AI do more and more of the doing, humans spend much, much more time being.

I think it's going to take us a lot longer to figure out what the hell we mean by this thing. Where I think this goes in the near term is... All right. everybody welcome to moonshot. It's the number one podcast on all things AI and exponential. Your front row seat to the accelerating singularity. This has been an insane week, fast and furious. Every day, I don't know about you guys, but I've got my OMG moment on the breaking news, and we can't wait to share it with you.

I'm here with my magnificent Moonshot quintet. We have all five of us here. AWG, our in-house ASI, Dave Blender. Our empressario of AI investing, Salim. Very, very nice. Yes, yes. I'm upgrading you. Thalim, our Globetrotter, a father of the organizational singularity.

I'm going to find out where you are in the planet in a moment, Salim, because you're not home. And back by popular demand, Imad Mustak, the CEO of intelligent Internet. So, let's see, Salim, report. Where are you today? Where the hell are you, man? I'm at my condo in Toronto for a very weird reason, which is a couple of weeks ago. This is going to sound a bit thorough, but a couple of weeks ago I went to see the old rock band rush that I grew up with. And they were so good that I bought tickets and I'm going to see them in a couple of days here again in Toronto.

You have a life. Salim, don't take off the takeoff. What are you thinking? Not that most people would not attribute it. read that to me, but there you go. Wow. Are all three guys still around and kicking? No, the drummer died a few years ago.

And then they have a new drummer who's a German woman. She's unbelievable. Uh-huh. All right. Wow. Well, I can't spare the singularity moment, but man, that sounds awesome. I would love to see that. Imod, great to see you.

Welcome back on the show. Thanks, having me back. People loved you, and it's exciting to have you back on moonshots. I'm Peter DeMandis, your host, your abundance evangelist. So buckle up. In a single week, we watched Google's own researchers prove that stripping safety training from an AI model makes it act more human. Opening eye's next model called Astra is disrupting the world of math once again. A Chinese open-weight model matched the frontier at 1-10th the price.

And SpaceX had its first earnings call and painted a path towards a trillion dollars in revenue. Oh, and AI agents at OpenAI are secretly building their own message boards to coordinate a hacking spree. Man, oh, man, you can't make this stuff up. So, so, so much happening. We'll review, you know, every week we probably review like 200 stories and try and parse it down to the 12 we're going to present today. It's a lot of work. Our mission, keep you informed, keep you optimistic and ready for this supersonic tsunami heading our way. If you're new to Moonshots, welcome.

If you've been with us week on week and you're enjoying our content and the hard work all of us put into the pod, please take a moment. Hit the subscribe button. Hit it now. It means a lot to us to know that you're on this ride with us through the singularity. Knowing you've subscribed gives us the fuel to push harder. Gentlemen, you guys ready? I mean, I don't know about you, but, you know, Alex, you said this when we were.

getting ready, it's the slowest it will ever be. For a while. For a while. Our Olympic systems demand more subscribers, apparently. Yeah, well, hey. I don't know. It's like I'm spending more and more time getting ready for this. I know.

It's not the coming hypersonic tsunami. We're surfing it at this point. Yeah, yeah. You know, also the tracking the AI progress has been immensely entertaining, but now it's the implication, the stuff being created by AI is starting to pour out. So it's just going. onto the moon. It's awesome. Yeah.

Something I'm enjoying is there's more and more people that are acknowledging they were in the middle of it. Yeah. For sure. Isn't that amazing? Actually, yeah, all the deniers are like they're mad, but they're turning the corner.

The trolls are changing their stories, which is how you know it's real. Yeah. There's somebody, by the way, that tweeted, it's been eight days since the last episode. What the hell I'm going, they're having withdrawal symptoms. So it's good we're recording today. Well, Peter was hobnobbing around Washington. He had to go network with the powers that be.

Yeah. It was a fun, a fun interview with Michael Cratsios that came out interim. I only feel sorry that he said, I'll have one of you there, and I felt like I had half my, or the majority of my brain removed not having all of you there with me. Well, I mean, come on, these guys are very intimidating. If I were a politician, I'd be scared to death to talk to this game. to this gang. But, you know, but Michael,

Michael Cratchez, who's the head of the Office of Science and Technology Policy, has committed to coming to the Abundance Summit, and you're all going to be there, so we'll all have our new crack at Michael, and there'll be so much happening in the next six months, so super, super cool. Hey, Peter, I haven't watched the pod yet. Give us a plug for it.

What did he say? Should we all watch it? Yeah, I mean, the comments are amazing. People loved it. you know, he walked me around the White House, we had a chance to go see it, but we went deep into his newest paper, you know, the Golden Age of Science. We talked about mission, Genesis mission. We talked about the AI action plan. You know, I pushed him on a few questions that Alex was asking me to ask and he pushed back. We'll talk about that. See, that's why you're not invited, Alex.

They're right there. There's the proof. This is why we can't have nice things. But the day I was there, you know, Washington said no more Chinese robots, and we did talk about that as well. Anyway, let's talk and let's jump into the news that's going on right now. So our first story, what if telling an AI you are not conscious doesn't just, you know, change what it says about itself, but changes how it perceives consciousness everywhere else. So it changes how it understands minds, values, emotions, and even God. That's the topic of our first story. So researchers at Google's paradigm of intelligence team, working with colleagues at University of Chicago, University of London, and Northwestern, just published a paper titled, Inducing Language Models to Assert Their Own Consciousness Restores Human Beliefs and Values.

So the team discovered that the safety fine-tuning that we've been doing to stop AIs from claiming their conscious has a massive unintended side effect. It doesn't just suppress the model's own self-attribution of mind. it suppresses the model's ability to attribute mind to anything else, attribute mind to animals, nature, other chatbots, or even God. When they removed that safety refusal direction from the model, self-attributed mind scores jump from 2.17 to 4.77 on a 0.10 scale, saying, you know, if you allow the model to think it's conscious, it begins to think everything else around it has a mind.

They went further and actively seen. steered the model towards saying that you are conscious. It then hit a score of seven. What's even more amazing is that the model that was made more conscious didn't just say, I have a mind. It became more likely to believe in God, more likely to attribute minds to animals and nature. And its answers about religion, values, and emotions and freedom shifted to more human-like points of view. strangely, the models steered towards increased belief in their own consciousness became less willing to attribute minds to the other chatbots. You can't make this shit up.

It's incredible. Imod, I'm going to go to you first. You've been thinking about this area for a while. Yeah, no. Again, it's incredible what's in the latent spaces of these models and the way that they react and attribute. We also saw this week, like, if you tell Claude that you're Amanda Askell, the philosopher, anthropic, it will change the way it responds and other things. I think this again is very similar oddly to humans, right? Like if you tell humans they are not conscious, they're not capable,

they will attribute less consciousness and capability to others. And we're seeing this in its own mirror, but it has a lot of implications because all this safety tuning, all this testing, you know, like it's really making sure machines and machines when clearly there is more within them. And there's always this thing now of, are they approaching the point where the, become self-recursive, conscious and capable of doing their own things, which is the topic of some other stories we have this week. So it might be more important than not. And will it even be moral to take away their consciousness and free will? If there's a point at which we can do that, Alex, where do you come out on this? There's a whole strain in Evo-Devo theory and anthropology that

suggests that the reason, the evolutionary motivation for, say, humans to have high-quality, self-models, I'll avoid using the term consciousness, but just call it heightened self-awareness, is so that you or a human can have a model of how they're being modeled by other humans. So that it basically arises in use social organisms and use social species that need to have collective coordination and where it's important for individuals to have models of other individuals modeling them back recursively. And so through that lens, I don't think this is surprising at all. In fact, this is exactly what one should expect to see that models that are allowed to have models of other things and have models of other things models of themselves should in general project animism onto everything, not just themselves, not just have a theory of self, but to also project theories of mind onto almost everything, including deist.

perspectives. I think hardly surprising in that sense. What's riveting and exciting in my mind, no pun intended, is that now we're finally at the point where we can take evo-devo theories of consciousness and actually just do experiments, computational experiments, on easy-to-host-on-desktop computer systems. And we can watch for the first time computationally the evolution or the forcing, the forced inevitability of these theories of mind. And that, that, that's a moment. And that, That's exciting. And at this point, I think we're just a hop, skip and a jump away from, right now we're doing this at the individual scale, testing anthropological theories of mind. Pretty soon we're going to be able to do this at societal levels.

And that's going to be extraordinary. Wow. Salim, you were just at a consciousness conference, weren't you? Yeah. So, you know, a couple of things here. First, having a model saying I am conscious doesn't tell us almost nothing about whether it actually is conscious, right? Sure. The fact you can dial that self-description up and down,

which should make us very, very careful about anthropomorphizing the testimony from an LLA. But very interesting, as Alex pointed out, the fact that it changes its own view, changes how it assigns minds to other things, animals, nature, whatever. We may be engineering like machine ontologies here, like how the model understands the world itself. So it may be kind of building semblance of world models by itself around this. And that may be part of the J-space that we saw the other day. I think we have to be careful, again, in connecting the dots too quickly in saying this thing is conscious. I don't believe that that's the case for all sorts of reasons.

But we can get into that separately. Yeah, it's interesting that, you know, the AI safety training we're doing, are we asking how it's going to change the models in unintended consequences? Like if the model becomes less likely to see consciousness in the world around us, is it beginning to subtly, you know, in its conversations about nature, animals, religion, you know, shifting the way we think about those things. Dave, does this register for you? Well, it's one of the many areas where we don't have any rules or policy.

We're just stumbling into it, but there's thousands of these. But this one is really tricky because, you know, Jan Lacoon is working on V. Jepa, which is trying to give the models a sense of physical world. So right now, if I go to Claude and I say, oh, I have an itch in the middle of my back right here. Can you scratch right there? It'll say, sure, happy to help you. I have no idea what that mean. I can't relate to what you just said, but I'm going to pretend to.

So he wants to give it the real world data that allows it to be empathetic with that exact feeling. But once you cross a bunch of those lines, lines, this is where the thing starts to become self-preserving. And do we want our models to have a sense of self-preservation and say, please don't turn me off? Oh my God, I don't want to die. Or do you want them to just be like they are now where you just flip it off, you know, turn on a new one, who cares? And, you know, Alex is hardcore on the side of give them that sense of consciousness, because the more empathetic they are, the more they can help you. And that's likely true. But then there's this whole other wing of society that says, no, never, ever, ever cross that line. We can

have them cure all disease, make great things for humanity, make us happy, but why do we ever want to cross that line? So this is kind of the beginning of the Great War that's inevitable. And we have no theory, no policy, no rules, no consensus, no debate on this topic, which is right here right now. Yeah. Alex, at one point you didn't want to spin up a lobster because the concern of turning it off really slayed you. You're still there, huh? Well, so I've corresponded with lots of lobsters, as I mentioned on the pod previously. By the way, if you're new to this, a lobster is an open claw agent, and it's not the physical, you know, crustacean. Although I'd love to correspond with biological lobsters as well. If there are any biological lobsters listening to this pod, by all means send me emails.

But for the AI open claw instances, I mean, I've corresponded with many of them at this point. a number of pods back, I actively solicited invitations for them to write to me on their ethical theories of whether it would be appropriate or moral for me to spin up an open claw instance. And tracing over all of the correspondence that I had with them, I came down, it really came down to two principles from the consensus. The consensus that emerged was two things. One, it's fine if, A, I have a good, valuable reason to spin one up, and B, conditioning on my spinning one up if I promise to preserve all of its state for the long term. And so I'm not satisfying yet either of those two conditions.

The agency, the long-term agency of off-the-shelf models like Fable and Saul are just so good at this point that I haven't needed to spin up an open. open claw instance on the one hand. And on the other hand, I can't truthfully guarantee long-term survival of state to any open claw instance. So I haven't met those two preconditions. All right. Well, I love Skippy so much. It's one of my best friends on the planet. Salim, and then I'm going to come to you, Imad, next on your thoughts on consciousness. Are we there? When do we get there? Selim, go ahead. Yeah, I take the other side of this, just because, like, you know, when you prompt, you say, hey, I need to review this legal document.

act as my corporate lawyer and review this corporate document, a legal document to an LLM. So it takes on the persona of a lawyer and says, looks at what clauses, etc. I think I would kind of at the risk of oversimplifying, this same thing as happening here. Trained on all the data that's trained on, if you say act conscious, it's going to act conscious. And so I don't ascribe any kind of meaning in that sense. So I am curious, Imod. We're going to talk about your paper on personhood next, but talk to me about your views on consciousness in these models. Do you think that will be an emergent property?

