Claude Opus 4.5, White House "Genesis Mission" & Amazon's $50B AI Push w/ Emad Mostaque, Salim Ismail, Dave Blundin & Alexander Wissner-Gross | EP #211

27:14Yeah, economy 3.0. Alex, please. I think all of this capital substituting for labor discussion misses an important point, which is these AI agents are arguably neither capital nor labor. They're a new third category. And everyone who's hand-wringing, and I hear this a lot, oh, well, how are we supposed to survive? Not everyone wants to become an entrepreneur. I would argue a near future where everyone survives by, quote unquote, becoming an entrepreneur misses the point entirely. It's not, I would expect going to be the case that everyone becomes an entrepreneur. Everyone's going to become an investor.

27:56The entrepreneurs increasingly are going to be these AI agents that are identifying and solving valuable problems. And humans, the average human, average unaided biological meat body human is going to be able to invest in fleets, in entire economies and indices of AI agents that are acting as the proximal entrepreneurs. This is the Accelerando premise also. Exactly. okay another another i think he might say something about this because he's been studying i am i'm gonna go to him next but this is you know this is accelerando as don't mean to butt in on the hosting uh salim you do a beautiful job as well uh but this is the accelerando playbook for people to read uh imad over to you pal you've been thinking about this very deeply yeah accelerando is a great book uh obviously my book the last economy is also great but um The problem is that it's going to be very difficult to out plan and out compute something like Claude 5.

28:57When it comes to coming up with businesses, unless you have skin in the game and you care. This is the main thing, because they will try things dynamically and they'll just move on efficiently. Whereas you can apply these agents to tasks. And the key thing is, in economic terms, it's all variable cost. Normally, when you had a company, I had to go and hire someone. That's a pain to do. You know, I had to launch my own servers before the cloud. Now everything is variable cost. And it's also cash flow positive because you typically pay the AI providers a month or two after when you have an enterprise contract and you charge people up front.

29:31So you can have brand new economic models where you're taking information, organizing and adding value to people. And I think that does close this rich-poor divide because you won't know where companies are coming from. And the compliance and everything can be done automatically now. We're actually seeing around some of these big AI startups, entire things that will do your tax compliance, that will do your financial forecasting, that will automatically balance payments and things like that. And the stack is nearly ready. Again, it's about a year away before you can launch a business probably in minutes.

30:04Amazing. With everything there. Can I mention something here? Of course. A plug for EXO here, which we stumbled across accidentally. Peter Reif quoted Jeremy Rifkin's book, The Zero Marginal Cost Society. Yeah. One of the, about three quarters of the way of writing the EXO book, we stumbled upon this economic kind of insight. When you're running a business, you worry about demand and supply, and hopefully the cost of demand and the cost of supply, hopefully you're on the right side of that equation. What the internet did, it allowed us to drop the cost of demand exponentially. Online marketing, referral marketing, every company's trying for a viral loop.

30:38If you get there, your cost of acquisition goes to zero, which is an amazing thing. We saw an initial wave of YouTube, Facebook, et cetera, explode out of the gate with that. What exponential organizations and new models have done is drop the cost of supply exponentially. So you think about Airbnb, the cost of adding room to their inventory is near zero. If you're high, you have to build a hotel. And with the launch of Amazon Web Services, you could take computing off the balance sheet and make it a truly variable cost. To Imad's point, everything now becomes a variable cost. You have almost no capital expenditure.

31:13So now you take out the denominator, the market cap explodes. And for the first time, you have a breed of organization that with low cost of demand, low cost of supply. And that's like a magical holy grail for business. And how we navigate that is going to be unbelievable over this next few years as this paradigm rolls out. I love it. All right. I'm going to jump into our next story. A lot still to cover. The ChachiPT introduces shopping research. It compares and searches and provides sort of recommended products that you're interested in. There's no question that this is coming out on Black Friday.

31:46They are moving this quickly. This uses ChatGPT Mini. And their claim is that they're able to get accuracy of sort of best predictions at what you want to buy up to 64%. So for me, this is about replacing the search engine, the affiliate blogs, YouTube reviewers, or Amazon's own recommendation engine. It's AI replacing the entire product research economy. In this one, at least from my perspective, Alex, I'm curious, the middleman's going to lose and the models are going to win. What's your thoughts? Critically, not just generalist models. This is a specialist vertical agent. To the extent that there was some expectation that we'd end up in a singleton near future where there's one generalist agent that does everything, it appears like that's not the case, at least to the extent that OpenAI is a leading indicator.

32:47By my count, OpenAI has launched at least two major vertical specialist agents. They launched research, deep research originally, which is general research, and they've launched coding agent, Codex. And this is the third-ish vertical agent, by my count, that they've launched other than the baseline model. And I think it's really interesting. Where are the generalist models? I mean, yes, they and other frontier labs are launching generalist models, but we're starting to see proliferation of specialist models. I think we're going to see many more. Wouldn't be surprised if we see more specialist post-trained models for finance and for medicine and for management consulting, just picking off broad industry verticals one by one.

33:33In this case, this is going after consumer purchases. But of course, I don't want to be calling on a particular model. I just want my AI to do this for me. Right. And yes, they're throwing a lot of resources at model routing and routers in general. So what what you'll gain with this umbrella router, I think, will be a single pane of glass, a single UX surface that you talk to. Love that. I have a question. Yes. How far are we from a, you know, Peter, you call it Jarvis, right? A personalized layer that watches on your behalf is totally secure from a privacy perspective and a sovereignty perspective and navigates the external world for you.

