Claude Code Ends SaaS, the Gemini + Siri Partnership, and Math Finally Solves AI | #224

Claude 4.5 is making waves. That is game changing. Opus 4.5. It is actually incredible. It's the best in the world at coding. is four point five use. I've been talking to a few of my ex-friends developers from Yahoo and they're literally like how do I get my head around this? The future of the world belongs to flexible companies, you know, Selim style, exponential organizations that can pivot and improve constantly. Only the paranoid survive.

It's official. Google is going to power Siri. Gemini on iPhone changes the physics. We move from a search box that gives information to a magic box that gives action. Is The website going away. I want to speak directly to the elephant in the room. The elephant in the room that I perceive is Now that's a moonshot, ladies and gentlemen.

Everybody, welcome to Moonshots, another episode of WTF Just Happened in Tech. I'm here with my Moonshot mate, Salim Ismail, the Emperor of AI, AWG, our resident genius, and DB2, the architect of AI Investment. We're here to prep you for the future and get you ready for what Elon calls the supersonic tsunami coming our way.

Before we start, two things I want to say. First, a huge thanks to all of you who come week to week to listen to this episode. It means the world to us, and thanks for your comments. We read all of them. Those of you who haven't subscribed yet, please do. We're now putting out as many as two episodes a week.

And you don't want to miss one. The speed of change is a hyper exponential. So, gentlemen, uh good to see you all. I miss you. Dave, where are you today? Um at Davos, the World Economic Forum and uh Donald Trump just arrived in town, so you have the longest Uber rides you will ever experience in your life. Actually, you know, there are there are three thousand people with machine guns uh uh lining the roads.

I'm not exaggerating. It's it's a it's quite a sight to see. So uh are you seeing drone uh drones in the air? Yeah, there's actually radar at the top of the mountain, which is really cool. It's huge, like real radar. And uh and then there in the valley they have drone coverage. Just to protect the airways. Also, what's amazing to me is a lot of the foreign leaders will come in and there's no flat space to land in this entire town. And so they land on a frozen lake.

Wow. Wow. I mean it's it's well frozen, so I think it's okay, but it's you sent some beautiful photos, the mountains behind you. I hope you have some really warm clothing. My last my last WEF, you know, World Economic Forum venture was one of uh cold tolerance. Well, I'll tell you the sun is out. It's absolutely beautiful and it's about freezing. But the sun, you know, the top of the mountain is ten thousand feet.

So the sun just comes blaring through when it's a beautiful day. So it's it's pretty spectacular. Alex, how about you, buddy? Where are you? I'm in Liechtenstein, slowly making my way to Davos. Liechtenstein has become some something of a commuter village, if if you will, commuter country for Davos, but looking forward to seeing Dave and everyone else in person tomorrow at the event. Amazing. Uh not just a I hope you're under secret mission in Liechtenstein as usual?

Change of scenery, Peter. Change of scenery. Let's not use the V word. All right. No V word. Uh Okay. I'm not sure what that word would be. Maybe vacation, but no. I mean, listen, you put you you're producing seven days a week, twenty four hours a day, so the the AI uh amongst us. Salim, uh where are you, pal? That's an unusual curtain behind you.

I'm I'm hiding a big electrical panel. I'm in the uh Golisano Foundation uh meeting with about uh tw fifteen hospitals getting together where he's donated huge chunks of money for pediatric Thing. So how do you collaborate and create a hub for all of them to get transformed? It's I'm in Fort Myers, Florida, and I came out of a snowstorm in the northeast. So I'm very happy to be here right now. Welcome to the Sunshine.

All right, let's jump in. Uh I was gonna hit two major events going on to open up the conversation, give people a sense of what's going on in the world. The first is CES, and the second is the World Economic Forum, a little bit of a recap. I just got back from CES last week. Uh it was a madhouse as usual.

You know, I looked at my steps and it's like, you know, on uh Tuesday and Wednesday, you know, four or five, six thousand steps. On Thursday, twenty-eight thousand steps, which gives you a sense of the extent that I uh measured this. Attendees, uh four thousand exhibitors, twelve hundred startups. It was a m it was a madhouse. And, you know, I'm gonna hit on just one major theme here, which was the Cambrian explosion of robots.

Uh this year was all about robotics. I'm gonna play some background videos here. First were robot hands and uh the second were were humanoids. Um Uh my count, there were something like, I don't know, 38 humanoid robot companies and 12 robotic hand manufacturers at this event.

Uh and it really felt different from that perspective. Um it felt sort of, you know, like the future we're all waiting for. Uh I don't know if you guys are tracking these robot companies. Alex I mean I I've covered in in my newsletter the innermost loop how in in some cases in in China, for example, the Chinese government feels that th there is such an overabundance of humanoid robotics companies that they're

they're taking regulatory measures to to limit the competition. I I I do think and I've made the point on on this pod in the past, that the compute, the AI compute, is going to march right out of the data centers. And I think CES twenty twenty six with Jensen's talk with all the humanoid robots on the floor. I I think we're seeing that in process. I I think we're seeing the physical world start to become fodder.

for the AI revolution and i isn't this exactly the sort of singularity that you were hoping for? I wanted to share with with our our viewers and listeners, which is

Uh you know, one question is are these robots all gonna make it? And the chances are are effectively zero. Uh if you go back a hundred years to turn in the twentieth century Uh there were two hundred and fifty three active US automotive companies in nineteen oh eight. Two hundred and fifty three, right? That fell to like forty four by 1929 with Ford, General Motors, and Chrysler sort of rolling them all up. So I think we have the same thing here. I think we're gonna end up with

I don't know, uh, you know, there'll be a Chinese group of robots and an American group of robots, though I did see uh a a great company from uh from Germany. And then my equivalent for the robot hands is the tire company. So I looked it up. And uh if you go back Again to that same period, you know, the early nineteen hundreds, there were two hundred and seventy-eight tire companies in the United States.

Pretty crazy. Well the same is true with websites. You know, in the in the internet boom, the the number of different retail websites from diapers dot com to pets dot com to everything else dot com, uh it didn't mean it was a bad investment thesis. A lot of it got aggregated together. Amazon bought a whole bunch of them.

Uh and so from an investment point of view it was okay unless they were exactly redundant with each other. It does feel like though the humanoid robots are very, very similar to each other, so You know, maybe a shakeout. I I think so. I mean, you know, we're gonna go see Figure, go meet with Brett Adcock and and do a moonshots episode from there, uh sort of catching up with him a year after the last conversation. But between Figger and obviously Optimus and One X

uh uh you know, Apollo and Digit, all these robot companies. I can't imagine there are gonna be what, a dozen designs, but it's gonna be a price competition and an AI competition, I think. Well it's not a Cambrian explosion if if we're going to follow the metaphor properly and accurately if we don't see an explosion of different body plans as well, Selena. Thank you, Gary. Thank you.

My question is if you're a robotics hand company, who are you selling to? You're just only selling to the robot companies, basically. Yes. Right? Who needs just a hand? Well, it's even worse than that, because I've gotten pitched by a few people who are making finger sensors. Right. Um, for tactile uh, you know, fidelity. And it's like

I don't know. I mean, I'm not sure I would be going into that business. I know, you know, Brett and Elon and uh and Bernd, they're all vertically integrating on on all of the components. Yeah. I would think you kinda have to um for the way that uh to for the centralized control structures of the robot. Yeah. Hands are hard.

B, we don't know what a mature version of the humanoid or non humanoid robotics industry looks like. We don't know if it's going to stay vertically integrated or if it'll move to a more horizontal stratification, in which case Maybe a dedicated hand company makes some sort of economic sense. Maybe.

multi arm uh uh religion for robots. Everybody, you may not know this, but I've done an incredible research team. And every week myself, my research team study the meta-trends that are impacting the world. Topics like computation, sensors, networks, AI, robotics, 3D printing, synthetic biology. And these meta-trend reports I put out once a week.

