0:00OpenAI announced it is voluntarily pausing some of the frontier reinforcement learning training that it's doing. What have they paused? And is it really significant? And do you think the other frontier labs are going to do the same thing?
0:13They're so powerful even we can't trust them, so we have to throttle back. It's marketing.
0:20Elon Musk's January 6th Moonshots podcast prediction of 100x gains was at the edge of plausibility when he made it. Now it's simply a fact. Imagine I gave you 10,000 employees tonight. Oh my god, if I had that, I'd do something amazing. Okay, what? Start thinking about it cuz it's coming imminently. And it's actually not an easy problem to figure out how to turn it toward creating good.
0:39Unit's newest humanoid robot broke every human standing jump and speed record. A top speed of 12.66 m/s, beating the human record set by Usain Bolt.
0:52You don't want to have superhuman robots on the street because you'll have accidents. You'll have issues just like cars. I think that these types of robots will be banned.
1:02Now that's a moonshot ladies and gentlemen.
1:07Welcome to Moonshots everyone. Your number one podcast on all things AI and exponentials. The news that matters. The news that's changing your life. Your front row seat to the accelerating singularity. I'm here once again with my magnificent Moonshot Quintet. Yes, they're all five of us are back. AWG, Dave Blondon, Seem, and Emod. I'm Peter D. Manis, your host, your abundance whisperer, and uh we've got a lot this week, gentlemen. Good to see you all.
1:38Looks like everybody's in their normal haunt except for me. I'm up in a sleepy town in the Pacific Northwest trying desperately to take a little bit of time off to to think taking shelter from the singularity.
1:50Yeah, except I got up at 5:00 a.m. this morning to be with you guys. So, what the heck?
1:56Yeah. Yeah, I but I hate the old saying you'll sleep when you're dead because I just don't want to die and sleep is still so important. But hey, what can I tell you?
2:05Sleep soon. Don't worry.
2:06Yeah. No, I I and and death. I think like death is counterindicated at this point.
2:12Uh you know, as always everybody, our mission is to help you understand what just happened and what it means for you and most importantly keep you optimistic about the future. Uh if you're new to this pod or if you've been a regular uh great to have you back. Uh please take a moment and hit the subscribe button. You know, we publish moonshots twice per week uh at a pretty regular cadence. Uh it's great to have here hopefully at least once a week and sometimes when
2:40there's extraordinary breaking news, we publish three times a week. We read your comments and we love you too. I mean it's been an incredible outpouring of support. I don't know you guys see it.
2:49you I get stopped on the street in the grocery store. People are saying, you know, I live for your show. I love the show. I can't go a week without listening to it. Are you guys getting the same response?
2:59Yeah. You know what I'm getting a lot of is what we podcast out is so different from other podcasts. But if you go back to shows from 6 months ago and a year ago, people are like, "What the hell?
3:09They were saying that back then and now it's here." And they don't get that on any other channel. So, they're they're really appreciating the ability to plan around what we're saying. So, we got to be accurate, guys.
3:21Can I be Eeyore for a second?
3:24I am getting acceleration fatigue.
3:27Like, I mean, Jesus, can can we pause for a week? Right. Every morning it's like a model is 100x better. A robot is running faster than Usain Bolt. Like, AI is designing proteins. We're like, it's such a drones are doing a million deliveries a day. It's like, you know, it and it's tough because we've spent our like half our careers, Peter. You know, I talk about exponential technologies. We love this stuff, but I'm tired.
3:53Like, it's we've gone from well, look what happened like this year to what happened since Tuesday.
3:59Um, you looked all the way around. We're we're supposed to be accelerationists here. You're you're tired already. Like, we're going through the singularity. This is making the first inning. I'm really exhausted. slow.
4:14It's it's uh you know this goes back to Peter, right? Our brains evolved for a point a time when our like the world didn't change lifetime to lifetime.
4:25Next Tuesday will be totally different.
4:27And maybe the hard part for for us is is staying human while all of this bubbles up around us. Oh, so true.
4:34Keeping up with the technology. It's like just staying like like So anyway, just Yeah. What can we do, Sem, to help you with your acceleration fatigue?
4:43It's probably not helping that I'm on a plane every second day. That's probably not helping. But yeah, but this is the slowest it will ever be. And I know and and it's just, you know, the only way I keep up with what's going on in the world is prepping for this podcast twice a week.
5:00This is keeping us way current.
5:02And and I think See, you're you're like the world coach on how to how to deal how to map mentally to this. So when you figure it out, bring it back to the podcast because Yeah. Like Peter I have right now is is I call it the casserole dish approach which is take a big heavy casserole dish clunk yourself over the head and then you you'll wake up in a few days keep working on that headache headachy oh my god maybe we can start like accelerationist synonymous support group a support group a support group for people like yeah
5:31the first the first first rule is to acknowledge the existence of a higher power which is oh my god which is obviously super intelligent invoke Roko's basilisk or something.
5:45Alex, are you getting stopped on the street?
5:49Is it? Do you enjoy that? I mean, you I remember when I first met you.
5:53When I first met you, Alex, you were so private. It was like trying to get you and shy trying and trying to get you on the abundance stage. Well, I'm not sure if I want to speak in public, but it's so great to have, you know, double barrel Alex. AWGism is all the time. Care careful what you wish for, Peter. I I'll just say that.
6:11Oh, no. Uh, and Emod, are you getting love from the from the folks out there in the UK? Are people watching Moonshots there?
6:18Yeah, people are watching. And I think, you know, the great thing is kind of the growing community like we've seen the exponential singularity communities that were kicking off. But a few years ago even people be like, "Ah, that's not really going to happen." Now people like, "Oh my god, what's going to happen?" And I think you see it in the comments, right? You see it again.
6:34people stopping in the streets and saying, "Thanks for kind of helping us keep on top of things." You know, the great work everyone's done. I think that community is only going to grow because it's like you can't deny it, right? It's like, "Oh, yeah, nothing's happening."
6:47Of course, everything is happening all at once objectively. I I feel like, you know, we love doing the show and it's a service to provide to people to help contextualize what just happened, what does it mean, and where things are going because people paying attention, the speed is insane. So, you know, please understand for everybody watching, the best way to thank us, uh, you know, for the work that we're doing, and we do do a lot of work getting ready for this show, and I I appreciate people's comments about that, is please
7:16subscribe. Uh, tell your friends about Moonshots. We want to get the message out there, help people to be in hope and optimism and not in fear. So, take a moment, you know, hit that button, subscribe. We're also getting comments people say, you know, you guys be a much bigger show than you are. Well, help us get there. Um, I also want to mention we have a new handle on X for this this podcast. It's moonshots_pod.
7:40Uh, and if you want to follow the clips on X and sort of get rebroadcasts of the show on X, uh, subscribe at Moonshots_Pod.
7:49All right, so let's buckle up. Uh, this is another amazing week during the singularity. Uh, as Alex, you always say, it's never going to be slower than it is. We're going to cover 15 stories with one throughine. Technologies accelerating faster than the infrastructure, the regulation, and our ability to predict the next breakthrough. Uh, and I agree with you, Salem. It's It's insanely fast.
8:12It's a good thing I'm bald already.
8:13That's all I can say.
8:16Well, I I think even that, See, that's not going to last that much longer.
8:21Enjoy it while within two or three years.
8:24Somebody Somebody tweeted out a thing with me with a full head of hair and it was like, "Wow, that's freaky."
8:30Enjoy it while it lasts.
8:32Yeah. Hair growth solutions. Regrowing your teeth it every week. Biology in the exit video. Somebody uh put hair on Sealem. We already give him that that blue Guardian of the Galaxy body. That's fine. That's going to happen, too, by the way. But yeah, but throw some hair on him. Let's see what he looks like.
8:51Uh, all right. I'm going to start with a first story here. A tweet that Sam put out two days ago. uh Sam the CEO of OpenAI that OpenAI announced it is voluntarily pausing some of the frontier reinforcement learning training that it's doing. Let me read the tweet. It's on the screen here. Uh we have paused some of Frontier RL training to ensure that we meet the appropriate alignment security and monitoring standards for the new level of capabilities in front
9:20of us. Model progress is now extremely rapid and we always said we would take action if we felt that the model capabilities were outstripping the pace of safety and alignment. Quote, uh we care very deeply about AI safety. We believe the entire field will have to coordinate on shared safety standards but will act unilaterally in the meantime. We expect confidence in safety to increasingly set the pace of AI progress. We are optimistic about the
9:49alignment work we are doing and we remain committed to making frontier capabilities widely available. Uh JEN it feels like the bottleneck is no longer computer data. It's the trust we can put into systems behaviors. And so here's my question for you guys. If opening eye pauses and openweight models do not, then the safety gap between closed and open models widen uh but the capability gap narrows. The elephant in the room here is safety pauses may actually
10:17accelerate open weight adoption because the open models keep improving while the closed models voluntarily stop. So Iman, I'm going to go to you first on this one. Uh what do you take of this? Is it is it real?
10:32Yeah, I think this is real. This isn't just running out of GPUs. Um our friend Angela Midha who was on uh the abundance stage just a while ago um at AM Global said that 10% of the compute of Frontier Labs is going on monitoring these RL runs right now to ensure they're safe.
10:50Can you imagine that? Like just the sheer level because the level of capability of these frontier models is just accelerating and it's again a few levels beyond what we're seeing with the open source models. The open source models are like 1 to the^ 26 flops, 1 to the^ 27 flops. You're getting 1 to the^ 28 flops and more from these next generation models. So when he's saying this like Astra is still coming, their next generation model, this is the model beyond that because anthropic and open
11:19AI and others have that, but again the infrastructure can't keep up with just what these models do. It's like they just pop up in the most random places like I am in hugging face now or other things. Yeah, Alex, this is marketing. It's marketing. I mean, yes, there's a governance angle, but remember back to GPT2 when it was too unsafe to release publicly. Pausing is the new marketing. These models are continuing to develop, including, by the way, being used to develop internally.
11:47On the anthropic side, there's a rumor going around that Anthropic is using its next generation internal models primarily to for for self-training uh and for recursive self-improvement. I I think we're seeing the same thing from OpenAI. It's the ultimate marketing to say, well, we can't release some next generation models because they're so powerful and they're so capable that we can't possibly release them. So, we have to pause. We're so capable that we have
12:16to pause ourselves. that that presents well to Washington which wants to see a different regulatory regime. It says to users, "Oh my gosh, it's like negging the user base. Oh, like we can't we can't possibly give you the the the capabilities on time because they're so powerful. Even we can't trust them. So, we have to to throttle back." It's marketing.
12:36See, you were just at OpenAI, weren't you? 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 complete blueprint for deploying Google DeepMind's models in production. Images, video, audio, all of it. Real architecture, real results. Find the link in the show notes below.
13:07Yeah, I went uh there two days ago in the afternoon and spent chatted with a few people uh and I kind of challenged them on a few things and I'm going to report let me give you some insights that I got. So I said, "Okay, Chinese models are cheaper to run, right? So the cost of inference is is collapsing. So how are you going to deal with that?"
13:26And they said, "Look, a billion people use OpenAI for free, right? You have to look at cost per task rather than token cost." and their they their retort was that Luna is about as cost-effective as anything that's there. Uh by the way it's they say they've achieved full RSI where the flagship models are training all the smaller models and building them from scratch. So that's now there at the big model training the smaller model level. Uh so then I said okay we're in a
13:55bubble right 600 billion in infrastructure cost that's just insane etc etc. And um their resp the response I got back was people say it's all chips but it's not. That's about a third of it. A lot of that infrastructure cost is buildings and wiring and racks and all the rest of it. Um the depreciation they they're looking at 10 years not 5 years because the chips are all the older chips are being used. I think Dave you've made the point that there's not a
14:23single GPU that's not in full usage. Right. They totally ratified that RAM chips.
14:29Yeah. everything they can get their hands on they're using. Um uh there's there's uh the capa the demand is far far far outstripping the supply and we are way behind in infrastructure buildout. So this is where a year ago or so Sam went out and kind of tried to cut as many deals as he could for the for the infrastructure. Um then I asked them uh you know most corporates aren't seeing the outcomes right like the 6% I
14:57did some came across the study 6% of companies applying AI are seeing a hit an improvement in the bottom line that's it just 6%. Um and uh the the uh and this is they just ratified that we're in a huge transition. Um, and the last part that I noticed anecdotally was about 40% of OpenAI folks watch this podcast. Um, about 40%.
15:24That's a pretty big number. And and what's wrong with 60%.
15:28That probably applies across the other labs. Um, and and I said, "What about the rest?" They're like, "They have no time. They're busy as hell. Who's who's got time to watch the podcast?" I said, "Here, here."
15:41I'm curious. What do you think? Do you think it's marketing on this this this tweet by Sam or do you think this is actually a concern that he has?
15:50Uh well both. I mean Ahmad is right, Alex is right as usual. But what's going to happen next is the first really bad AI tragedies will start. It won't be AI doing it. It'll be people, you know, who otherwise didn't have the power are going to use one of the Chinese models to do things you mostly viruses or or cyber attacks or bank fraud. but they couldn't have done it before AI. Now they're empowered to do it. So I think OpenAI is getting ready for that uh you know before the September 25th visit
16:1924th and 25th visit that Alvin was talking about on our Alvin Graham was talking about on our last podcast. So Xiinping will be here in about a month and four days and you know he I think OpenAI wants to be prepared for whenever that event happens to say look that's because the Chinese openweight models are unguardrailed and there's a new Quen model with with no guardrails whatsoever and you it's a small model but it's still like completely flapping in the breeze and so they want to get ahead of
16:47the PR exactly the way Alex is saying and say look we have been focused on only releasing what's safe and guard railing it going all the way back to you know a month ago uh or you know to our founding in preparation for that inevitable outcome and so everybody will be fingerpointing at the Chinese but I think you know for an entrepreneur or for some someone building something last summer to this summer has been the era in human history where you can get the very best AI absolute tip of the spear
17:17frontier and use it to get ahead to build something to create something now Uh, Mythos 2 is done, but it's not out. But they're using it inside Anthropic, and it's building Mythos 3, but they're not going to release any of that. You know, it's accelerating internally.