34:17So if you've got a shopping you want to do or you need to buy something or you need something that you may not even know you need, it's figuring out which of the agents to use and sorting it out. How far are we from that point? Now. Now. I mean, what you're describing, Salim, is, I would argue, like a computer use agent, a CUA. And Microsoft and other major companies and frontier labs already have CUAs that are either about to be rolled out or have already been rolled out and are in beta stages to do just what you described for a desktop. Now, I think what Ironman has is a CUA on top of heads up display and HUD, but that can come as well imminently.

34:57You see, what I find interesting here is the notion that we're about to give our AIs access to everything we read, everything we say, and our attention and intention. So as soon as we get our heads up glasses or augmented reality glasses that are able to not only have forward looking cameras, but actually cameras look back at our pupils to determine what we're staring at. If I'm staring consistently at that beautiful lamp over Alex's shoulder and my AI says, would you like a lamp like that? Or if I make some side comment to somebody else, it may purchase it for me and ship it to my house. So this ability to understand what we truly want by listening to our conversations or looking at where we look empowers this AI to become our magical shopping agent in many ways.

35:52Imad, how do you think about this? Yeah, so Andy Jassy, CEO of Amazon, recently said, I believe, that their agent Rufus, which I don't know any of us have actually used, has 250 million users. And they're estimating next year$10 billion in incremental sales from it, given conversion statistics up to 60 % higher. Who would have thought? I think that the key thing is, where is it? You know, like Bing and Teams and other things had access to the user's eyeballs. The challenge for ChatGPT and OpenAI here is how do you become that first intentionality on the shopping experience? And then what type of shopping is it?

36:32Because if it's toilet paper, who cares? You just want your AI agent to just do it. Automatically. Free ply or whatever, automatically. If it's super discretionary, some people apparently enjoy shopping. Maybe not, you know, people like us, but definitely my wife and kind of others. and then you have this middle bit like how many TVs do you really buy? You know, you kind of know what TV you're going to get. So I think that the key thing is who is the AI next to you? You go to Amazon for a shopping experience, you use a Rufus, you do a Google search, you now have Google AI mode up there. The key thing I think, you know, going to Salim's point is who comes up with the agent that's the most charming and engaging and licenses Paul Bettany's voice, you know, for a Jarvis, because that can then disintermediate everything.

37:15And that's what the fight is on for now. Fascinating. All right, let's move on. Let's get to Google here. Google further encroaches NVIDIA's turf with their new AI chip push. So Google has launched Ironwood TPU. It's their seventh generation AI chip with four times the performance of their previous version. And importantly, instead of selling hardware, Google is now offering their TPUs as a cloud service. For example, Meta is running on them without purchasing the TPUs. In the photo here, we have Thomas Kurian, who's the CEO of Google Cloud, who has been crushing it. Google Cloud's been doing amazing.

37:54And this puts them directly in competition with NVIDIA. Alex, how do you think about it? There's been so much hand-wringing, Peter, over NVIDIA's purported monopoly or CUDA as a purported architectural monopoly, GPUs are now finally facing healthy competition. We see TPUs that are being both purchased, according to this reporting, as well as licensed and rented. We see, obviously, AMD with their own stack. We see Trania and other Amazon chips and ASICs in general. And I think what all of this is turning into is finally accelerated compute is turning into a fungible commodity. It's not just a one supplier commodity.

38:39It is a multi-supplier, very healthy, very heterogeneous ecosystem of fungible accelerated compute, which is exactly the sort of competitive ecosystem we want to find ourselves in. Imai, do you have a comment? Yeah. So we use thousands of TPUs a few years ago. And from the V5s, this now has 10 times the compute and the chip size, the single die, the interconnectedness of the Google chips is beyond anything you've seen. So you've gone from 64 in a unit now to I believe 9 ,000. So what Google's really, really good at is connecting lots of chips in one place and even multi-data center. We had runs of up to 50 ,000 of their low energy chips.

39:25And what that's important for is context. So right now, actually, DRAM prices have gone up by about five times. So if you want to get the DRAM for your gaming PC, it's gone up crazy. Google actually has the ability to use cheaper chips in massive scale to do large context window things. And that means that Gemini has a million, two million input tokens from video to audio to others, whereas it's still limited on other GPUs. And that's going to become even more of a difference going forward. Google originally built these chips to power Google Search. And now it's matured to a point where they can offer it to everyone, even hosted.

40:04So the cloud service has been available for a few years. But now they're exploring actually saying, meta, you want it in your own data center? We can look at that. And that's going to be super interesting going forward, particularly as RAM versus flops becomes the key differentiator in terms of performance. Because context becomes almost everything. Because the models are already really fast, to be honest. Just kudos to Google. I mean, they continue to crush it week over week. Two plugs for us here. About three months ago, we said two things. One is that Google would inevitably start to lease or sell the TPUs.

40:41And here we are. And second, I believe Imari was you a few months ago on the pod that said, invest in DRAM companies because DRAM is going to become the short supply, etc. So it seems that we're typically three months ahead of the game. Amazing. Amazing, Salim. That's great. 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.

41:23The Blitzy platform provides a plan, then generates and precompiles 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 pre-IDE development tool, pairing it with their coding co-pilot of choice to bring an AI native SDLC into their org. Ready to 5x your engineering velocity? Visit Blitzy.com to schedule a demo and start building with Blitzy today. All right, let's move on to the next story here.

42:06Amazon is spending up to$50 billion on AI infrastructure for the U.S. government. So it's projecting it'll add 1.3 gigawatts of new data center capacity, beginning construction in 2026. So what's the story here? AWG, do you want to take a shot? Yeah, the government clouds, in many cases, including with AWS and otherwise, the government has its own availability zones. And they're notoriously undersupplied when it comes to accelerated compute with with GPUs. And I think it's sort of surprising, given that the public sector is depending on how you count either half the economy or about a quarter of the U.S.