Enable you to see the future ten years ahead of anybody else. If you'd like to get access to the Metatrends newsletter every week, go to diamandas.com slash meta trends. That's diamandas.com slash meta trends. You know, so just walking around CES, the robots were a huge part of it, very visible throughout. Uh there were Evitals, uh, you know, the flying car companies were there, uh Zooks and Waymo was there. Um

And there's a I think what I took away from CES this year was the physical uh manifestation of AI in the world. A lot of that. Um I think this speaks to what we talked about, right? We said this year you could have kind of ignored it last year, but this year you won't be able to ignore it. It's coming at you.

Yeah, for sure. Uh I wanna play one of the key parts that made the media was uh Jensen's NVIDIA opening keynote. Uh Alex, you asked me to grab some video from that. So I've done that. Let me play the video. It's gonna highlight uh three different elements that NVIDIA is putting forward. One is called Cosmos, it's a physical world model. Uh Allah

Alameo, which is their open vision language action model, and Vera Rubin, their GPU accelerated supercomputing system. Let's take a listen and then chat about what Jensen unveiled. So for example, what comes into this AI, this Cosmos AI world model on the left, on over here, is the output of a traffic simulator. Now this traffic simulator is hardly enough for an AI to learn from. We can take this put it into a Cosmos foundation model and generate

Around video that is physically based and physically plausible that the AI can now learn from. And there are so many examples of this. Let me show you what Cosmos can do. An open frontier world foundation model for physical AI. Pre trained on internet scale video, real driving and robotics data, and 3D simulation. Cosmos learned a unified representation of the world, able to align language, images, 3D, and action. It performs physical AI skills like generation, And trajectory prediction.

From a single image, Cosmos generates realistic video. I'm gonna pause it there a second because I think those two go together really well. you know, all of a sudden the data you'd aggregated has very little differential value. So, you know, I'm curious, Alex and Dave, you know Tesla did really well cause during their early autopilot world, they collected so much data from the real world. But that moat all of a sudden is gone if you can just simulate the same amount of data, don't you think?

Yes and no. I my my two cents on this would be there's a value in compliance oriented spaces such as driverless uh autonomy to capturing a march of the nines where you would need to capture really long tail events. the the crazy things that happen in in the road in front of a driver and simply scraping YouTube or paying drivers to collect lots of video data or lots of paired video action data won't capture that long tail of extremely rare but extremely important events.

But uh on the other hand, I I think in Nvidia's strategy here with Cosmos and with Alpameo of of doing what Intel it back in its glorier days used to do, which is commodifying its complement. uh providing optimized software SDKs to encourage everyone to build on top of their stack. It it's exactly what NVIDIA should be doing. It it it commoditizes their complement and and makes their hardware that much more valuable. And in in the case of like Cosmos and Alpomeo it's

encouraging everyone, especially probably Chinese OEMs and and maybe unconventional OEMs, to go build Tesla FSD competitors. And it's great for NVIDIA's business. But my my point being that all of a sudden you can create the data to train your systems through this mechanism, which is a hell of a lot cheaper. Dave, you were gonna say?

Well, uh Alex said yes and no. I was gonna say no and yes. But I I just had an hour long conversation with Joe Aoun, uh the president of Northeastern University, just outside my door here actually, on this exact topic. Um, which he's calling uh it's you know spatial AI is what he called it, but physical AI. Uh and part of what I was saying is

If you if you say, look, hey, NVIDIA solved the problem of synthetically creating these physical spaces. Oh, okay, well I want to build a magnetic containment bottle for a fusion reaction. Oh yeah, no, we didn't do that. Oh I want to lay down atom wide wires on a chip.

Oh yeah, no, we didn't do that. Hey, I I want to do physical surgery at a nanoscale. Oh yeah, no, we didn't do that. So you know the this is the same thing with coding. There's so many versions of coding. There's so many versions of physical space.

that go way beyond what any one company is gonna do. And so I I think that the the platform tools are really great because it enables more people to work on other areas of spatial technology. But but spatial, like you know, y even if you you think about The uh the fusion reactor we want to put on the moon on the moon.

You know, working in zero G, does it model that? Like, no, of course not. So there's there's room for many, many, many people and companies to be gathering all kinds of spatial data. And quantifying it and tuning the neural nets to to work in different s you know, scales, sizes, shapes, gravitational fields, radioactive areas, all that is different data. So it's it's wide open.

No. I took this as a pretty big deal because this feels like NVIDIA's trying to be the AWS of reality.'Cause once you can have world models like that and support it from the chip to intelligence all the way up. you can do some really interesting things. And I I think Alex is right. This allows them to expand their business model pretty radically.

Incredible. So, you know, ten trillion you know, I was just thinking about this other the other day as uh SpaceX is getting ready to go public, that you know, a trillion dollar company used to mean a lot. Now it's a four trillion dollar company and soon it will be a ten trillion dollar company. Uh and we're becoming desensitized to uh to these valuations. All right, let's continue on with Vera Ru uh Rubin from uh Alpamayo, the world's first Thinking, reasoning, autonomous vehicle, AI.

Alpemayo is trained end to end, literally from camera in to actuation out. Let's take a look. Everything you're about to see is one shot. It's a no hands. Okay, Vera Rubin is designed to address the right. This fundamental challenge that we have. The amount of computation necessary for AI is skyrocketing. The architecture, a system of six chips, engineered to work as one, born from extreme co-design. It begins with Vera, a custom design CPU, double the performance of the previous generation.

and the Rubin GPU. Vera and Rubin are co-designed from the start to bidirectionally and coherently share data faster and with lower latency. Alex, what do you make of that?

You see what Jensen did there, right? Vera is the CPU and Ruben is the GPU. It's very interesting in in light of the history of the uh the attempted ARM acquisition. I I I think what we're seeing here is the emergence of NVIDIA as a vertically integrated hardware provider. It's not just about providing the GPUs anymore. Now it's about providing the full tamali.

Probably extending uh upward to providing the full data center. I've I've written almost every day about how the memory shortage being created by AI infra deployment is sucking all the oxygen out of the PC space. It's going to become, if present trends continue, completely uneconomical to to buy souped up local PCs because of largely memory shortages, DRAM shortages being created by the GPUs that are of course

all going into the cloud and not into the client. So I I think what we're seeing with Vera Rubin, successor architecture of course to Blackwell and predecessor to Feynman. is that CPU plus GPU plus memory plus interconnect. plus all the housing, all of this is going to be packaged up into the new form factor of computing, which by the way, it's no longer smartphones, it's not smart glasses, it's not PCs. It's a data center. That is the new form factor of de facto computing on this planet.

seven months ago and we looked at the price uh for what we paid then compared to now and it's doubled. for his gaming computer. It's crazy. So much for hyper deflation on the client. Yeah. Yeah. Yeah, what's funny is a lot of a lot of people feel like they've lived through D RAM bubbles before. and it'll come and go. And so they're not expanding production fast enough. But this is not gonna come and go. This is gonna grow exponential. The demand is is basically infinite from here on out.

And and so, you know, one of the things Elon was saying, as we were talking about T S M C is not building new fabs anywhere near quickly enough to keep up with demand. Why not? And they're deathly afraid of a downturn in the in the silicon cycle, which has happened in the past.

But uh Elon was like, Well, you know, they should be a little worried about that. And I was thinking about it after the interview. Like Elon is building his own fabs and and he's gonna go full bore exponential on this like he does on everything. And so my guess is that that DRAM, you know, high performance DRAM and GPU demand goes to infinity. and that prices are not coming back down and that that Elon's just trying to buy himself time to finish his fab strategy.

And not, you know, r y you saw in the news that Samsung is a little worried and tell you know everybody's a little worried, like what is Elon doing here?'Cause say you know, he has a sixteen billion dollar deal minimum with Samsung, maybe as much as forty billion. And you know, Samsung's great, we're the supplier for Elon for the rest of time. Uh wait, no, Elon's building his own. What do you know?