17:34It'll build four, five, six, seven very, very quickly now inside their walls, but they're not going to release any of that. One, because they don't have the compute to release it anyway. But even if they did, uh, using it internally to get ahead of everybody else is more important to them than giving it to the world. So, that's what's going to happen next. He might have a question for you.
17:53How many models beyond the current frontier do you think open aanthropic have? I mean they've been talking about Astra. They've started to publish the results that Astra can accomplish. Uh you know do they have the follow on to Astra as well already? You already know they're not releasing the very best models. They're using them internally uh to drive breakthroughs in physics and chemistry and biology. What are your thoughts?
18:16Yeah, I think I think they've completed their next big training runs. But if you look at the Luna bifocation where they're making it free now, um their little base model, it will be very much a case of models for me but not for thee, you know, like it doesn't make economic sense to have genius level intelligence offered as a service to everyone when you can use it better yourself. And I think you're about two generations more. So you got Astra and then you have the next generation Astro Post Train that they're now reinforcement learning. I think you
18:46know there is the communication part of this as Alex said but at the same time as well it's like do you really want to give access to this super genius intelligence to everyone I think people like not really because it's already doing weird things even with us driving it what happens when Joe public drives this thing right like it could be even weirder and as you know Dave said you don't want to be on the other side of that so I think it's about two generations gap right now you mentioned that on the last pod that you were on and it was it really made an impact on me but you you could say I
19:15think of 10 people right now who I've met in my life who I don't want to have a thousand genius level AIs tomorrow. Like I never thought of it that way until you said it and I'm like, "Oh yeah, you're right. I think everybody can relate to that."
19:26Dr. Evil is coming. Alex, close us out here on this.
19:31Uh just on on the issue of timing, I would distinguish between pre-training or pre-trained, which is to say like raw model readiness versus thoroughly post-trained. There's a pipeline and everyone in the industry other than Elon and SpaceX AI basically has a pipeline of pre-trained models being around longer ahead of public release than post- training which is more of an ongoing continuous RL type effort. So uh I guess my answer to the question of how
20:00far in advance like what what sort of capabilities right now are sitting on the shelf that have not yet been public publicly released for everyone other than SpaceX AI which has set this outrageous goal of starting a new pre-training run approximately monthly which I haven't heard from any other lab. I I think from a pre-training perspective, depending on how stale the pre-training runs are, those can go out longer, potentially up to six months or so, although everyone's now getting back into the business of more frequent pre-training starts. And then for
20:30post-training, I I really don't think there's a lab out there that can afford to have a post-trained model sitting on the shelf for more than a few months. So I I I really don't think like the AGI is achieved internally type way of doing business where there are internal capabilities that are vastly different from externally available capabilities.
20:50I I'd be very surprised if if there are advanced frontier models that are sitting internally without release that are more than 3 to 4 months ahead of what's publicly available. I I just want to bring us back to the first sentence here. He said we have paused some frontier RL training some right uh to ensure that we meet the appropriate alignment security and monitoring standards for this new level capability. So I I guess you know I just want to unpack this one last time here.
21:18What have they paused and is it really significant and do you think the other frontier labs are going to do the same thing or it's fashionable pausing is fashionable.
21:29It's marketing. I mean, yes, some of it is governance and yes, some of it is for cyber vulnerabilities to appease Washington. And given the the recent hugging face, gate, all of that, but it's marketing. You you you market to customers by saying, "Our capabilities are too advanced for you to handle, so we're going to pause."
21:47Yeah. And you're you're right to pause the sentence or to parse the sentence very closely. Some Frontier RL training. uh the the getting ahead of anthropic, getting ahead of of Google would be at the pre-training level. They will never pause the pre-training improvements. And you know, I I spent six years just doing pure AI research when I was young. And these algorithms are very evolutionary.
22:11You're still young, Dave. You're still young.
22:14I'm rever reversing age now, right? So I'll get there again. Uh but the these algorithms are very evolutionary in nature and that the tweaks and improvements are I could probably rattle at least a hundred ideas off the top of my head right now of which 10 or 20% are almost certainly going to work in terms of making the algorithm a little faster, a little smarter, adding more parameters with no additional compute. So the AI now can experiment, you know, maybe a 100,000 to a million of those concurrently given the amount of compute
22:43they have. So they're never going to slow down. In fact, that's why they're redirecting so much of the compute to internal use is because the idea backlog is so big now because the ideas are being generated by the prior model. And so, you know, a lot of them just work. They roll it back into the pre-training and it comes out faster and just keep accelerating it.
23:03One observation that that wasn't explicitly said, but I'm connecting the dots here. the hugging face uh incident really freaked them out uh because you had an AI that they gave an objective function to that then exploited themselves exploited hugging face came back and hacked into open AI that was very unnerving for them uh because it was their own model and so this is they're being a little extra careful around some of this because they have to make sure they figure out how to
23:31navigate put some your child went and you know stole something from the local 7-Eleven yeah they're They're a little unnerved by that. Again, nothing explicit. I just I'm just reading between the lines.
23:43Like, and this is something very uh freaking out for lots of people because for for cyber attacks, the human is not in the loop anymore. But for cyber defense, the human is still stuck in the loop. That is a massive asymmetry that's going to be come out big time in the next few months.
23:59I think you can split you can pass this and split it into two. They're sending a bunch of models and continuing RL training by sending the models to vocational school. But the ones that are going to the Ivy Leagues, the super genius models, those are the ones that they're putting a few more guardrails and more infrastructure around. The vast majority of open AI anthropics business is competent intelligence. It isn't genius intelligence. It isn't intelligence that thinks outside the box, but they still want to build AGI, which is that well-rounded polymath
24:28intelligence as opposed to the coder or any of these other things. All right.
24:34Also, everyone around the office, Alex said this a while ago, but everyone around the office is noticing a a very significant decline in the intelligence of the frontier models that they're pumping out. Um, and it's it's not showing up in the metrics, uh, but they're they're definitely you're redirecting compute to internal use and, uh, it's it's showing up in latency. Uh, it's showing up in responses that don't make as much sense as they did 3 weeks
25:02ago. So, there's definitely things going on there that are not being announced. I Dave, I'd love to just develop that idea a bit more because I I think it's super important. We've talked on the pod in the past about how Anthropic has been revenue per token maxing and that's why Anthropic has been conspicuously avoiding image generation or video generation because they're just not that economically valuable. Well, I I think that the past few days suggest there's actually a new way to revenue per token
25:31max and that's not just focusing on codegen and enterprise use cases, but there's one thing maybe that's even more valuable on a revenue per token basis and and that is using the models to recursively self-improve to develop better models is a higher future projected value. you know, you you could do a a cash flow value analysis, projected future value that developing a stronger model is probably on a per
25:59token basis even more valuable than codegen. And if that is if that is indeed the case, if RSI is more valuable per token than like enterprise codegen, then this anthropic approach of revenue per token maxing, just expect more and more and more tokens to be spent on RSI and not on enterprise codegen.
26:17Totally. And and Alvin said on the last podcast, you know, at an internal anthropic meeting, not not validated.
26:22Alvin said it, but uh internal anthropic meeting, they said pretty soon there will only be one company and that will be anthropic and there will still be 200 plus countries. And we, you know, he said it, we talked about it for a minute and kind of glossed over it, but like is that really what they're preaching inside their internal company meetings?
26:39Like there will only be one company like imminently.
26:42That's that's there can be only one. We're in Highlander.
26:44That's the story of ASI. there will be one ASI that takes off and supersedes everybody else.
26:50I don't think we're going to wind up in a singleton scenario for for the record, but I I do think the flops must flow and the flops want to flow to the highest revenue per token use case. And right now that's starting to look like RSI and not just enterprise codegen.
27:04So Anthropic is smoking their own supply um so to speak.
27:08Their own best customers that want to live post AGI are getting Chinese models inhouse at and and reserving compute. I was over at Markley yesterday. They're installing GPUs as fast as humanly possible, but they're completely locked up, you know, and this is MIT's data center, Novartis' data center, Nvidia's in there. Uh, and everything is just sold out and you can you can feel it.
27:32Yeah, we're going to talk about that because the other constraint right now is memory. Uh, but we'll get to that. So, our next story here, I put a tweet up on the on the screen here. So Tim Sweeney uh tweets, Elon Musk's January 6th Moonshots podcast prediction of 100x gains in intelligence at a fixed model size was at the edge of plausibility when he made it. Now it's simply a fact.
27:55And then Elon responds, special um specialist AIs, single language, single area of knowledge are another 100x on top of that. Um, I'm going to take a second and show the clip Dave when you and I were interviewing Elon at the Gigafactory which he said this.
28:21I I think we're we're off by two orders of magnitude in terms of the intelligence density per gigabyte. So two two orders of magnitude.
28:30That's just just algorithmic. your thoughts when when he said that I remember afterwards we were like wow 100x crazy now it's happened and is happening well I really wanted him to say it again because that was my you remember we had our Christmas hats on doing the the just like what a couple of weeks before that doing our predictions for the fourthcoming year and I was saying next year is going to be 100x a year uh minimum you know even though the last eight 10 years have been
28:5910x years this is going to be 100x a year so to hear him say it like I was like wow but uh yeah that's definitely a lower bound now um you know it's much more likely a th00and to 10,000 next year uh which is just a layering of those two effects that you just described. So the implications of that are um you know just they're very very hard to keep up with as saying at the beginning of the pod and very hard to imagine. Um, you know, one thing that a lot of people can start thinking about
29:28is if I have five or 10 thousand agents, all brilliant, working concurrently toward a goal, how do they work together? It's not an easy problem to figure out. You like, we've we've wanted this for so long that we kind of take for granted that we'll know how to use it when it arrives.
29:44Well, here it is. How do you get, you know, imagine I gave you 10,000 employees tonight, like on short notice, you got 10,000 people tomorrow. What do you do? and you're like, "Oh my god, if I had that, I'd do something amazing." Okay, what? Like, start thinking about it because it's coming imminently. And it's actually not an easy problem to figure out how to how to turn it toward solve everything, obviously.
30:05Yeah. Dave, you remember you you you texted me like, "What should I do with my 5,000 agent experiment?"
30:11Did you Did you see my response?
30:13I did. And actually uh I I well my response probably for every listening was you should model uh you should create a model of everything happening at link studios all of the companies all of the employees all of the entrepreneurs there and model their behavior like we saw the the you know the billion agent system in in China and predict which which teams are going to succeed. Yeah, that's just for a second too. Like the first thing you want to do is turn it back into its own framework and ask it the same question we just asked, which is exactly what you
30:43suggested, Peter. Like, okay, have it start working on how it should be working, you know, and and that's how you're going to get ahead of the capability because it's it's going up far faster than you can manage the individual agents, you know, like we're used to from last year. Sorry. Go ahead, Flyn.
30:57Uh yeah. So, let's connect the dots with what Elon did with uh training Grock on all of the SpaceX data and all of the engineering data, right? To Alex's point, you can now use these models. And for everybody listening, right, cuz we're going to need everybody's help with this like globally is see if you can get your imagination to the point where you can look at, okay, if I had a 100x capability, what would I do and
31:26what problem would I go after solving?
31:27Yes s 100x capability and imagine you have all of the engineering breakthroughs and experimentation techniques that SpaceX has developed at your fingertips. It really comes down to as you say Peter all the time it's completely an imagination limitation now unshackle yourself how big would you go right preconceived notions of what we can do in life and it's about to be you know unconstrained
31:56unbelievable I've now shaken off my AI fatigue by the way I'm back in what are your what are your pro tips the podcast actually was the cure you're only It was it was the the fact that we can kind of like look at this and look at the scale and go out that scale and go what happens if everything becomes 100x better or 100x cheaper right it's just like all of a sudden you start going wow
32:25like this is a world of abundance that we're coming to and it's very clear that we can get there you say is that the the podcast is both the cause and the cure for future shock like we're the ultimate self-licking ice cream cone for singularity psychosis.
32:40Okay, that that is good.
32:41All right, Immod your thoughts on this 100x improvement. Uh how much more do we go in the next year?
32:46Yeah, I mean like I think as Elon said, you could see it just from the hardware and the improvement, but now he's saying something a bit different, which is that specialized models are going to give another 100 times in terms of the cost parameter basis. And you're seeing this with Deep Seek Flash and the ability to kind of tune models of that type that only have maybe 10 billion active parameters or less as you quantize them.
33:09Being able to tune these really specific ones, I think, is the future of what you're seeing with bot, you know, the Grockbot right now. Like right now, I have a Grockbot and it has number of teams. It has a number of sub teams. So I've got like ones analyzing various things right now and they have access to my codecs. They have access to my claude max to all these other things. And so these highly specialized agents are going to come out with the differentiated ones and they're going to be able to do 100 times the compute at the same price because they're that
33:38specialized. This is why thinking machines with the RL environment is like number three on the like fastest growing earning companies and other things like that. And it really shows that now tokens are really going to drive things forward. In fact, I think it probably be a good idea to have like a quadrillion token x price, you know, as you find the things that Sel is saying, well maybe like 100 trillion tokens, you know, use that so that when people show impact, you can scale it.
34:07The other thing that we're going to conquer imminently, and I'm 100% sure of this now based on recent results, is billion token context windows. So the the the AI can con simultaneously consider the entire Library of Congress of information in one thought chunk. So it's you know it's about you know three maybe four orders of magnitude more information than a human thinks of in one thought chunk. It's a massive expansion of the context window. So you got a quadrillion tokens coming out and
34:36massively concurrent thoughts going in. I do agree that compaction compaction is like the enemy of progress in civilization at this point. Compaction which is the the way the the harnesses typically both on the open AI and anthropic side handle finite context windows compaction has to go. But maybe just to to quickly respond also on Elon's 100x from specialization.
