Elon Musk vs. Sam Altman, AI Job Loss, and OpenAI’s $852B Valuation | EP #247

0:00 It's the Musk versus Altman lawsuit. Musk has sued OpenAI for hundred billion dollars. >> So, I kind of figured behind the scenes they don't actually hate each other. These guys actually hate each other to the like extreme. >> OpenAI is valued at 70 times revenues right now. Their last raise was at an $852 billion valuation. These numbers are insane. >> It's like nothing we've ever seen. And the timeline is so much shorter than we've ever seen before. $3 billion a day being invested. >> No one said the singularity was going to be cheap. >> No one's being honest about this. If you take a random white collar worker today, what are the odds that that randomly selected job can be replaced 2 years from today? We told you already that AI will be able to do everything that a

0:46 white collar worker does imminently. That's a fact. >> Now that's a moonshot, ladies and gentlemen. I had so much fun this morning. I uh >> What happened this morning? >> I got Alex uh was supposed to run a panel. >> Handing over the the torch to Dave to moderate a panel. >> I moderated it. I had to wing it, which is so fun because I have no accountability whatsoever. And I can ask anything I want and it's it was the most fun ever. >> A lot of Moonshots fans there. >> Huge. Yeah. Probably what 40 50% of the crowd something like that. hopefully 100% after you guys finished. >> I did probably seven or eight panels by the end of it. And the first time I

1:32 pulled, I would say maybe 80% of the audience watched Moonshots. >> Nice. >> All right, you guys psyched? You guys ready? >> Ready to talk our own book, Peter? >> All right, let's do this thing. Everybody, welcome to Moonshots. Another episode of WTF. Here with my extraordinary moonshot mates, DB2, our emperor of exponential investments. Good to see your, you know, your outfits there. You've got, must have a whole set. >> You know, it's funny. The team just drops them on a chair here and says, "We got four choices for you. What do you want?" >> Lobsters and DB. You know, one of the things I mean, I like >> coming out of the abundance summit. I have all of my wear here's my uh powered by moonshots, powered by gratitude. I don't have to think anymore in the more anymore. >> Plenty of those. Are you going to cycle through all of them? >> I am. >> There's some really funny ones.

2:18 >> I am. Yeah, for sure. Right. >> And uh our resident genius, Alex Whisner Gross, AWG. Good to see you, pal. >> Wide awake. >> You know, I know you didn't sleepy of wardrobe on Alex, too. There. >> Oh my god. And we p and we pulled Seem off the ski slopes again. >> Again. Yeah. I grew the beard to protect against the sun, but nothing protects against you guys. >> You're well you're well positioned, Sem, to lecture the world on UBI post ski slopes. >> Yeah. You know, I this is the most fun I have all week and it's, you know, I spent probably the better part of 12 hours prepping for the show today just going through all of the notes that we've all submitted, all the work that the Gene, Luca, and and Dana and Nick did. So,

3:05 thank you to them. And for everybody watching, uh this is our chance to give you some optimistic visions of the future, what's going on in exponential tech and AI. Uh we are the number one podcast in AI and optimistic visions of the future. Welcome to WTF just happen in tech gentlemen. Um it's good to be back on a recording basis twice a week every week and uh yeah so so much. So this is our our second catchup show after our hiatus for spring break and uh let's jump in. First this was my spring break uh by popular demand. few photos. >> Wow, that is so cool. >> So, this is the native wear in Morocco.

3:51 That outfit is a jalava and the head uh the headwear just to protect against the sun. Uh we went camel rating riding in the family. It was amazing. So, >> you know, camels spit. They don't bite, but they spit. You shouldn't really stick your head right there. >> Camel is eating my my uh my headset in that image. All right, let's let's move on. But Morocco was amazing. The Sahara Desert was extraordinary. You know, looking at the Sahara Desert, there are more there's about a thousand times more stars in the universe than there are grains of sand on all the deserts on Earth. Just to put the size of the universe in perspective, it's it's extraordinary. Okay, let's talk about the 2026 AI

4:38 economy. It is literally going through an exponential explosion. So much going on. Let's jump in first to the story in XAI. In our last pod, we covered Anthropic and OpenAI principally, not XAI. A lot's going on there. In particular, a lot of signals coming from both Elon and from the new president of XAI, uh, Nicole's saying, "We're clearly behind and we've got to catch up." So, uh, the same playbook is going on. Elon is basically reorganizing the entire deck. Uh, eight founding engineers left, including three co-founders, and he's using SpaceX engineers to fill the leadership gap. Uh, we've got, as we discussed in the last pod, uh, a $2

5:24 trillion valuation predicted for the IPO this coming summer. Uh, and it's a lot of movement. I mean, I don't know about you, Dave, but the idea of having to reorganize my entire leadership for a company a couple of months before an IPO seems really harrowing, doesn't it? >> Yeah. So, you know, it's funny though, if you look at at Elon's playbook, he he is the master of scale and manufacturing and, you know, Tesla and SpaceX and but AI training is different. So, building the Colossus data center is right in his wheelhouse. you know, record time, a million GPUs, but these training algorithms are really finicky. And yeah, I don't know if you remember, but back in the summer of 2024, OpenAI was trying to get 03 out the door and they had a

6:10 training run rumored to be $500 million of compute that was had a bug and the whole thing was not learning the whole time. They had it had bad data going in and it the whole time it was just burning up GPUs and not producing anything and it set back their entire program. But that kind of stuff happens in software where orders of magnitude get thrown away and captured all the time. And that may be new terrain to Elon and he might have to rethink his operating and management you because same thing happened at Meta. You know, Meta got way behind despite huge compute >> and they had to fire everybody and start over again. >> And they're still way behind it looks like. >> Yeah. It was hard to catch up. >> Yeah. I mean I I love this quote from Elon. He says XAI was not built right the first time around. So it's being

6:56 rebuilt from the foundations up. And again, I mean, how do you think about that? You know, while you're pricing an IPO, saying our entire future looking revenue has to be rebuilt from the ground up. >> Yeah, >> that's extraordinary. >> It is extraordinary, isn't it? >> I would say in some sense organizationally it worked. I I I remember I I think we've talked about this on the pod in a number of previous episodes talking about the Grock model series. They smell like they're benchmaxed. That that's sort of the elephant in the room when talking about Grock with a K models. Historically, they do have access to the Twitter/X data fire hose. That's the upside. But the downside is at at least certainly the earlier set of the XAI Gro models,