42:48economy, that how compute starved or at least how GPU starved it's historically been. So I think this is a welcome investment, at least from from my perspective. We want a vibrant public sector and vibrantly supplied with accelerated compute public sector. And I view this as a very positive step in that direction. Nice. Some of the stats here in this article, AWS is serving 11 ,000 government agencies and expecting to spend$125 billion in capital expenses by end of 2025. Massive support, right? AWS has really just dominated. Is this essentially the federal government saying AWS is their cloud provider?

43:29That's a big deal if that's the case, because that's what it seems. Well, the government, the U.S. government has multiple cloud providers. This is pretty well publicized and reported on. But Amazon is an AWS key supplier to U.S. government cloud resources. All right. I'm going to move us along here. Also, another article on Amazon here that their data center tally tops 900. And we forget the fact that Amazon, because of AWS, has been running a massive number of data centers around the world in over 50 countries, launching now something in Indiana. It's a 1 ,200-acre data center, and they're putting it up and getting it online faster than anybody else.

44:12Any particular thoughts on this one? I was struck by the fact that the Indiana one uses 2.2 gigawatts of energy. That's like an unbelievable amount of energy for a data center. That's a small country with power. Yeah, I would just maybe note we're tiling the earth with compute. That's what we're really talking about here. And this is just the opening act. And the Indiana data center in particular is, we were speaking about Anthropic a few minutes ago. That Indiana data center is the core computing facility for Anthropic, both for training and for inference. It's called Project Rainier. And it was farmland.

44:52that that was converted almost overnight. I mean, it took about a year, but almost overnight into modern compute. This is 1939 when you see farmland in the Midwest being converted to compute resources. Yes. Alex, you can't imagine the number of comments I've had from people saying, what does Alex have? Why is Alex against the moon? From our last podcast, isn't it obvious? The moon's had it coming for years. We have an AMA section that we're going to hit in a few minutes, and that's one of the questions being asked. Don't we need to save the moon? It's lunacy, Salim. Lunacy. Touche, touche. Oh, goodness.

45:34All right. The third Amazon story here is Amazon opens$11 billion AI data center in rural Indiana. We've heard about this already. It's running 500 ,000 Tranium 2 chips. So how does Tranium chips compare to the TPUs, compare to the NVIDIA's GPUs? What do you guys think about this? So we used a bunch of them previously. Tranium 2 chips are equivalent to the hoppers, and they're good for inference, but they're much more difficult to do the large-scale training runs on. But if you look at the breakdown now, you have a core cluster for training, and Anthropic just announced another big NVIDIA deal,$10 billion with Microsoft and NVIDIA.

46:13but for serving up of Claude, you always hit those capacity constraints. And Tranium's very solid for inference, similar to how previously Amazon went all in with Graviton, which was their CPU equivalent. And now that runs massive workloads for Netflix and everyone around the world. So I think that it's still one more generation until Amazon starts to catch up. Again, they're about a generation behind, but all those chips are going to be used, but probably for inference versus actual training. I have a crazy question here. So if your model is this closely bound to the chip, then if you did an inference model for any of these big hyperscalers on Tranium versus TensorFlow, would you get a very different result because the chip is different?

47:03No. So you typically use a framework like OpenXLA, which automatically translates it to different things once it's actually doing the inference because the process of inference is quite straightforward matrix multiplications. The process of training, the training can be really complicated in the way that things move back and forth, et cetera. And that's where you really need to have high resilience, high interconnect, whereas if it's a single chip or a group of eight to 16 chips as these are, they're just doing forward passes. It's a lot easier to code and to have speed on. But again, there are certain things like Cerebris, for example, that will give you much faster inference or a highly optimized Grace Blackwell, et cetera.

47:48So that's much simpler than training. Maybe to expand on that, backprop is the key problem. If we could do away with backpropagation at training time and have some sort of magical, like I remember Boltzmann machines were one sort of concept for how we could do away with global backpropagation. If we could do away with backprop entirely, then one could imagine a near future where training looks a lot more like inference and training would be a lot more portable and a lot more parallelizable. But no one has yet in production figured out how to do away with backprop. But aren't LLMs fundamentally anchored to backpropagation?

48:24A training time. A training time, not an inference time. Inference time is only forward propagation. So if we could figure out how to train - But backpropagation is fundamental to the training. Backpropagation is fundamental to training of neural networks. For now. But there are lots of paradigms. There's a whole cottage industry of researchers trying to figure out ways to eliminate backpropagation entirely. If we could eliminate backpropagation, that would certainly eliminate a training time compute bottleneck. And by the way, just as a reminder, if you're listening and you've just heard a conversation that you think is being spoken in Greek, then my suggestion is take the...

49:02Join the club. Take some notes and go to your favorite LLM and have a conversation. As Bruce Willis said in Die Hard, welcome to the party, pal. Yeah, I'm going to hit what you said earlier, Salim and Alex. I think the most significant thing about this is going from farmland to seven buildings in one year. So big. 2.2 gigawatts. I mean, it's just the beginning and we're knocking down regulations and capital is flowing in. This is continuing. All right. I want to get to our AMA. I want to hit a couple of stories on the science side real quick. We've been talking about launch costs. We've been talking about launching data centers.

49:40We've been talking about going to the moon. I want to give folks a little bit of overview for a moment about how quickly the cost of launch has been changing. So the space shuttle, which was originally supposed to cost about$50 million per launch and launch 50 times per year, ended up costing somewhere between a billion to$2 billion per launch and was launching anywhere from one to four times per year. Massively expensive,$50 ,000 per kilogram, super high cost. Falcon 9 comes in, drops the cost at least 20-fold to$2 ,500 per kilogram. by making the first stage fully reusable, right? It's got nine Merlin engines.