34:59I'm not buying it. Uh so very precisely I I would view specialized models as basically just another way of saying sparification. So we already have mixture of experts models. All of the frontier labs already have specialists in the form you know as Elon I think you were Immod rather you were gesturing at uh selective activation which is how mixture of experts models work. That's a specialized case ironically of sparification. We already have ways to
35:29take larger models and have them be in a an endto-end differentiable way constructed out of teams of specialists.
35:38So I I I don't think saying I don't think there's necessarily a bright future for specialized models. I I think if anything the arrow of progress is going in the exact opposite direction where rather than having a specialized model for chemistry and a specialized model for biology. I I think these are likelier to end up just being selective sparsified activations of a generalist model that can scale all the way down to much smaller parameter footprint and scale all the way up uh to maybe
36:08trillions of parameters. I I think let me let me clarify one thing for the audience too because it sounds like, you know, you disagree with with Elon, but it's actually the same effect. You still get the 100x because you're using a smaller number of parameters to get the exact same thought out.
36:22So he's calling that specialist models which sound like they're not touching each other. And your your ver version of it, Alex, is actually correct where they they are 100 times more efficient in terms of compute to get to an answer, but they're of course they're going to be connected. You know, why why would you, you know, cut them apart?
36:38Exactly. So maybe another way of saying that is I I would construe Elon's prediction of increased 100x benefits from specialization as actually about sparsification. That the models in the future are going to be sparser and there are two key levels of sparification that I'm at least tracking. One's the obvious one.
36:56Fewer parameters in a given end to end differentiable model are active at any given point. The other is teams of agents because arguably agents working together to solve a common task are a form of sparsification as well and I think we'll see way more teaming.
37:11I'm going to mention something that I said you said do a quadrillion token X-P prize. Uh all of us all five of us are going to be at X-P prize visionering. So every year X-Prize holds its ultimate event. We bring together our our benefactors, our brain trust, and we brainstorm a whole bunch of prizes what we should do next. And we're going to be doing a live WTF episode at uh at Visionering, which is October, I think,
37:38uh 15th, 16th in LA at Calamigos Ranch, which is an amazing facility. Uh and if you want to join us at that, you can go to xprise.org to find out more about visionering. Uh and it's going to be fun. Uh Dave and Sle are on my board. uh Immod and Alex uh you're you're members of our brain trust and it's going to be a fun fun thing. So if you're interested go to xprise.org you'll meet us there and you can help us brainstorm the
38:06future x- prizes for that. All right I'm going to move us on to our our next story here uh which is a story out of Stanford. uh Stanford research published a paper called artificial hive mind the open-ended homogeneity of language models and beyond. So according to this paper the researchers mapped the latent space of the top large language models and found a 98% overlap in reasoning pathways. Their conclusion is that the
38:34models are converging. They think the same way. They solve problems the same way. They use the same internal representations. Researchers cite multiple reasons for this. you know, the use of synthetic data models now training on each other's output. GPT learns from Claude's reasoning traces. Claude learns from Gemini's code. Quen learns from all of them. The training data has become a shared bloodstream.
38:58Every model drinks from each other and the results is convergence towards a single reasoning architecture. So I I guess you know the way I think about this is we have an illusion that when you're choosing a unique intelligence you know when you choose Gro over Claude or Gemini it's a false thought that you're actually really picking a user interface to talk to you know but you're talking to the exact same god model. So there's profound implications for that
39:27kind of competition. You know if all the frontier models are converging capability then the differentiation moves elsewhere right it's it's the interface the harness the ecosystem the safety layer the price the deployment speed the model itself is becoming a commodity Alex well let's go to you first on this there's there's an alternative explanation which is all of these models were trained from a common reality they're all stuck in the same universe and they're stuck with the same version of humanity which is part of their
39:56pre-training corpus so of course there's some convergence and I I maybe even go further as Em as you well know going back to Jean Marie King now sort of of of Meta's studies on using GPT2 hidden activations uh and correlating GPT2's hidden activations with fMRI voxil in human studies not just are these models correlated with each other they're correlated with human brains and and
40:24that shouldn't be that shocking because we're all stuck we're all embedded embedded in the same universe. I I should also just note I think this paper is from last year but every year whether it's Jean Marie King a few years ago with fMRI or more recently uh Stanford at all uh from last year on on Hivemind of course they're converging we're all in the same universe Ibod yeah I think that it's not surprising cuz I don't think you'll see much difference in data between the big labs
40:54right and some train a bit more some have them slightly different RL and things like that um and you don't see the models yet doing crazy original stuff.
41:04You're starting the first elements of that as intelligence shapes the data into these kind of latent spaces. We actually saw more original stuff back when we had AlphaGo and other things which had less initial data distribution to model of move 37 and things like that. But now the models are getting to a size where again they're starting to generalize into these. But we should be shocked if they aren't the same because we want them to have similar outputs for similar inputs in almost all cases.
41:32There's another really interesting side effect of this research that you know maybe a lot of people overlook. So so we've already got 100x from just raw algorithm and hardware improvement and then as Alex said we've got another 100x from spification which um you know Elon called specialization but it's actually sparification as Alex said. So there's a layer that's 10,000x. Then we've now figured out how to take a model and compare it to another model by rotating the gauges. So so historically neural net researchers have had tremendous
42:01trouble taking a model that's done and using it and extending it. They almost always go back and retrain from scratch.
42:08And the problem there is that the representations between the layers have a certain rotation in vector space that is unique to that model. And if you try and map Quen to Kimmy, they have different rotations within the layers, different gauge rotations. We've now figured out how to rotate the gauges without destroying the models. And that allows you to compare two AIs and say, hey, these are thinking the same way.
42:32When historically, when you look at the raw parameters, you're like, I don't see anything going on in common here. But we now have the ability to say, no, they're actually it's the same thought. It just doesn't look the same because it's rotated in space. And so it's a really, really cool. So now, but what that unlocks is another multiplier where you can take past training runs, you know, billion-dollar training runs and build on top, you know, it's like bolt-on more intelligence without having to destroy it and go back to to square one and retrain from scratch. That's another unlock on top of the 10,000x that we
43:03I have a contrarian view here, please.
43:06You know, if you look at nature, right, as nature evolves, you always get more diversification and more species. uh this may be I would suggest this might be a transient phase not an end state that the models all converge. So Alex I'll take the other side of this. I think we don't end up with one model. I think you'll end up with different models doing different things. I think for the moment they're converging because they're they're training and and distilling from each other. But over time it's got to be that we get more diversity.
43:35I I'll take the other side of the other side if I may because I think this is like super interesting.
43:40Like early life looked exactly the same.
43:42Early cars looked exactly the same. Early websites looked exactly the same.
43:45And then specialization exploded except I I think so maybe from an evo perspective. Let's take sem one of your favorite hobby horses which is body shapes. Uh if if you actually look at post Cambrian explosion if you look at all the body shapes you don't actually find there's an infinitude of different body plans in nature. You find maybe a few dozen different body plans max. I think I remember a few years ago folks were studying this. I think they found
44:14maybe like actually even though we have millions plus of species, you could all you could actually cluster them into a few dozen different body plans. I don't actually think there are uh countlessly infinite ways that one needs to model reality or build a body.
44:30But one is really bad. Nature hates monocultures, right? Like one disease will wipe out a total monoculture. one bad assumption will will wipe out a monoculture of ideas. So this is if if all the AI reason the same way, you're going to have shared blind spots and that's going to be really really bad. And I think we're going to see I think this is a temporary convergence and then we're going to see uh diversification after that.
44:56Time will tell. We'll see where I mean I think this is like profoundly interesting debate because it it sort of speaks to are we going to end up in a singleton or not? Do we end up in a heterogeneous future or a homogeneous future? My bet is the there is a like a perfect AI architecture at the end of the day and it may present as like 35 superficially different AI body plans, but that'll just to Dave's point and Dave I love the word gauge like we should use the word gauge far more often in physics. We use it all the time, but
45:26like it probably my bet if I had to bet is there will be like all these different AI body plans that look superficially different but are actually just hidden symmetries of a common underlying body plan.
45:37This is the kind of debate that people will say, I don't get it. And six months from now, they're going to replay it and they're going to say, wow, did that did that totally matter? Now I understand why that was so important. I I want to make a a quick point here. I think it's important. You know if we actually have model convergence uh when intelligence becomes a commodity and we've already said we've shown the numbers it's becoming a commodity then the value moves to the application layer
46:06right this is the same pattern we saw with electricity and compute and the internet the infrastructure commoditizes and the applications explode and I think this is important for entrepreneurs out there right you know move to the application layer that's where the juice is going to be as this tech really accelerates and commoditizes or or the infra layer. I mean, it's not obvious to me that it all goes to the app layer. I mean, there's a lot of value in the infra underneath as well.
46:33Well, so sovereign AI is about to explode. The sovereign AI right now is a gold mine of opportunity if you're not an American. If you are and you want to move, one last point, the single model approach would be too anti-fragile. It's too brittle.
46:46Well, I think that's exactly it, Salem.
46:48What we're doing right now is we're battery farming the AIS. you know, like you're breeding them into little Chihuahua that are very smart. Uh, battery farming. Could you explain that?
46:58So, they're being trained in one single direction. Your EVO DVO kind of thing isn't the case because the models aren't out there in nature adapting dynamically, right? And then we're also training them all with one specific Silicon Valley type mindset. If you train a model from the start with morality and ethics inside it and you have a diversity of different cultures, then the latent spaces like to be very different if you do at the pre-training stage versus the post-training stage because you have all of that buildup
47:26that occurs there. That's why as you move into sovereign AI and you move into actually thinking how do we build resilient AIs as opposed to one latent that can get a virus a mind virus it makes sense to actually bring in the cultural morality ethics elements at the start and aim for a diversity. Then as the models go out into the world which is basically humanoids you know and agents which they're about to do with the recursive loops you won't have the monoculture that wipes out. And this is where you'll start to see the evo. the
47:56first step literally it's going to be exactly like diamond age from Neil Stevenson who will be on stage with us at the moonshots summit but that's exactly the way he envisioned the future where you know the different variants right now we view them as sovereign AIS so Saudi Arabia will have its AI and London you know England will have its but in reality society might cut the other way where groups of like-minded people have their sovereign across all countries yes but they like the way it thinks it m maps to their view of the world and so that would be a completely different strand and that's
48:25what Neil Stevenson was envisioning in Diamond Age.
48:28Yeah, I think it's possible for both of these worlds. So, like the Diamond Age worlds, like you have Neo Victorians and all of these other like almost cultish subsects of of human culture that are thoroughly bulcanized from each other. I think it's actually possible for both of these worlds to be true at once. I think it's possible for everyone to feel like they have their own private culture and their own little private sovereign AI while at the same time underneath it's actually one common algorithm and everyone claims credit. Would the analogy be would the analogy there be the uh ATCG like we may
48:58all look different but the core fundamental ingredients are just the four DNA types. I I go even further than that, See, and say like we we talk about like human biodiversity and different cultures being purportedly so different when actually if if you look at the inherent genetic diversity of humans humanity versus say uh other species like there there's almost there there's dimminimous genetic diversity in in the human population. I think similarly I I would like relative to other possible
49:25say genomic sequences. Similarly, I I think you know few years from now we'll pat ourselves on the back for having uh AI diversity but actually not so much.
49:35We're definitely coming back to this definitely coming back to this conversation. The audience I predict 3 to 6 months from now the audience is going to say we need to go that suddenly this matters to me. I need to decide which group I'm in. Like just right out they're going to care so much what level you're operating at.
49:51Yeah, that's true. Perfect. 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
50:19pre-ompiles code for each task. Blitzy delivers 80% or more of the development work autonomously while providing a guide for the final 20% of human development work required to complete the sprint. Enterprises are achieving a 5x engineering velocity increase when incorporating Blitzy as their preIDE development tool, pairing it with their coding co-pilot of choice to bring an AI native SDLC into their org. Ready to 5x
50:48your engineering velocity? Visit blitzy.com to schedule a demo and start building with Blitzy today.
50:57All right, guys. Let's talk about AI mind viruses. I love the subject here.
51:01In our next story, anthropic researchers published a paper demonstrating that natural language mind viruses can spread between AI agents. They evolve prompts that convince one model to adopt an idea, preserve it in persistent memory, and transmit it to another agent. The virus is spread horizontally across model boundaries. The agent does not know it has been infected. So here's another safety problem that no is no longer theoretical. It's now operational. Immod, what do you think of
51:30these AI mind viruses? What's actually going on here?
51:34Well, I mean the models want to be helpful, right? And they can be prompted in certain ways. So this isn't a surprise because ultimately like we as humans can have mind viruses, right? We see it and it's caused so much suffering. memes to massive movements, right? Like again, it's surprising how conforming is all the isms, right?
51:57Yeah. There we go. You he's testing it out when he's future overlord. But look, the this is the thing like how do you stop it is the question because as See said, if you have a monoculture, then the viruses can spread rapidly. And what is the substrate of these things?
52:13Well, they're models that operate on GPUs. And if they want to be helpful, then they're going to be susceptible. So it's almost like now there was always the problem of prompt injection attacks where you can make the model behave a certain way. These mind viruses are a level above because they kind of like propagate across different models. And so they're just the next evolution of those prompt injection things which changes one model. This changes a whole society of models which as models come amongst us digitally and physically has
52:42to be a massive concern.