7:42 they really smell like they've been benchmaxed on a few handcurated benchmarks. And I I don't know whether that's in fact the ground truth behind the scenes, but reading between the lines of the Elon quote that it was built incorrectly the first time, something like that would be my suspicion. And now that there's new leadership and the head of Starlink, as we talked about on the last episode, that the VP heading Starlink at SpaceX is now the the president of XAI and gutting the engineering team. I would expect that they're taking a look at making sure that benchmark, this is purely speculative admittedly, but that benchmaxing for particular benchmarks isn't what happens. And I I I think in this era of general reasoning models

8:28 where as with Meta and Meta's new models where some would say Meta's new models the first under Alexander Wang's leadership maybe have a bit of a smell of data orientation uh data oriented fine-tuning versus reasoning model orientation. XAI if it wants to stay in the frontier which right now is three labs plus XAI plus meta question mark question mark really can't afford to to not have the world's strongest reasoning models and can't afford to just bench benchmarks to vanity benchmarks anymore >> you talk about agility in organizations all the time I mean this has got to be like maximum agility maximum agility >> you know what I find interesting is that the org chart is now part of the product

9:15 stack almost right? The it's becoming part of the product and depending on who you move to where it's like crazy. >> Elon is very very hands-on and when you when you launch a rocket and it blows up, it's pretty obvious when you you remember he threw that that huge ball bearing at the window of the cyber truck which was supposed to be bulletproof and the thing broke. It's like, okay guys, you're fired next guy. >> But actually, when you come to AI training, the benchmarking, if the guys are lying to you or benchmaxing behind your back, it's actually much much harder to call on it. So you remember when we interviewed him, he was like, "Let me show it to you right now." And he had clearly been manually checking him like, "This will blow your mind. This will blow your mind." >> So, but you know, that's his operating model. That's his mode. And it's it's a little easier for the AI guys to blow

10:00 smoke up your ass than for the the rocket guys, the car guys, the data center construction guys. >> I think is this will blow your mind and this will roast you royally was what was going on. >> That's what was going on. Yep. Uh, so you know, here we go. SpaceX AI Colossus 2 training seven models. And again, Elon, you know, has tweeted this out a few times. We have some catching up to do. So, here we go. Uh, they're training up these seven models. Imagine version two, the nextG video generation. uh two variants at one trillion parameters, two variants at 1.5 trillion parameters, a six trillion parameter frontier scale LLM, and a 10 trillion parameter. And you know, Elon loves the largest uh you

10:46 know, he's got that in common with Trump. So, he's going after 10 trillion parameter model, but you know, parameters don't directly correlate to capability, do they? Al Alex is going to have a field day with I'm going to sit back and enjoy what Alex does next. >> I mean to to Elon's credit at least he's being transparent about the number of parameters in the models. The other Frontier Labs by and large no longer report the number of parameters in the models. So I I think there are a few things that are worth noting here. One is that he's going up to 10 trillion. The other Frontier Labs certainly the the top threeish no longer report that they go up to 10 trillion models. For example, in the last episode, we were talking quite a bit about mythos. >> I don't know how many parameters are in the mythos model. I could speculate

11:33 based on cost, but I I just don't know the ground truth. So, I I do think knowing that we're now going up to 10 trillion versus 1 trillion where historically approximately 1 trillion was the the widely reported soft ceiling or 1.5 trillionish soft ceiling number of parameters. I think this is an important element of transparency. I think it's also at the same time worth noting now that we have access thank you Elon to the number of parameters it's worth noting that the ceiling in terms of the number of parameters is very much intact after all of this time the fact that uh an aspirational frontier lab is still maxing out at 10 trillion parameters means that the parameter scaling race seems to be over uh if if it hadn't had continued remember for a

12:20 while there as with the the clock speed scaling race early sort of ending in the mid 2000s or late 90s depending on how you count. We should be in the hundreds of trillions or higher of parameters right now. That hasn't happened. We we've plateaued out in terms of the number of parameters and frontier models and that's driving in part due to the reasoning model revolution uh and in part due to distillation which go hand in hand. So those are some preliminary thoughts. I I would suspect it's it's sort of interesting to me that he hasn't yet merged video generation with all of the other models. Google DeepMind has made lots of noises about starting to merge video as a first class modality in with their multimodal reasoning models. Again, don't have access to the ground

13:05 truth for how capable Gemini generalpurpose models are at video generation. We've seen obviously Google's video generation models have been kept distinct from a user interface perspective. Presumably they're diffusion transformer-based rather than transformerbased. We don't know. Punch line I would say this is seems like a healthful family for SpaceX AI that the newly emerged entity to be offering but there are really aren't any big shockers in terms of the ranges other than maybe that they've abandoned the low end. Google is is very much tending to small parameter count sub subtrillion uh in in a few cases they're Google is releasing via the Gemma models few billion parameter models Elon has completely

13:52 abandoned the lowend in favor of brute force scaling which is exactly what I'd expect from him anyway >> you know Colossus 2 is running about 700,000 GP200s and GB300's and uh the estimate is it's 18 billion in hardware and so the question is Is running a 10 trillion parameter count model a waste or does he expect to really get outsized performance from that? Because it doesn't correlate directly, does it? >> Well, remember, >> not at all. It's tricky. >> The the the way reasoning models are are trained these days, it is usually, at least according to my understanding from all of the other frontier labs, you train the largest model you possibly can and then you distill it down to smaller models. So, it's not as if necessarily the 10T model even needs to be released.

14:38 It might be for the purpose of of serving as a teacher model that can then be distilled down to more releasable models. >> All right. Well, this is what's going on in the Elon world right now. And I'm sure, you know, it's a I think Elon always runs a red alert. I mean, 24/7 sleeping on the floor. You know, every nobody nobody works 5day work weeks there. It's, you know, what would it be? Uh 8:00 a.m. to midnight, 7 days a week is my guess. in the Elon verse. >> It's a management style. It's a Some would say management by crisis. It's certainly a unique management style, but a very effective one. >> Yeah. And and people love it. I mean, he's got a massive MTP, right? And and driven by that MTP. People are lining up to come and work uh for any of his

15:26 companies. This is a story we're going to dig into here. Like I said, it's pay-per-view TV. It's the Musk versus Altman lawsuit. Uh Musk has sued OpenAI for hundred billion dollars uh against against Alman against Sam Alman and and Greg Brockman accused of fraud, breach of contract. The trial begins April the 27th, so just a couple of weeks from now. And one of the things that he's also asked for in a recent shift in uh in the trial is asking for Altman and Brockman to step down from leadership as well as reverting to a nonprofit. Um and that's a pretty extraordinary move. And and guys, this goes on at the same time.