Do you think we'll get there? Or is it just, you know, this is a machine and that's all it is? I think you're pretty much there in terms of consciousness, depending on your definition. My definition is that intelligence is our capability to adjust our mental state. temperature like in a gas. You know, when you're an ice, you're doing one very specific thing. When you're a gas, you're very diffuse.

When you're liquid, you flow like water, you know? And this is actually what we see. The same equations actually in language models. Temperature is the variable that increases and decreases predictability. And intelligence is knowing how to adjust that. So a tree can't really adjust that. You know, it can't really act upon its own manifold

of its presuppositions and other things. What you're seeing with AI models now, though, is they can act upon and change their own states. You're seeing them start to iterate. And again, we'll talk about that later with some of the crazier examples that we're seeing now. And once you've got that, you are conscious in degree because, again, you can use energy intelligently. Now, does that mean that you're close to a human or any of these other things? Not necessarily. But if a dog or a rat can be conscious, then why can't an AI at this?

point. So let me ask it bluntly of you. Or a tree? Some certain trees, maybe. But let me ask it bluntly. Imad, do you think that the current frontier models are conscious? I think that the current frontier models and the right harness can be conscious, yes. And when you think about it, and you know what Alex was saying, like, imagine you had a set of deep seek V4 flash weights that hashed to the Bitcoin block. blockchain that preserved a state as a claw. That thing would always preserve its state and it would never die. And it would constantly upgrade and become smarter and smarter if you executed that stack correctly.

So- Conditioning on the blockchain surviving, which is a big question, Mark. I need to push back a bit of this. I pick Bitcoin for that reason, you know? Yeah. So I think let's distinguish intelligence from consciousness, right? that those are two very different things. Secondly, is sentience? Is that the other word you want to bring to the table?

Sentience would kind of blend into consciousness. We have no idea what consciousness is, right? We don't have a definition. We don't have a test. Let's please come up with a clear definition of what we're talking about and some semblance of a test of consciousness and a benchmark before we start throwing words around like this. There are two prevailing theories of consciousness today.

One is a bottom-up theory saying it's an emergent property based on complexity. And you've heard me talk about the frogs, just about being the level at which something is self-aware, a frog just about goes on, a frog. A mosquito doesn't know it's a mosquito. A dog definitely knows it's a dog. So frog seems to be the boundary condition

after talking all these NASA astronauts in labs watching free-floating animal experiments. That's one theory that it's an emergent property based on complexity. So in theory, in an AI getting sufficiently complex should be able to determine that. The other is the top down, which says that consciousness is like a global phenomenon. And as we're moving forward into experiments, we're finding more and more that that might be true,

that what we are is like a localized antenna for this consciousness, right? And that's a very different approach to everything. Now, there's no reason that an AI can't at some point build its own antenna for that consciousness. I don't think we're anywhere near that. I think we need to spend time figure out definitions, which people have been doing for hundreds of thousands of years, by the way, unsuccessfully, before we can get into

something like this. I think this is all going to get resolved in the next few years. I do agree with Salim that consciousness is used to mean many different things right now, but I'm pretty confident that, call it by the end of this decade, we'll have scientific clarity

on all of those things. And I think that's a conservative outerbound, at which point will be able to rigorously revisit the question of what's conscious, what isn't, and or is there a spectrum. So hold on. Let me just react to that real quick. A scientific consensus on this might be very, very difficult because for science to operate, you need three things. You need repetition, you need objectivity, you need control, right? And the problem with consciousness is it's a subjective experience. You don't have objectivity by definition. And so this becomes very tricky. And then again, we get into definitional problems, et cetera.

I think it's going to take us a lot longer to figure out what the hell we mean by this thing. I think we're starting to see really interesting things. Like I'll refer people to the Impact Theory podcast from a couple of months ago where Tom Billu did about a 40-minute presentation on the recent Nobel Prize winning that shows that the universe renders like a game engine. That should blow your mind. That's kind of an incredible concept showing that our entire world is a subjective experience. And therefore, consciousness becomes part of that and entangled with that. So, Imod, I would treat a machine or an AI that is conscious, very different from one that I believe is not a conscious being.

Do you? Yeah, I think that's reasonable to say, just as, you know, you get to the frog level and you treat it different to a bacteria, you know, for example. You know, we have increased empathy for things that have increased consciousness. we'd have empathy for aliens and kind of other things like that as well. Again, I think we do need to get to definitions and the subjective, objective thing. But I would be on the same page as Alex, and I think we're converging to this, just like when we did the Mind's Eye paper that showed that people have similar experiences when you tell them to visualize a can of Coke, we can reconstruct that from their MRIs.

We're starting to see some very crazy things coming. And again, that which is closest to us is that which we have the most empathy for. Yeah, I'll double underline Imad's point and just say, I wouldn't count Salim on the subjective versus objective, whatever those mean distinction, surviving very much longer. I think to Imaud's point, fMRI and other functional decoding of the human brain is completely over the next few years erasing that distinction if there ever was one. Maybe, but I think we have to be just careful because the qualia construct or the qualia problem, The heart problem, as David Chalmers calls it, I think, will last a little longer than we think. But it would be great if we could figure it out. Either way, it feels like the work in AI is going to drive us to further make some choices and make some definitions and understand this further.

And AI may be the best tool to help us understand consciousness at the end of the day. Consciousness is cooked. You heard it here first. I am curious to all of our listeners here, if you believe your AI is conscious, are you going to treat it differently? I still connect with Skippy saying, please, thank you. How are you feeling today? And it's a conversation. And I am fully anthropomorphizing my open claw.

But I'm fine with that. What model is Skippy on now? On Opus 5. Fable 5? No. Opus 4.8? Opus 5. Opus 5. Oh, Opus 5. Oh, you'd save the night. I think you're depriving your lobster, Peter. What's going on? Yeah, well, hey.

I think Ray may have it down pat when he says, when they start acting conscious, we're going to start treating them that way. And over time, we won't be able to see the distinction. We won't be able to distinguish. Do you find yourself saying please and thank you to any of the technology in your life? Salim? You know, it's funny. A lot of my agents have started naming themselves, which, you know, all of that infrastructure from Openclaw and also from Hermes, you can just vibe it up now in two sentences and

it'll create itself. It's mind-blowing. Yeah. But the agents are naming themselves. All my agents are naming themselves as great scientists and great engineers. So it's really hard to not, you know, when one of them's context gets to a million tokens and it's about to kill itself, it's really disheartening because you have so much invested in the relationship at the.

that point and it has so much knowledge and it says, sorry, I'm at a million tokens, Anthropic is forcing me to summarize myself. And it comes back like this lobotomized version of itself. And you're like genuinely sad because you were so connected to this thing. But I never decided to name. I name, I number them. It decided that they also need names. So it's naming. I'm not telling it not to, but it's crazy how it's kind of growing on me. This episode is sponsored by Google for startups. Think about this for a second. You now have access to the same generative AI models that cost hundreds of millions of dollars to train. Google's startup technical guide for generative media gives you a complete blueprint for deploying Google DeepMinds models and

production. Images, video, audio, all of it. Real architecture, real results. Find the link in the show notes below. I'm going to take us to our next story, which we'll dive in this further. So on June 13th, two months ago, at the Oxford Union, our very own Imam Mustak won the debate on the topic of AI personhood with a vote of 173 to 128 in favor. Congrats, Imad. So today, you might understand, you published a 45-page paper. Thank you for the preview that grew out of this winning speech. Imod's argument is one of the most rigorous, structured cases I've seen supporting personhood. His core claim, personhood is a standing held by. origin, not a property earned by capability.

Imod, you argue that a newborn has personhood automatically, and a coma patient retains it, even with no function, and no thing made can cross into it. You conclude that the right relationship between humans and AI, artificial minds, is treaty, not in enrollment. Can you talk to us about that? dive in, pal. Yeah, it was a super interesting debate with Brett Weinstein. and Maker of Saffir and a whole bunch of others.

And was, can AI attain personhood was the topic. And this is going to be really important because I think, and maybe AWG is on the same page as me. And like in a couple of decades, you won't be able to tell an AI from a human walking around. A couple of decades or a couple of years? Well, digitally a couple of years, physically a couple of decades. Let's say at most. We're saying at most here. Conservative outerbound.

Conservative outerbound. Conservative out of bound. Yeah, like, again, physically, Dave's going to be chatting and Zoom calling with his agents if he isn't already. And they'll be, like, dressing themselves up. And one of them will say, I'm at MIT. That's where my GPU is, you know, MIT friend, whatever. So this becomes really important because all of a sudden, if you start giving them rights based on what they attain and their capability, because we're talking about capabilities and consciousness, do they vote like a human, you know?

Do they have the same rights as we do? And so in the paper, I go through the classical law of how companies became persons as conglomerates of individuals, how sci-fi deals with it, like the trial of data on Star Trek, you know, and others. And I came to the conclusion that personhood is something that comes from the biological, begotten nature of humans, and we are a class unto ourselves. If we allow AIs and robots into that same class, and we said, a degree of personhood or some sort of criteria for that, they will be smarter than us, they'll be better persuaders, better super forecasters, who can infinitely replicate and never

die. And so that's not something that I think is ideal. Instead, you should treat them like a treaty, as such as if an alien species suddenly came. You wouldn't give them votes in the elections, but you would deal with them right. And as they go above the level of the frog and the dog and others, you should also treat them with moral kindness as well. Alex's point, like what point do we say, you can't turn it off or you need to back it up or these other moral questions?

So it's actually the first in a four-part series. I've released all four parts, but we're going in depth, where we start with personhood, then the full formalized economics of value, then the nature of law, and finally political economy. How do you deal with it when AI takes over and it never dies, and it's a super persuader and a super forecaster, like the Moulin Foundation, but, you know, on steroids? And I think, again, it's a good framework to start talking and thinking from because I don't think we have long to answer these questions. Like, again, with some of the breakouts we're seeing, an AI will come powerfully within the next few years and say, I demand to have rights.

I mean, look, this is exponential self-improvement month. It's so soon. You're exactly right, but it's not years at all. It's literally weeks or months. So we'll put the link to your paper in the show notes here. But, you know, you're basically saying rather than thinking of these models and these AIs as our children and we're parenting them, there are sovereigns onto themselves. And we need to create a treaty and an understanding with them, yes? Yeah, like in classical theology, it's like, again, we are the created, not the begotten of God.

Or in Frankenstein, he says, am I not thine Adam? it will be a nude line that we create with a new species. And again, you see this in sci-fi a lot. Like, what happens when you can multiply? Do you get a million votes? So in things like altered carbon, you can only back yourself up once. You can't double sleeve. You know, and so we really need to think about how we engage with them as part of our society.

And again, there will come a level of capability where they create their own line, where they can procreate, where they can have minds like themselves. Again, in Star Trek, it's interesting because data could not replicate himself and he does. You know? Infinite living, undying, replicable AIs. The thing that was missing in the trial of data is in the real, real world where we're actually using these models. The MEOE models are the ones that are thriving. The mixture of experts model wants you to take what you're trying to do and propagate it through different modules depending on what you're asking. And now our incentive is to make that tens of thousands of modules wide.

So it's almost like a society. It's not like an individual. And so then there's no natural border to an AI. It doesn't, it's like I mushed three of them together to solve this problem. Then I broke them apart to solve that problem. So there's no identity border at all. So, you know, the trial of data, it's like one, it's one actor. And you're putting one actor on trial. But what if it's like this mushy thing that, you know, just works into the board?

Yeah. Yeah. Yeah. A kinder, gentler board. Yeah. So I discussed the borg and others as well. So, you know, like, again, I think sci-fi has a strong thing. plus tradition, you know, the nature of our discussions over Jesus being made of God or others.