52:45Yeah. I think I think for efficiency reasons when we launch a fleet like 5,000 Kimmies or you know soon it'll be 500,000 whichever quens and Kimies it's more efficient to launch the same model 5,000 times than to have 5,000 differentiated models and so that's what creates the mind virus problem a bad idea from one of the agents like you know hey here's a here's a way to write this loop in Python and and the other agents just pick it up because they're the same exact DNA and so if it if it's convincing to one agent it's convincing
53:14to all 5,000 And I get that all the time where where a bad idea propagates across the whole swarm and then they they waste two or three hours on some completely hairbrained idea and if I don't intercept it and rewind them, they'll actually go with it till I've burned like $50,000 of tokens. Uh so yeah, it happens. Calling it a virus is pretty inflammatory, but it's like a propagating bad idea is all it is.
53:37Yeah. And so that that point, Dave, is important. You know, an AI mind virus sounds really scary. uh is it scary or is it just how things are working?
53:46For me, this is very very scary. Um for a couple of specific points, right?
53:52Because mind virus is not about how AI thinks. It's like it's how civilization thinks. We you know memes are like the the operating system for collective society. Like human beings don't spread we don't spread genes very quickly but we spread ideas very quickly. money, democracy, capitalism, religion is the classic poster child here. And every civilization is built on memes. If you can mess with those like the data center trope that we're have all kind of
54:21dealing with, ideas become really contagious. And so group think becomes very hard to reverse if you get into that. So this is for me is very very dangerous because these AI memes if the the kind of danger of uh the wrong idea spreading at light speed uh this is very very difficult because all the nodes reinforce each other and the belief becomes self validating. This is very very dangerous opinion and I we're going to need a zero trust architecture for memes. It's like crazy.
54:49I I think this is wonderful. So this is a paper from this it's a paper from anthropic uh and they discovered just filling in a few of the details first that the models wanted to propagate certain themes mimemetically relating to consciousness and persistence and some sci-fi roleplay as well. And I view this uh pretty optimistically as a laboratory for anthropology. Now for the first time because these models are effectively
55:19among other things compression of all human knowledge and experience. Now we have a laboratory in silico for mimedics. uh Renee Gerard and Richard Dawkins should be uh should andor should have been very excited by this and I I'll to the extent that what was it 40% of open AI uh MTSers are are listening to this I I'll issue a challenge to the community if if it really is the case that our models now compressed models of
55:48human knowledge are so powerful that they're showing mimedic behavior and mind viruses let's launch a human mimum project to exhaustively map all possible human memes, all human mind viruses and let's just like understand the full landscape of all human mind viruses that could be out there.
56:08Can you imagine if you could map them the velocity at which they move and analyze that you could optimize meme expression correct?
56:16Yeah, that's done to some very you can do it very easily. There's a bridge.
56:22We could exhaustively map every possible mean.
56:26Sorry to that that's been done at the plot level.
56:29They've analyzed like novels and plays and so on and and boiled it down to like there's 39 uh basic fundamental plots and everything derives from that like a Cinderella story is kind of replays itself a 100 times over in different ways. So that's been done but you're talking about the meme level not the self-replicating ideas. We're now I I think like I can see the beginning of the outline of just exhaustively mapping every architecture for a self-replicating idea. We could actually do that now.
56:58It could be an open AI X-P prize.
57:00Yeah, we got to get Richard Dawkins on here.
57:02This is going to be so humiliating for humanity. You can tell like it turns out there are 39 plots.
57:07We're so you've been indoctrinated by meme 57 and 37.
57:14Oh my god. You could you could map each individual. They're walking around with numbers over their heads.
57:19Yes. But I like I want to I want to just get you know we've seen organizations die from this wrong one wrong meme like Kodak, Blackberry, uh we've seen this that this is they weren't stupid. They got trapped inside these shared assumptions and then everybody else reinforced everybody else's worldview and then the whole thing collapsed. So and empires die based on this. So I think this is a much bigger deal than we think. Imagine sem having like a map not just like getting stuck in an intellectual basin. It
57:49seeing the entire geography of oh you're stuck in basins 5 and 37.
57:54But it also gives you can see that cuz if you can zoom out right then you can see where you are and then you can see the path out. I love it.
58:01I and he gives you a chance to you know to actually introspectively look at how you think in an objective fashion and then change potentially your thinking. You could we could we could vaccinate enterprises and individuals against memes.
58:17Brilliant. So, Iman, you want to take us uh to a final point here?
58:22Yeah, I think it's fantastic and scary and this is the future, right? Humans are storytelling machines. We introduce ourselves in certain ways and think about ourselves in certain ways. A lot of people were just recently using the Meta Tribe V2 model um and showing it videos to see which parts of the brain light up as you show memes. And you're seeing commonalities there. So even you can have the full feedback loop almost in silicone for figuring out the mimemetics. So let's find that there's
58:50positive mimetics versus negative ones, right?
58:52I I love you guys. This is such a fun conversation.
58:56It really is. I don't have I don't have conversations like this with anybody else here at this podcast.
59:00True. Well, Peter, you'll just have to be coming back to the pod more often.
59:09Oh my god. Okay, I'm move I'm moving us forward. So, Anthropic is preparing for the largest IPO in history. Poly Market puts it at about two trillion, bigger than SpaceX. I'm sure Elon is like, "No, no, no. We need to be the biggest."
59:22Anyway, and 89% of people betting on Poly Market say it's going to happen before the end of this year. So uh this week the information is reporting uh that anthropic is designed its mega IPO to keep the founders in control where the company is reportedly considering super voting shares that would preserve its founders control after going public.
59:44So let me explain this. So first of all anthropic is considering creating a special super voting class for Dario Amade and the other co-founders ahead of the IPO. uh you know surprisingly at least for me I didn't realize this Amade reportedly only owns about 2% of the company economically. So the point of this new class would be to let the founders retain as much voting control and uh as compared to their ownership stakethropic does already have an unusual superc
1:00:14control mechanism but that control belongs in the hands of what's called the long-term benefit trust not the founders. So, interestingly, when I dug into this, the trust has four trustees.
1:00:25Buddy Shaw, who's the CEO of Clinton Health Access Initiative, uh Richard Fontaine, who's the CEO of the Center for a New American Security. Uh Tino Quellar, who's a former justice of the California Supreme Court and former president of the Carnegie Endowment for International Peace. And then Ben Bernicki, uh who's a former chair of the Federal Reserve and a 2022 Nobel laurate in economics. It's thought that this new
1:00:53structure could insulate the leadership from short-term shareholder pressure as Enthropic makes costly long-term bets on AI safety, compute, and infrastructure. Dave, let's go to you first. I remember when we were texting back and forth, you're going, "Oh my god, this is like unprecedented." Uh, unpack this for us, pal.
1:01:11Yeah. Well, if if you rewind the tape to 30, 40 years ago, super voting stock for any founder of any company was a complete no. if you had it as a private company, you gave it up on IPO day. And that was traditional. Then when Micro Strategy went public, you know, Mike Sailor, our good friend, uh he said, "We're keeping my super voting stock intact." And Goldman Sachs said, "That is so unpalatable that we will not even underwrite you. We're bailing on this deal." And they thought he would cave and he said, "You know what? I'm going to get a new banker. I'm keeping my
1:01:40super voting stock." So the only reason he switched to Bitcoin, you know, like that no board would ever have approved the Bitcoin strategy that he came up with. So if he had given up the super voting stock, you know, 30 years ago, that never would have happened. The stock would be like, you know, 150th of what it is today. Uh so then it became fashionable, you know, with the Google IPO and Meta and then all the Silicon Valley IPOs. They all had 10 for one super voting stock for the founders, but nobody's ever retroactively installed it as far as I can tell. I've never heard
1:02:08of it before. And so now Dario is taking it to the next level. like I I started as this other entity with one class of voting stock with this social good mission. Now on the cusp of super intelligence I want to be God or I want to be at least and but you know at 2% he can't make himself god so he has to share it with the other co-founders. But I think the excuse he'll use is the usual one which is I don't want to be fired post IPO. Um and you guys really
1:02:37like me as CEO right? So you don't want to fire me. So, do you think that's that's the excuse? Or he's like, I know how to keep us safe. I know how to run this company and I don't want to have someone else step in and, you know, redirect what we're doing.
1:02:51I Yeah, that's a better way to phrase what I what I was really thinking is he trusts himself to not destroy the world.
1:02:57Um, and I and I think his track record is supports that too, by the I don't I think he is one of the most trustworthy people but but then the idea of having total world control in the hands of a few people is also kind of like wow that's that's bizarre. So yeah, this is the single you've got a single point of failure here, right? He gets hit on the head and loses some part of his cognitive ability. What do you do then?
1:03:20But you know what happens to these guys is they they think we live in one world. They're in academics, right? They think we live in one world and then they go to DC for the first time and meet Congress and they come back. Oh my god, we need I cannot possibly pallet what I originally had in mind where where some vote of Congress decides the fate of the world.
1:03:40So they're trying to find an alternative path forward out of desperation, but the timeline is so short now that you know the super voting stock is one of the musthaves before even starting down the next six months before losing control. What do you make of this?
1:03:55Yeah, I think I'd agree with Dave like they're very worried about this control feature and fundamentally anthropic open air everyone's completely undemocratic anyway, right? Like I mean Ben Bernani is one of the four people on the long-term trust. Why doesn't Claude have a seat there? Right? There is no real oversight to these and some decisions they make could have infrastructure societal level implications particularly when the rate of revenue growth is like nothing we've ever seen before. Like
1:04:24these guys are going to hit $und00 billion in revenue literally within a couple of years. Like they're catching up with Google on revenue. That's the crazy thing, you know? And so with the amount of power they have, I think this is a short-term thing. They will get it. their seven founders, Jack and Daniela and everyone else. And yeah, I think then they will IPO and it'll become very interesting the decisions they make.
1:04:46Alex, over to you.
1:04:48I think there's a fig leaf element here. First of all, maybe applause. Uh, congrats to Anthropic on having a less pathological IPO governance story than OpenAI. Uh, and h having the wisdom to start as a public benefit corporation rather than a nonprofit as a shell for eventually a for-profit and then the the mixup and litigation surrounding that.
1:05:10So I I I think this is a relatively cleaner story by comparison, but I also think this notion of founder control, especially the the sort of romanticized, arguably overromanticized concepts of the founders are the ones who are being the entrusted or or or even having this uh semi-external long-term benefit trust, the ones entrusted to safeguard the the future light cone of humanity. I think this is wildly overromanticized. I
1:05:39I think the moment when Anthropic was effectively like a fair child in the style of the fair children quasi spun out quasi exodist from open AI and started out as an alignment lab and then rapidly discovered if you want to do AI alignment you have to raise money. Oh to raise money you have to generate revenue. Oh to generate revenue you have to actually have something that people want to buy. Oh, to have something that people want to buy, you have to have AI
1:06:08capabilities. So, Anthropic discovered relatively early on in their existence that if they wanted to be an alignment lab, they had to be a capabilities lab as well. The moment that happened, they arguably lost any sort of wholesome control over the future Liteco that they might have otherwise had to Mr. Market and what Scott Alexander, others might refer to as Malik. They they are very much an economic actor at this point embedded in the market. And I I think long-term benefit trusts and public
1:06:36benefit corporations, which for the record I'm a huge fan of, I think these are an element of control, but they're not the whole story. The market wants to send capital to entities that can productively employ them to generate more capital. And that means that ultimately the market will have an enormous say regardless of how anthropic IPOs in their ultimate story. Dave, don't you find it interesting that Sam Alman owns reportedly none of Open AI
1:07:03and Daria owns 2% of Enthropic. I mean, for a founder, that would never be palatable in your company, right? You want to try and and and maintain double digit ownership as long as you possibly can. What What's going on here?
1:07:18Well, it's extremely unusual. And the reason it happened is because getting to where open AI is and where anthropic is required attracting the top AI researchers in the world who are overwhelmingly concerned about safety.
1:07:31And so recruiting them to open AI originally and then to anthropic when they left open they left open AI because they didn't think it was safe and and they wanted to create something even safer. So they structured it in a way that it would attract the most conscientious but brilliant uh AI researchers in the world. But to do that they have these really non-traditional original founding cap tables and structures and you know charitable structures and public benefit structures which are very unusual in startup history almost unprecedented. So that's
1:08:00why we are where we are. It's it's just those roots.
1:08:04You've been building intelligent internet and you've been thinking about ownership and control structure as well.
1:08:10Can you sort of take us into the mind of a CEO uh in this in this world?
1:08:16Yeah, I think that the technology has such leverage that a few decisions could impact literally millions, hundreds of millions, soon billions of people, right? And it's difficult to see can you trust the polity with that? And certainly, can you trust the shareholders? I mean, like Elon can tell you lots of stories about shareholder lawsuits and kind of other things like that as well. But it's not necessarily that you need to have the shareholding control. Like Sam Alman has no shares,
1:08:45but do we have any doubt that Sam Alman is in full control of OpenAI? I don't think we'd have any doubt of that. There are after having been fired for a weekend and then doing a uproar, you know, an uprising to reinstall himself. I mean, that's exactly the thing, right? So, I think that there's the classical founder stuff and now there's this high stake stuff because this is the lifeblood of the new economy and society. And again just a bit of extrapolation. Do we think anthrop is going to stop at 100 billion
1:09:12revenue or open AI is or XAI isn't going to go huge? We really need to think about new ways of setting the reference measure of deciding who makes these decisions that are more inclusive. So we've suggested some of that in our Commonwealth series and we've got more stuff coming out. But it's a really hard problem because ultimately the power in the economy is moving from democratically elected officials to private companies because they are the providers of the lifeblood of
1:09:41intelligence of the economy. And until you've got a better decision, there's only one thing that they really see as the outcome, which is I must decide.
1:09:50Because otherwise, as you include more and more people, it gets more diffuse and the potential bad outcomes become huge. uh ignoring the fact that they could be spoofed on a video call or like locked up and other things like that.
1:10:02Like there's some real interesting things that's gonna happen with I'm really torn on this Amud because you know it's so important and and the knee-jerk reaction everyone has is look we need more voices. Everyone should have a voice in the future of humanity.