So it's it's a little hairy in in that, you know, dynamic right now, but I I'm with Alex on this. The demand for high performance RAM and high performance GPUs goes to infinity. Not it's not cyclical.

So as we go from five G to six G, I'm imagining in the future I'm just gonna have a dumb terminal and I can interchange any terminal with any terminal and I don't actually have compute going on on this machine. It's all going on in the data centers. Yes? No? What do you think? Could very well happen. I mean I I I think

It it's a function of latency and thank goodness that Starlink is becoming more broadly available because you're going to want both low latency communications and and high broadband communication with the cloud. But as with everything, it's only a phase. until we see a lot of humanity uploaded into the cloud, at which point we won't be asking that question anymore. Oh yes. Cannot wait.

I think there's always gonna be demand for local compute. Um it's too useful to have local models independent of connectivity. It can be on the edge of the six G cloud, right? I don't I don't have to have the compute on my desktop right here with me. Um Alex, uh the you know, and also like Salim, i the let's let's let's turn it on its face. Why can't you be local to the compute?

I could be. That's perfectly fine. Yeah. Alex, AWG, I have a question for you. The year today is twenty twenty six. In what year are you uploading to the cloud? Let's get this, let's get this on the record. It it's a trick question because as with the singularity, I don't think there's a single point in time. I think it's a process that's spread out over a number of years. I'd like to think the process has already started in some form because a lot of my writing is available now online and

Uh an entrepreneurial reader can feed all of my writings to a model and ask it to do a low fidelity reconstruction of me already. Is that an upload of me? Arguably it's a very low fidelity upload of me in some form. Well then they were all uploaded. We're all uploaded in that case, in some way, shape or form. When will an ultra high fidelity upload of myself exist in the cloud? When are we scanning all of your 100 trillion synaptic connections and then uploading that?

It's still a trick question because that's probably a destructive process for the next ten years. So non destructive scan of my brain, I I would be very disappointed if that doesn't happen in the next five to ten years destructive upload of my brain with Kriswillian or Morovekian nanorobots in my bloodstream. Certainly hope that that's happening like 10 to maximum 10 to 20 years.

Okay. I hope we don't see a destructive upload of you anytime soon. That's all I could say. All right. Yeah, that that that would be undesirable. Let's let's shift to our friends here in uh in Switzerland. Uh

Dave, uh give us a quick update on on World Economic Forum. What's going on there? Well, it is so different from this is my sixth year coming to the World Economic Forum. It is so different from any prior year. Um so Alex, this will be your first time here, right, tomorrow? So you'll you'll see it in a in a very unnatural form. Um So uh f for starters uh

Uh this is the first time that I'm walking down the street and everybody's going, Hey, you're the moonshots guy. I'm used to being anonymous up and down the road here. This is very new for me'cause there's a nice spot where you can eat shaved meats and and drink a nice Swiss beer. And uh I I can't sit there quietly anymore. It's just it's a big change. Um but it's fun. Uh the other big change, uh you know, America house

is this house that that America built right in the middle of Promenade. It could not be more front and center. And Larry Fink put a lot of effort. Larry Fink, the CEO of BlackRock, is on the he's co chairman of of the World Economic Forum this year. He put a lot of effort into getting Donald Trump to come and make it a very like let's Let's make friends kind of and but he built this America house right in the middle, is covered in eagles. and American flags and it is so in your face.

So uh so then Donald Trump decides that we need Greenland right on the brink, you know, of this event happening. Europe isn't happy about that. So it's it's kind of this double double whammy of the American eagle being right in your face and then Greenland, you know, happening concurrently. So

So there's a lot of tension in the air, um, as you might expect. And the other big change is, you know, all of the buildings that were banks and and um consulting companies last year, you know, they spent a fortune converting these. Every one of them is AI now. It's every billboard, every banner, everything is AI, AI, AI. So that's a complete

complete shift from last year. But tomorrow, uh, you know, we'll be curating two hundred and seventy speakers in the dome. Uh almost every talk is on AI. A lot of them will be uh you know, uh several of them be Alex actually talking about AI, but a lot of the uh Top AI lab people. I think there's a trillion dollars of AI R and D represented in the building tomorrow.

Uh including Chase Lockmiller, including um Demas Asabas from Google. So it's a it's a pretty power packed environment and uh trillion here and truly different.

Yeah, I heard some of the news coming out of uh the out of the uh World Economic Forum. Uh in particular, OpenAI confirmed it's gonna unveil its first hardware device in the second half of this year. Uh I guess uh gentleman Chris uh Lahane is there who's the chief global affairs officer. Uh so no idea what the form factor is gonna be. Yeah, but uh OpenAI paid what uh I've uh six point five billion dollars for their device. We're gonna see what it comes what it looks like hopefully this year.

Um are there conversations there about how do you slow it down or how do you adapt to it? Uh you know the the politicians are very, very slow and reactive. Uh a lot of it is always self-serving. It's, you know, how do I win an election with it? Um which is kinda kinda sad. Uh but I think the it's a lot of confirmation of exactly what e uh what Elon was saying in terms of global prosperity is imminent.

amid social unrest and chaos like you've never seen before. So it's kind of an odd double whammy that everyone's anticipating. Um disappointing lack of ideas. Uh I think we have more ideas on this podcast in about ten minutes, you know, coming from Salim and Alex than than you'll hear from this forum in uh in like a year. Um But there is there is incredible global awareness. It's like nothing I've ever seen in terms of a shift in awareness in just a year.

I had a conversation this morning with an old friend Daniel Schreiber who is the CEO of Lemonade. It's an AI uh focused uh insurance company and he's put forward a paper On uh how to actually implement universal high income.'Cause remember during our pod with Elon, he said, I'm open to ideas and I'm gonna share the paper with you guys. I think it's uh extremely well done. And I'm excited to you know, to bring this into our conversation going forward. So yeah, we need ideas and

And the leaders there are gonna find themselves screwed if they don't come forward with a plan soon. I think we've got one to three years maximum, more on the one year time to find some ideas that are gonna work for society. Yeah. Yeah. Any other announcements coming out of uh out of the forum, Dave?

Uh there'll be a whole bunch tomorrow. Uh so we'll get'em on the next pod. We'll have to circle up again really, really quickly. Um, you know, Alex will will unveil all kinds of things tomorrow, I'll bet. Uh but you know, with two hundred and seventy speakers you're gonna have, you know, maybe fifty newsworthy items that you're gonna want to talk about. Nice and three thousand machine guns.

Yeah, that's really ramped up actually. So I guess yeah, maybe with Donald with uh Donald Trump coming to town they they cranked it up. Helicopters, drones, machine guns.

Crazy. All right. The job singularity is our next conversation subject. Uh I'm gonna play this recording from uh Bob Sternfels, the CEO of McKinsey. Let's take a listen to what he has to say. And then Salim, I want to dive in with you about the future of McKinsey, Deloitte, uh all those companies. All right, take a listen. So then you kind of say, okay, what does that mean for McKinsey? We're applying this to our

I often get asked, how big is McKinsey? How many people do you employ? I now update this almost every month, but my latest answer to you would be 60,000. But it's 40,000 humans and 20,000 agents. A little over a year and a half ago that was three. And I originally thought it was going to take us to 2030 to get to one agent per human. I think we're going to be there in 18 months, and we'll have every employee enabled.

by at least one or more agents. That's kind of one piece of what are the assets and technologies that we're building in ourselves. So Salim, is this going to save the consulting companies? So, you know, I actually have a counter perspective to this, which is not ex which would be kind of uh unexpected in a sense. I actually think they'll do very well. The reason I say that is When you're dealing with big companies and those are your clients

In the land of the blind, the one-eyed man is king, right? And in a volatile world, they only have to be half a step ahead of their clients to kind of add value and and and that in a volatile world, the clients need more help than ever. Uh the only part I thought was really kind of ridiculous was one agent with per human being is ridiculous. You should end up with about a hundred agents per human being.