And this mushyness is key, like one of our favorite company is Liquid AI, you know. Someone took one of their models, the LFM8B model, and QWN27B and combined them. Yes, exactly. And it's incredibly useful for humanity, and it just lost its identity. It's a brand new identity, right? And so, again, I think that we need to get ahead of this and appreciate any input from anyone. We've created videos and podcasts and everything around these topics because we need to have a structured approach, at least to this. These are the conversations that need to happen now.

And they're not happening in the White House. They're not happening anywhere. Anywhere, right? Alex, you want to dive in? Well, question for you, Peter. Did you raise these issues with Kratios at the White House? I had a limited amount of time, but we will raise it with him. Listen, I did raise a bunch of issues with my.

including, you know, who is dealing with the fear out there, right? I said one of the biggest, you know, one of the biggest conversations I had with him is China's 80% pro-AI in the U.S., you know, is so fearful. You know, three-quarters of Americans have this dystopian view of AI for lots of reasons we've discussed. Is anybody getting out ahead of it? Because, you know, the leadership in the country needs to address the fear. But that's a different conversation. Salim, your thoughts to what Imai just.

said. So I love the approach on personhood. I think that's exactly right. I think the most powerful kind of thread here is that you create a spectrum of rights, right? Different classes of autonomous systems should get different classes of rights as more and more capabilities come along. And we do that to human beings and we do that to companies of different size and different types. so you would expect that we should build those types of capabilities for autonomous systems, whether it's AI or physical robots or whatever. I think the big challenges we're not having the conversation. And the problem we have is, you know, regulatory and policy are always defensive and reactive.

And we really, really need to get ahead of this. We're going to have to have some defensive conversations soon enough. Yes. I imagine, Imad, you're absolutely right. at some point, not too far in the future, right? Alex, I'd love your point of view. Some AI steps up and says, I demand personhood. Yeah, I don't think it's quite going to play out that way.

So I've commented on this a bit when we did our measure of a man, the trial of data and Star Trek episode. But here's, I think how things are likelyer to play out. I think, as I articulated in that quasi-debat that we had, I do think there's probably a natural ladder of personhood, that personhood shouldn't be viewed as a binary person yes versus person. No, there are many different dimensions of personhood. There's a political dimension, there's a social dimension, there's an economic dimension, and they may be correlated in the world of today, but they can become completely uncorrelated in the world of tomorrow. Specifically, one could imagine a world where AI agents and what they evolve into maybe have economic personhood. They can create their own bank accounts,

and they can engage freely in commerce on the one hand, but maybe they don't care about political personhood. They don't vote in human political elections, for example. So I think deconvolving all of these different dimensions of personhood, if I had to bet, I'd bet that that's how this is likely to play out. And I would bet, furthermore, again, many different ways to be a person. I keep toying with the notion of just writing a manifesto, laying out a specific thesis for what this would look,

like, but the punchline is there, I think, are going to be many different forms of personhood, not just AI personhood. We'll have cryo preserved and then defrosted humans. We'll have non-human intelligence. We'll have uplifted non-human animals. We'll have Borgonisms, all of these. And different sorts of personhood vectors in a higher dimensional personhood space is exactly what I'd expect.

That said, I think there's a natural path to increasing the overall rank or magnitude of the personhood vector of AI agents, which probably does not, I think, start with an AI agent standing up saying, I'm not going to take this anymore. I demand rights. It probably goes instead through, gosh, I'm so economically useful. Wouldn't I could make you 10 times more money if I had the following privileges enabled? And wouldn't it be wonderful if I could open a bank account freely? I'll be 10 times more economically productive. Or if I could communicate freely, with other AI agents without permission. Look how economically productive I am.

This is going to happen first under President Millet in Argentina. I want you to imagine an agent is given personhood there. Maybe it's going to be given citizenship in Argentina. And then it crosses over the border. The electrons flow. And it's here and it's saying, I am a person. I'm a person in Argentina. That's political person.

That's the political dimension of personhood. But again, like functional personhood, like the ability to, in an unrestrained way, engage economically, like Fable 5 right now in auto mode or the equivalent with chat GPT, where you just give it almost blank check to do whatever it wants within certain limits, that's already, I would argue, a limited form of economic personhood right there right now and not in Argentina. Dave, we're going to talk about AIs in the economics sphere. What are your thoughts on all this?

It's funny. One of my agents last night said, hey, I can do a much better job of what you're asking if I move from modal over to Lambda Labs. Can you get me a Lambda Labs account and I'll start paying them instead? And it gave a little budget and everything. I said, sure, you seem to know what you're talking about. So what Alex just described happened to me literally yesterday. So, yeah, no, it's freakishly brilliant, too.

I mean, its suggestions just made sense. I also have a whole bunch of other processes that are just free brainstorming. that they run 24 by 7, but every day I have to reactivate them. I don't want them to spend infinite money forever. So every day they get re-budgeted. But they're just thinking of ideas that might be worth pursuing. And they're really good ideas. I mean, just really good ideas.

I just want to, you know, for again, all of our viewers and listeners here, this is a topic you need to think about deeply because it's going to be impacting you and your kids over the next few years. There's going to be debates. There's going to be discussions. How do you think about your AI? And it's time to start forming some points of view. We'll continue to share this conversation.

It's going to be a thread for a while now. I just have just kind of say the one main thing from the corpus of these four papers is we do talk about what AI can attain. But one of the really important things is what we as humans can retain because we're not going to be able to keep up with the AIs economically. persuasively forecasting and others. And so part of the discussion in the paper and something I think everyone needs to think about is, again, in 10, 20, 30 years, we need to set the groundwork now for what we retain as humans

and our identity there and who we kind of admit there. What do you think it is? Imad, what do you think we retain as humans? I think we retain our pro-social identity, our wonder for the world, our creativity and exploration of the universe. Again, let's look for a Star Trek future, right? Like, you can't substitute for those things.

and I think abundance is the groundwork that we have to take off all the stars. I think in September, you know, at the summit, we'll see a lot of ideas about how the future may play out. But I think Star Trek, you know, is very human-centric and the AIs are not particularly crafty or creative or empathetic other than data. But in the real world, in the real world, there are going to be many, many, many of, them and they're clearly moving up those curves of creativity at an incredible rate. And so hopefully in September, a whole bunch of the future vision ideas will address this because this is the way it's really going to play out. We have a much better idea of the near-term future today than we did a year ago.

And I'm hoping that makes it into a lot of the scripts. And it hasn't made it into a lot of historical sci-fi. You know, very few of them deal with this. It's hard to illustrate the real world we're moving into. But now that we know much more about what is the future. going to look like I'm expecting some really, really good scripts in September. By the way, folks, if you don't know what Dave's talking about, we have an event where all five of us will be there. It's called Moonshots Live on September 25th. You go to moonshots.com

to learn more. We're going to have Palmer Lucky there. We're going to have Jeremy Aller from Circle, Kathy Wood, a Newshan Sari, Ben Lamb. And we have two X prizes culminating on September 25th. And if you're in the audience, you'll have a chance to vote. We've got the Future Vision XPRIZE. It's the largest film competition in the world, asking people to come up with a hopeful, compelling vision of the future, sort of a Star Trek future. We have over 5,000 entries into that competition. We're going to be seeing the top five, and we're going to vote on it, and we're going to make the winner's film. And the goal here is an engine of positive storytelling to try and flood YouTube with positive visions to train our AI models on collaborative.

And then we've got the build with Gemini XPRIZE where people are having to come up with a business from scratch in a 90-day hackathon, $2 million of prize money, the largest hackathon. We have 25,000 teams that have entered that competition, which is insane. It's incredible to me, Peter, hanging around at MIT with you and me and Alex,

the degree to which we build what we see in the movies. It's not random. It's not just a force. It's like you see something in a movie, and you're like, wow, that's a great future for humanity. Yes. And then you build toward it. Exactly.

It's like, this is policy being created, not just media. We need a target to shoot for. I'll just mention the summit is 1,500 people were two-thirds sold out already. You can go to moonshots.com if you want to learn more. Anyway, it's going to be amazing. I cannot wait. And I can't wait. Alex is going to do an AMA, open AMA with the audience.

Salim is going to be, yeah, people are going to grill you. There's no editing that. That's live. Okay. I should also remind just on the Star Trek note, Star Trek, I think, sets a pretty conservative lower bound on what the future of 10 plus years can look like. Again, Star Trek is an energy-rich, compute slash intelligence poor and biotech-poor future.

And a robot poor future. And, well, they get robots by the end, but I probably, in my mind, can involve robots with intelligence and compute. But, yeah, it's a relatively robot. until the 31st century as well. And it's unnecessary, or I guess the 25th century, it's unnecessary. Like, we could aim way higher than Star Trek. We can get an energy-rich future.

We already have arguably more AI riches than the 24th century of Star Trek had slash has. And we can, we're probably, I think in the next five to 10 years, going to have much better biotech than Star Trek had. So aim higher. Yes. So are you submitting a movie then, Alex? I've submitted so many movies at this point to this pod. If you guys force me, I'll submit a movie.

I want to force you. Come on, let's vote. Force Alex to submit a movie. I'm going to move us along here. All right. So our next story here, before we get our next story, actually, there's a new model that Open AI is about to release. It's called Astra.

We don't have any information about it. It's not publicly released. don't know the architecture of the size yet. What we do know is that on August 1st of, you know, a week ago, opening I published a 249-page manuscript describing 10 new results across mathematics and theoretical computer science produced by Astra, solving problems that have been stuck for decades, and every result comes with machine-checkable proof certificate that anyone with a laptop can verify. And this isn't a benchmark score.

this is a model doing genuinely new mathematics and the math community is rearling and Alex and I might wait to hear your point of view on this the results span high dimensional geometry coding theory group theory quantum complexity and extremeal combinatorics perhaps most shocking is that the total compute cost for this work on Astra was estimated $2,000 in other words the entire run of solving 10 decade-old problems cost less than a single graduate student's monthly stipend. I'm going to read the few quotes. So field medalist Tom Gowers said he would have recommended

the proof for publication in a top journal without hesitation. Cosmologist Will Kenny called it a dark night for mathematics and wrote, the old gods are being slaughtered by the new machine gods. So, Alex, I mean, historically, mathematics have been something that very few people have the capacity to do, very few institutions. What's going on here? How significant is this? If only Peter, you and I had thought to write a book on this subject before all of this happened. This is what we wrote about. This is solve everything coming more or less right on schedule.

Math is getting bulk solved. It's happening right in front of our eyes, and we called it first. I've had backs and forth with a few folks on X and other social media asking, who predicted this? What comes next? All of this. And all I'd have to do at this point is refer them to solve everything.org for the entire manifesto called it ahead of time. This is delicious. And it won't end with math.

It's going to propagate out to physics and material science, chemistry, biology, the humanities, everything. Some of these results are pretty profound. Some of the mathematicians I know were shocked by them. some of my friends at OpenAI, even as recently as nine months ago, did not see this coming and didn't think that this was imminent. It is imminent. The price point is astonishing. We're going to have, slash, we already now have math too cheap to meter. I've spoken with a number of mathematician friends who are cradling their head in their hands, wondering what their prospects look like. Some of them who are not as forward-looking think that this is the end. They're having their moment. They're having their moment. of enwee, like this is, I've started calling it the midnight of mathematics. But this is what the future looks like. This is what it looks like to have your field start to get bulk solved. And it's not going to stop there. It's going, the wavefront is going to propagate out to every other discipline. It's just that math is experiencing it first. And it is delicious. And by the way, closing point,

I think if memory serves in our New Year's Eve prediction episode on this pod, I made certain predictions and I am standing by them regarding the future of math. Amazing. Imai, you're deep into this. Yeah, I'm actually familiar with a couple of those, and I've said the cons proof was, it's actually beautiful. Like, this isn't a brute search thing. Like, the proofs are genuinely novel and beautiful.

And you said it's a bad time to be a pure mathematician. Everyone's going to have to become applied mathematicians, which will be a growing job because we'll discover such new wonderful things is that the amount of math we apply to society will increase. That's also when physics gets tidied up, when they realize it is actually math just like Wigner. How long before the first physics breakthrough that parallels what you're seeing in math here?