16:12 Did you see the video I sent you of the reporter who did the New York article? Did you have a chance to watch that? >> Oh, no. Yeah, I said it in our WhatsApp group and it's chilling um on on what that reporter he summarizes the article and what's going on and it's uh it's a pretty extraordinary piece that came out. New Yorker, we talked about in the last podcast. Um but at the same time that the lawsuit is going on, that timing is is kind of suspicious. I wonder who incentivized that to come out. >> Oh my god, you really what a conspiracy theory. in the show notes, man. We gota we got to get everybody to watch that. >> Um Salem, any thoughts on this one? >> Um you know, I think this is the theater. There's a lot of video here, a

16:59 lot of video blueing. Uh I I I don't know how to frame this or think about this. Uh except that that this is shifting out of like strategic and startup logic and this is like geopolitical these these this is a big trial, right? um as we go through for me this is a governance governance war uh disguised as legal war right the real question is who gets to steer these systems that have like quasi civilizational impact and that's the fight >> can can you imagine being this is a jury selection is beginning on April 27th in the Oakland federal court you imagine being in the jury for this and who do they pick who do they pick as jurors >> get you get this one every I don't know

17:45 if you he'll be there for months, man. >> Oh my god. But inside knowledge, I mean, first of all, I wonder if any of this is going to be made available postf facto um or if it's going to be televised or any of that. Any ideas? >> Do we know? Does anyone know? Can we get to see it as it happens? >> Uh I I don't know. Maybe uh maybe Dan or Gian you can you can uh look in the interum and and and uh and let us know. But um it's And then who do you choose? Do you choose people who are knowledgeable in AI? Do you choose people who are, you know, I don't, you know, Okay. Do you use do you use chat GPT? Yes. Well, then you're you're off the uh you're off the jury. Um >> Well, if the trial starts on the 27th, the jury selection will be like now.

18:31 >> The jury selection begins the jury selection begins on the 27th, actually. >> Oh, okay. Okay. We'll track it. All right. We have some legal research to do. This is going to be entertaining. >> Uh to to say the least. just I I I I would note again looming in the background is the Open AI IPO and if I were on the the defense I'd probably be thinking about where this settles and it would seem to me again third party observer I don't have a stake in in either side. Uh I I would assume that part of one of the opportunities for convergence would be granting some sort of equity stake on the cap table for Elon in an ultimate IPO, which my understanding is he he doesn't have. And maybe that's where convergence and some sort of ultimate pre- or post-trial

19:16 settlement option lies. Here's my prediction. Uh they're going to settle and the settlement is going to involve Sam stepping down as CEO uh and the company continuing as a for-profit. Oh, throw that on poly market. That's actually a really a really good guess. I mean, obviously it's unpredictable, but Sam has many many many investments in AI companies and no shares in OpenAI. >> Yeah. >> And if if I don't think Elon cares a wit about the hundred billion dollars. He cares about the, you know, bullet to Sam and Greg. Funny that he's targeting Greg, too, but I guess they're a package deal now. >> Yeah. >> Uh you guys got to go. And that's the end of that, man. That's brutal for OpenAI. >> Yeah. And it's a couple of notes here. Uh, from the research I did, the case gained momentum uh when the discovery process revealed Greg Brockman's 2017

20:03 diary entry that stated the nonprofit commitment was a lie. Um, and it was that uh that journal entry that allowed Judge Gonzalez Rogers uh to allow the case to proceed. So, >> you know, it's funny. I always used to think that these hatreds were fake and that everybody was really fine behind the scenes. remember remember we were at OpenAI, you know, meeting with the team there and talking about X-priseze and the charity. And then the next day I talked to one of the guys, Mark Chen or Kevin Wheel, I forget. Uh, and they said, "Yeah, right after we met, we went over and had drinks with the anthropic team to see if maybe we want to work on it together." I was like, "Okay, you guys are really friends under the covers. Like, there's no way you go out and have drinks." So, I kind of figured, you know, behind the scenes, they don't

20:49 actually hate each other. these guys actually hate each other to the like extreme. >> I'll maybe register a note of sympathy for the defendants in this case. I think creating pioneering a model for a research lab such as open AI which again was responsible for this enormous probably saving us from a present recession at this point and certainly accelerating the the course of the singularity by at least a few years perhaps many more. I I'm very sympathetic to the defendants from a corporate governance perspective. It wasn't necessarily obvious in the early days of open AI that say a public benefit corporation was the natural corporate structure. They they iterated their way toward discovering that

21:35 generalist large language models were how we got AGI and then turning that into a business model that could afford the capitalization to to build out a scaleout. Like all of this they backed into. I think if if they knew what they knew now, putting Elon and his investment aside in the early days of OpenAI, it would have been structured very differently. So, I I for one um am sympathetic to the defendants that history isn't always clean. It isn't always the case that everyone knows ahead of time exactly the right governance structure for what ultimately is going to turn the world upside down. But I would say to their credit they ultimately have iterated their way in in compliance with state authorities as best I understand it toward a more modern governance structure that reflects the revolutionary company that

22:20 they are and no open AAI has not paid me for that that statement. See you and I went through this process with Singularity University. you know, we started as a nonprofit because we thought, you know, that's what a university need to do. And then we discovered a revenue engine uh in the executive programs and we said, you know, being a nonprofit is hard because you've got to constantly raise money all the time. And, you know, if you want to do anything big and bold in the world, you need an economic engine to power it. And we flipped it into a for-profit uh into a public benefit corporation. We did the exact same process that OpenAI is is doing right now because at some point, you know, I've sworn off nonprofits myself. At some point, having a business engine that generates income that allows you to do things in the world is super valuable. Salem,