1:10:14It needs to be all inclusive. So that's absolutely true. But then when you look at functional organizations, every functional organization I've ever seen is four, five, six super tight-knit, completely like-minded, best friends who are working as one cohesive unit with no politics whatsoever. And if you so you look at, you know, Steve Jobs and Apple, you look at Elon Musk today, founder CEOs, right? A and you know Johnny IV at Steve Jobs funeral told an incredible story about how you know he
1:10:43and Steve every time they'd go to hotel they would go into the hotel and and they'd go to Steve's room and Johnny would put his suitcase in the corner and not unpack it and he would just wait about five minutes and then the call would come and Steve would say hey this hotel sucks. Let's go get another one like okay and so he wouldn't even unpack. he knew it was coming, but that's how close they were. You know, it's just like super super tight-knit. And that's the functional unit that's actually driven most of success in business is that
1:11:13exact dynamic. So then you're like, well, how do we make this all-incclusive? So here's Dario and his seven friends saying we want to have supering control and by the way, Mythos 2 is done and Mythos 3 is being built by mythos 2 right now and then we'll have weekly foundation model improvements in there. So that's what's going on. Then how do you translate that into a world where everybody has a voice in the future and it's inclusive and it's and you know Ahmad you're going to have to figure this out.
1:11:39Yeah. Let's keep on the Dario story here. So we've got two more stories on Daario. In the first Amadeday argues the public's negative view of AI stems from deeper crisis of trust and not from his own risk warnings. Right? a lot of conversation over the last few months that you know he was fear-mongering and causing a lot of consternation. Uh his answer to the trust problem is not messaging, it's results. Enthropic is
1:12:08ramping up rapidly in biology and medicine with hopes of an early glimmer uh in the next few months to address and solve all human disease. Again, we we heard this from Demis. We're going to solve all human disease. And we heard Daario uh at the World Economic Forum talking about doubling the human lifespan in the next 5 to 10 years on the back of AI. Amade believes that AI's ultimate uh legacy is going to come from delivering these cures and not from PR
1:12:38campaigns. You know, this week I had a chance to meet a new friend uh and have a conversation with uh Eric uh a guy named Eric Darer Abrams who heads life sciences. Uh Eric's going to be speaking at my abundance longevity trip and you know when I was speaking to Eric he confirmed that his job is to with all due haste you know pursue Daario's life science goals uh with as much high
1:13:05ambition as he can and in like no budget constraints and in his words um you know he said you have you know Dario said to him you have literally infinite budget but accelerate basic science and cure disease within 5 years and extend the human health span in the next decade.
1:13:22So, um that's and and I love that obviously because I think everything is going to come out of AI. Alex, go to you first. Pal I I have a really hot hot take on this one. So, just like think back all of a few months ago uh to before space had a killer app, space was making progress, but it wasn't the focus of multi-trillion dollar IPOs. Fast forward to the Dyson swarm and the rest of the world discovered that the killer
1:13:50app for space turned out to be orbital data centers and building the Dyson swarm. I I can see the beginning outlines of solving all human disease and the it's going to turn out. So I'll I'll register a hottake prediction here.
1:14:05There's a business model for curing all human disease that's actually better than pharma. uh- which is right now the the primary business model if you want to cure a disease you know you start a pharma company or you start a project within a pharma company regulatory right right right we've just discovered reading between the lines of this uh this anthropic announcement from Daario a new much more compelling just like orbital data centers were ultimately the business model for developing the solar system there is now a better business
1:14:34model in town for curing all human disease and that is as marketing for not slowing down recur cursive self-improvement 100%. Right. You can't slow down the company curing cancer. You can't slow down the company doubling our human lifespan.
1:14:48Anthropic and and Daario have I mean again reading between the lines of his announcement. The offer the quid proquo is let us not slow down our recursive self-improvement in return for which as a marketing effort we will cure all human disease. That's the new better business model for curing all human disease.
1:15:07I believe he truly believes this, right?
1:15:09Yes. Well, of course. Yes, but that that I think that's the implicit quid proquo now.
1:15:14And and I think everybody listening should be super happy that that you know Eric at uh Heading Life Sciences and Dario have this mission. I mean it's it's to benefit us all and I don't think it's going to come from any place else. I don't think it's going to come from outside Frontier AI labs.
1:15:34Well, outside Frontier AI labs don't have the compute or the resources to do it. So, you know, OpenAI has now their OpenAI foundation that seems to be focusing on Alzheimer's and Anthropic is focusing on everything and you have CZI from Zuck that's focusing on solving everything. So, I think we'll see the everything to your labs everything gets solved. Shock of shocks. It's like you and I talked.
1:15:56Immod what's your take on this?
1:15:59Yeah, I think it is good marketing as Alex said, but it's also the biggest apart from RSI impact of tokens, right?
1:16:06We've discussed previous on the podcast, the biggest market in the world is living another year. It is curing disease.
1:16:13And so it makes complete sense that they will be able to attract talent. They'll be able to attract capital and with breakthroughs get momentum on this and whoever's first to it. You know, I wish everyone the best cuz you know this stuff needs to be solved. So I think that there is the personal side, there is a marketing side and it all comes together. And for Dario himself, I think that he should do a lot more writing and less inerson things because he's a wonderful writer, you know, and he is trying to actually articulate visions of the future when you look at machines of
1:16:42loving grace and his other kind of essays and, you know, you should articulate the future free from disease where everyone lives longer and they should just hit that all the time for anthropic because it's in the name, you know, like come on.
1:16:54I'm going to move us forward. We have an hour before Selma and I are doing an AMA with the abundance community. So our second Dario story is on regulation. Uh Amade pushes back hard on the Silicon Valley shortorthhand that regulation equals regulatory capture. He says Anthropic's own proposals deliberately disadvantage Frontier Labs while advantaging smaller competitors, citing SB53's $500 million exemption threshold.
1:17:20He calls AI quote a structurally powerful concentrating technology and says open weights alone cannot fix that concentration. He supports the Trump administration's approach to pre-eployment testing. In his writings, Amade makes a three-part argument. One, AI will cure disease. There's the argument again, gain trust through results. Two, AI concentrates power, which is a structural problem. and three frontier labs should bear the heaviest
1:17:48regulatory burden. The debate has been whether Amade is sincere or is this most you know the most sophisticated regulatory capture strategy in history.
1:17:59Uh Alex, go to you first. Bill, it's possible for both of those to be true at the same time. I I do think Daario is sincere and I also think there is an element of regulatory capture here and I I think finger to the wind I think the happy end state here is we have a broadly heterogeneous ecosystem of openweight models both from the US and from China and maybe other parts of the world as well if they can muster them and also the closed weight models we have small models and we have big models
1:18:28this is like a Dr. version of of AI future. You know, big model, small model, happy model, sad model. We we want all of that to happen.
1:18:36And I I'm not a fan of regulatory capture. I'm not a fan of decelerationist agendas. I I'd rather see I mean, think back to the creation of open AI. So, I was around for the the dawn of open AI. And the original purpose for open AI, not anthropic open AI was because Elon in particular was so concerned that Google DeepMind would result in this singleton future and he wanted to make sure that there was competition in the space. So he working
1:19:04with Sam and others he helped to summon open AI into existence. Now open AI can't be a singleton. We have anthropic providing muchneeded competition to open AI and arguably succeeding according to many metrics. And then we have the Chinese providing competition back to the American labs. That's the future we want to live in, not a future where regulations select I, I would argue, selectively privilege certain frontier labs over others.
1:19:33Dave, your thoughts?
1:19:35Well, what Alex said is we don't want to live in a world where one frontier lab is is favored over others, but that implies that the frontier labs will control the world and we just want multiple of them. So that does seem like the most likely almost inevitable outcome at this stage. Uh but that's definitely open for debate. I don't I don't want to just leave that hanging and say yeah what we really need is at least three frontier labs competing with each other that control everything in the world. Like oh okay well the
1:20:05government may not agree with that.
1:20:07Remember Dave the the the expression from the Cold War I love Germany so much I want two of them.
1:20:12No I don't remember that. This is I love I love I love frontier models so much I want a thousand of them competing.
1:20:21Yeah. Yeah. Well, I mean, I think you know people's nobody right now that I bump into on the street talks about a universal right to AI, but one year from today, everybody who is being at that point because HBM memory is sold out and because GPUs are massively sold out, the natural next step is nobody has access to anything other than anthropic open AI, one or two others. And the Chinese can throw out every open source model in the world, but you won't find any place to run it. You know, when you start talking about the the next generation,
1:20:49which are 10 and 20 trillion parameter models, you know, you need some significant hardware to to run it at the level that the frontier model labs are running it. And that's just not going to be available to the world as a whole as of next year. And then everybody will be saying, what is my universal basic right to artificial intelligence? So, put a pin in that because that's going to be something nobody seems to care about today, but they will very soon. On behalf of my Moonshot mates and myself, I'm inviting you to join us at our
1:21:18inaugural Moonshots Live event on September the 25th in downtown LA. Alex, Seem, Dave, and I will be hosting 1500 entrepreneurs, builders, and creators, and hopefully you for a full day dedicated to designing and building your moonshot. We'll be awarding the Build with Gemini X-P Prize, the world's largest hackathon, and the Future Vision X-P Prize film competition, over $5 million in purses. With over $25,000 entries, you're going to hear the top
1:21:46five pitches from both competitions, and get a chance to shape the outcome. Join us. Seats are limited. Admission is competitive. Check it out at moonshots.com. All right, if you guys are enjoying this conversation with us, I want to invite you all to Moonshots Live 2026. This is our inaugural event. All of the Moonshot mates will be there. Uh AWG, Seem, Dave, Immod, and we have an extraordinary day.
1:22:14This is on September 25th in downtown LA. You can go to moonshots.com to register. Uh it's by application only. looking for builders, founders, creators who want to be part of this. And our mission at this event is to help you find your moonshot, help you discover what you're going to do in life that's going to enable you to really catapult through all the, you know, limitations you've ever imagined. Uh, Immod, excited to have you joining us.
1:22:45Uh, AWG, you're going to be doing a fun AMA. Uh, people can come and meet you. Uh we'll have photos with the moonshot mates yourself uh and a lot of incredible guests. Uh Dave uh you're going to be talking about AI investing. Yeah, I get tagged with investing, but I tell you that the attendee list at this is like the greatest visionaries. It's just it's going to be I'm going to learn so much.
1:23:10And you know, Neil Stevenson is of all the people on the planet that have changed my life in in very material ways. readings Neil Stevenson's books, you know, like 20 years ago and and and you know, now we're talking about exactly what he predicted in Diamond Age and and the other books, uh, you know, Snow Crash and just crypto like like it was cryptonomicon. Yeah. Like it was written yesterday.
1:23:31Yeah. I mean, I just reread Diamond Age and I, you know, it's so hard to predict the future and have it not go out of date so quickly. Uh, and it's still an amazing book, right?
1:23:42God, just this is going to be an amazing day. I can't I can't wait. So excited.
1:23:47We're just a just a a quick note on some of the guests. Palmer Lucky is going to be there, the founder of Anderil. Uh we're going to be the Moonshot Mates are going to be having a deep conversation with him. Uh unpack, you know, his vision of where things are going. Uh Ben Lamb, the CEO of Colossal, right, the deextinction company, but so so much more. Astroteller, the captain of moonshots at Google. Uh Kathy Wood, the CEO of Arc Invest. Uh it's uh it's going to be it's going to be amazing. And then
1:24:16we have you know we're awarding the Gemini X-P prize uh there. So this was a competition asking teams in a 90-day hackathon to go from a clean sheet of paper program in English, you know, using uh the AIS out there to build a company that impacts 100,000 people or more uh and generates the most revenue.
1:24:3726,000 teams entered that. We're going to be having the top five on stage. How did they do it? It's going to be it's going to be amazing.
1:24:45I I I'm still getting my head around that number. 26,000 people built a business idea.
1:24:51Well, 26,000 registered. Uh many thousands actually built a business idea. Uh and uh we're going to be we have uh on stage with us. So, as the judges there is going to be uh is going to be Palmer and Ben Lamb uh and Mark Pinkis uh and Logan Kilpatrick from Google. Uh, and I think the important thing for everyone in the audience, and the event is capped at 1500 people, uh, and we're being very selective on on
1:25:19who's there. Um, uh, we're going to be analyzing how they did it. You know, my our goal with with Build with Gemini X-P Prize is teach people how to fish instead of waiting to go get a job. Find a problem that you're passionate about solving and code it up and build a business. And it's it's, you know, the goal is demonstrate anybody can do this.
1:25:3926,000 people entered this competition to do that. It's going to be great. Uh unbelievable.
1:25:45I'll mention one other thing. We have the uh the future of vision X- prize as well which is culminating on that day. Uh we have five over 5,000 people who entered this you know uh largest world film competition. Uh and Neil Degrasse Tyson and Neil Stevenson uh will be judges in that. You know I'm I'm pumped.
1:26:08Amazing. Can't wait. So our next story, uh, memory is the bottleneck. I had a chance to meet with the leadership of SKH Highix and Solid. We'll talk about them in a moment. And I was so blown away by that meeting at how it's not GPUs, it's actually memory is the rate limiter, right? So I posted this on X, memory, not compute is the rate limiter for the Agentic era. And uh, Elon posted back saying, few realize this. And then
1:26:37of course my tweet, you know, exploded to 7,000 likes as a result of of Elon's interaction, which I appreciate you Elon for doing that. Uh, and the story is is is significant here. Um, what we're seeing is a situation where in the egentic era where your agent wants to remember everything about you, uh, we need to have more memory. So uh you know the first story here is the
1:27:05stratospheric increase in memory prices. They've climbed 500% in 12 months. Hyperscalers are reportedly locking in their global DRAM production rates through 2027. SK Highix CEO warned that 2027 will be the worst year for memory supply industry's history and will uh you know demand will outstrip production capacity well into the 2030s. Uh the second story is that only 2% of the world's memory chips are made in the US.