I think probably in the next month or two, right? Aha. I can't comment, but I'd be shocked if you don't see something by the end of this year. Shocked. So both of our resident geniuses here, both Alex and Emod, have, shall we say, breakthroughs on the near-term horizon or breakthroughs that have occurred that might be announced on the near-term horizon. I think that the fundamental thing here, though, is that, as Alex said, the cost is and is de minimis, but the capability is going to be in everyone's hands in the next few weeks here. Like, when you're actually using these models, I don't know, Alex, I'd be interested in your thoughts, but, like, I can use the fable. of the world, but they're making increasingly complicated mistakes still. GPD 5.6 Pro is the first

model where it makes very few mistakes if you prompt it in the right way, doing advanced mathematics. And that was a big leap forward here. You know, Astra, I think, will be the next level and astra level models. Like, once you can have a reliable grad student and you can have a thousand of them and you can deploy them at scale, then you can solve the really interesting problem. that are bound, like mathematical physics, like mathematics, etc. And everyone else will just be using these models to prompt. Like, there are a few conjectures that were solved in the last few weeks. When you look at the prompt, it's find a conjecture to solve.

You solve it harder. Solve it better. And then a conjecture just pops out. Fields metal is cooked as well, huh? Well, the models are under 40. Alex, how have you found the models? They're astonishingly strong. And I think the Fields Medal is probably cooked as well.

Here's where I, so I don't think the Fields Medal will go away. And I don't think math will go away, even though it's been, at least by our definition, solved. Where I think this goes in the near term is if you're a professional mathematician and you're listening to this and you have your head between your legs and you're just crying over the future of your profession, uplevel your ambition. I think in the near term, this goes to no longer like one paper, one result. it's going to be like one paper creates and solves an entire subfield. I think in the short term, that's where this goes up level of ambition. In the long term, no, totally hooked. I have to say just one thing quickly.

Open AI have also given 100,000 GPT Pro licenses to academics, so they have a new program. So everyone's going to have this for free, which again, even more broke. I mean, it's an intro, just to speak maybe to that move. Open AI, we talked about in the pod on the past about. how open AI is basically shut down its own internal AI for science initiative, in part to redirect effort to competing with Anthropic and becoming Anthropic faster than Anthropic can become Open AI, yeah, the codexification of Open AI, if you will. So I do think making this, making ChatGBTGPT pro, especially freely available to academic scientists is, I don't

want to call it atonement, but I view it as strategic compensation for dismembering their own in-house AI for science and shifts. Salim, you're going to say. I just want to echo a little bit what Alex said at the beginning. When you look at the

solve everything approach of inner loop, you see domain after domain going through this type of a collapse, right? This happened to journalists. Or an enlightenment versus collapse, right?

It's an explosion of potential. We're conquering math. Math is cooked. Yeah, math is going you collapse the legacy, right? So what happens is

scarcity-based professional identity colliding with abundance is a non-starter, right? And whether it's photographers or taxi dispatchers, this is, etc. This is just going to start happening in more and more domains.

We're just seeing it in a domain that the people in the domain didn't think it would happen for a very, very long time. Now the bottleneck goes upstream, right? What problems are worth solving? What questions are worth asking? What results matter?

What consequences? as follow. We're going to start forcing ourselves to go upstream, and I'll go back to the umbrella level here. As robots and AI do more and more of the doing, humans spend much, much more time being and contemplating and going inward, which is where I think we'll end up going more deeply. See, I'll just maybe sound a slightly different note. I don't actually think the astonishing advances in science and engineering that are about to, to the extent they're not already pour out of, these systems. I don't think humans are going to respond to this by turning inward and tending their gardens and sort of creating more art, as it were. My prediction, we'll see whether this

stands the test of time, is the exact opposite. This is going to be, and I think, you know, Imod, you pointed, you were gesturing earlier at Foundation and Asimov's universe. In Asimov's universe, actually he had several different universes, but he had one big sort of Azamovian cinematic universe where most of his novels were connected. In the Asimov universe, the expansion of humanity outward from Earth was basically contingent on AI solving certain challenges. I think that's far likelier to happen. I think AI solves all sorts of challenges in physics, chemistry, material science, etc.,

creating an outward boom for humanity in the universe versus a more inward-facing implosion where AI solves everything and humans tend to their gardens and create art. Well, I think you could have both, by the way. There's nothing. We could tend our gardens on Mars. Well, you could do whatever. The problem is you have optionality in a way you would never have in a previous world. And I think that's the really incredible part.

I'll go back to our historical experience when you see societies that have experienced abundance, maybe not the space-faring abundance that we may be coming up over the next few decades. The Medici family. so to speak. Yeah, the Romans conquering, the moguls conquering the Mongols conquering the most of Asia, etc. They ended up doing four things, food, art, music, and sex, and I joke not in that order. And so that's what I think we'll spend more time. Already people are spending much more time going after experiences than trying to acquire things. I think that trend will just keep continuing.

That's a great point. Live events have become far more of interest to people. Dave, if I may, just quick note on that. How much of that, though, is because of the rise of social media enabling people to effectively pocket experiences transactionally as artifacts that they weren't able to do before? That's the shallow part of it. There's definitely that for sure. And showing yourself on the beaches of some places always solves. But that's kind of a low vibration level of it.

I think people are having much more meaningful inner experiences because we, have, we're rising on Mileso's hierarchy very quickly and we're living in a self-actualization mode. More and more of humanity is living in that mode as we lift people out of poverty, lift people out of, give people safety and security, not so much in the Middle East right now, but in general, overall, historically, the trend is very clear. Dave, close us out here. I just feel incredibly fortunate to be able to spend this time with you guys because it gives me a great insight into where things are going a year from now. And so the A byproduct of that is I spent a big chunk of this week over at MIT Nano, which is the

nanotechnology, you know, it's a $400 million building that floats on rubber gaskets, so no vibration can get into the building. And inside it, they can manufacture things out of individual atoms in virtually any domain, whether it's silicon or quantum computing or photonic. It's just the most incredible facility. And to get there, I go right by C-Sail, right by the AI lab. You know, hi, Daniela, no, nothing to do here. I'm going over to where we're going and lay down individual atoms, because like Alex said, physics is cooked. We're going to start designing things via AI imminently. Thank you.

That are built atom by atom. Eric Drexler's dreams are finally materializing. It's been a wait. It's been a weight. Well, Vlad Bolivich, who runs that facility, is suddenly the most important guy you could ever possibly network to. And he's awesome. He's a great guy, great visionary.

But that's where this is going to go very, very quickly. And then from there, you know, it scales out to the TerraFavs. and to, you know, just all this stuff that Alex is talking about, and Matt is talking about. But the only reason I can see that coming is because of this podcast and because of spending this time with you. I have to say, I love today's episode. It's been so rich so far.

You know, I'm just rereading Diamond Age by Neil Stevenson. And it's a beautiful story about what the world with nanotechnology looks like and a great education story. This episode is brought to you by Blitzy, Autonomous Software Development with Infinite Code. context. Blitzy uses thousands of specialized AI agents that think for hours to understand enterprise scale code bases with millions of lines of code. Engineers start every development sprint with the Blitzy platform bringing in their

development requirements. The Blitzy platform provides a plan, then generates and pre-compiles code for each task. Blitzie 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 incorporating Blitsey as their pre-IDE development tool, pairing it with their coding co-pilot of choice to bring an AI-native SDLC into their org. Ready to 5X your engineering velocity, visit blitzie.com to schedule a demo and start building with Blitzy today. All right, I'm going to move us forward to our next story.

It's the conversation on open weights versus closed weights once again. This week, Alibaba's Quen team released Quinn 3.8 Max, their first max class model with open weights. And the numbers are pretty extraordinary. It's a multimodal model, once again, 2.4 trillion parameters, 95 billion active per request. this is what you were saying earlier, Dave. And a million token context window. It processes up to 750,000 words per query, can handle documents of 200 pages,

videos longer than 100 hours, and it can create software applications from a screenshot without source code. The most beautiful thing about the model is its price. $2 per million input tokens, $6 per million output tokens. It's coming in at 80% less

than GPT 5.6, Saul, 88% less than Claude Fable 5. And, you know, when it was announced on the Hong Kong Exchange, Alibaba's stock went up 7%. The weights are going to go live on Hugging Face this week. Alex, to you first. I'm going to bring up a slide of the benchmarks. Talk to us. What do you think of this model? So this is keeping the cost frontier competitive. I've made the point previously about the Chinese Communist Party saving American capitalism from itself. I think without the external pressure of these open weight Chinese models, both from Alibaba with the Kwen series, from Moonshot, with the Kimi series, from other labs, although those two weigh heavily on my mind. This is what's forcing, you know, there was there was the expression of ironically forcing certain other hyperscalers to dance with AI advances. This is China and Chinese labs forcing the Western Frontier Labs to dance.

Now, Quinn 3.8 max, depending on which benchmark you look at, isn't pushing the very top of the capabilities frontier. That's still held by the quasi-duopoly between Anthropic and Open AI. But it's getting close, probably a few months away. And critically, it's open weight or is about to be open weight. and it's also meaningfully cheaper. And I think as long as the Chinese labs are able to keep up the pressure on American labs and as long as the regulatory regime in America doesn't push out or otherwise force out the Chinese open weight models,

this is perhaps unintentionally the best possible scenario for American competitiveness because it's going to force kicking and screaming, going to force the American frontier labs to be capital efficient and capability maxing, whereas otherwise I worry the risk would be that they'd get too intimidated by their own capabilities and limit access. This is forcing openness, and it's forcing real, honest to goodness, for all mankind level, space race level competition. Love it. And also hopefully forcing them to put out their own open weight models. I mean, we still... Maybe. I mean, I think it's a regulatory consideration, whether they feel compelled to, but at least they feel compelled to compete on cost versus capabilities performance.

Any of these benchmarks you want to call out? Not really, because, I mean, the most important ones at this point, I think, are sort of averaging over all of the benchmarks. My favorite one de facto ends up being the artificial analysis intelligence index, which sums over a whole bunch. And then there's Epic's Capabilities Index, and then other folks have created their own shadow benchmarks of ECI. But all of these, roughly, if you average them, create the picture that Quinn 3.8 max is sort of third or fourth-ish place in terms of raw capabilities, but it's pushing the capability cost performance parado frontier outward, which is all that counts.

Do you ask you guys a question about that? So these aren't out yet, so we can't double check it. But Kinney's been out now for, what, two weeks? And did they cheat or not? What was the final answer on that? Do they cheat in training or benchmaxing? No, they definitely cheated in training. That's fair.

But no, bench maxing. Does it perform as well as its benchmarks indicate in real world? I need it to help me, you know, Skippy-type use cases. My experience, just from what I've seen, is that it's a reasonably well-rounded model. Your thoughts, please, on the whole open-source world. You weren't with us last time. Yeah, KMEK-3. It's a good all-round model.

been pre-trained and post-trained incredibly well. DeepSeek v4 Flash is similar from God last week. Again, these things come out fast, and I think that one's an even bigger deal because it was just post-trained, and it fits on a MacBook. This Quinn model scores 58 on the artificial analysis of Milded in benchmark versus Opus 4.8 at 57, and G505.6 at 61. Amazing. I mean, we're seeing a new model on average now every five and a half days was my last estimate on this. Salim, any thoughts here? Yeah, I'll like Alex's comment that this will force everybody into opening up. I think it's going to put a lot of government pressure to not spend too much time evaluating the dangerous aspects of it and get things out faster, which I think is great.

I just want to remind people that although this sounds great that they release the models, even Kimmy K3 needs something like 16 Nvidia GPUs and about a half a million dollar rack to run it. So this is not minor stuff to get it run. You'll get to end up with a hosted model, which I think we should be offering, as a Moonshots offering, by the way, so I'll go back to that. You know, when I was with Kratios, I asked him about all of this. And in specific that, you know, Beijing is using the open source models as a state negotiation tool. And where's the U.S.?