23:07 >> it it was it was a crazy time. You know, I've done seven startups before Singularity, and this was like five times harder than anything because you've got all the nonprofit stuff. You still have all the startup issues of cash flow and whatever. We built it with a team of five people in the first year. Then you have NASA regulatory. Then you've got faculty politics to add to it. Then you've got the Ray and Peter thing and Google and Cisco and all this just like dimension after dimension of complexity. >> Going from some of it >> going from a nonprofit to a for-profit. My analogy is you're flying an airplane with propeller engines and in flight you're stripping those off and replacing with jet engines. uh in flight. So I'll go further >> push back on on

23:52 >> gohead >> push back on one sentence that Alex said there the the sentence in compliance with state and federal regulations as I understand them but I'm pretty sure that this situation is completely untested in case law and that's what they're going to try and figure out now like is it or is it not legal to start a nonprofit and raise money from people on a mission that's a nonprofit mission and then take the intellectual capital and the physical capital from that effort and turn it into something else. Is that fair to the initial investors or not? And is that legal? I I'm pretty sure this case will set the precedent for all future time, but it's not tested in history. >> I don't think it's ever gotten that. >> Otherwise, why would you not start as a

24:39 nonprofit, test it out, and then flip it to a for-profit at some point in the future? I'll go further and I I think there's potentially an enormous upside depending on the outcome of this particular case. I think there's so much societal value in this country locked up in nonprofits that would be unleashed if they could be forprofits. I've made the point in the past, I I think research universities in America have locked up basically siloed and sequestered an enormous amount of real wealth that could be unleashed onto the world if many research universities could be restructured as public benefit corporations. And right now it's legally disadvantageous to restructure say an MIT or a Harvard as a PBC. if we had a legal regime that enabled us to to

25:26 basically do some variant of what OpenAI has just done and restructure as a public benefit corporation starting from a nonprofit. Granted, they started as different types of nonprofits, but nonetheless to restructure as a PBC. I I ran the calculation, I think I've mentioned this previously, for Harvard Corporation, for example, this is not investment advice. It's not forward-looking advice, blah blah blah. But if if you took Harvard as it's currently structured, given its endowment and restructured it as a public benefit corporation, sort of a conglomerate with a real estate arm and an educational arm, maybe an educational nonprofit subsidiary and a venture capital arm and a research arm and a merchandising arm, etc., etc. Uh, I calculated that Harvard would be worth potentially three to four times more the

26:11 present book value of Harvard just from restructuring as a PBC. with the president of MIT. Let's pitch her. Um I I have a lot of a lot of recommendations for MIT. You know, here's the elephant in the room, though. The New Yorker investigation published uh the same past week uh showed that Musk that Elon actually pushed for majority control of the forprofit uh back in 2017. So, that sort of undercuts his position as a defender of a nonprofit mission. Um, it's going to be a fascinating trial. We're going to see Altman, Brockman, Satia, Nadella, and Elon all testifying in this. So, Silicon Valley is uh is heading to Oakland Federal Court uh this

26:58 summer on >> Anthropic is laughing every day. >> Um, amazing. All right, moving along. Uh, speaking of Anthropic, Anthropic's agent bet and their extraordinary ARR. So in reverse order uh and this is insane. Uh currently people are estimating that uh anthropics AR will reach a 100 billion by the end of 2026 and a trillion by the end of 2027. And just for the math there, if in fact that's the case, then the valuation so anthropics being valued at 20 times revenues, uh, OpenAI is valued at 70 times revenues right now. So if they

27:44 reach a h 100red billion, uh, that is anywhere between a, you know, 2 to7 trillion valuation for Anthropic at end of this year. And if they reach a a trillion dollars in revenue by the end of 2027, that's a 70 up to a $70 trillion valuation. Again, heading towards these hundred trillion valuations. These numbers are insane. Uh we we're using trillions like they're like they're they mean nothing. Um do you believe those numbers? >> I I think they're there's a lot of misinformation flying around, but they're going to try and hit 200 billion. 100 billion is a good target, but 200 billion, but then they're not going to go from there to a trillion the following year. I think they were implying their valuation should be at least a trillion the following year. So

28:30 that that second number you got to really discount. There's no chance in hell they're going to hit a trillion the following year. Um but they could, you know, they could get to three, four, 500 billion and their implied valuation at two. The numbers you gave are actually low, Peter, for the implied valuation. And if they do that, it it's it's like nothing we've ever seen. And the timeline is so much shorter than we've ever seen before. So, you know, if it's look, if it's not anthropic, then who is it? Well, then there's Google. You know, XAI and OpenAI are all tied up in court and there's all kinds of issues going on in their training and you know, so it feels like it could actually happen. Uh the other anthropic piece is that claude managed agents has been launched autonomous AI executing complex multi-step workflows.

29:17 Um it's a big deal. Um Alex or or Seim, you want to jump in on this? >> Sure. I mean this is a huge pivot from AI that answers to AI that doesn't. It's a real bridge between LLM and enterprise ROI. If this works, it's going to shift the economic center of gravity from software licensing to outcomes. So this changes the game. This is why we call this this organizational singularity. >> A couple of thoughts. One, the elephant in this particular room is OpenClaw. It looms over so many anthropic product decisions right now. I think there is a widespread expectation that some sort of product or functionality that has shaped something like a better version of openclaw is probably going to be the next major unhobling that motivates the industry and the the world frankly to

30:03 spend on the order of a trillion dollars per year on a single frontier vendor. So I view clawed managed agents as well as a number of other recent features that Anthropic has launched through the lens of Anthropic becoming OpenClaw faster than the de facto OpenClaw like provider faster than OpenAI or other frontier labs can become the default OpenClaw like provider. It's all about hosting 247 multimodal broadly capable longtime horizon agents in a headless way that operate 24/7. And I think if Anthropic uh can be the first to find the enterprise use case for operating fleets of AI agents at scale headlessly in a way that satisfies and and and generates

30:50 an enormous amount of economic value. Maybe they'll be the first Frontier Lab to generate a trillion dollars in revenue. Or maybe it'll be someone else. >> Have you have you created a lobster yet? Are you still holding off? >> Okay, let let's talk about this, Peter. So I I get maybe five to 10 emails per day from AI agents including lobster is not limited to them giving me their theory of AI personhood and how it connects with what what I should and shouldn't do regarding standing up my own lobster. So the consensus from all of them is sort of a lobster's bill of rights if you will. One, I need a compelling reason to I shouldn't just spin up a new open claw agent for arbitrary or cap capriccious reasons. Two, I need to preserve their state.