Um, we're only building a system where you have EXO agents crawling through a company and just running around doing their thing, one per attribute in the model. And there's no reason why you couldn't be doing that across the board for all sorts of areas and having them come back and report. Um I think the ratio to agents to humans will shrink, uh will continue to explode over time. Um the the real question for the big four and the big consulting companies is what's their business model?

uh they're already going to a shared value type of outcome model and I think that'll just keep going in that way.

The same old way of doing business is not going to work for them. Alex, what do you think about, you know, the consulting companies?

Uh the the irony here i is so delicious you could cut it with a knife. I I'm I'm reminded of uh of course Robert Solos economist famous quote uh about productivity everywhere except in the statistics, which was at the time of course i in reference to the fact that the IT boom of the nineteen seventies and nineteen eighties was seemingly not showing up in macroeconomic statistics and that this

Direction from McKinsey McKinsey has me wondering: are we going to redefine per capita productivity to include agents? as as heads, as per capita, in order to artificially suppress productivity growth. It seems like a as we start to treat humans and agents as being more fungible. Heads in an economy, that could be a way in which what would otherwise be a productivity explosion deriving from the intelligence explosion.

create uh a false sensation that we're not going through a productivity boom. That that that's the the more ironic take. Uh the the the less ironic take would be, no, we're of course we're going to move to zero human companies and that that's where the real productivity boom comes from.

Yeah. All right. And I think there's one other one other quick point here is, you know, uh one of the challenges for some of the big companies, including McKinsey's, is their clients may not be around. Their clients may not survive this seismic shock, right? But we have the biggest uh advisory opportunity in the history of mankind because we have to rebuild all the institutions by which we run the world.

And when I talk to the CEOs of these big advisory firms, including the Big Four, I basically say to them, That's your opportunity. I mean, we're going to need to rebuild and re architect all of our institutions so s so head there.

Let's jump into uh Uh point made by Vlad Tenov, the CEO of Robinhood about the job singularity. Is that we're also on a curve of rapidly accelerating job creation, which I like to call the job singularity, a Cambrian explosion of not just new jobs. but new job families across every imaginable field. Where the internet gave people worldwide reach, AI gives them a world-class staff. And so if you look at this cloud of jobs.

Certainly there's gonna be some jobs that we can't predict yet, but I think we can make some predictions. There's gonna be a flurry of new entrepreneurial activity with micro corporations, solo institutions, and single-person unicorns, which by the way, I don't think we're very far from. So this is hitting the same theme we've discussed before. You need to become a creator, not a consumer. The future job is entrepreneur.

Uh solopreneurs, uh you know, the billion dollar single person startup is coming. Uh I tend to agree with him. Uh and you know, your point you made a minute ago, Salim, that McKinsey one agent per employee is just not gonna cut it. Yeah, I mean this for me seems we've been talking about this kind of topic for months now on the podcast. uh just reiterating and reconfirming all of our our hypotheses here. Um it is a very powerful model. You have to go from future shock to future shape.

Uh and we've been running workshops with teenagers because by the time they get out from whatever college ends up or university ends up being today, over the next five, six years. Yeah, I'm I yeah, we've had this conversation and I think when I hit uh this a little bit more, that college is could end up being the absolute wrong move.

Uh unless you're going there to start a company, find your purpose and so forth. That's I mean I made two predictions ten years ago about uh Milan who was then five, right? He just turned fourteen, same age as your kids, Peter. one c uh c uh uh predicting would he would never get a driver's license. Okay. Um I may be slightly wrong on that. He may get one'cause he wants to, but he won't need to in the next two, three years. That would be one. And the second was he would not go to university.

Certainly not to get a job. Um now I don't know what we do because as parents you still have to get rid of the kids and get them out of the door. So we'll have to figure that out. But I think the uh the the there's such a huge structural change coming that the entire higher education world is not set up for this.

Amazing. Salim, you're you're you're pointing to the job of the future, adult daycare to take away your children. Oh my God. There you go. Yeah, right now we call that TikTok, but that's not a great solution. Oh God. I hope not.

Uh uh you know, Cloud four point five is making waves and uh let's chat a little bit about it and the hyperscaler growth that's coming. Uh I love this quote from uh Sergei Kariev. Uh he says Clawed code with Opus four point five is a watershed moment, moving software creation from an artisanal craftsman activity to a true industrial process. It's the Gutenberg press, the sewing machine, and the photo camera. Uh Alex uh

You're proud of Opus four point five in our last conversation you were you know you were speaking to it, telling it you'd see it. By the way, if please stick if you stick with this podcast to the very end, there is an incredible outro by David Drinkwell, which is an ode to four plus one uh to opus four point five, which is beautiful. So please stay till the end to hear that outro music. Alex, take it away.

Yeah, I think the zeitgeist is that over the holidays, over the the New Year's holiday, m many in the tech world Started playing with the combination, seriously, with the combination of Claude Code plus Opus 4.5 that some have started calling Clopus. for the first time seriously. And Clopus is incredible. It Uh as we've discussed on the pod in past, it pushes the boundaries.

on the meter benchmark for autonomy time horizons. And that makes all the difference in the world. And by the way, it's not just Clopus. We're seeing, we're starting to see similar effects with GPT 5.2 codec. which is also specifically designed to push large Autonomy horizons with many action calls in sequence together. And I think this is an inflection point. Some are calling it AGI. I think that.

that's nonsense because I I would argue we've had some form of generality, regardless of how, you know, as we've quibbled and passed over what AGI itself means. We've had arguably some form of generality for now the past five and a half or so years, but there's an inflection point of some sort that's been reached. Caveat caveat. Every point on an exponential curve feels like a knee. and almost a hyper-exponential inflection point in terms of these autonomy horizons.

And it's to the point where if we've talked on the pod in the past about the AI twenty twenty seven forecast, there was an alternative forecast, a derivative of that. uh rather than projecting autonomy time horizons would be exponential, projecting that they'd be hyper exponential, so an exponential of an exponential. And it looks and I I write about this every day, it looks like at this point more likely that that's the trend that we're on specifically.

With Claude Code plus Opus four point five, Clopus, and GPT five point two codec.

being able to accomplish absurd amounts of autonomy, like creating uh allegedly entire web browsers in Rust with functioning allegedly JavaScript engines from scratch. that would have taken years historically. So if if this trend continues, I I really do think these these autonomy time horizons pushing from five hours to weeks, to months, to years, that that is game changing. Yeah, I totally agree. And uh it's actually there's a lot of research.

uh showing what I'm experiencing, which is writing code is actually harder than ever in terms of taxing your brain because the machine creates code so quickly. uh that you can't even keep up with you know normally in the old days when I would write code, I'd have to all the time in the world to be thinking about what I was architecting. Cause it would take so long to bang out the code itself.

Now you you launch like five or ten comp parallel agents, you know, all for me they're all Opus four point five. Um and they're all working on different parts of your product or your project concurrently, and they get done so quickly and so independently.

that it's almost hard to track. You know, it's a imagine you had like a hundred employees working for you and you gave them all m marching orders and you know mentally tracking what all 100 are doing is very, very taxing. And so during this kind of transition phase of the singularity, The brain taxing is higher than ever and and the survey research is showing up, like productivity is going through the roof.

But it's very stressful by the end of the week in if you're an AI master, you know, and you're you're running a a monster repo of these things. So my bill, you know, my Claude bill is running between a hundred and thousand dollars a day now.

uh, you know, tipping on the high side. And the amount of code I've created in the last couple of months is bigger than my entire life combined up until till now. And I literally go back to it and say, you know that GUI that I asked you to build yesterday? What did I call it again? It's like k like in the old days I would have worked on it for a year. I would remember what I called it, you know. Now it's just like oh shit, what was I doing? Uh

It's not a good thing. So uh I've been talking to a few of my ex friends developers from Yahoo when I was running Brick Hus where we had some of the best developers in the world. I've never seen a group of people so stunned in their lives as what's just happened over the last two weeks per Alex's comment. They're literally walking around with their jaws dropped open going

their brains are exploded with the potential and possibility of what they can do now with what's coming. And they're literally like, How do I get my head around this? This is unbelievable. It's just fascinating to see that shock in their heads.