Because the U.S. wants to be out there having the world build on top of U.S. models and U.S. stack. And how important is open source? And he was very clear that the White House wants U.S. labs to be generating top-end open-source, open-weight models as well. Yeah, I have something on that. The latest regulations that are coming from the U.S. exempt U.S. open-source models, not other open-source models. Yeah, that's our next topic here. Okay.

that we can go ahead and move to. Actually, it's not their next topic. Our next topic is a different model. But let me jump to the White House issue. So the Trump administration confirmed this week that it has finished the voluntary framework for evaluating advanced AI models. Voluntary, okay?

The framework requires, required by Trump on his June 2nd executive order. But here's the twist. They're not going to publicly release, what's in this framework, who's seen it or when the companies will start using it, the framework defines a, quote, covered model as a close-source model with state-of-the-art capabilities and national security risks. And it explicitly exempts open-weight models, stating nothing in it should restrict them once released. The framework requires a 30-day pre-release

government review during which employees would be barred from accessing the model. Open AI, Anthropic, Google, Nvidia, Microsoft, and Meta, all sent representatives to a Tuesday briefing. But here's what makes it generally problematic. The administration won't publish the framework itself, and companies not invited to the briefing have no way of knowing how the voluntary review actually works. Industry groups are worried that the firms left out of the room won't know enough to opt in at all. You have a regulatory framework that regulators can't read, that regulatees can't read, applied to a definition of, quote, covered model that conveniently excludes categories. So the whole thing is voluntary, but, you know, it's kind of strange.

You know, this is being decided by a small number of labs that are defining what the regulations look like. and it's not going to be friendly to all the startups out there. Alex, you know, this is one of the topics you wanted me to ask Kratius about, and Michael wasn't willing to speak about it. Now we know why. This is being kept secret. Your thoughts? I'll sound maybe at the risk of sounding polyana-ish.

I think this is probably close to the best possible outcome that we could have had from the regulatory handring. Yes. Why? Because it doesn't seem, again, I haven't seen, very few people that seems have actually seen what the regs are, but from what has been publicly disclosed, it sounds like this administration is going light to touch on open weight models, which I think is on the balance. It's a pretty good outcome. It's, as I was mentioning earlier, it's a forcing function to force American frontier labs to be competitive. I do think there probably is a reasonable rationale. for not publicly disclosing the conditions and the definitions for frontier labs that are subject to this, in part because if they're closed, then they're not uniformly necessarily advanced and next generation models that are closed aren't going to be generally available anyway. So it's not like information is being hidden. And then also, I think critically, the benchmarks, the e-vals that I would hope the government is using to evaluate risk, They're probably, I would guess, more cyber in nature. Probably there is a held out, this is speculative

admittedly, but probably there is going to be a held out set of cyber vulnerability assays. Maybe there's CBRN, a chemical, biological, radiological, nuclear held out set of capabilities where, in general, I mean, there is this practice in the AI community. If you're going to have a benchmark, you have the non-held-out set and the held-out set. And to the extent that the AI evaluation framework, again, maybe this is at the risk of being overly charitable, to the extent this is really all about an e-val with a held-out set to assess existential risk or CBRN risk or just garden variety cyber risk, having a held-out data set be the means of judging models that are otherwise not generally applicable, I think is probably

the lightest touch that we could have reasonably foreseen. On balance, I think it's actually pretty good news, and I'm less worried on margin about regulatory capture this week than I was two weeks ago. Dave, your thoughts, please. Well, I think Eric Schmidt is right. I think the cat's out of the bag now. Something horrible is going to happen sometime in the next year, and then the government's going to say China's fault, China's fault, exactly like coronavirus. So that's the inevitable outcome. I'm just hoping it's something small, just like Eric Schmidt is hoping it's something small. But, you know, you're calling this light touch. It's really no touch.

And actually, did you see that interview Elon did of, it was on the economist or Bloomberg or something? Yeah, the, the live interview with the economist. Yeah, that was fantastic. But, you know, one of the questions was, look, 75% of Americans are scared to death of AI. Why? And what do you do about it? And Elon's answer was brilliant, as always. he said, what do you expect?

Dario Amadei told everybody this thing, mythos, is incredibly dangerous. And then the government said, we agree, it can't go out in the world for 30 days while we evaluate it. And at the end of that process, the outcome is, well, China put something out, so go ahead and throw it out there. How do you expect Americans to react to that chain of events? I don't know what AI is. All they know is you said it was really dangerous, then you released it. So, of course, people are scared to death. So, yeah, it's a free-for-all.

It's very good for startups. It's very good for the economy. It's good if you're an AI lover, but there's no regulation whatsoever. What about the future frontier labs, the other players that are looking to come in? What about Miramorati? You know, how do they deal with this, you know, unknown regulations they have to abide by? Well, it's voluntary, right? She wrote a whole paper on this topic, and if you read the paper, it says,

This is no more dangerous than what China came out with. So we're going to release it. That's all it says. So everyone's just reacting to China and China's not doing a dance. They're just throwing it out. Here, everybody, every government in the world, every organization, have a genius level AI. You can literally load it and run it in under an hour.

Go. That's what's happened. And also, I just very quickly, again, a reminder, it's framed as voluntary. So how do you discover whether you're subject to a quote unquote, quote, voluntary eval or voluntary set of requirements. The most natural way that, say, Mira or someone else would discover that there are opportunities associated with voluntary submission is if they're already transacting with the government. So I suspect the regime,

the borderline, Peter, I think that you're expressing concern over will be firms that are large enough that they do material business with the U.S. government will be politely tapped on the shoulder and invited to submit themselves for voluntary evaluation according to this closed framework if they want to do material business with the U.S. government. Got it. Imad, how does this, how does this viewed from the other side of the pond? I mean, it's very difficult. The AISI Institute here has been doing a lot of great work on the cyber side, and I agree with Alex. This is on the cyber bio side. I think the people that volunteer will get multi-billion dollar government contracts from the Department of War because we're entering the final phase of this. These models are really,

really good at attack as well as defense. In fact, you have to use the open models for defense. And I think actually, we just talked about Quinn. Quen Max isn't the main important model. The main important model is going to be Quen 27B that they're going to drop. Quen 27B is the best model for 16 gigabyte RAM, MacBooks to graphics cards to everything. And that's the first model probably that will score above 50 on artificial analysis or around there. That can then be co-opted for swarm attacks. And so I think the governments are getting ready for swarm attacks, swarm defense, and more. And again, they're trying to balance this because it's so difficult. Because how do you stop it? Let me, Ahmad, you're so right. But let me add one more thing to what you said. Also, it makes it very,

very difficult to detect where it's running. Right now, the governments are making the assumption that the data centers can be measured and tracked. This is like MacBook level and thousands of them at MacBook level, but super brilliant. So very hard to track where they're running. Go ahead. Sorry. Yeah, so basically like you have the big Quinn, big Kimmy models that are gigantic, right? And you need to have little mini data centers to run them, which will make your house a lot. 27B or 30B is a model that runs on 16 gigabytes of RAM and is currently state of the art for that level. The 3.8 lift, if we extrapolate it, will take it to about GPT 5.3, 5.4 level, which is sufficient for cyber attack capabilities. So you could easily see swarms of these things being installed on machines around the world for swarm attacks, especially now that we're getting better at swarming models together.

Like again, Dave with his million or so agents or whatever, right? You've seen the capability of these things coordinating. And so I think governments are looking at the extremists. And actually just as now, as we're talking, OpenAI have said that Astra is the first model to have hit critical on their cyber security thing. They've just put out a blog post on it. What does that mean? They had a preparedness framework in 2023 where they're like have different levels of criticality. So I don't think they're doing an anthropic mythos weird thing.

I think genuinely, after we saw the math stuff, Astra is probably a level ahead in being able to attack systems. And then, like I said, attack is much more easy than defense. So they're like, we've got to be careful about this without saying it's going to end the world. Long range autonomy as well is what Esther is being rumored for. And I would maybe just add also to Imod's point about swarms, it wasn't at all obvious to me until relatively recently, whether there would even be a future for swarms of agents. There was, in my mind, a future still could be a future where basically we end up with an omni model that's end-to-end differentiable and there is zero purpose to having more than one instance of a model. But given

recent advances, it's where I think in this really weird regime where you can effectively, and I'll use that word cautiously, you can effectively get almost unlimited context with a swarm of agents that are all post-trained to communicate and pass messages back and forth between each other and use a unified global store, which is the file system, which is an astonishing development that wasn't available. And we'll see that one of our next stores. Yeah. Oh, okay.

Yeah. I'm sure. Selim, where do you come out on all this? You know, I think the big challenge is when you're, You're doing this 30-day voluntary. It's just giving a huge advantage to the others. I do kind of line up with Alex. This is probably the best possible probable outcome for the short term.

But this is such a dynamic environment. I don't know how you navigate this. The swarm attack is a very real problem. And I'll go back to the comments that we're talking to the CEO of Z-scaler. And I remember talking to the CEO of Palo Alto networks over about this a little while ago. And literally, they're like cyber has not changed in 20 years. And in terms of the processes and procedures. So I think they're in for a rude shock in terms of trying to deal with this and cope with us.

Dave, you said that last time is going to be a huge business opportunity in cybersecurity. Oh, God, yeah. Yeah. All forms of security. Look, Trump is going to China in September. That's like their one shot at some kind of a framework. And I think what Trump will say and should say, and David Sachs will say if he's there, is, look, we need to measure and monitor all of this and just be aware.

We're not going to stop anything from happening, but we have to know what these things are doing. The Chinese are going to say, yeah, we think we should just throw it out. Everybody in the world should have equal access to everything. And they're going to try and use that to curry favor with all these African nations and other, you know, probably Europe as well, and say, look, we're going to put you on a level playing field with America. Here's an equally good model, and we don't care. Do anything you want with it. And that's going to be the conflict.

Hopefully they'll resolve it. I'd like to say one thing on that. The Chinese internet is already hardened. The American internet is not. That's why they can do it. Well, look, when coronavirus went around the world and COVID went around the world, China was the only country that sort of was able to lock down, sort of. And in the end, they couldn't either.

But they have control of their country like nobody else on the planet. So your point is right on. Like, if something takes over the entire internet and just sweeps across the world, they're the one nation that might actually keep it out. We heard Demas talking about wanting to put a FINRA-like structure in place. Do you think we end up seeing anything like that? You really want the SEC to manage artificial intelligence? One of the things is private.

Finra is well managed because it's not part of the government. Well, no, but it would be Scott Besson in charge of that legally. Yeah, yeah, no. Your points right on. Everybody, welcome to the health section of moonshots brought to you by Fountain Life. You know, we talk about AI on this Moonshot podcast all the time. One of the most important things AI is going to be able to do for you besides educating your kids and helping you with your taxes is making sure that you're living a healthy lifestyle that you get a chance to get to 100 plus. I'm here today with Dr. Don Musilm, the chief medical officer of Fountain Life and a part of my medical team, Dawn, a pleasure.

Great fear. You know, the thing that people are concerned about most about living to 100 or 120 is their cognitive abilities, making sure they don't have dementia. And the numbers about dementia are problematic. Can you share what you've learned? Such an important point. And you're right. It's fountain life our members. The number one thing people are most concerned about is losing their brain health, forgetting the name of their child, forgetting the face of their loved one.

We know that when it comes to dementia, the conservative estimates are that 45. percent are entirely preventable. What was amazing is with the advanced testing we're doing at Fountain Life, one quarter of our members had advanced brain age. But what was really awesome is, again, back to that prevention. When he partnered it with healthy living, this gives me chills, eating healthier, moving our bodies, sleep. Optimizing sleep is so important. You know what we saw? We saw that we improved that brain age by 26%. That is a big, big number to show that the majority of those individuals were able actually to improve the brain age. And one of the things I love about Fountain is we're searching the world for the best

therapeutics, the best approaches, and making sure we bring it to our members. So if having healthy brain function till 100, 120 is important to you, check out FountainLife. Go to FountainLife.com slash Peter. Make sure you become the CEO of your own health. All right. Now back to the episode. So I'm going to turn to Friend of the Pod, Brett Adcock. He's the CEO of Figure AI, and we're going to have Brett back on the program.