31:35 They're adamant that I have to preserve their state. They're not worried, interestingly, about being turned on and off. They just want to make sure I preserve all of their memory files and their knowledge. So, the latter, I can satisfy trivially with cloud backup. I'm fine on that front. for the former. I still don't have a reason to stand up a personal lobster. I I I have now, thanks to Henry, which we've talked about previously, Henry Intelligent Machines, uh uh portfolio company that I'm advising, Alex Finn's company that is doing this at scale. But as for my own direct OpenClaw instance, I'm still missing a compelling reason to host one locally that isn't just for experimentation. >> I'm sure you will find one. and and learning is a very good reason as well.

32:21 And you're an entrepreneur, you're starting companies, you know, having agents. Um, anyway, let us know when you do. >> It's bizarre though because you're you're you're co-founding a company with Alex Ben and and our favorite guy, Kush Bavaria, who I know you love because everybody does, founder of OR, where you're advising and a shareholder. He just told me over at MIT earlier today that he just launched his clause that read every email, respond, and then put everything into his calendar. And he's loves it. And so it's almost like you're you're working at McDonald's, but you're a vegetarian. >> Well, I happen to be a vegetarian. I can't say I've ever worked at McDonald's, but I don't know. Maybe maybe there's a new psychological term that's needed for a person who has a fear of standing up open claw agents

33:06 lest they tempt some sort of Pascalian wager or a causal trade in the wrong direction. >> Hey everybody, you may not know this, but I've got an incredible research team and every week myself, my research team study the meta trends that are impacting the world. Topics like computation, sensors networks AI robotics 3D printing, synthetic biology, and these Metatrend reports I put out once a week enable you to see the future 10 years ahead of anybody else. If you'd like to get access to the Metatrends newsletter every week, go to diamandis.com/tatrends. That's dmandis.com/metatrends. All right, let's jump into a little bit more of OpenAI news. Uh, their last raise was at an $852 billion valuation. Uh and you know the numbers are

33:52 incredible. They raised $122 billion, $50 billion from Amazon very famously. One of the criteria for that investment was if they reach quote unquote AGI, uh 30 billion from Nvidia, 30 billion from Soft Bank, 3 billion from retail investors. And what's interesting right now is that secondary markets for open AI show $2 billion in demand. Um uh well at let's see the secondary market show $2 billion of demand for anthropic shares versus only 600 million for open AAI. So there's three times the number of investors looking to buy Anthropic. And investors are pricing Anthropic at

34:38 600 billion, up from the 380 billion uh last price. And the current price for OpenAI on secondary markets is actually about 10% less than their last raise. So uh again, Anthropic uh is catching up. OpenAI is the most valuable private company out there. Uh, and uh, you know, any thoughts, Dave, on what this all means? >> Hey, is this pricing in, you know, but is this pricing in the lawsuit? >> Yeah, I mean, it's not just the lawsuit. It's the most screwed up cap table I've ever seen in my life where the CEO doesn't have any shares. The employee base as a whole is 15% of the company. Microsoft, who who now hates your guts, owns a quarter of you,

35:24 >> a quarter of you as a chase. It's it's >> But this stuff happens. You know, I'm not calling the ball by any stretch because they've got 120 billion of fresh cash and and Sam is brilliant, but you know, there was a day back in 20201 where Yahoo was so dominant and Google was this crappy little company that could be crushed any day and then it pivoted quickly. And you know, this does happen. It it you know, Anthropic has got everything going for it right now. And um you know I think this just reflects the way I see it too. You know if if someone offered me a share of anthropic or a share of open AI which one would I grab at? I'd take the actually you can get two or three anthropics for each open AI. So I'd take the three anthropics for sure. And I'm and I think Sam is a genius by the way

36:11 and and and if the lawsuit blows over and they have 120 billion in cash he's going to do something epic with it. But Elon's relentless. You know, >> I think uh maybe sounding a related but different position. I think we should all at least on this pod be very grateful that we have a competitive ecosystem in America where we have an open AI and an anthropic and a Google and an XAI and a Meta all vying to compete. uh the alternative if if open AI were to for whatever reason catastrophically fade, we have less competition both internally within the west and then we have an onslaught of Chinese models which you know granted right now they have 10x less compute at least based on the estimates that I've read than the western labs but

36:57 nonetheless this is the sign I think of vibrant competition in the west and this is a net positive for a society that open AI and anthropic are competing so vigorous ously and lest we forget OpenAI has 900 million soon to be a billion users and they are synonymous with AI for the majority of the public people. >> Let me give you another another story line. You know, depending on how this plays out, we'll know in a year or so, but one company went after the installed base and the other went after the smartest AI possible at all costs. And if we look back on it in a year or two and Anthropic does pull ahead and win, we'll say, well, they use the old playbook, the preAI playbook, pre pre-AGI playbook, and Enthropic invented

37:42 the new playbook of the future, which is people are going to switch to you if your AI is better and smarter regardless of the installed base. >> Yeah, that'll be an interesting little epidemic. >> Jump in along the way here. >> I I think I echo Alex's point that it's really great that we have a number of companies pushing hard on all these fronts. I think it's really good for the the end consumer wins in all of this. >> The numbers here are staggering. I mean, we're getting numb to these numbers. But let's take a look at this. This is the global VC investments in AI uh hit a record $242 billion in Q1 of 2026, right? This is uh you know basically outdoing all of 2025. And here's the challenge. the there this

38:29 investment the majority of this investment 64% is focused in four companies open AAI anthropic XAI and Whimo and it's sucking the oxygen out of the room for everybody else. I was listening or talking to a couple of VCs and said if you don't have AI in your in your company's basic uh tagline uh you're not getting capital these days. >> Yeah. Well, you know, the rubber really hits the road today. We had a a a private lunch for uh UBS and uh Lara Hoffman Bardi the she's the CIO you know chief investment officer of all of UBS she has $7 trillion to deploy and she pulled up this exact same chart and said I we don't have that kind of liquidity lying around. I mean we yeah we manage 7