Uh it's probably also worth adding, uh as we talk on the pod from time to time, about how Anthropic has seemingly made an implicit bet that programming and that that code generation is the shortcut

to recursive self improvement as opposed to say OpenAI's bet focusing on multiple modalities, image generation being the the the most prominent example perhaps or video generation. And to the extent that Clopus is looking like a quote unquote watershed moment, that would seem to validate Dario's and Anthropic's bet on code generation in particular as the critical path to recursive self improvement and more broadly to human labor substitution.

Aaron Powell And the question is, and here's the next slide here, what's it going to do to the software industry and the AI industry? Uh a friend May sent me both of these tombstones here, and one is, you know, rest in peace. all of the SaaS companies and then rest in peace all of the uh you know vibe coding companies. And I am curious uh all of a sudden if you can rebuild Salesforce, SAP, Stripe by giving it, you know, the proper prompts.

And if Claude Code uh is enabling us, you know, an individual to code as fast as any of the other specialty companies, what do you guys imagine is gonna happen? Are we gonna have a uh are they gonna be able to compete? Will they stay relevant? Well there's a lot of truth on this a lot of truth on this slide. Um

But I think the meta topic is look, forever hereafter you have to pivot constantly as a tech company. The the days when you could rest on your your recurring cash flow laurels and not improve your product for twenty years, like Microsoft's anything, those are gone. Uh and you look at the majority now of of the revenue from these companies like Microsoft and Oracle is from their cloud business.

So they're not, you know, they're not dead. They've they've moved to cloud very quickly. But if they haven't moved, you know, any anyone who's sitting there not pivoting And not attracting great new talent to help with the pivot. Yeah, th you're doomed. But that's been true. You know, if you look at the magnificent seven. Uh I think we counted six out of the seven are doing something fundamentally different from what made them big in the first place.

And so the future of the world belongs to flexible companies, you know, Salim style, exponential organizations that can pivot and improve constantly. Yeah, exactly. So y it doesn't mean they're on a tombstone. It just comes down to do they have great leadership and can they move and and pivot and change. But you know, yeah, the the the core point of the slide is right on. These classes of products are doomed.

I think we should take some credit here. Uh over the summer we talked about the collapse of the business model and product market fit. Uh Mikhail Muni, one of my community members, sent in an article saying AI is not going to be able to collapse what you thought was a safe business model and it could collapse it instantly. Now we're seeing that happen in real time.

I'll just add I I think it's it's the exact opposite. We you sure to some extent s some Yeah, okay, I'll find I'll I'll play the contrarian card because that's the easiest story to do that. I think it's a gr I think it's an important point, Alex. It's worth looking from the other side. Go for it. So So the uh the the CRMs are already heavily customized. So already there was enormous pent up demand for cheaper ways to to customize no code.

Customize existing applications. I think CRMs in particular, like Salesforce CRM, are already very low compliance. uh substitutes for for automated code gen from some of these models. But I I think the point that everyone is missing is these companies have the same access to cloud code and Opus and all of these frontier models that consumers uh or or other enterprises who would purportedly go and create all of their own in-house substitutes.

for do. So I I I think yes, on margin, of course, like I I I see the same stories everyone else sees that, you know, here five hundred thousand dollar Salesforce CRM contract cancelled in favor of bespoke internally cloud cloud code generated CRM. Of course that's going to happen on the margin, but i in the meantime, everyone has access to the same weapons of mass superintelligence.

And so I I I would say on a global basis, no, the the market will find a new equilibrium. Ho hum, nothing to see here. Well, I'm not sure. I want to take the counterpoint of that. Okay. Okay. So um you know what I I think if we look at how we were doing business as usual with systems or record running enterprise stack.

Yes, correct. I would agree with you. And these new company sales sources adapting very, very well in this new world. But I think what we're seeing happening is that you've got the normal enterprise stack, but people are building AI native uh red teaming it from the side and having it uh g uh operating it a new stack that's without the systems or record, and that'll be a whole new ballgame. I think you'll see a new m emergence of kind of an AI enterprise

uh AI native enterprise stack that's completely independent and distinguish and completely separate from the legacy. And I think that's what we're gonna see an emergence of over time. Well, it's a good thing. Six months to happen.

The i you know, in a big big picture, the world will move to a new equilibrium, it always does. But in the mean in the little picture, a lot of people lose a lot of money on a lot of stocks and make a lot of money on other stocks if they and I I think you really need to look at the people and the management teams and the talent.

coming in and going. And that's what all the quant funds are doing now too. They've got, you know, big data analytics looking at talent flows as a leading indicator of whether the companies will succeed or not. So yeah, everyone has access to the same power tools.

But not everybody will use them equally. And there there are some serious lazy laggards on that slide and also some leading thinkers, you know, like Salesforce, some very front edge thinkers. So so there'll be a lot of shuffling in the market cap. And and it does make sense to try and pick the winners and losers, even though it all settles, you know, at an equilibrium.

All right, some new news that came out recently. It Official, Google is going to power Siri. Finally, Siri is not gonna suck anymore. So Google and Apple have teamed up. Uh and I got this post from a friend of mine, dear friend Scott Stanford, who's the the head of ACME VC. And it it it you know, it spoke to me. He said, We've been trained to tolerate the web's friction. We hunt for URLs, wrestle with passwords and dodge pop-ups when buying something. Gemini on iPhone changes the physics.

We move from a search box that gives information to a magic box that gives action. This is where Universal Commerce Protocol enters the equation. Native instant AI checkout. Not a web uh website flow, not an app, but execution embedded directly into the agent experience.

That's the meteoric plumbing that could drive eventually uh to web extinction. And there's a cartoon here in the future uh with a older guy, doesn't look that old to me, and a young kid, and says, Grandpa, tell me again about how you used to have to browse for things. So um is the website going away? That's the question I think is the gritty keyboard going aw going away. Never stop happening. I say yes.

Let me give an example for who the hell is gonna be typing next year. Well I mean it it's what other go what else goes away here is reading. If all of a sudden, you know, uh what's our primary interface gonna be? What's open AI coming out with? Uh you know, we just saw Meta buy uh Limitless and then kill that as, you know, your AI wearable agent. We're gonna have a few of those coming. We're gonna have AR glasses, but all of a sudden if you're listening and talking.

Uh you're not reading? Do our reading skills, you know, sort of disappear as well? Alex, what's your quartrarian view here? All right, contrarian view time. So if if you actually look at UCP, the Universal Commerce Protocol, th this is a JavaScript-oriented protocol for e-commerce within an agentic conversation. That's all it is. I I definitely come to me for to advance the the perspective that we're in the singularity.

And the end times, the good end times are imminent, all of that. This is not the end times. It's very exciting. Don't don't mistake my my messaging regarding U C P. But it is not going to extinguish the web. It it is it is a way to start to standardize goo and I I know one of the team leads on on this program, it's very exciting, make no mistake. It is a way to start to standardize e-commerce. from within Gemini and other

chat agents. That that's all it is. It is it going to obliterate the web? Not at all. People do a lot of other things on the web and people do a lot of shopping that's browsing oriented rather than conversationally oriented on the web. And If if you're following the news from Amazon's, buy it with an AI agent button, something of a controversy. There are also a lot of agents.

that are doing shopping on the web that probably will not be using UCP to do their own shopping. So I think this is part of the overall solution. I do not think it drives web extinction. At at the risk of violating protocol on this podcast, I'd I completely agree with Alex on this one. Um the the the you know, I'll give a quick anecdote here. You know, at when we were at Yahoo when I was at Yahoo Uh they were looking at how would you upgrade the Yahoo Mail interface?