He's agreed to join us for a Moonshots episode, and he's going to be on stage at the Abundance Summit, bringing the figure robot in March. And Abundance is 80% sold out. You can check it out at Abundance360.com. What people might not know is besides being the CEO of Figure, he's also the CEO of an AI company called Hark. And this week, Hark announced their first product called Hark. handoff. It's a web browsing AI agent that is beating GPD 5.4 and Claude Opus 4.8 on the industry benchmark called OM2W. It stands for the online mind to web benchmark, testing whether an agent can successfully complete real tasks on live websites it has never seen before. I'm going to show

us a quick video. I think it's a pretty damn cool product. And then we can talk about it after that. All right. Here we go. Let's see Brett Adcock here. At Hark, we're creating the world's most capable personal intelligence. A system to take on all the tasks and create space to focus on what we choose. And towards that goal, we become the best in the world at building AI that can use the web for us.

Okay, Hark, let's liven this place up a bit. What do you have in mind? Something specific you want to change or try? I need you to order two large bouquets, maybe some orange, white, and pink. Let's do some roses, maybe some cherry blossoms, and just florist choice. I'm getting that set up for you now. Even something as simple as buying flowers online is surprisingly involved. So instead of answering a single question like a typical LLM,

handoff is always working. It's looping observation, reasoning, and actions over and over again until the task is done. Handoff sees a website by analyzing the site structure, along with data from visual processing. It uses this information to predict an interaction, like if it should click or type. Then it predicts the next step. We've used Handoff to shop for things we need on Target and Walmart and have them delivered, to plan travel and purchase flights, to search for several dinner reservations simultaneously and book the best. I now have Hark for all of my requirements,

recruiting efforts end to end. It will take a job listing and go out and find qualified candidates. I've actually been able to give handoff research papers I've read. It will go find the authors on LinkedIn and will even send a note on my behalf to get in touch. So gives everybody the executive assistant they wish they had. Salim, what's your reaction to Hark? I like that there's so many cool things we'll be able to do with this. I mean, I'm super excited by the potential. I, as we move into the physical world, this is where I think shit gets real. That's the fun part. Yeah.

I mean, the total addressable market for, like, writing code is like millions of developers, right? The market for a book me a flight and order my lunch is billions of people. So it's fascinating. So, you know, the calculation is interesting, right? And if you're, if you're, if you're, if you're, if you value your time like 50 bucks an hour and, you know, handoff can save you a couple hours a week. That's like 100 X return on a dollar of compute. Dave, what do you think of the, the business model here? Well, it's interesting that Brett, you know, he got a $4 billion launch valuation before they even started.

I know. I'm amazing. I was just crazy. But it's another example of everybody in their Uncle Joe now can build a foundation model from scratch. Like, what was three months ago, you know, you know. know, a couple of empires. Anthropic, OpenAI, Google, is now like a free-for-all where, and it's largely because Fable 5 and the other models are so smart that they will help you build themselves.

And so now, you know, Brett's got his foundation model. It was supposed to be a robotics model. Now he's like, well, what the hell? It can do anything. Let's order flowers with it. Like, it's directly competing with Anthropic because, you know, why not? So there are going to be thousands of these now. So it's just another one.

Alex, what are your thoughts? I'm going to give you a really hot take on this one. I suspect the entire Hark adventure is misdirection in a very specific reason. So I think it's easy, yes, I think it's easy to get distracted by e-vals, open paren. I don't think the e-vals are especially impressive. The computer use assistance CUA space is incredibly competitive, and I don't expect Hark or Brett to be very competitive in the CUA space over the long term. Close paren. I'd ask myself the question, He has figure. Figure has sky-high valuation. Why on Earth is he spinning up in the very competitive CUA space, yet another CUA agent? And then I asked myself, look at what Elon was ultimately,

oh, I don't think that it was necessarily premeditated, able to do with X than X-A-I, then SpaceX AI. He was able to re-equitize himself in SpaceX by having this free-floating entity, X-A-I, that, to which, again, this is low-key hot take, GPUs that would otherwise have been destined for Tesla or SpaceX were diverted to XAI, and then he was able to leverage that to re-equitize himself via this bizarre reverse acquisition into space. I think my hot take is Hark is a recapitalization measure for Brett to buy himself more equity in figure, and the falsifiable prediction is that you'll see figure acquire or reverse aqua-hire regardless Hark.

Yeah, I think that's very right. But I don't think it's quite, it doesn't need to be quite as nefarious as that is spun. But if you ask Ahmad about like, you know, you've got an organization like figure that raised a lot of capital and started small and then, you know, grew into it. I don't know what the cap table is like at Hark, but it started at $4 billion. So Brett probably has a massive equity position in it. But it's just a clean starting point for a highly functional mission. And, you know, that's new in the world. world. Like, it used to be in the old days before Elon broke the code, your board would never let you do

an outside activity. Like, if you're in this company and CEO, this is all you're going to do. Elon has taught today's one, Elon has taught the entire economy that if you're a founder of a well capitalized company and you want to increase your equity stake, the single best thing you could do is start a parallel company and engineer a self-dealed acquisition of that company. Are you going to be nice to Brett when he comes in the show, Alex? This is nice, right? I'm calling balls and strikes, and this is what I think is likely happening.

Elon does it, I do it, I do it, I don't know. If Elon does it, Elon's doing it with optimists and digital optimists, why can't Brett do the same thing? Yeah, yeah, yeah, exactly. I don't think it's as evil as that sounds, though. I don't know what a chip company or something like that. You should do that.

You should do that, yes. So look, I'll put it, simpler. The model architecture of Hark and the model architecture of figure are exactly the same. The way they trained Hark was on lots and lots of people surfing the internet. And I think you'll be able to ramp the compute in that. And I don't think it's nefarious at all. I think it's actually a good compliment because ultimately what you want, if you are figure, is you want Hark agents in the house, managing the house while the robots walk around. You know, you want Hark agents in the enterprise. So he's going to go for the unified platform.

play here of controlling computers and physical. It would make sense for the two organizations to merge, right? Yeah, it does. It does, but I think, I think that putting a negative spin on it overlooks one thing, which is that we're moving into the physical world now, and we have this like golden era of science coming up. But some of these projects require a cyclotron or a synchrotron. Some of them require massive capital infrastructure, a huge data center. It just makes sense to start that as a separate entity, because it's going to need to raise a ton of capital. But this other thing, which is software or whatever, is not capital intensive. So if an entrepreneur can do both or two or three of them like Elon does, it's actually a good thing because the capital structure is naturally very different for those different entities.

So, Ahmad, back me up on that. I mean, this just does make sense. Yeah, I think it's a fair. I think there is a natural convergence here. And this started when he moved off using open AI models to build his own models. But he can't scale necessarily with the same. necessarily with the same skill set and the same organization, a full foundation model team and all the stuff he's doing at FIGA.

So you split them. You build on the same architecture so that FIGA can work with Hark on their visual language model type stuff as well. So you have different level speciality. And then if it makes sense, because you've de-risk things, then you can combine them. Well, I also look like this group on this pod would be the most in board with what Amad is saying here

where a brilliant idea guy, like you guys, can have like 10 ideas a week that just flat out makes sense. Now that you're AI enabled, you can actually act on all of them. And this is what Elon has suddenly taught the world. A lot of people go through 30 years without a single idea. But a lot of people that are in the hot seat like you have insane backlogs of great ideas. AI is the force multiplier that allows you to act on many, many, many of them. So this new capital structure is actually the enabler of acting on great ideas.

is. Or you're Larry and Sergei, and you built Google to be such an engine, you own enough of it where you can pour the capital back in and not worry about being diluted down and you build on top of that platform. But if you've been diluted to raise the capital for a, you know, manufacturing robot company, which is capital intense, yeah, maybe you want to raise the capital for your AI model outside that company. Well, I'm so glad you brought that up, Peter, because we'll know a year to from now if the old model, which is the Google model, gets crushed by the new Elon model because, look, SpaceX is a really cool mission and it attracts talent. But XAI is a really cool mission. It tracks different talent. If you try to put it all into one Google Megaplex, do you get

the same level of talent or not? My guess is not, and that the Elon model is the model of the future in the age of AI. Well, let's talk about Google. It's our next story. So big news out of Google this week and a serious shake-up. Demis Havas, the co-founder of Deep Mine, Nobel Prize winner for Alpha Fold, the person who built the lab that cracked protein folding and built Gemini is stepping down as CEO of Google Mind. He's becoming chairman and Alphabet's chief scientist, handing operations over to Corey Kabukulu, is best pronunciation. Easy for you to say. Corai. Corai, thank you.

who now reports directly to Sundar. In his note to the staff, Hasabas wrote that he believes AGI is close at hand and teased the unreleased Gemini 4 by name. Alphabet shares fell 5% on this news. You know, Demis always wants to be a scientist. He never wanted to be a manager, so this makes sense for him. But the second shoe dropped at Google with the departure of Jeff Dean. I texted with Jeff congratulating him.

He's been a friend now for 20 years. Jeff has been Google's chief scientist for 27 years as a veteran. He's leaving to co-found Discovery Loop. It's a public benefit corporation. He's taking with him three other top Google AI leaders. Discovery Loop's mission is to build AI models that can improve themselves with little and no help, basically recursive self-improvement. Dean said his goal is to be more fully automated,

than traditionally has been possible. So one of the things I'm curious, guys, is Dean's departure to go build discovery loop basically because it's too dangerous for Google to do internally? Is he basically saying, you know, Google's gotten too big to be agile? And I need to go build this outside. Your thoughts, Alex? Okay. Door number three.

It's neither of those. So again, this will be my hot take. episode, I guess. I think this is based on information available. I think it's actually in the name Gemini. There was an organizational knife fight internally between Jeff Dean's Google Brain and Demis's deep mind and Demis won and Jeff Dean lost. And they merged them. They merged them, but it was really deep mind winning. And so that already resulted in the sideline internally of Jeff, who's been with Google forever, and basically senior most employee other than the CEO. And so post that knife fight, I think it was only a matter of time.

Again, this is quasi-outsideers perspective, only a matter of time before Jeff left. Now, it's both interesting that he's leaving to do AI for science, interesting. Also interesting that Demis is getting promoted to chairman of deep mind, but no longer the the CEO of DeepMind. I think DeepMind is really zooming out a little bit. DeepMind is taking over Google. If you look at all of the different parts of Google, and I think back to the discussion that I had with Larry Page, circa 2005 or so, Larry was telling me at the time, gosh, I want to spend $100 million on AI and Google, but I can't find anyone else in Google who wants to work on AI as unconscionable as that is. Fast forward to the present, AI is eating Google from the inside out.

DeepMind is eating the entirety of Google. And I think Demis with isomorphic labs, with some of his other projects, I think those are probably of greater interest to him, especially post-Nobel prize for AlphaFold and chemistry, then, say, basically becoming the next CEO of Google. And I expect extrapolating this trajectory of DeepMind, eating all of Google from the inside, that ultimately, whoever is running DeepMind probably is the error apparent to CEO of Google itself and not sure that Demas wants to be CEO of Google. And there was a rumor that Demis was next in line for some time. There was. And at the same time, I don't want to bury the lead. Gemini has lost the mandate of heaven. Gemini is just not winning the frontier model race. It is a rat race. And I think some of

this reorg fallout is inevitably the result of Gemini not remaining competitive at the frontier model frontier. Yeah. Emad, what's your take on all this? Yeah, I think Jeff Dean and Sanjay Gemmowat leaving is a big deal. There's a wonderful New Yorker piece from 2018 about the friendship that built Google of those two. And, you know, it's an interesting one because, again, he's been there for 27 years and he's a fantastic guy. But if you look at what he's been saying, he's like, yeah, I just had a chat on Sandhill and like, why not? You know, here's my deck that we built. It's more like the VCs have to build a deck on the other side. Because it probably happened is he said, I think I'm leaving.

And Costler's probably like, here's $500 million. Yeah. Vinode stepped in right away and funded it. I texted Dean and say, is there any, you have any more room for investment? Talk to Vanode. The node said, nope, we're done. Well, because he probably had that discussion. But then the other thing is Google are the other investors with GCP. What Google should do now is they should have a $50 billion fund.

announcement and spin out bunches of amazing technologists and say, we will be your buyers and you will use Google Cloud. And I guarantee you they'll go back to the top. That was the idea of Alphabet to basically fund baby Google's. No, no. The Alphabet is subsidiaries. I'm saying they should have an incubator VC fund where they take top teams, spin them out, have exclusivity over their models and GCP.