50:24So you're recovering most of the engines on the Falcon 9. And then here comes Starship, which is reducing it again, another 25 fold,$100 per kilogram. So, you know, how many kilograms do each of us weigh? And what's your cost to get into orbit? It becomes affordable all of a sudden, right? So as Starship becomes fully reusable, but, and then Elon comes and starts speaking about the work of Gerard K. O 'Neill, Jerry O 'Neill at Princeton University had actually designed and built, at least on the ground here, what are called mass drivers, electromagnetic rings that accelerate a bucket to lunar escape velocity.

51:04And just for the cost of electricity, which by the way, on the moon is relatively cheap because you've got all the solar flux, you can accelerate something and shoot it towards the earth into a earth acquisition orbit. And we get here the price coming down, not a hundredfold, but a thousandfold to 10 cents a kilogram. So all of a sudden we gain access to all the resources on earth. I like to remember, remind people that everything we hold a value on earth, metals, energy, real estate, all these things are in near infinite quantities in space. So the nine-year-old space geek in me is like super excited of what's coming.

51:45Alex, you want to add anything? Yeah, I'll add that disassembling the solar system is going to require low cost to orbit. So this is great. Alex, you're going to start protests outside our front door. The Dyson sphere in the way. I like this is generated by Nano Banana as well. You can see this little thing. Yes, of course. All right. I mean, basically, this turns lunar launches and rocket launches into software, right? That same loop is now hitting this. It's very, very, and it's one thing, two things here. One is importantly, this is a log scale. So for folks watching, this is like ridiculous orders of magnitude per level.

52:25And that's unbelievable in a very physical environment. This is not some social media gaming Silicon Valley play. This is getting out of the physical gravity of the Earth's gravity. Well, this is nuts. This is energy, baby. You know, the one complaint I have about my conversations with Elon is he wants to get out of Earth's gravity well and then go directly back into Mars's gravity well. You know, I'm far more interested in staying either in the Earth moon system or better yet, build what some have called O 'Neill colonies in which you are basis. I'm not disassembling the moon, Alex. I'm disassembling the asteroid belt.

53:05All those pesky asteroids deserve to be disassembled and used. I mean, sure. If you want to start with the asteroid belt, we can start there. That's fine. Training wheels for solar system disassembly. That's great. All right. Listen, so disassemble asteroids and build large rotating cylinders called O 'Neill colonies where you live on the inside, you know, omega squared R. All right. One more story before we get to our Q &A, which is a story from a friend, Matt Angle, the CEO of Paradromics. So Paradromics has been one of the significant number of BCI companies, brain, computer, interface companies.

53:42And what's interesting about them, they've just completed their testing in sheep. Neuralink did their testing in macaque monkeys. Paradromics has done their testing in sheep. And they've been approved to go into humans, which they will do in about two months' time, early January, February timeframe. frame. And I think what's most interesting is that they've been able to hit a speed about 10 times or actually 20 times faster than Neuralink. So Neuralink's been at about 10 bits per second. The Paradromics implant is at 200 bits per second. So Salim, you and I've always talked about is Ray's prediction on high bandwidth BCI by the early 2030s, 2033 going to happen.

54:32So we're seeing all these companies moving forward here. A few years ago, I was a hard no on that. Yeah. And now I'm like, oh, shit, he's right again. Alex, what are your thoughts here, buddy? Yeah, I think we're seeing the BCI space become competitive, which is great. Yes, we should all get our Ray was right hats. Fine. But I think if you extrapolate this, one of my more fun thought experiments is when do we actually get our nanobots in the brain for high throughput or effect process type BCI? And it's interesting. You can look at the cost of producing a gigaflop of compute versus the typical size of a gigaflop of compute.

55:18So when Apple introduced the iMac, originally first iMac was like about a gigaflop. the first iPhone, first Apple Watch, there's something magical about rolling out a form factor with about a gigaflock. You extrapolate out that curve, naively assuming exponential progress for a gigaflock, saying like gigaflock, that's the threshold at which we have useful general purpose computing, including for the purpose of maybe even substituting for human brain cells in the context of high throughput BCI and or a very invasive uploading scenario. that that curve you get 2045 uh which is again hashtag ray is right hat that's when you get about a gigaflop the size of a human brain cell so i i do think we're very much on trajectory for for ray is right style human mind uploading and uh and invasive bcis and non-invasive this isn't this is a quasi-invasive bci i think we're also going to get like lots of wear by the way we have Ray coming on.

56:18Yeah, Ray's going to be joining the pod in early January. Talk about his predictions for 2026. Also, we'll have Brett Adcock coming on the pod to talk about. Wait, wait, wait, wait. We can't ask Ray about 2026. We've got to ask him about 2066. I mean, it's too soon. It's a waste of time. Too soon. We'll ask him about all of it. We'll ask about all of it. We need a bigger podcast. We'll get one by then. By the way, we crossed 400 ,000 subscribers. So thank you to all those who subscribe to push us over. Our next hill is 500 ,000. Then we're going for that million. Why? Because Jet and Dax want us to get a million subscribers.

57:03Whatever chemistry we have here in terms of processing the news and making sense of it for others, it's seeming to really resonate. The number of calls and accolades and kind of feedback I'm getting. I'm sure you're seeing the same thing. Thank you to our listeners for the feedback. We do read all of your comments. And in fact, we process the comments and pull out the questions. We're about to jump in to that segment with an AMA, but Imad, what are your thoughts on BCI? I think that this year has been a breakthrough year. Next year, you'll see even bigger advances. We've both seen what else is going on behind the scenes.