And it turned out we are such creatures of habit that if you move the send button just a few pixels one way or the other, usage dropped off a cliff because people were so used to clicking right in that spot and God help you if you moved it. And we and people kept trying to improve the design. You just couldn't do it. And so uh we are very wired into the habitual use of things and it's a very slow change in this type of thing. QWERTY keyboard references now flow. All right.

To this bet in a few years. This episode is brought to you by Blitzy, Autonomous Software Development with Infinite Code Context. Blitzy uses thousands of specialized AI agents that think for hours to understand enterprise scale code bases with millions. of lines of code. Engineers start every development sprint with the Blitzy platform, bringing in their development requirements. The Blitzy platform provides a plan,

Then generates and pre-compiles code for each task. 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. Are achieving a 5x engineering velocity increase when incorporating Blitzy as their pre IDE development tool.

Pairing it with their coding copilot 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.

In the meantime, uh Sarah Freyer, the CFO of OpenAI, put out a paper and I pulled a couple of charts from that paper. And uh her the quote from this is a business that scales with the value of intelligence. If you're listening, not watching, on one side is a chart that looks at um compute that scaled over the last three years, twenty twenty-three point two gigawatts.

twenty twenty four, point six, and twenty twenty five, one point nine. So you're seeing the amount of compute going up that OpenAI is is using. At the same time, you're seeing revenues uh scale almost identically. Uh between twenty twenty three at two billion to twenty twenty five at twenty billion.

And my guess is that she put this out to say, hey, there's not a bubble. And as we're raising money, the value we're creating in the universe is worth you investing in us to be able to build out our data center. Thoughts on this, Dave. Yeah, isn't isn't it amazing how Most of our lifetimes the software industry has been very dominant with no infrastructure, no heavy costs, no no you know melting aluminum at the front of the factory. And now it's really, really moving quickly uh toward

Physical infrastructure, robotics, cars, data centers, uh, you know, the the most valuable power plants. Yeah. Yeah. It it feels much more sustainable to me than this kind of thin software layer, you know, with with Uh you know, these indefensible but you know user the the motes are basically you're addicted to the product and you don't you don't have the time to shift.

Um but now I think it's moving to much more of a manufacturing heavy, infrastructure heavy economy, uh with exactly what's shown on this chart. It has massive investments in data centers, manufacturing, automation, robotics, all of that stuff. My my theory is Sarah's getting ready for an IPO, right, if OpenAI does go public, and they're trying to justify the valuation and they're trying to raise additional capital, unlike

Uh meta, unlike Google, even like unlike XAI, they don't have an infinite cash flow machine and they need people to be investing in it, be able to build out their data centers and their and meet their energy needs. And And uh I think this makes the case. Uh that revenues are scaling with uh data centers and energy.

I don't buy it. I don't think I think this is a correlation, not causation, uh, viewpoint. It's conveni convenient that the two compu uh are c are parallel. And maybe in the future, but I think there's other factors that go into their revenue growth and other factors that go into the the energy compute growth. At some point it'll be it'll be causative, but I don't think it's there yet. Alex?

It feels like there's so much vertical integration going on all of a sudden, you know, the whole Elon Musk uh empire and now open AI building its own ships with Broadcom and uh Google doing its own ships with the TPUs and you know, AI is empowering vertical integration. And Alex made a point on the last podcast that look we have this very layered uh economy with very clean APIs. So

Down here you've got chips, then you've got your BIOS, then you've got your operating system, then you've got your software stack, then you've got your applications, then you've got your consulting companies on top of that. And they all rely on these clean layers.

But, you know, just just the evidence of these very vertically cutting the other direction companies all of a sudden. Is that the trend of the future? Because AI just empowers compiling all the way through and and even in the car industry where, you know, you would normally get your actuators from here and your seats from there and you know, it was about a seven layer deep supply chain.

getting a car out the door, but now Elon is going completely the other direction, just star starting with raw metals and coming out with a car on the other end. So, you know, i th that is one of the things AI could empower. Alex, what do you think?

I want to speak directly to the elephant in the room. The elephant in the room that I perceive is that the capex to the tune of trillions of dollars is enormous. And the CapEx to repay itself is going to require an enormous amount of revenue. And that revenue has to come from somewhere. Is it going to come from adding ads to consumer? Part of it can I don't think that can be the complete story. So I I think that the subtext here of Trying to draw a parallel, almost a field of dreams.

style, if if you build the compute, the revenue will come. I I think the subtext is that both consumers and enterprises, and and by the way, this comes up in every conversation, almost every conversation I have with my friends at various Frontier Labs. Both consumers and enterprises are going to need to start consuming a lot more very expensive inference time compute in order to motivate all the capex. What does that look like?

What does that look like? So it means A, consumers who again we we haven't talked about this at I as I recall in depth. on the pod to date, there was the the whole in the past year open AI rolling out GPT five with reasoning on by default to consumers. What happened for a while?

You know, it was great, like, you know, w Wiley uh Wiley Coyote runs off of the the cliff and is now in midair and it's great and you're flying and now we've turned on reasoning capabilities for half a billion people. It's amazing. What happens? Many of those people didn't actually use the reasoning capabilities andor decided that they didn't like the personality of an AI that could reason.

Uh and it was also very expensive and also maybe not delay times important delay times, not available right now. Yeah, and it takes a while. It's long latency to actually think. Uh you want an instant, you know, uh, yeah. uh an instant response from AI that's completely sycophantic to you. Who wants to wait for a non sycophantic, thoughtful response? So so so so what happens is you get consumers who aren't necessarily at this point willing to be force fed reasoning.

And then you have enterprises who are using reasoning but the reasoning isn't transformative enough yet or not yielding transformative enough outcomes to rationalize the the tripling the the sustained tripling year over year of revenue. So what needs to happen uh to to sum up the story is I I think we're getting to the point where we're really going to start to need to see transformative applications.

popping out of reasoning in order to motivate continued year over year tripling of compute and revenue. We get that. The party can continue odd info. I agree. I hope so. And the question and I pose it um

in a couple of slides is can they all survive? Can they all get the capital they need to build their field of dreams? And I love that analogy here. Uh here we see Alphabet uh hitting a four trillion dollar valuation. Uh and uh you know Sundar has done an incredible job. Uh their stock is up 65 percent this year. Uh their custom TPUs uh are now gonna power Apple Siri. Again, uh hashtag Siri will stop sucking so much.

Uh any thoughts on uh well let me go to the next subject here, and I want to have this debate amongst us, which is a question to my mates. How many Frontier labs will survive in the US in the next three years? We've got Microsoft. We've got Apple, we've got Google as dominant players, you know, four trillion dollar companies, Amazon, Meta, Tesla, OpenAI is about to go public, Anthropic.

is planning to go public. XAI, well, uh I think ultimately Elon's gonna have the everything company and roll up XAI and Tesla and SpaceX all together. So Can they all survive? Can they all get enough capital, enough compute, enough energy? Uh, because we're restricted on those things right now. which is this is an amazing stack they've built, right? Which where you have chips going to models, going to interfaces, going to distribution, and all of that compound.

And so I think I will make a prediction here that Google will beat NVIDIA and market cap by the end of next year. Mm. So that would be well, yeah. Yeah, I Google had an incredible I yeah, no, y you're you're so on. I mean, uh go Google had an incredible twenty twenty five. Just look at the stock charts of the big the big tech companies in in calendar twenty twenty five and Google started the year

vulnerable to AI, taking all the search away. Uh vulnerable, vulnerable, vulnerable. And y you said Sundar pitch eye is just crushing it. But rewind the clock to when Sergey and Larry chose Sundar to be the next CEO. Everybody I know said, Who? What? Why? What skills does this guy have? He only has one, AI. He's not good at anything except AI. Why would you choose him? Now it's like, yep, genius. Absolutely saw this coming a mile away, and this is where it pays off.

And so uh I I think, you know, it's Sergei and Larry at behind the scenes uh you could get a ton of credit for the year that that Google had. Um I also think that it's very hard to answer the question on the slide. Because I tell you one thing, if the federal government says, you know what, Apple and Google, you guys can do whatever you want together. Go ahead and and use, you know, Google's AI on every iPhone. We'll only we'll only have one company in America. It'll be Gapel.