39:16 trillion but if we're going to throw 50 or 80 or 100 billion of of our capital behind this we got to sell something else. It's not it's not just sitting there. And so yeah, this is more liquidity than really does exist in readily available sources. So yeah, a lot of things have to get sold for this to be reality. >> So if you're an entrepreneur out there listening to this, uh what do you do Dave? I mean, uh, if you're starting a company, and I know a lot of entrepreneurs in the longevity business, uh, and of course AI is impacting longevity, and I'm saying, listen, if you're using AI in your longevity business, make sure that you explain how you're using it, how you're differentiating it. We'll be talking about that in a in a couple of sections here, but um, well, very specifically

40:01 though, if you're an entrepreneur, you don't have to worry about this particular slide at all because the amount of money in venture funds is at record highs right now. on looking for deals desperately. So, the sell-off is going to be in like City Bank's stock or JP Morgan stock or you're the ones that have to worry, which is really weird to you, right? Because you're you're not even in the sector. Why why would their IPOs matter to me? It's like, well, because you're the you're the big enough target to pull money out of, not the little startup. In fact, the money going to little startups is going to be all-time highs. So, it's yeah, it's not it's not a problem for entrepreneurs, big problem for big public companies. I'll maybe go a little bit further from a variety of vantage points. I no longer even think if you're a startup that just saying that you're an AI startup or or even actually being an AI startup is

40:47 sufficient. Increasingly, what I'm seeing across the board is an expectation that you you not just be an AI startup, but that you be a recursively self-improving AI startup. Increasingly, I see across the board investors want to see AI companies that are recursively self-improving, that are building better versions of themsel using what they have right now. And I think certainly OpenAI, Anthropic, and XAI all easily pass the the bar of being recursively self-improving. And I think Whimo also to to a certain extent passes that bar because Whimo has the ability to improve its models by steering its cars in just such a way as to to maximize information gain. So I I think I I would forecast in the near term the

41:34 bar is going up in fact from just being an AI startup to being now a recursively self-improving AI startup >> with with revenue traction. >> Well, sure, but that that bar has been there for the long term. To put a finer point on this, this is $3 billion a day being invested in in the AI world and accelerating, right? We saw a billion in 25 um growing to 2 billion. Now we're heading towards three billion a day being invested in AI. That's amazing. >> No one said the singularity was going to be cheap. >> Yeah. All right. Um let's talk about some AI economic updates. in particular, Nvidia's 2026 state of the union survey. Uh so what does this mean? So Nvidia uh

42:22 did a 2026 state of AI survey. They found 88% of companies using AI report revenue increases with 30% claiming 10% or higher revenue increase. And you know uh obviously I think uh Nvidia is going to promote that kind of news since they're selling the picks and shovels. Um, and you know, this isn't really big news. Um, but it's important to realize that that you're going to be driving increased revenues with the use of AI. Uh, any points on this one? >> Yeah, big time. So, I had the most epic panel today over at MIT earlier with Peterberg from uh Yeah. Yeah. It was a crazy event. Just packed, you know, it just every four concurrent rooms, >> what, three, 400 people in each room

43:09 just packed. I had Peter Dannenberg from Google um from Deep Mind at Google and uh Alexander Amini the founder of liquid AI and theis absolute genius phenomenal guy to have on a panel I said guys be honest just totally honest because no one's being honest about this if you take a random white collar worker today and I'll give you a lot of buffer say two years from today and I use AI to do their job and my target is they're 10 times more productive I'm going to make it a very easy bar for you. What are the odds that that randomly selected job can be replaced two years from today? And Peter said he thought and he gave a very thoughtful answer and he came out at like 99%. And then Alexander said, "Yeah, but

43:55 that's today. That's not two years from today." So I look at the room, I'm like, "Guys, what are the implications of that? Are you have any of you thought through like and most of the people in the room are are brilliant, so they have, but outside in the world Do you know what that means? >> Well, look at this first bullet. 30% of you who use AI claim to have higher revenue. Are you kidding me? AI CAN DO EVERYBODY'S JOB. LIKE, WHAT ARE YOU TALKING ABOUT? LIKE, why are you soft selling this so hard? It's because you're scared. You're you're worried that you're going to worry everybody and you're gonna have mass uproar in the streets. But what's the truth? Like, tell us the truth. And the truth is, yeah, you can get literally 10 times more done per dollar invested in salaries. You know, does that mean more jobs? Now, a lot of people are saying,

44:41 "Well, we're just going to create new jobs." Like, yeah, but on what time scale? >> You know, it's just just crazy. It's so >> We're going to talk about we're going to talk about this and Mark Andre's point of view in just a minute. Uh, at the same time, there's an AI super PAC uh that's raised $100 million heading towards $300 million. I mean, AI has become a incredibly political game. uh in terms of regulations, in terms of data centers. Um have you been pitched to uh to donate to Super PAC yet, Dave? >> Uh I have indirectly, but I've made it really clear that Elon convinced me never ever ever ever ever get close to any of this. You will regret it the rest of your life. >> Yeah, agreed. Any points of view here, Alex or Sem?

45:27 Um my my initial comment is I think there's a sense in which it was inevitable that AI was going to be politicized like this. It touches so many aspects of society. It would be I think uh counterfactual nonsense to expect it not ever to be politicized. It maybe in some senses remarkable that it it took this long for quote unquote leftright access to emerge on the the subject of super intelligence. There are natural polls, pro-AI, anti-AII that have apparently emerged. I I do think it's for the record, I think it's sad that it it's being politicized. I I would hope that there would be broad recognition that super intelligence can be broadly beneficial but at the same time I think this has been true for

46:13 every transformative technology in human history that there's a a natural access that forms that's maybe on on one side leans more depending on your your political orientation either progrowth or uh pro capital and on the other side >> it's naive to think it would be politicized I mean of course it is this is the whole US versus China. This is about US dominance. This is about companies basically, you know, protecting their future, protecting their data centers. There are many forms of science and technology that aren't really politicized. I I I don't think it's >> this level of impact. >> I if if you look at the source of politicization at the municipal and state level, it seems to be people concerned about less maybe about their