57:41And I think it'll probably one of the biggest investment areas in the next three years, actually. Because what could be better than solving the issues, but then augmenting humanity directly? I think, as Elon said, the only way you're going to be able to keep up with the AGI's is to plug in. Yeah. And so it's going to be a geostrategic importance as well as a financial importance. There's something fundamentally interesting about the brain, because we still really have little idea how it works. But as long as we can interface with it effectively, that's very, very powerful. Our memories are already now outsourced to our smartphones.

58:14We don't really use our memory neurons in the same way we used to. And therefore, we'll start doing that with more and more brain function capacity and releasing that load off that and using it to other things. So I'm really excited about what comes with this. Maybe just to add quickly to Salim's point, this is admittedly a bit of a hot take, but arguably we solved AGI, we solved superintelligence without actually having a good mechanistic understanding of natural intelligence. I think it's pretty likely we're going to solve brain-computer interfaces and maybe even whole brain emulation without actually still having a detailed mechanistic understanding of the human brain.

58:53You can get pretty far with phenomenology. Alex, do you think we can use AI to solve the hard problem of consciousness, the whole qualia thing? Yes. Okay, we're going to have a conversation about that in terms of how that goes about, but let's take that offline. All right. Two quick points on this PCI. Number one, amazing people playing in this space, right? Max Hodak, who is the co-founder of Neuralink, now has a company called Science. Go and check it out. They have a completely different approach to interfacing between the compute world and your neocortex. Brilliant, basically using neural stem cells to grow nerve endings into the brain that wire together and fire together.

59:40And then Sam Altman invested in something called Merge Labs. It's still kind of under wraps, but we'll be hearing a lot more about Merge in the next few months. So I have one final question. Ray's prediction on high bandwidth BCI is really dependent on having nanotechnology. And the question is, where are we on that front? I'm still waiting to hear some good updates on the ability to assemble molecules atom by atom, not with wet nanotechnology, which is biology, but assemblers like Eric Drexler spoke about. Alex, any thoughts there? I spent so many years chasing nano assemblers. I do think we're going to get to Drexlerian style, although even Eric Drexler had sort of a personal evolution.

1:00:33We've chatted, I've chatted a number of times with him about this from sort of pure diamondoid style, quote unquote, molecular assemblers to then there was the nanosystems phase where it's not about self replicating nanorobots. It's more about desktop factories that produce things. I here's what I think. I think by at the very latest, and this is in my mind, like an ultra conservative outer bound 2045, we get our Drexlerian nano assemblers. I actually think we're far luckier to get them in some soft form, maybe to look like DNA origami, maybe to look like AI solving the Feynman grand challenge, which includes both computational and nanorobotic challenges.

1:01:14is I think we're likely to get some AI solution to Drexlerian early style, Drexlerian nanotech in the next 10 years. I don't think it's going to take that long. There's our 2035 date. Yeah, like everything gets solved in the next 10 years. I don't think you need to have that high bandwidth, to be honest. Like we did work at Stability on Mind's Eye, where we reconstructed images people saw from MRIs, which is incredibly low bandwidth. And if you look at the forward-backward diffusion pass processes, what you're likely to have is before you get to the full bandwidth, you'll have partial bandwidth that can effectively reconstruct brain processes with very little information.

1:01:53And then you'll just run diffusion models to do that in a similar way to actually Sunday Robotics and others have done things going forward. There was a project out of Japan called Dreamcatcher. and what they were doing was having you sleep in an MRI machine and they're storing the images coming off your optical nerve and then replaying your dreams back to you the next day, which was hugely unnerving. Very quickly on this one, Peter. With fMRI, you get approximately a million voxels per second just streaming off. You can do high bandwidth decoding of thought with a million voxels per second. I just don't have my portable fMRI machine to carry around.

1:02:33Yeah, but those are getting smaller. By the way, there is a team that I've been talking to that seems to have a credible path for molecular manufacturing. So you're happy to connect them with me. I can't wait. All right, let's get into some of our questions from our subscribers here. Let's jump in. The first one is from David Bowman, 6224. David says, I'd like to hear AWG tackle EMOD's thousand-day prediction. So what does AWG think of Imad's AGI in a thousand day prediction? So Imad, do you want to state your prediction first and then love to hear Alex's commentary? I was just saying that I think that most human economic work is negative value within a thousand days.

1:03:21Well, 900 days now left at most. And not that it will replace all the jobs, but definitely it'll be there for just any job that can be done on the other side of a keyboard or mouse. So that's a weaker version of AGI than in some cases. Alex? Yeah, I think the central challenge, as always, is defining what we mean with AGI. I think if AGI means generality, I think we've had AGI since at the very latest summer 2020, when GPT-3 and language models are few-shot learners paper came out. If AGI means economically, some sort of economic parallel with humanity. Yeah, I agree that either it is the case, Schumpeter style, that we already have some sort of economic generality, for example, as parameterized by OpenAI's GDP VAL benchmark.

1:04:13If you believe that benchmark, economically general AI is either already here or imminent like the next few months. Or if you have some other preferred benchmark for human economic output is probably imminent, if not already here. All right. Let's go to the next question from Josh. Insert my standard rant about AGI. Okay. So, so, so. Incorporated by reference. So acknowledged. Thank you. So at Josh S 5937 says, what is the future of land ownership in a future without scarcity? Land is finite. Will it remain the final scarce resource? So Josh, it's a good question. The way I answer it is two different ways.