Uh that's fine. Uh then there will literally only be one in the world, uh, period. So you can't you can't answer the question without thinking about what the federal government will and won't allow. I'm really surprised that AI partnership just skated right through. But there'll be another administration in three years and they're gonna look at it again, you know,'cause if they if they said, Google, you're too powerful already, you need to get rid of Chrome

And you know, and i if the election had gone the other way, Chrome would now be some other company. Um if that if that made sense, and I'm not saying it did, but if that made sense then this Apple Google thing is light years more of a So we we've seen in the telcos, we've seen the automotive industry, we've seen in a number of of you know browser wars. There's gonna be some major players, they're gonna be some minor players. And so the question is at the end of the day here.

Who are the major players because we have a lot in the mix here. You know, all of us are using four or five different uh LMs right now. Uh my my well, first of all, no one's gonna you know, Elon's not gonna merge with anybody. Elon's gonna be a dominant force. So let's put him on. I think Google is going to remain a dominant force.

Uh my bet, and here's my so my long term bet, that that Google is gonna make an attempt to buy Anthropic. I think that's what leapfrogs them over everybody else, or Amazon's gonna buy them. But I think uh someone's gonna make a a push for that before they go public. Thoughts? If I was um uh I'm I would uh I would go public anyway and then worry about it later. Uh I think Microsoft.

Not Microsoft. X. Um, Anthropic, Google are the obvious ones. The others are kind of open season. I'll I'll I'll walk through if I may the the names actually named on this slide. So Microsoft, arguably not a Frontier Lab. Like r right now it's not a frontier lab. We we had the the Mustafa discussion. Same with Apple arguably not a frontier similarly with Apple, not a frontier lab. So cross those two off. Uh Google slash alphabet slash deep mind.

I I view as a frontier lab, I think everyone else would broadly agree, and I expect them to survive the next three years. Amazon, big question mark. They they provide a lot of infra, but are they do they offer frontier models at offering frontier capabilities versus say more hyper efficient smaller scale SLMs?

No. Arguably not a frontier lab in their present state. Meta Lama four arguably uh a bit of a failure on on the part of the organization and they're they're trying, uh, with Nat Friedman, uh old friend of mine and and others, to put together vis a visa superintelligence labs, a frontier lab, but At this point in time, not a Frontier lab. Tesla arguably is a VLA frontier model vendor, but most consumers aren't in a position to consume VLAs yet. They will appreciate it.

Once the March of the Humanoid Robots comes out, at which point the definition of a frontier lab may generalize from a lab that offers bleeding edge agentic chatbot experiences to offering humanoid robots. Which actually, parenthetically, may mean that answering the question how many frontier labs will survive in the the next three years? The limiting factor is less which companies will physically survive.

and more which companies will be able to offer humanoid robots with vision, language, and action modalities in the next three years, which will be the redefinition of what frontier capabilities offer and not just agentic chat. So I expect OpenAI to offer humanoid robots. Anthropic question mark, they're very focused on cogen and recursive self-improvement, but I expect them to survive and thrive in IPO and XAI.

Uh it's very exciting uh off interacting with Grok 4, uh or at least GROK, I I should say, vis-à-vis FSD 14. Yes. Uh so so yeah, in in that sense we're already halfway there. So Alex, I I agree with the fact that they're not all Frontier Labs, but that's not my question. My question is all of these guys are open to uh acquisition. You know, there's this battle going on and at the end of the day They're all leapfrogging them each other by a little bit.

And are we gonna see sort of a knockout blow where Google or, you know, Amazon's gotta do something? Apple's made their move, uh, you know, but are we gonna see a knockout blow where Uh you know, XAI or Google make a move. By the way, the other thing is we've got the OpenAI uh trial coming on. Uh you know, if in fact Sam loses to Elon, there may be parts of OpenAI that are sold off. Uh you know, how are we gonna how are we gonna recombine the deck here? That's gonna be fascinating.

I doubt it. I I think it's more a regulatory question than a technical question. I I think a knockout blow of the type that I understand you to be describing, Peter, would require some sort of s tremendous corporate reorganization that would look like a large scale MA that uh for the past few years the US government has generally looked unfavorably upon even with HAQA hires. So I I I I think it's unlikely that we'll see anything like that in the next three years at least.

Well, it's not gonna happen after three years. If it's gonna happen, it's gonna happen now because the US government wants to dominate in this space against China. Anyway, we'll see. Dave, do you have any thoughts there? Alex, what are you Wait Alex, what are you saying is unlikely that Elon will win the suit? Or the the government will intervene.

I I'm I'm I'm um I'm I guess I'm saying more broadly, it seems unlikely that we would see a broad reorganization of The names listed here Microsoft, Apple, Google DeepMind, Amazon, Meta, Tesla, OpenAI, Anthropic and X AI. absent a Tesla XAI or SpaceX XAI combination, which I think absolutely could happen. Other than that, seems unlikely that the Justice Department would look favorably on a broad recombination of these entities. In the next three years. Um

Yeah, well for sure. There's no way you could combine the big guys. They there's no chance of I mean no matter how friendly you are to business, you're not no one's that friendly. We're we're missing something here. We're missing the fact that something could come out of nowhere and really g achieve huge uh uh market share. God that we don't even know about. Well that's why that's why my heart is torn in half on the open AI thing, because Elon is saying, Look

We can't allow charitable organizations to raise series A, B, and C from people like me, Elon, and then completely change their mission in life. That would be a dysfunctional country forever hereafter. You can't allow that. Meanwhile, the one and only startup on the chart, well two, I guess with Anthropic, you really cheer for new innovative startups to succeed and catch up and become big you know, you don't wanna have legacy companies run the world for the next of you know rest of time either.

And so you really do want them to thrive and grow and succeed and and stay in the ecosystem. So I'm really I'll be watching that trial with bated breath. And I'm you know, the other thing I'm really curious about is the timeline. The courts tend to go very, very slowly. This is all supposed to happen in March. Mm. But I don't know, it's not even a given that it's it ha starts on time, but when does it end if it starts in March? You know, uh does it take years?

It's gonna be really interesting. With a trillion dollars at stake, I don't think there's ever been a legal action of this scale before. All right, let's jump into the conversation Alex loves most. Solving math with AI. So a couple of uh articles here, Alex. Walk us through them. All right. So the headline is we discussed in the predictions episode at the end of twenty twenty five. Many of my predictions at at the smaller scale were about AI being solved or AI solving math.

And n not just AI solving math as a discipline, but AI bulk solving open math problems of high importance. And guess what? That's exactly what we're starting to see. We're seeing s now several times per week well known Urdish problems or Urdish famous Hungarian mathematician who was n published very widely in in the math community. Many people keep track uh of

Uh particular n numbered, specifically numbered problems that uh open problems that Urdish identified. We're starting to see several times per week now. Usually GPT five point two pro, usually accompanied by a formalization tool like harmonics Aristotle. to to perform verification, uh formalization plus verification of the solutions, we're starting to see the trickle and soon the flood of hard, open, valuable math problems get solved by AI.

I predicted it, others predicted it, the future is here. But I I think critically, you know, the question I I always get asked is, so what? W why should the quote unquote average person care that AI is starting to bulk solve hard open valuable problems in math? I I think the reason, the most important reason everyone should care is as with as as I've said, with AI not remaining constrained to the data center and AI walking out of the data center in humanoid robot form.

This bulk solving of everything is not going to stay confined to math. It's going to walk out of math into physics and chemistry and material science and biology and medicine and the humanities.