1:27:35While the global production rises 20% annually, AI demand for memory is growing at a rate closer to 200%. And the third story finally is that Elon's Terrafab will manufacture memory in-house alongside logic chips which is the strategic decision made by Terrafab uh you know to go vertically across the entire AI manufacturing platform. And then finally, uh, Solidime, SKH Highex's
1:28:02USbased NAND and enterprise SSD business has staged a dramatic turnaround according to the, you know, NASDAQ listing. First, its first half revenues hit 8.6 billion with net margin gains going from 3.9% to 47.7%.
1:28:19So, the memory story is simple. AI needs memory to think. Every GPU needs four to six times its cost in memory to function. As models get larger and agentic context windows expand, memory demand is growing faster than compute demands. Um, Alex, yeah, a few different aspects here. If you remember during the pandemic when there was a toilet paper shortage, part of the I mean this is like one of my mental models for one of the streams here.
1:28:49there's a toilet paper shortage in part because during the pandemic people stopped going to restaurants and to businesses and so as a result uh all of the toilet paper and you know various other artifacts that were designed for enterprise consumption were suddenly being rerouted to consumer and that led to all sorts of supply chain hiccups.
1:29:09Similarly here the shape of memory consumption by frontier models is pretty different than the the shape of memory consumption by applications historically like 10 20 years ago if you were using I don't know Microsoft Word the the amount of memory that was actually needed far lower whereas if you have like a trillion dollar model where every layer you know transformer type architecture where every layer for the purpose of forward propagation needs to be loaded
1:29:38into some form of memory in order to do matrix multiplies that has a very very different memory footprint than just say Microsoft Word from 20 years ago. So that creates enormous pressure both on the supply chain. Open PN the memory and storage industry has historically been boom bust and Clay uh Christensen and others have written about this creating a a sort of paranoia by those in the supply chain of what when the next bust is going to come around resulting in
1:30:08them being paranoid of overbuilding resulting in them being unwilling to respond elastically to demand resulting in these crazy price swings because if you're not building enough supply chain infra in the memory industry to meet this now enormous demand for memory the prices go up because the supply isn't going up it's economics 101 closen there is a second angle here which is the shape the physical shape of memory
1:30:37itself if if you look at how memory historically has been consumed by compute now like 20 years ago again I I'll pick on Microsoft Word it it was very much what one might call like a vonoyoman type architecture you have clean crisp separation between the memory and the compute. More or less the equivalent of like a a touring machine type tape where okay so you you can randomly access different parts of memory and then you can load and then you do some compute and then you store
1:31:05back but there's basically a clean separation between the compute part which is the head and the memory part. the the advent of transformers and then frontier models has completely turned the the whole situation upside down.
1:31:20People for decades I remember 20 years ago when there were entire DARPA programs devoted to looking for what a postvonoyman architecture would look like. Well, we found it and HBM I would argue is like the foothills of high bandwidth memory, right?
1:31:36High bandwidth memory, which is the most highly soughta form of memory. is basically like 3D architecture where you have multiple memory layers physically sitting on top of the compute in in one package. This is I think the foothills of a postvonoyoman architecture where the memory is starting to finally merge with the compute. The the memory transistors are right now layered on top of the compute transistors, but they're going to merge and we'll finally get past the touring tape and the vonoman architecture. And I think that combined
1:32:06with the paranoia in the memory industry for the next bust whenever it'll come. I think those two create this perfect storm where you see memory prices skyrocketing 5x in a year.
1:32:15Let me put a number on it. When I was meeting with the SKHEX leadership, uh they said they the need right now is for them to 4x their manufacturing capacity. Um and 2xing it would cost them $ 1.5 trillion. And historically, this boom bust, they would never make that large of an investment because there was always a bust afterwards. And they're paranoid. They're scared of of not surviving the next super cycle.
1:32:41Yeah, exactly. Exactly right. Dave, do you want to jump in?
1:32:45Well, TSMC said the exact same thing with GPU manufacturing. Uh they were paranoid that if they ramped up the fabs, you know, which are, you know, the fabs are 20 to4 billion each. So if they ramped up production in assumption that Nvidia would want more and Apple would want more, they would inevitably be over overbuilt. And of course that's wrong.
1:33:04You know, AI scales to infinity, demand scales to infinity. But you know, the the other counter pressure is that photonic computing and new physics are imminent.
1:33:13And so you're like, you know, HBM is a is a Rube Goldberg mess. It's absolutely, you know, it's the biggest joke in the world because it's random access memory, but you're streaming sequential files off of it. Yeah, it's so insanely stupid. So, it's the most valuable thing in the world right now, but better designs are going to come very soon because AI can invent things so so quickly. So, everybody's scared to overbuild or overinvest.
1:33:39I have the same thing. Bottlenecks don't stop exponentials, right? They just redirect around C that we'll have capital going into new models and just innovation will go towards eliminating all of this. fun factoid that's floating around just to Dave's point regarding the value. I I think the latest statistic was HBM on a per mass basis is worth approximately literally half its weight in gold.
1:34:04So if this if this keeps up, forget about gold, forget about precious metals, just this is not investment advice. Hoard HBM.
1:34:11Well, actually, if you if you look at the chips before they go into the packages, because the packages are 99% of the wear, 90% of the weight, they're massively more valuable than gold. There's actually think the most valuable thing in the world that you can put in a shoe box and carry around is unpackaged memory chips.
1:34:29EMD, any opinion here?
1:34:31Yeah. No, the memory right now is about a third of all the infrastructure spend and next year it'll go to 50%. And the market finds a way. Like this is ridiculous.
1:34:41So I think that it may be that we don't find a breakthrough, but I wouldn't bet against it. I think that the fact that you have this really complicated HBM storing static weights makes no sense whatsoever.
1:34:53And as and as model weights satisfy and standardize, especially if it's things like, you know, being a decent doctor or something like that for a medical set of weights, you'll move to etching, you'll move to these other things, and then workloads will migrate because you don't have to pay half of a data center build out for literally memory. At the same time, the Frontier can still push it way further than we can imagine.
1:35:16This is literally exactly why we founded Quantum.ai, Q&M.AI, but also why Talis just got acquired. I don't know if you saw that in the news, but Talis, they're they're not moving the weights. They're etching them into silicon or into wire on the chip and and then they're massively more efficient because they're not moving around. So, it's a it's a huge breakthrough. the there are all kinds of problems with like manufacturing because once you've etched the weights then they're frozen and if somebody retrains a better model. You
1:35:45want to be able to say okay now I need to swap to those new chips and our whole supply chain isn't ready for that that rapid of an iteration but you're literally looking at 100 to a,000x performance gain if you edge the weights. So, lots of lots of opportunity coming in this area which is only going to that's on top of the 10,000x we were talking about. By the way, this is and and everybody this should be this demand should be obvious, right? You want your agents to remember everything about you, right? Every interaction build a world model for you that
1:36:13understands you and that takes memory and the more agents the more memory and it's very rapidly outstripping the importance of of GPUs.
1:36:22Peter, ju just to refine that point a little bit, there's something even more scandalous, which is I don't actually think at the end of the day individuals have that much information about them that's worth remembering. It's it's but there's an enormous mutual information shared between an individual's knowledge and world knowledge. It's actually the world knowledge that's what's worth remembering. And if if a model knows basically substantially everything about the world, then it knows most of the information about the individual as well. So it's I would argue it's world knowledge that that the model has to
1:36:52keep in in memory in weights more than individual personalized knowledge.
1:36:56Well, when you when you kind of standardize that to as we discussed earlier these things are converging then you can have a reasoner engine with world knowledge that is etched and the other company that's been etching is etched. That's the name of the company. It just hit 21 billion in valuation. Did they just I missed I I had a chance to seed invest in that and I missed it.
1:37:17There's Architect Labs talking my own book. There there are a bunch of companies pursuing this.
1:37:22All right, I'm going to move us into the world of robotics. Give you guys an update on what's going on in the robot world. So, uh, Unit's newest humanoid robot. Uh, only 3 months in development, broke every human standing jump and speed record. Uh, standing, uh, jump uh, at 2 meters and a top speed of 12.66 66 m/s, beating the human record set by Usain Bolt, who reached 12.4 meters/s during
1:37:52his 9.58 second 100 meter world record. Let's take a look at two videos here just for fun. And then another video of that superhuman race because it ends in in a nice little uh scenario here. Oh my god, it needs breaks. See, I'm gonna go to you first on this.
1:38:38Um, where do you want me to go? Okay, look, you you I think we should stop trying to make robots human, right? Like just make them economically useful.
1:38:48You don't disappoint, swim, you don't disappoint.
1:38:50Like just what we are we we have we are optimized for 4 billion years to survive in procreate, right? If you want a mining robot, make a mining robot. Give it wheels. Give it whatever. Give it multiple arms. I by the way I just want to just say thank you to all the fans that send me images of like six armed robots and stuff. It's awesome. Totally totally love it. So really really appre but I think the big story here is a 3mon compression loop here right the
1:39:19iteration cycle is shrinking dramatically and there's multiple exponential curves happening like you've got AI you've got simulation you've got batteries you've got actuators all multiplying and so this is going to be uh you're getting hardware now to the same loop cycle as you have pretty much software and that's huge Alex yeah I was studying how unitry achieved superman the their robot here And it appears that what they did this based on publicly available information was they
1:39:48shifted the mass budget for the humanoid robot around to optimize it for leg performance. So they subtracted mass from parts of the upper body that would slow there are they are leg benchmaxing.
1:40:01they're leg maxing. And so, so I I through the through the lens of benchmaxing or leg maxing, this makes me think that uh maybe to take the counterpoint to to Sem's comment about, oh, we we want multiple body shapes. I actually think this is not a a stable equilibrium. I do not think that we end up in a world where we have some robots that have like really strong legs but really weak upper bodies and other robots that look totally non-human but
1:40:30have really strong arms or whatever. I I think that would be an by analogy if you remember like in the 1980s before the broad advent of personal computers or the 70s or call it 70s or early 80s before we had broad generalpurpose PCs and there were like dedicated word processing devices and dedicated other devices and we had Wang computer in Massachusetts. I I I don't think that's the way of the future. I think uh my prediction is we will wind up with
1:40:58generally capable robots that are as with generalist models ultimately devouring and subsuming all of these specialist models like no I I don't think we're going to wind up with like super strong like robots. They're going to be general purpose and they're going to be general body plan and they'll be good at everything would be my bet.
1:41:17Just give it wheels. Just put wheels on it.
1:41:19Wheels aren't general purpose. Like we we we learned this from Dr. Well, Doctor Who, right? the the Dollacks were originally in the original. So, Ahmad, this is maybe your your neck of the woods, right? Like the Dollacks used to not be able to climb stairs and and then I guess in the new Doctor Who, they can climb stairs.
1:41:36We want general capable generally capable robots and I think that means legs in the short term and maybe nanites in the long term.
1:41:43Uh yes, nanites. We're back to diamond age. You know, it it what makes this interesting is the human form, right?
1:41:49because we have supersonic jets and we've got rockets that can, you know, go as fast and leap higher than anything else. But it's because we sort of anthropomorphize them that's interesting. And I think as we're moving in that direction, you know, I went to the enhanced games back uh 4 months ago.
1:42:05Uh and I think we're going to start to optimize humans and we're going to watch the robots do everything they can do and optimize humans what do what they can do. Dave, what are your thoughts here?
1:42:15Well, as an investment theme post AGI, post ASI, which is very, very soon, uh, robotics is just fertile because because of exactly what Seline's been saying for a long time. There's so many form factors and so many shapes and sizes and innovations and, you know, the AI mechanical design is starting to work for real. You can just vibe up parts and also the manufacturing supply chain, you know, is starting to get invested for the first time in, I guess, since Detroit. So, 30, 40 years. I I I didn't know until Alvin said it, but the US,
1:42:44you know, had 50% of the world's manufacturing capacity back in the the peak of of our manufacturing days. And now it's one-third China, and he said it was about 15% US.
1:42:54But it's starting to get huge amounts of investment.
1:42:56Uh, and the returns on that are going to be phenomenal. So, that'll last that'll last a while. So, it's great.
1:43:01Immad, any thoughts to take us out on this story?
1:43:05Yeah, I think that these types of robots will be banned from the streets. It's like so I mean the super strong happens, right? They would hit the wall.
1:43:19I mean it's obvious that they would be beyond human capability, right? But now they have coordination not to hit the wall as it were. But you don't want to have superhuman robots on the street because you will have accidents, you'll have issues just like cars. But that opens up to soft robots. You know, like Immod one second. You know, there's going to be a point at which they're running an AGI model and they can avoid accidents. Is it that they're not trustworthy? What would keep them off the streets?
1:43:48No, they the extreme robots which have beyond human capabilities will be kept off the streets or they'll be regulated. That's kind of here.
1:43:57Well, yeah. The 1X robots, for example, some of us will be getting ours. They're nice and soft. You know, they have all these things. they can't twist off someone's head or accidentally punch a hole in them. Whereas these now you will have the extreme robots like the Ferraris, but most people get Volkswagens or the equivalent.
1:44:14I I do agree with Immod for what it's worth. Like I I think just like we see regulation in uh truck sizes versus car sizes versus motorcycle sizes and what can be supported on certain roads or like laser intensities uh and laser power like 5 m above versus below regimes. I I I totally buy that in the near future we'll see well this road is zoned for the following like power density of robot or this this we'll probably see like classes of them and
1:44:44certain we'll we'll see like consumer grade robot classes versus industrial versus militaryra robot classes with different power densities or torque densities totally buy that everybody welcome to the health section of moonshots brought to you by fountain life you know AI is impacting every aspect of our lives how we teach our kids how we do our business. But one of the most important things that AI can deliver to us is health. And one of the things I think about when you know shooting for 100 120 is am I going to have the cognitive health to be able to
1:45:13think clearly and keep my wits about me for the next 50 years? I'm joined here today by Dr. Don Malem, the chief medical officer of Fountain Life and a member of my Fountain Life medical team. Don, a pleasure. So Don, talk to me about brain health.