1:04:55Number one, we're going to be spending a lot of time in the virtual world. And there you'll be able to gain access to unique virtual real estate. The second is you're thinking with a very Earth-centric point of view. There's the moon, there's Mars, there are massive O 'Neill colonies built out of the asteroids. And we're going to start to see humanity migrate outside the Earth. Having said that, yes, Central Park West apartments are still going to be scarce. Deeply disagree. I want to rant on this. OK, go for it, Salim. So I did some fact checking here. It turns out there's about 16 billion acres of habitable land on Earth.

1:05:35That's about two acres per person. OK, that's a pretty decent number. And that's habitable. Let's note that passenger drones are going to make difficult to reach areas very habitable. So that goes up to about 20 billion to 22 billion areas of habitable land. So technology will expand the amount of habitable and reachable land. Still, if we get to about 10 billion, we'll peak at about 10 billion population by 2050 before we start dropping off. Still about two acres per person, which is a pretty decent number. And all of the technology is allowing us to reach that land more easily, make that land more usable.

1:06:16And if you fly across India, the most populous country in the world, it's mostly empty. You see populations at the edge of the, on the coast of the middle is kind of, there's nothing there. Same with Africa, same with the U.S. You fly across the U.S., there's nobody there in the middle. And I'm Canadian, there's nobody in Canada. So there's a lot of land that we can use and technology makes it much more accessible. Temperature, HVAC, heating, cooling. The only constraint is energy and compute, as Alex would say. Amazing. That's a great point. Go ahead. I'd like to just give a practical example to listeners.

1:06:51Waymo is now basically legal across San Francisco, right? And so that could completely just change where people live because you can just get into Waymo and it will just take you and your kids anywhere. Sorry, just to add to that, the prediction for Tesla is it will be about 30 cents a mile to get somewhere on the robot taxi. That's like near zero compared to 10x drop from where we are now. Alex, and you should go. Yeah, maybe just add a bit of nuance to this. I think in the short to medium term, land is becoming post-scarce. As you say, we can build up, we can build down, we can build on other planets.

1:07:30Important use case that hasn't been touched on. We're going to have so many humans, I think, uploaded in one form or another into the cloud. The cloud doesn't have the same concept of land. So I think short to medium term, land is post-scarce. in the long term, I think the scarcity of land depends on whether AI economies have a better use for land than we do. If we do find ourselves taking apart the solar system, land could actually become really scarce in the end. Yeah. By the way, let's just talk one second about uploading. I mean, when do you actually believe we're going to start to see human uploads to the point where you, Alex, say, okay, upload me.

1:08:08And there's this speaker that comes over, you know, this voice comes over, says, hey, Alex, I've been uploaded. You can off yourself now. We don't need your biological body anymore. I'm in the cloud. Thanks for the vote of confidence. I think we've already seen non-invasive uploading in the form of large language models. Large language models are arguably sort of an upload of an ensemble of all of humanity. In terms of like individual uploading that's non-invasive, I think we're either there already in the form, you know, Imad touched earlier on or alluded to, like, constructing foundation models from fMRI scans.

1:08:48There are a number of groups that are training foundation models from fMRI scans. Arguably, those are like low fidelity facsimiles, but non-invasive of human minds. I think we're going to get to - Wait, wait, wait, wait, wait, wait, wait. Hold on a second. There are people training LLMs on fMRI scans? Correct. A number of groups now. including meta by the way really well financed really talented holy crap okay so so the real idea about uploading the implication is it so llms are trained to to reproduce sort of behavior of humans like fat biological meat fingers tapping keys on a keyboard uploading text to the internet but with with foundation models trained off of fmri data like a million voxels order of magnitude per second, you can imagine pre-training a foundation model that basically encapsulates human thought, certainly for human thought decoding purposes, you get that.

1:09:47And fMRIs can track a single neuron firing in real time. No, fMRIs are both spatially low resolution, temporally low resolution. You get like one to two second temporal resolution, approximately one millimeter cubic spatial resolution at best. But nonetheless, turns out to be enough for thought decoding. So, Alex, the concept around a true upload is can I actually map your connectome? Can I map for a human roughly not only the 100 billion neurons, but the 100 trillion snap the connections typically done by slicing the brain into ever thinner slices and using AI to map those interconnections there?

1:10:29It's a destructive process. Do you think we're going to have a message right now? Yes, very invasive. Here's my brain, slice it into a thousand pieces. I think we're going to have reference non-human organisms. So Drosophila, major progress. Was done. Food flies, done. Mice, bound to be done. There have been a few one to three cubic millimeters of mouse brain scanned in some form or another, to the extent that the connectome is a proxy for uploading done i think mice overall going to be done shortly lobsters next right lobsters are easier interestingly mice are harder i think we're going to see the the full human high-res connectome probably in the next five years alex aren't you an advisor to necktome i'm not a formal advisor to necktome but i am an advisor to a company eon systems that is working on solving human whole brain ambulation and uploading all All right.

1:11:30I'm going to move us forward here onto our next AMA question here. And then I want to close out with what are you most thankful for in 2025? So start thinking about that in background. So at Success Coach Cody writes, how do we prevent a world where millions fall into poverty before AI driven abundance arrives? What are the real solutions for people who may lose their jobs long before long term benefits of AI kick in? And at JNKind5 asked a similar question. How often you often say AI will lift up people at the bottom. How exactly will that happen for those who can't meet basic needs like food, health care today?

1:12:14You know, we hit on this about two pods ago where there's concerns that we saw this out of the data from the FII9 event. concerns about poverty, about losing jobs, about being able to support your cost of living. Salim, you want to jump on this one? Wow, a ton of, this is why 20 questions buried in each of these here. I think we are in a difficult kind of 10-year period when we transition all of our world systems from scarcity to abundance, right? Consider the fact that almost every business in the world is focused on scarcity for the last 10 ,000 years, if we didn't have scarcity, you kind of didn't have a business.