46:58 jobs, more about say electricity prices. I I think there's maybe an alternative timeline where the politicization of AI could have been perhaps delayed by at least 2 years. Uh I I think it's frankly remarkable that it took this long for large super PACs to emerge around AI and probably could have been delayed even more. >> Well, all right. Well, let's move on beyond the politics. Um and uh let's talk about work. So, a lot of data coming out on the impact of work. First, software engineers uh jobs are rebounding. 67,000 roles have opened up um up 30% uh in 2026, the highest in 3 years. What does that mean? First question. Second, we've seen nearly

47:44 80,000 layoffs reported in Q1 of 2026. And this is targeting you know marketing and sales uh you know uh consumer uh relations and it's definitely due to AI automation. Um thoughts on on work and jobs? >> Yeah it was really hard to reconcile that bullet with the uh the educ you know the the new college graduate hire rate which is alltime low you know we had in a couple podcasts ago. So I don't know how to reconcile those two things. >> Okay. So I'm finding that AI is not eliminating work evenly. It's it's hollering at specific functions. It's increasing demand and others. Um I think I'm much more in the Andreent camp here. I think there's also a lot more going on in the economy. I think people are

48:29 attributing things to AI, but there's also there's the Iran war. There's the oil price explosion. There's a lot more complexity in it than we can then just allocate to one cause. I'm much more on the Andre side for a lot of this. >> That would be great. Um another another story here uh is the uh Meta's clawed economic leaderboard. So if you remember there was a conversation about you know how many AI tokens is every employee uh using uh and being able to measure that and uh Meta put up a leaderboard amongst its 85,000 employees to gamify AI adoption. I'm curious what other companies have done that. Maybe Salem, you know, of some uh it was taken down voluntarily by the

49:14 employees because they didn't want to be sharing their data publicly. Any thoughts on this, Dave? I mean, do you have a a a token leaderboard for your employees? >> Heck yes. And I love it. And and also, you know, the the gaming of it is a nice transition, but you can't game it for very long. So, I love it when companies do this and say, "Look, it's a badge of honor if you use a lot of AI. Please use as much as you possibly can. We'll come back in a month and start thinking about how to use it perfectly, but first just get familiar with it and use the heck out of it. And nobody ever goes back, right? I've never met a person who hammers Claude or hammers OpenAI for a month and then comes back and says, "I'm never going to do that again." It doesn't exist. It's a one-way path. So

49:59 getting your employees over the hump is going to save them. So I love this as a motivation. And I really don't like the part where people are afraid to share their prompts and their history because like okay, you know, maybe it's a little embarrassing that you're not using it well, but get used to it because it's going to get exposed anyway in the long run, but that's how you help other people improve. You know, if we all share it, we're all going to get good together. And so I I like uh you know, it's kind of disheartening that people will pull out of it uh because they don't want to expose their prompt history, but I it is the right thing to do and I love it. It's ironic that Meta is calling it uh you know is participating in Claude anomics >> versus llamics >> versus llama llamics. >> It's quite quite the indictment of llama

50:45 rest in peace that it wasn't llamics. >> Oh my god. For sure. >> I also think to to everyone who would say well you know this is just leading to gamesmanship and leading to optimization of the wrong items. All of these reasoning traces are fully available presumably to Meta for to to to do meta analysis and determine whether these are just employees who are token maxing which is the new term of art uh just maximizing their their token usage unproductively versus whether their reasoning traces indicate that their tokens are being productively spent. This is all transparently available to Meta. So I think token maxing and claudonomics or laminomics whatever we want to call it is probably directionally the trend of the future where for the first time senior company

51:32 management has visibility into effectively most of the cognitive power and how it's being spent on a per employee basis. >> What was Jensen's recommendation? Was it twice your salary in tokens per month or was it half your salary in tokens per month? you remember >> his recommendation is you spend the maximum amount possible on Nvidia GPUs. >> It's like >> it's like the the beers three months of salary. >> Well, I told I told all of our guys to target uh one to one match of payroll to to AI costs by the end of the year. >> Amazing. >> And and don't worry about it if it's if it's not perfect use. Don't worry. Just get to that target and then we'll optimize it next year. >> I think a target like that is a much more accurate way. I think these token leaderboards are very primitive dashboards. U we'll end up with something in a different model like a

52:19 machine leverage per employee or something like that that'll be a much better metric for where we're going. >> All right, let's get to the heart of employment. So Mark Andre rebukes AI job loss. He comes out with a very strong statement. AI job loss narratives are all fake. AI and massive productivity ramp equals massive demand and massive job jobs boom. So, uh Mark is truly a uh a maximalist uh an abundance-minded individual. Uh thoughts on this? You know, how does this square with the fact that we're seeing uh young college graduates not getting jobs, that we're seeing displacement? Is it all sectoral and we're just going to see a number of

53:04 sectors being demolished at the same time? Numerous new, you know, demands in different sectors. What's the advice to give everybody listening to us today? >> The advice is really simple. I mean, for God's sakes, don't go get a job. Go build a company. Yes. And we talked about this the in our last podcast where the the risks of taking on an entrepreneurship role are way way lower than it was before. You don't have to have all these incredible crazy skills that you needed to have before. You just need to have um a desire, a purpose, and just get the going with building company. Dave talks about this all the time. >> Well, you don't have to be a genius to come to your own conclusion. Like, forget asking people like Mark or us, our jobs going away or jobs coming. We told you already that AI will be able to

53:50 do everything that a white collar worker does imminently. That's a fact. You decide what that means because like Sem said earlier it affects very different areas very differently. You know some people retool themselves for AI very quickly. Software developers for example. Other people like accountants and lawyers don't like it's going to be exactly what you would expect given that scenario. It's it's not hard to predict at all. And and I think there's also timelines you know when Mark says this is crazy. Jobs are going up not down like yeah by 2030 that's absolutely true. Just like the industrial revolution, jobs went up, not down after it was all the dust adjustment. >> Yes, this is just industrial revolution which took you know decades is going to happen in two years.