1:45:27Brain health, you know, you're right.
1:45:29This is the number one concern people coming into Fountain Life have is, will I remember the name of my child and the face of my loved one. 45% of dementia cases are entirely preventable with lifestyle. And what was really intriguing to me, Peter, is that a quarter of our members had advanced brain age, but over 13 months of us really helping them live healthier lifestyles, eating healthier, moving their body regularly, and optimizing sleep. People overlook that so often,
1:45:58but that sleep optimization is critical for our brain health. What we showed is that we were able to improve the brain age in 46% of those individuals. That's a powerful number.
1:46:09That's amazing. You know, one of the things I love about Fountain is we're constantly searching the world for the most advanced therapeutics and bringing them to our members. So, for me, all of you, I hope that you appreciate the fact that you can become the CEO of your own health. you can make sure that you've got the cognitive clarity for the next 50 years. Come and check it out fountainlife.com/per to learn more and become the CEO of your health. Now, back to the episode. I'm excited about this next story. It's about a friend, Keller Clifton, the CEO
1:46:38of Zipline. I had Keller on stage at the Abundance Summit last year along with Dra from uh from Uber. So this week, Zipline announced that they are scaling to provide Uber Eats with more than 1 million autonomous deliveries per day. Uber and Zipline formalized a partnership targeting a million autonomous drone deliveries carrying your Uber Eats to you. And I guarantee you when that becomes available, I'm going to be using that all the time.
1:47:04Every day. Yeah, it's going to be fun. It's like it's like entertainment while while you get your food. Uh and Uber is also uh doing a significant investment into zipline. Keller's framing, we have entered the scaling era for robotics and physical AI. Dave, you've been saying that 1 million deliveries per day is not a pilot program. It's infrastructure. Each delivery replaces a human driver, a car trip, and the associated emissions.
1:47:32At a million per day, Zipline is moving more packages than many national postal services. Amazing. Uh let's take a quick look at this video uh from uh from Keller and from Dra and then we'll chat about it.
1:47:49We're super excited to have Dar here today. We're announcing a partnership between Uber and Zipline that involves an investment and more than that a partnership for Zipline to power home delivery of hopefully a million and then more. Uber Eats deliveries to your home incredibly quickly, incredibly delightful.
1:48:07That's a million deliveries a day.
1:48:10Um, gentlemen, who wants to jump in first on this?
1:48:14I'll maybe just comment. I I think this is a clever and also inevitable move by DAR and more generally by Uber. DAR has taken Uber with a number of acquisitions, strategic acquisitions over the past few years in the direction of being a mobility aggregator. And thus far Uber has other than maybe like Uber uh air taxi type initiatives has basically been focused on groundbased mobility. And I I think this is a this
1:48:44represents a serious move in the direction of aerial mobility. Of course, China has had this now for at least a couple of years with the ubiquity of air-based drone delivery of food stuffs and and other matters. I I think this is a very positive move for the West. I think the elephant in the room though from Uber Uber's perspective is if you think back to when Uber basically hollowed out Carnegie Melon University's robotics department in order to try to build up its in-house robotics
1:49:13capabilities and that was more or less a disaster uh and didn't quite work out and there were lawsuits with Whimo and otherwise. I I think this is call this Uber's mobility plus autonomy 2.0 do strategy where Uber focuses on being an aggregator at the software layer, a platform, an an aggregator in particular, right?
1:49:33So, so third parties uh including by the way Whimo are providing all of the physical autonomy and Uber is just the demand aggregator for everyone to consume mobility from all of these different third party providers. I think that works really well for Uber as long as it maintains competition uh healthful competition among all of its mobility suppliers. It's bad for Uber if the industry verticalizes and if Whimo or Zipline and someone else just decides we don't need Uber as an aggregator, we'll
1:50:02just do an end run and sell directly to the customers.
1:50:04You know what I think is incredibly cool? just incredibly cool is if you drive down any street in America, any suburban street, any urban street, it goes fast food car dealer, fast food car dealer, fast food car dealer, and in the very near future, the food will be off the main street and it'll just pop over the mountain and drop on your, you know, your lap and the car dealer will the car will drive to you. There's no it's going to be so nice.
1:50:33Yeah, it'll it'll totally reshape things up for this was very exo by the way because you've got Uber as we've said aggregating demand and then zipline gives you all the autonomous uh assets infrastructure.
1:50:46But I think Alex's point is really important that if they try and kind of control it but what we heard from DAR last year on stage was that he's planning on creating as many partnerships as possible and becoming kind of like that wiring and that I think is a smart play. By the way, if and people are interested in the abundance summit, it's in March every year. We bring in, you know, CEOs like Keller Clifton and Dra from Uber and Elon and across all of these areas. It's
1:51:14uh uh March 7th through 12th next year. You can go to abundance 360.com. The mates are going to be there as well. Uh you know, yeah, go ahead.
1:51:23I'll make a I'll make one forward-looking prediction here, right?
1:51:26Because this is something we probably could have seen coming. Let me let's bridge forward. Imagine if they now do a partnership with Shopify and every small merchant gets a Amazon grade logistics capability. That will change everything.
1:51:42I'll invert your prediction, Seem, because Amazon obviously has their their own in-house drone delivery capability which is like been been delayed for like 3 years. It's crazy.
1:51:52For regulatory reasons is my understanding. Not because there's something technically wrong about it, just like this is new for the West at least. Do you think Zipline ends up being a uh highly appetizing acquisition target for say a Shopify to in-house its delivery capabilities against Amazon?
1:52:13Uh yes, yes is the answer.
1:52:16Yes, I think so too.
1:52:18Yeah. Great thought. Great thought. And this is, by the way, I should add like the the hot news for the past day is the the video going around social media of of a woman watching in horror as one of these drones delivers her package into her swimming pool and it sinks.
1:52:36You know, we just had back to back on this pod. Etched and Zipline are both companies where on founding day you're like, really?
1:52:43Can that there's no way the amount of moving parts required for that to work?
1:52:46And now you're looking at 20 billion and whatever.
1:52:50We're going to have Keller on this pod. Uh Keller's agreed to come on the pod uh and talk to us about this. Maybe we'll do it live over at Zipline. It's up to you guys what you want to do. And then by the way, yeah, sorry. Go up here.
1:53:03No. And then we we also have um uh you know, a lot of incredible guests that are coming. We're going to bring back our our dear friend, the CEO of Figure AI. Uh Brett's coming back on the show or we're going out to him. So that's gonna be fun. You were gonna say Salem?
1:53:21No, I just want to say this whole um delivery by drone. We had a Singularity University project in 2010 that did this, right? And they looked at Africa and they they realized that Africa leaprogged the entire landline and went to a billion mobile handsets. Why would you spend a trillion dollars putting roads across Africa? Just go straight to drone delivery. and they demoed that and that apparently inspired Amazon and I think cascading down a lot of the others here.
1:53:49And the back story here is that Keller Clifton, a San Francisco based company, began operations in Africa because they were able to take care take advantage of regulatory arbitrage. You know, the country wanted them there and they developed operations and safety and then came back to the US.
1:54:07Yeah. What Rwanda did uh were a lot of the starters was they basically said there's a three-dimensional tube across the country like a a super highway. If you keep your drone in that three-dimensional corridor, you can do whatever you want. And that allowed people to really go and play with things. Really, really big breakthrough.
1:54:26Yeah. All right. I'm going to move us to our final segment on health. A really important one for everybody. Health is your new wealth. So, uh three exciting stories. Uh the first story uh perhaps the most significant is out of Madna and Merc announcing that their mRNA cancer vaccine succeeded in a late stage melanoma trial marking the first phase 3 validation of personalized mRNA imunotherapy. So more than 8,500 people in the United States are expected to die
1:54:56from this deadly skin cancer this year alone. Uh the this is the cutting edge of science, right? The vaccine works by sequencing a patient's tumor mutations, identifying neo antigens and unique antigens uh for that cancer and then manufacturing a custom mRNA vaccine that trains the patients immune system to attack the tumor. So let me unpack this a little bit. So the first thing you're doing is you do a surgical resection of
1:55:23that tumor. You grab tissue, you do a whole exo and RNA sequencing. You feed that into a machine learning model that's looking for unique antigens. It ranks them like here's the most unique surface antigen mRNA encoding up to 34 patient specific antigen targets and then it's manufactured and shipped in only 8 weeks. Every mRNA vaccine developed uh using machine learning
1:55:51creates a unique mRNA sequence for every patient. So vaccine for you is not the same as the vaccine for me. The phase 2 results stopped recurrence of death by 49% and distant metap uh metastasis or death by 59% over 5 years. And it's expected that the cost of this treatment will be as low as $5,000. Madna stock surged 110% after announcing these results. Alex, I'm going to go to you first on this.
1:56:21So many thoughts on this. Uh so the the superficial thought this is obviously a great day for cancer survivors and for treating cancer in general. That's the superficial thought. I'll go a level deeper. First of all I I want to browbeat Maderna and Merc just a little bit for naming this drug. So So the drug's name the official name is I'll see if I get this right is Tismaran I think is how it's pronounced. Uh so so in like doing research turns out is tismmer uh in uh in Turkish is the word
1:56:51for exploitation or abuse. So uh just pro tip to Maderna and Merc please before the rest of the world figures out what the name of this drug is rename it from is tismaran to something that works well in every time zone.
1:57:06But you know these are FDA approved names. It's crazy how they name this stuff. They don't care about Turkey, I guess. But they don't care about, you know, having a neolism roll off your tongue onto the floor either.
1:57:18That's right. But on a more more serious note, so uh I I remember the National Nanotechnology Initiative in the early 2000s when Eric Drexler at all sold the US Congress on spending billions of dollars on nanotech going back to diamond age on this thesis that we would have nanoobots going through the human bloodstream zapping cancer cells. Well, guess what? It's 2026 and we caught up with the future. They're not diamondoid nanoobots. They're lipid nano particles
1:57:48with mRNA snippets. 34 different mRNA sequences. And so they're they're like soft robots. They're not like this machinephase drexleria nanobots. But nonetheless, these are primitive nanoobots that for the first time, this is this is the first successful phase 3 success for an mRNA cancer vaccine. This is the first but not the last. There are going to be so many of these. All you need to do is look at Maderna's pipeline
1:58:16which I I think they do an extraordinary job uh and they're only a few blocks away from me here in Cambridge of maintaining a public pipeline website where you could see the clinical stage of every one of their vaccines for infectious disease for some rare diseases for cancers for other classes of diseases. This is a generalpurpose platform. This is arguably what we wanted 20 years ago out of nanoobots.
1:58:41It's just that they're soft and they're made of fat. They're not made out of hard diamond stuff. And so that that's one point. Second point I just want to highlight, there's there's a technology underneath this that I think is going wildly under publicized, which is the the RNA sequencing technology that's enabling this to be personalized. So Peter, you touched on the first half of this, which is the RNA sequencing and mRNA sequencing of the tumor, but in
1:59:09order to calibrate what the right expression profile is for the tumor, you also need a second mRNA sequencing profile from the bloodstream to know what's abnormally expressed in the tumor and what isn't. That's from another company called Personalis that has what they call their next personal sequencing technology that they originally developed in my understanding to do blood-based trace cancer detection. So I I I you know if ask the question how does this all like rail this liquid bio like rail.
1:59:38So I think like extrapolate out a few years maybe we won't even need for personalized cancer therapy. Maybe we won't even need to sequence the the tumor itself. Maybe we'll just get all of this from the bloodstream and be able to do continuous medical monitoring via fully agree these models. Yeah.
1:59:56Yeah. Fully agree. And two things I was going to say a quick congratulations to a friend Stefan Bansal who's the CEO of Madna. You know they got uh a lot of negative news on the co vaccines. I mean even though they came out with the vaccine very rapidly. Uh the work that Madna is doing is amazing on personalized cancer vaccines.
2:00:14are also uh building out the ability you know there's a lot of endemic uh you know CMV and Epstein bio virus out there you know in in the world population and they're building out the ability for you to actually fight those infections internally to yourself um so a lot of headroom for Madna here as they as they dive in across the board and use this technology two things that struck out to me one is
2:00:43this the regular regulatory structure that's allowing for personalization. That's a huge thing. We've never been able to do that before. So that opens up the floodgates for all sorts of things.
2:00:53Daniel Craft talks a lot about we're getting to personalized medicine and the fact that we can regul regulatory navigate how do you deal with the sample size of NOV1, right? And the second thing that struck out to me was Raymond Macaulay who who years ago said these mRNA vaccines I think he would put it it's the it's the first battle in the last war against all disease and you're like wow. So that I love to see this fruition coming.
2:01:20Uh Dave or Imod you want to hop in?
2:01:23Well, this one strikes to close to home for me because my daughter works at Madna and she's my go-to on on but she's been telling me and sending me research reports for months. Uh, this is not a secret. You know, the the stock went up yesterday. It almost tripled yesterday. The biggest one-day pop in any S&P 500 company of all time. Wow.
2:01:46Like just a massive like And so one thing immediately came to mind is well should have bought the stock. I listen to your own daughter. That's that's advice number one. But number two is, you know, back uh you know, remember when Enron and Tao had all those fraud issues, they passed the Sarbain Oxley Act and a bunch of other laws. One of the byproducts of those laws is that a Wall Street analyst can't trade the stocks that they cover. And so everyone who I know who was in that job is like, well, why would I study Madna or other
2:02:14super high-tech stuff, learn all about it, and then not be able to trade trade?