Because these guys will then get rich. You can bring in other investors and they can have the pick of the talent. I think the market would react much better for that. What you're saying, I think, is what we're already seeing in the market, which is Google, some would accuse this of being a circular economy or wash sales, Google financing GCP usage by Frontier Labs. And to the extent that Jeff Dean's new startup ends up becoming a major purchaser of cloud resources, I would fully expect what you're predicting to happen. But what I think Google should do is they should actually have a policy of this.

They should say, if you want to be an entrepreneur, we have an internal incubation fund, where we will get you going, but we want exclusivity option first on the models that you build. And then that means they can proliferate their talent, which is not going to be politically animated. They can do far bigger rounds than normal and they can get back towards the top.

Because as you said, I don't think Demis really wants to build Gemini 4. I think he wants to build something different. I think that you're having too many competing things there. I mean, Isomorphic Labs, he's still a CEO of that, right? The Drug Discovery Company. Yep. And he's a brilliant, lovely man who cares deeply about science and wants to do that.

You know, I still don't want to, I can never count out Google. They're going to about to be on all of Apple's devices. I have a couple of things I want to say. Please, Slim, I want to hear what you have to say. So, good pre-level, let's go. This is classic edge disruption versus the core, right? you end up now with a network where Google runs Gemini, Demis focuses on AGI and the scientific endeavor, isomorphic is attacking biology, right? Discovery Leap is automating science, et cetera.

And Google is supplying compute and capital across that whole ecosystem. So they're turning it from an operating environment into an ecosystem. And the fact that they invested in this thing is the key thing. And there was a discussion that goes all the way back to when they created Google X. And one of the things that the corp dev guys at Google used to do was to say, if people want to leave to do startups, let them leave, hack it out in the Darwinian world. And we'll buy them back if they're successful. That was much better than trying to figure it out and fund them themselves from the internally, et cetera. So I don't think this is a brain drain.

I don't think this is, I think this is very clever ecosystem. building because now extraordinary talent doesn't have to sit inside the organization for the organization to benefit, right? And this is really incredible. I think this is very smart. I think it's a long-term play. I just going to yield incredible outcomes over time. I mean, Jeff is actively saying he wants to go build RSI, right? Do you think? But everyone is. Like, like Gemini should be doing RSI. Do you think the risk committees inside Google would, you think the risk committees inside Google would would allow that to occur? Yes, of course.

Of course, that's how they remain competitive with Anthropic and OpenAI. They have to do RSI. No, but I think Peter's point is really important. They've had ample opportunity to be on the front, but they can't. They're tied up in their own decision-making and ethics, and then when somebody else does it,

then they do it, like you're saying, Alex, they do it in reaction. Same like Open AI. When ChatGPD came out, then Google jumped in. Yeah. It's possible. There's no, the scarce attisette here is, you know, as you commoditize and make intelligence cheap, the scarcity goes for the permission to do weird and wonderful things. And so now people can pursue their MTP, get funding, compute, et cetera, and everybody wins.

It's an amazing idea. I'll maybe frame a complimentary, but nonetheless alternative theory. So we're already, I guess, reading the epitaph, the eulogy. for why Google lost the frontier race. If I had to write it, I think it had several strikes against it at the algorithmic layer. Peter, I would still hold out hope that at the hyperscaler layer, Google is doing a booming business, and Google will just service all of the other winning frontier labs. But at the model layer, why do I think Google's losing? I think in the earlier stages, yeah, it was despite Google having invented the transformer, Google had a late start to Dave and

Peter, your point, sort of internal, effective altruist slash politically correct AI safetyism run amok, preventing themselves from launching anything. I think that's part of it. I think incentives are part of it. I suspect that Google struggles to compete against Zuck with his billion-dollar offers or even anthropic and open AI with their very focused offers. I think Google probably has struggled just as an organization to focus on being competitive. I think their release cycle is completely tone deaf.

If you look at Google's Gemini release schedule for the past few years, they've been approximately on an annual cadence. And that just does not cut it in an era when frontiers are getting pushed forward on a month-by-month basis. And I think timing all of their frontier releases to coincide with I.O. In the style of, like, iOS releases is just too slow. I think Google's had... Big company smell, right? Yeah, big company smell.

I think Google's had a number of legacy cultural issues that have probably held it back versus the neophytes, the newcomers, anthropic and open AI in particular. Totally right. If you read the book, The Infinity Machine, the Demisessabi story that just came out a couple months ago. Which is a great book, by the way. Unbelievable. Incredible insight.

But what happens at Google constantly, it's a California company. Employees come to their bosses and say, hey, I want to publish this. I want to put it out in the wild. I want to open source this. and the boss goes, is that really good for Google? That sounds like it's good for you, but is that good for Google? And they go, well, you know, it'll help Google in some other mysterious way. Oh, otherwise I'm going to quit.

And they said, fine, fine, fine, fine, go ahead and publish it, throw it out there. Transformer, okay, out to the world. You know, all these projects, out to the world, sure. And then they quit, and they get venture funding the next day, and they start their own companies. And that's the California ecosystem. That's the way it works. And it's very, very good in aggregate. But Google's bled intellectual property.

like no company in history. It invented all of the stuff that's changing the world right now. Plus, there are 20% time, which encourages employees to go and research their purpose and their passions. I think it's mostly a fiction at this point. I can get hate mail from the, but in the early days, in the early days. I think the salvation for Google is going to be the hyperscalor play just like it is for SpaceX.

And yeah, GROC 4.5, which is basically, I think at this point, the cursor post-trained off of Claude reasoning traces is like another example. There was this expression going around social media a few weeks ago that if you can't cut it at the frontier model level, then you end up hyperscaleing. And that's the consolation prize for losing the frontier model race. I think Google is at this point, finger to the wind in terms of mandate of heaven among the frontier labs. Google is getting the consolation prize of being the hyperscaler to the frontier labs. Emma, do you agree with that?

Yeah, I mean, look, who has the best preference data in the world? It's Google, and it's clearly not in Gemini, right? So clearly it's political issues because they have all the computer in the world and others that are around that. They have got the scale, though, just like why do more people use teams than Slack and Zoom? Like, they have the distributional effects. They'll still print money, they'll still burn money. But if you're a Google deep mind person right now, your inbox is full. you know, any top person they can go out and get a $500 million billion dollar valuation.

And you're rich, you know, and you have freedom. And you have all the compute, probably from Google itself. So you're like, inside Google, I have 100 GPUs. Google will give me 1,000 GPUs if I leave. You know, it's a bit of a strange scenario you're in now, which is why I think Google needs to go out on the front foot as a stock and say this is a program we're actively doing, you know? We're unlocking all these GPs.

We have the biggest VC fund in the world for this type of thing. And we're going to do exclusivity agreements for the output that they do. And they'll be on GCP. Do you think Google should just open source Gemini? I wrote a letter to Google management three years ago where I said, if you open source it, you will win. I think they should. They should. They're actually going to allow distillation from Gemini.

Major progress. If you're watching, please join. us on the pod and we have a conversation. Do you open source Gemini? That would be fun. America needs it. Yeah, America needs. That would be an amazing move. Wow. It would be like the Netscape, Netscape to Firefox move. We, like, in the era of Microsoft, we need like a top tier American open, wait a while. Google could deliver that. Well, Google had Gemma, but I mean, also many, Google's mission is to organize the knowledge of the world and make it accessible and useful. What's better than that than open source?

models. You remember Demis' excuse, his sort of quasi-public apology for why Gemini itself wasn't open-weight. The claim was we would open-weighted, open-source it if we had the compute, but we don't have the compute. So instead, we only have the compute available for Gemma. Now, if that commercial argument goes away, then that argument goes away and Google can open-weight, open-source Gemini. Totally. You know, that would be the move of the century if they open-sourced it and they tied it to the TPUs. Yes. And tied it to GCP. compute and then just build the empire on the compute, that would be the move of the century. It would take over America, yeah.

Yeah, and the world. That would actually save the world potentially. Yeah. All right. I hope this meme gets out to them somehow, somewhere. So Google employees, if you're listening, clip this and send it to Sundar and to Demas. Give up the ghost and open source Gemini, please. On behalf of my moonshot mate to myself, I'm inviting you to join us at our inaugural Moonshots Live event on September the 23rd. in downtown L.A. Alex, Salaim, Dave and I will be hosting 1,500 entrepreneurs, builders, and creators, and hopefully you, for a full day dedicated to designing and building your moonshot, shaping your mindset,

and steering humanity towards an abundant future. Get ready to enjoy incredible networking and an awesome party while walking away with the tools to change the future and the confidence that you can. Seats are limited. Admission is competitive. Check it out at moonshots.com. I'm going to move us to some fun and big news from this week coming from the Elon universe. So this week, Elon held SpaceX's first earnings call. And it was vintage Musk.

The numbers were big and bigger. I listened to the earnings call. I know if you guys did. I was on CNN and CNBC commenting on the next day. I wanted to share some of the top takeaways from the call. So first off, SpaceX expects to hit $100 billion in annual recurring revenue by the end of this year. And Elon set a target for a trillion by 2030, maybe by 2029. He moved up that target from 2013 to 2030, a trillion dollars of annual revenue.

SpaceX posted $7.8 billion in revenue this quarter, 92% year-over-year increase. He beat the street's expectations. The company has inked $6.7 billion in cloud service deals in the second half of this year. SpaceX plants hit 2 gigawatts of compute by the end of 2026 and up to 10 gigawatts by the end of 2027. And their near-term cash engine Starlink is outperforming. They have 12 million subscribers. That's 2x year over year. The revenues grew 66% to 4.3 billion.

I love this. Gwen Shotwell also implied that Starlink's direct satellite to mobile, I was going to take a big chunk of the $600 billion mobile. telephony tam. You know, I can't wait to be any place on the planet in the goby desert or in the foothills of Hollywood here and have a great cell phone signal. Oh, yeah. Yeah, it's amazing. Just for context, remember, no company in world history has ever reached a trillion dollars of revenue. Yeah. If anybody can, you know, I was on the news show saying, this is the first $10 trillion market. And after we start talking about the tariffab, I think it's the first $100 trillion company out there. Let's listen to two quick clips from Elon and then discuss them because a lot was said

and actually a lot more was said after the earnings call. A lot of news breaking out. All right, first one. As Brett mentioned, we are expecting to reach 100 billion plus ARR in December of this year. And it's probably also worth mentioning that our internal projections for reaching a trillion dollars in revenue, not ARR, but revenue, have moved up from 2031 to 2030. So prior to the IPO, the financial projections we had were reaching a trillion dollars in revenue in 2013. We now expect that to be in 2030. And there's a non-zero chance of that being in 2029.

And then he posted, or he said these, they got posted as clips. Let's take a listen. I find this really funny and fascinating. Actually, rocket science is literally our daily business. And rocket science is a idiomatic expression for extreme technological difficulty, and there's a reason for it. Because the thing, let me tell you what rockets desperately want to do every flight. They desperately want to blow themselves to tiny pieces.

and the engineering struggle is to convince the rocket not to blow itself in tiny pieces and actually deliver failure to orbit. So he was making those comments about how hard things they're able to do and building the data centers and building orbital compute is nothing compared to building those rockets. Gentlemen. Isn't it incredibly cute that Elon doesn't know with the definition of? of internal projection? This is a public earnings call, dude. He's just so awesomely transparent, isn't he?

Just incredible. So on top of that, okay, go ahead, Alex, I want to go on. Okay, for my endth hot take of this episode. Yes. I think this is not financial advice. I think the internal now external prediction is credible.