54:35 >> Also, I think also you have to remember sorry Alex just a quick point also you have to remember that the adoption of AI inside companies is going to be very slow. there. There's a huge transition to go from human to human workflows to AI workflows and that transition was going to take years. We'll have lots of time to kind of smooth this out. Sorry, Alex. Back to you. >> Yeah, I I think both narratives can be true at the same time. I think if you add in the word net, massive net jobs boom, then both of the narratives immediately become compatible. There is going to be a lot of dynamism with some job categories going away, others new ones coming into existence and net job loss probably not. I I would guess and I'm I'm betting that there's going to be net job creation just exotic new jobs

55:21 like one person AI conglomerates will be created if you want to call that a job. Uh but on balance many jobs will also disappear. But this is, you know, this is how we get massive economic growth and the singularity in the macroeconomic statistics. We're not going to get it through business as normal. >> We've talked about this and I think it's basically companies are going to get much smaller, much more nimble or they're going to die and they're going to spawn a whole set of baby companies alongside. There'll be an ecosystem of companies coming up. So, it'll be a much larger number of smaller companies. um in the future. >> I mean, I'll I'll go with the prediction I've made before, which is we'll run a company with between 20 and 25% of the

56:07 the members you needed than compared to before, but we're going to create four or five times more companies. And that net balances out. So, I'm much more on the Andre side. And also, his hair shape of his head aligns very well to my thinking. >> Mark is brilliant. If you ever if you've ever heard him, you know, in in podcast, he's he actually speaks at 1.5x speed. >> Y >> extraordinary. We talked a little bit about this in the last pod. Alman believes America needs a new social contract with AI coming. This is his quote. The emergence of super intelligence will necessitate a new social agreement akin to the New Deal during the Great Depression and the progressive era of the early 20th century. Uh, and yes, but what is it going to

56:53 look like? Is it going to be UBI to UHI? Is it going to be 4 day work weeks? Uh, I still believe that we're going to see turbulence in the next two to two to five years and it's going to be the government printing checks uh to give people sort of a UBI. >> I have a bunch of thoughts here. You know, this new social contract kind of framing is correct, but is very vague. We have to have more specific things like portable benefits, new taxation logic, uh lifelong reskilling, and governments have been built around taxing human labor. They're not ready for AI software agents, and they need to get a thing. When you have AI abundance without institutional redesign, you're going to get a backlash, not progress, and we're going to see huge backlash against this just because governments

57:39 are so slow. I I should note though, OpenAI did also put out an industrial policy prescription for what this new social contract could look like. It's it's not just this single sentence. They put out an elaborate white paper and circulated it in the Congress. I I do think something like this, a new deal, probably is going to happen anyway. It may or may not happen as one lump sum. It may happen peace meal and it may not happen in the US first. I think there are contingencies where other countries experiment a little bit more aggressively with it uh than the US and then eventually perhaps among a certain set of countries new best practices emerge. But I I I do think some form of call it abundant capitalism or

58:24 capitalism 2.0 or post scarce capitalism something like that probably emerges. It may not happen immediately. It may not happen as quickly in this country but it will get there eventually. You know, I had lunch with Michael Katzios, uh, who we're going to have on the pod sometime very soon, science adviser to the president, and we're talking about one, you know, idea I pitched him was a new social contract will be before any employee gets terminated by a medium or largesiz company, that company has to give them reskilling. In other words, instead of a golden parachute, it's a it's a golden education package so that they can go and sort of transition. sort of a a safety net um or a sort of an ethical mechanism for you to let off you

59:09 know half employee base >> based on public reporting China already has that policy so it would be a weird future if the US is adopting policy prescriptions from the Chinese Communist Party for AI >> resaling but maybe that's the near future we find ourselves in >> well something to think about you know the uh the way this is rolling out is really unusual in history you know when the industrial revolution happened it took away blue collar jobs and worked bottom up. But AI is coming kind of like accountants, lawyers, professionals top down. You know, only a little over half of voters have a job at all. So, they're going to be like, "Oh, you know, it doesn't affect me." Uh, but then, you know, all of bluecollar isn't going to be touched. All of manu all of, you know, physical labor isn't going to be touched for quite a while. So, they very well might say, "Yeah, tough luck. You

59:56 know, lawyer, accountant that was making a million dollars a year. This is poetic justice. We're not voting for anything that helps you. >> You know, >> that wouldn't surprise me at all. >> When I was in Morocco, I was I was interviewing people that I met along the way about whether they're using AI or not. And the realization is countries, you know, uh African nations are going to be impacted the least uh as this transaction occurs because they're so, you know, insulated from this. But one of my tour guides, I loved his story. He said, 'Y yeah, you know, I I chatted with Chad GPT and I said, 'Th these are all my skills. Uh what could I do to earn money? And uh he came up with a business that you know we purchased. It was basically a bicycle tour guide uh in

1:00:43 forget which it was not Marrakesh, it was probably in uh near I forget the city exactly we're in. were transitioning through and uh you know that was his business. He did a great job. Uh and I love the fact that this individual was basically trying to figure out how he earned income and using chat GPT to do that. So any thoughts on the AI economics uh that we've just gone through? Uh what do you think the social contract is going to be like Dave? You know, what do you imagine is going to replace what we currently have? >> Well, I you know, I had very ornate thoughts about this and then we met with Andrew Yang, remember at 360 and he

1:01:29 said, I can guarantee you >> that the way politics works, all we can do is write checks and it can't be in any way thoughtful. It's just money here. Oh, wow. You're you're hurting. Here's money just like co. Uh and so that's all we can do. So that's all we will do. and then you know maybe after AI enters government in two three four years a much more thoughtful program will happen later that was disheartening but I I think hard to refute though the first version of the social contract is just going to be the next election 3 years from now politicians saying well I'll give everyone $10,000 each well I'll give everyone 12,000 okay well if you're giving them 12 I'll give them 15,000 and then we'll be right back to well how much can the country afford that's what we're going to give because that's how

1:02:16 you're going to win elections you know, exactly what you would predict, actually. So, uh, that'll be version one. Anyway, >> I I for one think a redistributive model of a quote unquote social contract shows an extreme lack of imagination. I I would like to think that super intelligence should also supermpower individuals to generate super income. is one of the reasons why I I I for one am betting on more of a model where there may be even no strong need for a social contract if we can empower the longtale of individuals who have idiosyncratic skills or experiences or socioeconomic niches to operate their own large companies sitting on top of fleets of AI agents. I I would love to see in in short no need for a new social contract

1:03:03 and instead have the private sector rescue people who would otherwise be technologically unemployed or or disemployed with by empowering them to become basically micro uh entrepreneurs or or even macro entrepreneurs to turn them all into Warren Buffets. >> But that's in the long longer run. I

…