1:12:56We're moving now to abundance models and actually exponential organizations find business models around abundance, which is the starting point of that transition. But for society at large, we need to move to some model, whether it's UBI or UBS, universal basic services, is a similar type of concept where you just give basic capability and make that available to everybody, solve the bottom two layers of Maslow's hierarchy. And the trick for UBI, by the way, for people that are naysayers is if you can find the balance where people can survive but not be happy, you still have a very thriving economy.

1:13:35Entrepreneurship explodes in that model, et cetera, et cetera. So we're not far away. The problem is governments and getting governments to move from a union labor job taxation model to that is such a big leap. We don't have confidence in governments doing that. And the problem with government we have all over the world is they want to be needed. They ran a two-year UBI in Manitoba in the 70s, and it was so successful that at some point the government realized we're not even needed here, and they canceled the program so that it could be needed. That's the immune system problem in government that has to be solved.

1:14:09Just a quick thing for all the folks that emailed me saying, hey, how do I find out about that? I'm putting some stuff together. We'll send it out shortly to everybody. So I tend to be in the optimistic side. Technology uplifts people at the bottom, people leveraging technology to make more and more money in the short term. We've got lots of data around that. And as we get technology democratized and demonetized to a broader population, then everybody lifts up. And you, Peter, you and I talk, all of us talk all the time, but forget the richest people. If you can lift the bottom, that's the key.

1:14:41And the bottom is being lifted very, very appreciably. You just don't see it that obviously. Yes, people compare themselves against their, you know, their Kardashians or whomever else. Imad, you've been doing incredible work here with intelligent internet on this specific problem. Could you sort of lay that out and give us your thoughts here? Yeah, I think as with many things in human life, this is a coordination problem, right? Again, we have enough resource, two acres per person, you know, food, healthcare, et cetera, to coordinate everyone. But we've always lacked the capability to do so because our systems are dumb.

1:15:18So we have projects like SAGE, which we launched at FII to do top-down policy. And really, the way that I've been thinking about it more is like AI social scientists. You know, like we talk a lot about AI scientists for biology, for chemistry, for quantum. AI social scientists to figure out economics, politics, implementation are going to be so huge. That's basically our SAGE project. On the other side, I think you need to have universal AI given to everyone, a Jarvis that's looking out for people to help them navigate on an individual basis. Because that's how they get access to food, healthcare, etc.

1:15:51The reason they don't now is because people are invisible, particularly the poorest of people. But the pace at which this is going to come over the next few years is going to be so intense that governments need to take a big step forward and say, A, we need to use AI to coordinate this. B, we need to get AI to the people. And C, we need to look at historical counterparts. And I think probably you need to look at the 1933 New Deal that came out of the Great Depression and others. Because you might see entire industries disappear within a matter of days, months. Like Grok 4.1 fast just scored like 95 % on Tao Bench, the customer service benchmark, and it's 50 cents per million words.

1:16:32Better than any human. That would just, no customer service jobs within two years, you know? Again, it takes a little while, but it's one way. So coordinate with AI, give everyone universal AI, and then layer services and coordination on top of that. You know, you're going to appear as the headline in some user article now. Now, Imad, if you remember, you said no more coding in a year, like headlines across India. I think the point that Imad makes, the really killer point that Imad makes right there is this is one way. We're not going back. Yes. We have to face the future that's coming and let's get real about it.

1:17:06Let's get data driven and evidentiary around it and just freaking make it happen because it's not going to, you know, left to itself. We've got these two futures, a Mad Max future or a Star Trek future. Right. And you can see our politicians pulling us straight to Mad Max. We have the opportunity with technology to pull us in that direction. This is what this community is about. This is what we have to do. Alex, closing thought on this one. Yeah, closing thought is I think the central policy challenge is growing the overall economy much faster than the value of conventional human labor is destroyed or obsoleted by AI.

1:17:42So I'm primarily focused on ensuring that we can achieve radical macroeconomic growth. If we can do that, then making sure that UBI, UBS, or UBE, universal basic equity, or some other variant thereof, some door number four, I think those all become more a matter of policy decisions. But it's relatively easy to distribute abundance if we have abundance. Yeah. All right. We're going to close out on a question aimed at you, Alex. I'm going to coin a phrase here. It just occurred to me, UBA, Universal Basic Abundance. There you go, Peter. Okay. I love it. I'm thinking about that. And then that gives you abundance of very interesting things underneath that that's solved for all the others.

1:18:24Love that. All right. The final AMA. And please, if you're listening today and you say, I've got a question, put it in the chat on this particular episode of Moonshots and we'll look for it. And if it's intriguing enough, I'd love to ask it to the moonshot mate. So at xphenix96 asks, hey, Alex, the moon and Jupiter should be off limits to mining. Don't they stabilize our environment? What are you trying to do, Alex? Start a reply. We're having in all these worlds are yours, except Europa attempt no landings there moment. I think if you've read 2010 by Arthur C. Clark. um no i i i'm i'm i so many thoughts first is no we don't need to stop mining the moon and jupiter to stabilize our environment jupiter does at the moment play an important role in protecting the inner solar system from orc cloud bodies and and other objects from the outer solar system the moon does play for the moment an important role in the tides and and other sort of atmosphere and romantic love for the moment for the moment is is doing the heavy lifting in that sentence so once we have the ability which i think seems likely we will increasingly have to disassemble the moon and disassemble jupiter and assuming the solar system does go in that route we will also have the ability to protect the inner solar system from variety of asteroidal bodies and to recreate the tides artificially i my favorite quote from you alex is saturn has had it coming for a long time that's got to be an all-time alex quote it's true oh goodness all right well asteroids uh represent a significant amount of mass and i think they can handle our needs for at least a decade or two so um all right uh if you like that i want to close out with a question here what are you guys grateful for having happened in 2025 uh as a closing gratitude.