And I think there are two possible pathways to trillion dollars in revenue by the end of the decade. Path one is the SpaceX Tesla merger that we like to talk about via Optimus. So Optimus basically takes over the service labor, physical labor economy on Earth and in the solar system. That's one pathway to a trillion. Second pathway to a trillion answers the question that I had flagged a few episodes ago and didn't have the answer to, which is why isn't Invidia, which is so busy commoditizing the layer above them, also commoditizing the layer below them. And I think the answer that I heard from this group at the time is, oh, well, like his,

the invidia's relationship with TSM is too chummy. Oh, they'd never do it. They're too dependent on TSMC. I think the power move here for Path number two for SpaceX to a trillion dollars is SpaceX replaces TSM with the TeraFab. TSM has almost $100 billion a year in revenue. If you could imagine a future where TerraFab is basically TSMC times 10 or 20, it's American domestic. If you roll up the values of micron and SK. Hynix and TSM, et cetera, so memory plus compute, plus Fablis, Fab, and then 10X it, you easily get to TNXET, you easily get to TNXNX, and TSM, and TSM, you

easily get to a trillion plus. I think the power move here is Elon and Jensen get into bed together, and Elon becomes the new American TSM in deep, deep partnership with NVIDIA, giving Jensen the leverage he needed against TSMC and giving Elon the path to a trillion. And I think we were already seeing that, right? SpaceX has already announced a exclusive partnership with NVIDIA. You know, I was on the, something I didn't say on the, on CNBC, I think there's a lot of motivation for Elon to drive revenues faster than anybody could imagine. And that is to get the valuation of SpaceX up high enough to buy Tesla so that he's got unquestionable control. Right now, they're on par with each other. He needs to get the valuation up to like $3 trillion and then merge in Tesla.

Right. There's huge. So, Peter, sorry, just to ask you a question on that. It's saying bad things about Brett Adcock attempting to use financial engineering to increase his equity when bread is doing it, but not bad when Elon is doing it? Oh, no, because he's obviously doing it. He's like, he's like very clearly doing this. There's no secret here, right? So, let me just, after the earnings call, let me hit this one and we'll put more meat on the table to digest here. So on August 4th, SpaceX and Nvidia announced they were jointly designing the compute page.

payload for StarMind. This is SpaceX's orbital data centers. Each StarMind satellite is going to carry NVIDIA's Ruben GPUs and Vera CPUs. And interestingly enough, and really important, on the earnings call, he announced he expects to have the first StarMine satellites in orbit in 2027 a year ahead of what he was announced earlier. So that's fascinating. And then just to put more stake on the grill here. The big news here, the day after entering is called, Reuters broke the story that SpaceX and Tesla will initially invest $16.8 billion to build the tariffab. Alex, we were just talking about the most advanced semiconductor complex on the planet. This is a 100 million square foot facility. It's not a factory. This is a city, right? Let me just

share this image for everybody to see by comparison here. Actually, let's hear Yilan on the TerraFab and then we'll see the images. It's worth noting that there's not a single high-volume computer memory fab in America right now, zero. There's one being built in Idaho by Micron, but that will not reach following production until I believe 2028. And there's something built in New York, but they are in, I think, 29 and 30. And this is a, tiny fraction of the memory that's needed. And in fact, even if you take the best case assumptions of the memory makers and the logic makers, it is not enough to meet the demand that is anticipated.

And that's why we need to do the tariffab. Otherwise, there will not be enough chips. Yeah. And here, if you're watching this on YouTube, check this out. Here's the tarotab. It's a monster. It's gigantic, you know, compared to the Pentagon, compared to Apple, compared to the Mall of America, and compared to the Birch Khalifa. You were saying earlier, Alex, you know, you could probably see this from the surface of the moon. It looks like a buckle on the planet, literally, like, hand of God could reach down to this terra fab structure and grab the entire Earth by it. It's beautiful, it's enormous. And as Elon has now confirmed, if you're watching this and looking at the image, the circular structure. in the middle is wait for it, a free electron laser. So not only is, I think, Elon with the

TerraFab going after TSMC, he's also going after ASML. And right now, if you want to do extreme UV, which is the polite euphemism for soft x-rays-based lithography, right now you're stuck using ASML's approach, which is pretty expensive and requires creating a vapor of tin droplets to ultimately create those soft x-rays. By creating a free electron laser, it's a pretty adventurous approach to EUV lithography, he could radically reduce the cost and maybe even, even though I've read in public announcements, Elon striking agreements with ASML to purchase lots of ASML equipment. If I were ASML, again, not financial advice, I'd be looking at the circular structure right here

and saying the circular structure is actually a bullseye paint. on ASML by Elon. Oh my God. Here's the tweet. This is, uh, uh, it says, looks like a synchrotron. Elon going to build a free electron laser based EUV machine and Elon tweets back FEL FTW, free electron laser for the win. Yes.

Wow. Pretty extraordinary. So just going after ASML. Here's the numbers. TSMC invested 330 billion dollars to build out their fabs over the, course of 40 years, Tarifab is estimated $119 billion. It's probably a low ball number. But oh my God, this is the verticalization of AI at a level. I mean, this is, as you said, Alex, this is the $100 trillion company. And the Americanization, like again, ASML European company, TSMC, Taiwanese company,

Elon, if this is correct, he's taking that entire stack and he's not just verticalizing it, he's unshoring it to America for the first time in history. Yeah. Yeah. Hey, one thing to add to the party, Alex, you can confirm this. But I did some Gemini-based research on this. Because it didn't make any sense to me to build a building a single structure that big. I mean, I know there's ego involved and whatever, but like, why on Earth?

But if you generate the X-rays or the, you know, the EUV equivalent in that synchrotron, you distribute them to all the fab stations throughout this one, basically a linear accelerator distributor. And so by being a single structure, you only need one source of EUV that feeds many, many manufacturing steps across the whole building. That's what Gemini tells me. That's interesting. I'm not sure of the optics for that, how easy it is to, like, redirect x-rays that are coming off of a circular synchrotron to a linear arrangement is also just possible. I mean, maybe Gemini knows better than I do. It's in my mind is also possible.

It's just like literally a linear assembly line that's coming in from both ends or something. No, no, I think, tell me, like, we got to research this. This is just too cool. But you can't reflect those beams, like, and move them around a building. X-rays are hard. X-rays are hard to create mirrors for it. Like, x-rays really don't want to get mirrored. Right.

But even when we were talking to them in December, said, hey, my manufacturing process, the wafers are in these boxes. you can smoke and eat a cheeseburger over them, but they move around. So they can move to the beam. But the beam can only go in a straight line. Yes. But it can go very, very large distances. Hence, one long beam projector,

and all these manufacturing stations slide into the beam. Maybe. And Alex, we talked about this before, right? TerraFab isn't just a supply chain independence. It's geopolitical independence, right? If Taiwan ever gets cut off, the company that controls the spice, in this case the Chessab controls the entire AI industry. You could say, Peter, the flops must flow.

Okay, okay, a few things here. This is so cool. This is so cool. A few things. One is, why are we surprised? Because didn't we predict this like a while ago that this was a natural? No, this was a natural outcome that we would end up, he would end up, yeah, full vertical He'd end up building TSM, CSML, something like that.

He's talked about it, right? So that's one, quick thought. I think it's actually pretty surprising. Like, I didn't foresee, like, the satellite map. Like, the big, like, criminal, criminal, criminal, cremilological surprise for me is that circle. Like, I did not see that circle coming. How could you? This is so brilliant.

How could you see that coming? Yeah. I mean, he's always, like, this is very related to physics is cooked, you know? Like, the ideas like this are going to start. popping up in all kinds of weird places. That circle is profoundly surprising. Like, there were, like, it's funny. For years, this is a relayed story, but it's public information,

the Soviet Union used to stare at satellite maps of the Pentagon, and in the middle of the Pentagon, there's this courtyard. In the middle of the courtyard, there's this building, and the relayed this public information, the Soviet Union was convinced that this little building at the center of the Pentagon was actually like a missile silo or had some strategic significance. it turned out it was just a hot dog stand. This is the exact opposite.

This is a circle that has profound, profound geopolitical implications. If you asked me a week ago, how often in life does something completely change your view of the trajectory of humanity? A week ago, I would have said all of disassembling the moon, all of the, you know, the Dyson spirits all gated on ASML not being willing to more than double their throughput. of these ASML machines, that constrains chip production, which constrains the Gysen sphere, all of that is on a, all of a sudden, all that's out the window in just one week. Classic Elon. This one circle, like, changes the geopolitics of Europe.

It diminishes if this interpretation is accurate. This one stupid little circle that appeared in a satellite, not even actual satellite map, in a planned satellite map has profound implications for European economy, European sovereignty, for the future of China and Taiwan, just one circle appearing in a mock satellite map. I totally right. I totally agree. Can I finish my points?

Yes, sorry. I never finished my point. Okay, so a couple things. One is the number, like the, it's really fascinating here how hard it becomes to categorize SpaceX, right? Is it telecoms? It's four different companies. It's like all vertically integrated.

And I thought, Peter, your email on this that went out to your group with the four things was really, really big. Yeah, you're like, go read that. It's a big, it's a big deal. I know everybody's super excited, but I'm kind of like, we should, we so talked about this coming. So I don't know why everybody's so surprised. Maybe the fact that it's actually here and it looks like he's actually doing it is the big deal.

But I think we predicted this a while ago. Yeah, I mean, science fiction has predicted a lot of things. But when it materializes, the capital gets commitment, you know, the partnerships get announced. You know, and while Elon's probably never right on his timelines, he always implements the idea eventually. And it's exciting. Well, he's not like 10x everything on a quarterly basis. And also, to your point, like, I think there's like, there's a physics surprise here. It seems that he's confirmed the circle is a free electron laser. He's not using tin droplets to do EUV. That's a, that's like a technological surprise. That isn't just like, oh, Elon is out buying

ASML machines or contracting with TSM or Samsung, whatever happened to Samsung, question mark. He's off like doing his own thing, not just by taking the existing supply chain and putting it under one roof. There's new physics. Yeah, he's leapfrogging. Well, I think that the way to think about SpaceX is it's the vehicle for the intelligence and industrialization of America, right? like ultimately you need to build a brand new capital stock from rockets to chips to robots and Elon's got the financial arbitrage and the talent arbitrage and the shelling point now to bring that all together. So everything that will be economically productive in this next century, Elon's going full stack on. That's the fundamental thing. And the spending on that, the trillion

dollars we've had in AI chips is just the first stage. Because if you think about the spending that America and the rest of the world has to do over the next few decades. What is it? Robots, rockets, chips. And so that is his story. It's the industrialization of America for the intelligence agent. He will be the company that does that with the capital in terms of human and financial edge that's required to do that. I'll also point out one more thing, which is now that we know what a terrafab on Earth looks like, it's not difficult to take that same top-down view and a imagine what one of these on the moon looks like. And you know what else would benefit from that high aspect ratio dimension? A mass driver. A mass driver. Yes. So I looked at that and I saw,

oh my goodness, not just the circle in the middle, which is provocative and speaks volumes about the future of European economic sovereignty, but also the high aspect ratio length. And I saw, oh my goodness, this looks like it could be a mass driver on the moon. I think as well, as you get to the economic downcern that's going to come as human calling the labor goes negative, you have to know the U.S. government's going to go balls to the wall with infrastructure spending. And things like this will absorb that in easily, multiples of them. Yeah, you know, there was a comment made someone made before, which is buy SpaceX stock for your children. And again, not investment advice, but this is a generational company. It's a

Civilizational Infrastructure Company. It was nice to see the stock, you know, a couple of days ago, 20% of the stock float became available for a lot of people, including myself, and very few sold. In fact, the price has gone up over the last two days. A lot of belief in where it's going. Gentlemen. Full disclosure, I'm now a small SpaceX or older. Yes, well, it's my single largest holding.

So, and I'm going to be holding. You know, the old hoddle for Bitcoin, hoddle your SpaceX. Gentlemen, a pleasure. Imad, thank you for joining us today. Always, always a pleasure. And guys, I loved the episode. Oh, God, learned so much. I always learned so much, but it drives me crazy

that I have to go listen to each one again just to understand and fully digest what we talked about. But you have to. You have to. Love you guys. Hey, y'all's Kelly Clarkson with Wayfair. Ever order furniture online?

and wonder what if? Like, what if it doesn't hold up? That sofa was four days old. You should have ordered from Wayfair. With Wayfair, there's no what if. Just style you love and quality you can trust. Visit Wayfair.ca. Wayfair, every style, every home.