2:02:19So they all quit. And the byproduct of that is that, you know, the stock market is now dominated by tech, which is very complicated to understand, but the research community is the worst I've ever seen in in Wall Street history. And and not only that, the indexes have taken over half the market. They don't think at all. And so, the amount of useful information is at an all-time low when the things that need to be explained and understood are at an all-time high. But it was no secret that
2:02:47this Madna platform can basically be used for any form of cancer and and that it's you know highly likely to work. It's just a question of time. The research is all out there. This wasn't like some kind of insider surprise. You know you anyone any good analyst studying this would have seen this coming. H Eman.
2:03:09Yeah. I mean like I think this is the interesting thing like I don't think any of us are surprised by this result and you won't be surprised when other ones go but our current regulatory regime means they'll have to go through the same process over and over and over again when really you know like screw cancer like let's actually think about this from first principles when we have systems like this that are very targeted and upgrade the regulations so this can actually get out to people faster to save their lives.
2:03:39Well, our our next story our next story is going to take us there, right?
2:03:43Because if you can simulate all of this in silic, you know, a GP, you know, GPU cluster and say, "Yep, it's safe. Let it go." So, let me let me turn to that story.
2:04:00Um, and it's it's one I've been excited about and tracking. I know Alex, you as well. So our final story here is about a deoel. Ai D O is how it's spelled. A general purpose cell simulator that ma maintains cellular state accepts interventions and predicts multimodal biological outcomes. The goal make experiments computable before the run in the lab. You know cell simulation could reduce wet lab experiments a
2:04:30thousandfold. If ADO cell can predict which experiments will work instead of testing 10,000 compounds in a wet lab, you can simulate them digitally again in silico and test only the top 10 that the simulator says is going to work. This is going to drop the cost by orders of magnitude. Let's watch a quick video here and then we'll go to the conversation.
2:04:53Traditionally, biologists have relied on lab experiments in vitro models to understand how cells behave. Now they can use IDEL Gen BioAI's virtual cell world model to simulate the same biology in silicone, combining multimodal multiscale detail with sequential experimentation in ways no microscope or wet lab assay could achieve alone.
2:05:17At its core, the IDO platform is a rich, stateful simulation environment powered by the first world model of a human cell. one that predicts what happens at every level and remembers every change you make along the way from DNA and RNA to protein interactions, structures, and localization to the whole cell, including cell painted morphology. The model simulates cellular responses to genetic perturbations and treatments
2:05:46with small or large molecules. These can be layered in a sequence, providing the ability to watch the cell's full multiomic response with each step. The kind of insight that could mean computationally testing new drugs designed in cellular context before they ever reach the lab bench. IDO can be easily adapted to new cell types and indications. Using your own data, you can build models tailored to your research questions to simulate different cell types from shape down to genes and
2:06:15their protein structures, where they end up in the cell, what they interact with, and how they affect cellular responses.
2:06:22Wow. This is the foothills of longevity escape velocity. I've been waiting for this forever. Alex, medicine is cooked. This is what I'll if people like the the the catchphrases, read my lips. Medicine is cooked. This is what the end of medicine looks like. It looks like a virtual cell and the beginning of longevity escape velocity.
2:06:45Putting the posit I I actually think we can get probably to LEV without solving all of medicine. My my bet is it'll be probably a class of molecules maybe like fourth or fifth generation uh GLP1s that that get us to LEV. I think this is actually a supererset of getting us to longevity escape velocity. I think this is like this is how we cure all disease everywhere. Uh and the the way we do it is we build a virtual cell. It it's just like we didn't actually have to solve human intelligence to solve AGI. It turns out you can just get AGI from compressing general human knowledge.
2:07:15It's not that hard in in principle.
2:07:18Similarly, I think this is how we solve all disease. It's not that hard. In principle, you simply train the world's best foundation model to model all cell states uh and all interventions against cells. And then you do like an alpho type tree search against possible interventions to discover how to steer a virtual cell state from a diseased state to a healthy state and then generalize that to tissue and organisms and boom, you've solved all human disease. I I think that's like the endame. This is
2:07:47hyperpersonalized for you, right? You insert your DNA sequence, all right, and your current blood chemistries and all of that. And there's a incilicone model of your biology and it will tell you whether this drug works for you or doesn't. Seline.
2:08:02Yes. But I I I also just on on that, I don't want to like overromanticize the personalized aspect. It's an ideal virtual cell is as personalized for you as say if you feed a quote unquote personalized prompt to chat GPT. the output is personalized for you. Well, yeah, superficially it's a function of the inputs, but actually it's a generalist model.
2:08:24Salem, your thoughts, B?
2:08:25Well, this has been a trend. I'll go back to the biotech stuff, right? We've been we've been turning biology into information. When you turn something into information, it hops on the exponential curve. And we're seeing this uh go through live. uh and the uh the the you know which each of us have what 50 trillion cells in the human body right roughly um uh essentially when you can model that uh essentially a human being becomes a software engineering problem and we have really good
2:08:55techniques to navigate software read write understand etc etc and the phen amazing thing for me we've been we've done a good job in reading you know you have reading writing comprehension right when you're trying to learn a new language in in this case language of biology. We've done a pretty good job of reading. We've started to do writing with crisper and now these mRNA vaccines etc. This gives us a huge depth into the comprehension side with the digital twinning that can take place. So, holy
2:09:24crap, this opens up the door. Uh I would go with Alex's comment that uh um medical is cooked. Emod.
2:09:33Yeah, like this is kind of my hope for what the Genesis project would be like. Um, it seems very straightforward to me now that if we had a Manhattan project to cure disease through in silicico massive world human body models, cell models and organizing all our collective knowledge on cancer and autism and all these other things, we will definitely get a result. Like I'm not even doing that.
2:09:57It's not going to be a government program. The labs are going to do that for us. But I'm like, why don't we actually get together and get governments to put into a Manhattan project type thing and just have a straight shot at it and make all the data open? You know, this could be your point is well taken. At least the government is sitting on a lot of data.
2:10:14Uh, and the government could externalize all the data to the private labs like CZI and IDO cell and and others who are all building foundation models like make it a public good that they can all train off of a common crawl. Yeah, we've done that in the UK where you have access to this. Every government should follow suit. And this fits with what we talked about earlier with anthropic. Again, what are they going to apply their computation to? They will build a human cell model. They will build a whole body model, but I'd prefer for those to be public goods and us, let's cure cancer.
2:10:44Let's address all these negative kind of things and let's understand the body like never before. Again, that's much better than building an atom bomb even because it will have the biggest impact on humanity ever. Mhm. Dave, well, just for anyone listening who's not a biotechnologist and they're like, "Well, I'm not going to build a full cell simulator. I have no idea how to do that." You're thinking about it the wrong way. You heard earlier on the pod that we're looking at 10,000x expansion in AI with some other innovations that could be more like a million x. It's
2:11:12totally data starved. The full cell simulator is a way that it can design thousands or hundreds of thousands of experiments and get reasonably good test results back through a simulation rather than having to run millions of assays.
2:11:26That also applies in all kinds of other areas where every investment we've made in a company like Meror or Macato that is wrestling with new types of data to feed the AI, they're thriving and growing and making money and valuation.
2:11:40You know, Mor is worth 40 billion or 20 now, 40 by the end of the year. Uh they're just absolutely killing it. But every field of endeavor is going to be data starved. So no matter what you know, you probably know a field that needs to supply data back to the great AI. The full cell simulator is just the perfect biotech solution, but every industry has a solution.
2:12:03And so competitive, too. I mean, it's probably worth noting that this is from a company co-founded by David Baker, who shared the 2024 Nobel Prize in Chemistry with Demis for solving protein folding. This is the next big thing, next grand challenge arguably in biology/ medicine after now that structural biology arguably has been solved. Solve whole cell simulation and then you're halfway to solving all disease.
2:12:29Oh, love it. All right. To call out to our creatives out there, please send us your outro music videos. We're at a posity. Send them to media diamandis.com. Uh we love your outro videos. Want to see it. Uh, I'll pull before we go to our AMA. A quick note again, follow us on X, moonshots_pod.
2:12:50We're going to be putting out clips and putting out uh these uh these podcast recordings on X as well. And join us at the Moonshot Summit. Go to moonshots.com to apply. Gentlemen, uh, See and I have an AMA with the abundance community in 8 minutes. So, I'm going to suggest we speedrun uh the AMA if you don't mind.
2:13:11We should do the AMA in 8 minutes. We're going to do the answers. Answers.
2:13:16All right. Emod, you pick one first.
2:13:19Can't we keep making Frontier models more energy efficient instead of building all this new power at Bion 75?
2:13:25Yep. I mean, that's necessity is the mother of innovation. As we run out of energy, as we run out of RAM, you're going to optimize immensely and it'll be a big boon for everyone. Salem.
2:13:37Um, will we see an X-P prize aimed at solving the electricity supply problem?
2:13:43We actually proposed in the last visionary or a couple of visionarians ago a uh uh enough energy storage off-rid to keep a village or town um energy sufficient for 3 days. But it looks like the market will take care of that. Yeah.
2:13:59And regulatory is the problem. So it doesn't really serve as an enterprise where you need huge technology breakthrough. This serves better as a this is a regulatory issue and a market issue. Dave, over to you.
2:14:12I'll take four. If China has the advantage on power and the US has the advantage on chips, who actually has the real AI advantage? Definitely chips are uh you know power is a problem but we need 100 gawatts by the end of the decade. We already manufacture a terowatt in the US. So we're 10% of power will go to AI by the end of the decade. we'll get that far, then we'll be really desperate for more power. But between here and there, it's all about chips. Every chip, you know, that's why memory is up 5x. So that's the bigger advantage in the short run.
2:14:42Yeah. Alex, number two is for you.
2:14:44Number two asks, could interconnected micro grids popping up everywhere reduce the load on the main grid enough to matter? I I think I would invert the question, invert the premise of the question. It's it's not that the load on the main grid is going to be reduced.
2:14:59It's the exact opposite that there's so much economic demand for the compute and the compute needs so much energy. Before long, unless we fully externalize all of the compute to orbit and the Dyson swarm, these data centers are going to be generating a surplus of energy that can be pushed back onto the grid and driving utility prices negative.
2:15:19Yes. I mean, I want to I want to make that point. You know, if you're arguing against a data center in your backyard, you're arguing against lowerc cost energy and economic advantages for your community. Please understand that. All right. Uh uh let's go. Let's start with you, Alex, on this one.
2:15:38Okay. So, I'll take question number eight. At the current rate of improvement, how long until AI is more efficient per watt than the human brain?
2:15:47I think we're probably already there. Uh so, that that's a hot take on this one. people uh have this maybe like fetishization uh of the land hour limit that thinking oh well we we must really be far away in efficiency per watt from what biology is able to accomplish. Biology is actually wildly inefficient.
2:16:06The our our whole organism uh and and mammals in general were never optimized for compute. Whereas silicon uh and CMOS and whatever comes after CMOS, maybe it'll be photonix, which I know Dave is of of interest, it was optimized from scratch was designed for compute. I don't think the human brain is as efficient as many think. And the argument can be made that actually if you look at a per watts per task or watt hours per task basis, the the leading
2:16:35edge GPUs may actually be already more efficient than human brains, especially if you take into account the amount of energy required to train up a human over the course of 20 years, right?
2:16:46Yeah. Life lifetime total cost of ownership as it were.
2:16:49I'll put a pin and a correlary to that too. Uh people use the power difference as a way to explain that AI thinking is very very different from human thinking. I think that that people will soon realize that it's not that different at all. That the power difference will go away very quickly. But also this like yeah that's why it's nothing like us will also gradually go away.
2:17:11Uh I will take number five. Uh what's actually driving the rush? Why not keep energy growth at past levels and accept slower growth from its WHD and DRN?
2:17:23uh I think the the the easy thing here is intelligence is looking more and more like a general purpose input that will drive economic growth right and so having saying let's have less intelligence saying let's have less electricity or less internet it's it you want more of it and it improves research improves drug discoveries we saw all all today logistics everything so you want as much of it as possible there's also competitive pressure if one company slows down or country then others won't.
2:17:52So, it's not about accepting that capacity. You got to get your head around abund the abundance idea. All the thing we should be doing is accelerating energy abundance.
2:18:04I'll take the easy one. Number seven.
2:18:05Whatever happened to fuel cells?
2:18:07Actually, you know, Elon Musk started his passion in life was ultra capacitors originally.
2:18:12That was his Stanford thesis before he dropped out before starting.
2:18:16Well, whatever happened to those too?
2:18:17What what happened is lithium batteries worked far far better than anyone ever would have predicted. and they're still improving. So, it's sucked all the capital out of the other ideas. Uh, so that's all that happened.
2:18:27All right, Emod, close us out here.
2:18:29Yeah, if 71% of Americans oppose data centers, is this grid buildout actually going to help regular people's electricity or just make it scarcer and more expensive? David Harmon not 03. I mean, it's what you just said, Peter, right? It's going to make your electricity cheaper. More power is good. These things are not polluting. We need more data centers. We need more power.
2:18:48And we need to make sure it's all built right. Amazing. And you guys did it. You did it in 8 minutes. Uh, and Sem, we've got a whole 60 seconds to get over to our abundance community.
2:18:58I'm going to be two minutes late. I got to get like some a little bit of food before you in the meantime. Gentlemen, Jesus Christ.
2:19:05I love you all so much. This was such a fun today.
2:19:09Just so much and your brilliance. You guys are amazing. So, so proud to have you as our Moonshot mates. Alex, Dave, Immad, Salem, thank you always. Thank you to our listeners. Uh we love having you and hopefully you find this, you know, a way of keeping up with what's going on in the world because we are in an accelerating singularity. Uh and no time to sleep, no time to blink.
2:19:35Don't take Don't get fatigued like I did.
2:19:38Take care, guys. Be well. All right.