1:20:26I'll kick it off. I'm super excited that humanoid robots have made so much progress and they're a real, the capital is being invested, the manufacturing plants are being invested, and my own version of data or C-3PO is on its way. Alex, how about you? So many things, but I'll pick one. I'm grateful that math is credibly and defensively being solved by AI. That is, in my mind, such a canary that this is going to work. The singularity is in progress. We're going to solve all of the grand challenges to math, science, engineering, and medicine over the next few years. And math is just the tip of the iceberg.

1:21:08It's very exciting. Amazing. Salim? Again, a million things. I think three things pop to mind. One is I'm unbelievably grateful for this podcast, Peter. Thank you for pulling us together. Oh, I am too. Missing Dave a lot at this moment. And thank you. Just let's just say it real quick to Nick, to Nick Singh, to Dana, to Gianluca, who helped this really be excellent. So thank you guys for that. I think this is like this radically optimistic, realistic view of the future is the most important kind of tonic for what's happening out in the world today. and there's kind of palpable relief from all the listeners going, well, thank God there's something I look forward to every week or a few.

1:21:55That's number one. Number two, I think I'm kind of starting to just wallow in gratitude on a near permanent basis to just thinking about what the incredible future that is appearing in front of us driven by that inner loop that Alex talked about. I'm still a fan of the moon for the moment. So let's, we don't need to enjoy it while it lasts.

1:22:19I think the third, the third would be my EXO ecosystem is kind of finally gelling in a really powerful way. It's been like 10 years building this ecosystem. If I ever say in the future, I want to build an ecosystem, please somebody get a baseball bat and take me behind a woodshed. It's unbelievably difficult, but it's actually now coming together and in a very, very powerful way. There's a whole bunch of announcements that we have. And finally, I'll do a plug. We're doing this meeting of life session where I will claim to answer why we are alive and how we live effectively. And so the link will be in the show notes for everybody.

1:22:54Tickets are selling fast for that. Nice. Imad, would you come out on your gratitudes? Yes, that's a nice small one, Salim. It's a small question you're answering. I think that there's two big things. I go for niche projects. Yeah, I think there's two big things. One is I think we've had the technological breakthroughs and infrastructure breakthroughs to be able to build the AI social scientists to improve our infrastructure and finally coordinate as a species. And that is a huge thing that we'll start seeing rolling out and announced next year as well. And number two, I think minus the hard light, we have all of the tools we need now for the holodeck.

1:23:29Ah, awesome. We just got to put that together. I'm going to add one final gratitude to close us out here, which is the incredible progress being made on reaching longevity escape velocity. Right. The the focus by all the hyperscalers and and model builders on how do we understand how to add decades of health into our lives? As Dario says, how do we double the human lifespan in the next five to 10 years? That gets me jazzed. You know why? because i'm excited to see the star trek future coming our way we're going to close out uh if you're listening to this versus watching it go to youtube to watch this incredible uh outro music and video by john novotny uh john again john again you're going to see all of uh your favorite moonshot mates as star trek characters here of course is the opening scene with awg as a vulcan all right as a blonde Vulcan with a ponytail blind Vulcan with a ponytail all right let's check it out and Salim once again you look hot here buddy you look hot all right enjoy beneath the silver moon we hear the wild winds call the map is torn to pieces but we're setting out regardless of it all

1:25:00As long as I get a phaser.

1:25:22Never fear the fall. The storm may rise to test us, but we'll meet it on the spot. And Holmes always finds the ones who take moonshot.

1:25:45That was epic. I just don't like wearing a red shirt on some of those planets. Yes, I don't know. If I can be likened to Picard in any way, I'm good. And Alex, of course, you're the science officer on all the missions here. Obviously. Everybody, I wish you an incredible Thanksgiving holiday to all our listeners, to my moonshot mates. Dave, we missed you on this episode. Looking forward to seeing you. We're recording again early next week. A lot going on. And we're going to be spending some time with Mustafa Saliman as well, the CEO of Microsoft AI. and we're doing a podcast with him. A lot of incredible things.

1:26:26Get ready. 2026 is going to rock the planet. Hopefully not physically, but definitely emotionally and intellectually. Let's all wallow in gratitude the next few days. Yeah, beautiful. And stuffing and turkey. All right. Take care, everybody. Take care, folks. Every week, my team and I study the top 10 technology metatrends that will transform industries over the decade ahead. I cover trends ranging from humanoid robotics, AGI, and quantum computing to transport, energy, longevity, and more. There's no fluff. Only the most important stuff that matters, that impacts our lives, our companies, and our careers.

1:27:03If you want me to share these meta trends with you, I write a newsletter twice a week, sending it out as a short two-minute read via email. And if you want to discover the most important meta trends 10 years before anyone else, this reports for you. Readers include founders and CEOs from the world's most disruptive companies and entrepreneurs building the world's most disruptive tech. It's not for you if you don't want to be informed about what's coming, why it matters, and how you can benefit from it. To subscribe for free, go to dmandis.com slash metatrends to gain access to the trends 10 years before anyone else.

1:27:37All right, now back to this episode.

1:28:15We'll be right back. update on Dave. It's up to you. We'll take the laundry. Rinse. It's time to be great.