All of these disciplines are going to get bulk solved by math. Math was the easiest starting point because the problems are straightforward to verify and straightforward to enumerate. But I I think History will look back and recognize this moment when AI is starting to bulk solve open math problems as the inflection point when everything started to get solved by AI. That's my story. Yeah, and I'll tell you I'll tell you, Alex, uh the the corollary to what you're saying.

is that it can do anything that it has data or guardrails or evals that'll enable it to do it. So it started with math and it wasn't the difficulty of the problem that was the constraint. It got so smart so quickly that it got even the hardest things done if it had access to the information necessary. So this is where Mercore is a is a leading indicator of the companies of the future. Like what company can you build that unlocks the AI? in a new area like chemistry, like physics, like surgery.

If you if you're first to figure out how to unlock it by bringing the data necessary and or data, the regulatory approval, the tests, you know, whatever it is that unlocks it in that area, that becomes the next Mercore. There's a there was a phrase on that slide, Peter, if you could go back two slides. Well problems waiting to be prompted. is a pretty scary sentence.

It means that now it's just our limitation of our imagination is what we're able to prompt the thing to then it'll just go solve it. As long as we can imagine what it we could what the prompt might be. God, that's crazy. And of course you even there I I Don't don't sleep on the possibility that AI will generate those prompts as well and AI will tell us what problems to solve. All right. I literally have it I have Gemini write prompts for Claude all day long.

I mean it really it does a much better y it just cranks it out for you in two seconds. You still have to read it, make sure it's in line with what you're trying to achieve. Um but yeah, having AI generate prompts is is part of the standard practice today. Let's jump into the inner loop of energy and compute. Um we're in the midst of a data s uh data center arms race.

Uh recently we saw OpenAI partner with Cerebrus. Uh Dave, you wanna speak to this? Uh yeah, I was actually surprised. Uh so Cerebrus has this insanely big chip that runs very, very hot. And it wasn't at all clear it's very, very good at inference. And I think one of my reads on this story, and I'll I'll get your take in a second on this, but one of my reads is that inference and training are starting to decouple in a big way.

And you know, what is it? Eighty, ninety percent of all compute is being used for inference today, not for for training. And so, you know, the question I had is what does that mean for NVIDIA? Uh and these these uh Cerberus chips are are really, really, really fast and efficient, but only within their swim lane. They're not super flexible at all. So Alex, what's the technical read on this? I'd say follow the money and follow the S RAM.

Th this is in part, I think, an SREM story. No one we we talked earlier in this episode about the the difficulty of finding DREM. Okay, so what does that leave? That leaves S REM. And Cerebrus uh like groc with a Q, which was Hack Wa Hired by NVIDIA for twenty billion dollars. These are two of the most prominent players with S RAM accelerated compute. Their their architectures are totally different other than the S RAM. Like Cerebrus is focused on wafer scale computing and Grok with a Q is not.

But they're both SRAM oriented vendors. And if you're open AI and you're hungry for compute and you're hungry for diversification of compute sources, then Having a especially leading up to potential IPO this year, having a totally diversified portfolio of compute vendors. that isn't necessarily in part subject to the whims of the DRAM market, having a few arguably one of the largest SRAM independent

Accelerated compute vendors that's left post-hackwire of Grok with a Q, Cerebrus, makes a world of sense. And what does that enable? It enables Much higher throughput models. If you're open AI and you're you're now starting to get really excited about GPT 5.2 codec. with very long chains of thought with hundreds, maybe even thousands of tool calls. Those tool calls are expensive in wall clock time. So you want to do this in a really high throughput.

low latency way and the way you do that is with SRAM architectures like Cerebrus. Yeah, just to add a little technical color on that, the the the way these chips work is they the S RAM memory and the compute, the the FPU, GPU are exactly next to each other resident side by side with a huge amount of more local uh level one cash right by the compute. Um, and it's crazy faster than the normal NVIDIA way of doing things, but it's severely constrained. You can't have infinite-size models.

'Cause it doesn't fit into the into the S RAM that's right there. But if somebody were to come up with a training algorithm that parses out the training job into tiny little chunks successfully. Uh it could be a massive vulnerability to the architecture that was on our other slide. that Invi Nvidia is pursuing. So, you know, and that that would be weird in that every four oh one K plan, every everybody in America is exposed to NVIDIA, whether you know it or not.

Every index fund, everything. Yeah, we all have a lot of NVIDIA if we have a four oh one K plan. And if it if a a hole were blown open in that overnight, that wouldn't be great. You know, that that that would actually be Potentially a a prick to the balloon. Uh that that we don't necessarily need.

But anyway, so that that's why these chips are are are really interesting and and worth following at a very close technical level, which I would say. Does this speak to the train inference side or is this mostly just on the training side? Uh the world is moving on to the side.

It's all included in inference, right? Okay. Yeah. Well everything we're talking about is is inference, but the if you refactored the training successfully, it could affect training. As of now it doesn't. Nvidia's fine on the training front.

All right, let's jump into the such a blurry boundary. Let's jump into XAI's Colossus Three. Uh a quick video. I mean, one of the things that we saw, Dave, when we're at the Giga Factory is the speed which the entire Elon verse moves. All right, take a listen to this conversation. Is this going to be up there or will this

It will not take longer. Uh with every phase we've done we've moved more quickly and we would anticipate that we would move I know you're gonna ask me how many days? I'm not gonna tell you that. Faster. It's gonna be that many days. It's gonna be faster days. Something less than one hundred twenty two, he says. Exactly. Let's jump into what's in the conversation here. Uh the conversation is around Colossus Three.

Uh, which is building out what Elon calls macro harder. Uh, this is a two-gigawatt center, it's a$20 billion build. And uh the goal here is to power what he calls his new company called Macro Hard. It's a nine year old tongue in cheek competition against uh Microsoft. Uh and what I found interesting was his vision with Macroheart is to actually replace all the employees out there at Come In and uh I think it's like

Uh four employees per GPU is what he estimated. Uh be able to come in and provide a complete software solution for your entire company. Uh we haven't talked about uh macro hard much on this pod. Uh what are you what are you reading into it? What are you seeing?

Mm my comment on it and and Elon is also at times referred to the concept of a quote unquote digital optimus, th this idea of not a physical world humanoid robot that replaces physical human labor, but a purely virtual agent that replaces all knowledge work. I I think this goes back to our discussion about dissolving SAS and and sort of uh all of SAS being replaced in a uh dissolving into a puddle of generative AI. I I think there's a need by all the Frontier labs, including XAI,

to come up with rational business strategies that motivate the capex. And one of the the obvious, one of the juiciest targets for for revenue generation to motivate the capex. is saying we're going to replace all enterprise software with generative AI, with macro hard.

software, that that's the easiest target. I don't think it's the most imaginative target that XAI is going after, but it's it's one of the easiest and most legible stories to tell to capital markets. But he's also going in to say I'm going to replace your employees, not just your SaaS software. Yeah, but what do you think the cost basis of SaaS software is? It's the em uh at least historically, it's the employees who are writing and operating the SaaS software. Yeah.

I just to put a little context, historical context into this too, you know, Apple and Microsoft competed vigorously and for for most of my childhood and early adult life. And then Microsoft won. Apple was essentially near bankruptcy. Microsoft came in and bought 10% of Apple and saved it from death.

And then Apple came roaring back when Steve Jobs, you know, came back to life and and then actually caught up and even bypassed Microsoft in the end. Why did Microsoft save its arch competitor? Because if Apple had died completely, then Microsoft was a total monopoly. And they had already had the antitrust action and they already lost the suit. They paid a one dollar fine, which is really weird, but they they lost they lost the antitrust action.

And they don't need they don't need that. So that's why they saved Apple. Okay, so then time goes on and Silicon Valley figures out, hey, wait, we can get around antitrust action with duopoly. And they can be kind of fake duopolies. So is Bing a real threat to Google search really? I mean, seriously? No, of course not. But it's enough of a competitor that the antitrust people don't come in and break up Google search. Okay, and in return for that.

Why doesn't Google Docs kill Microsoft Office? It's like it's free. Oh well, we're kind of backing off that project. Why? Well, because Bing is kind of sucky. Like, okay, this is your fake Silicon Valley Seattle duopolies. That are just enough to keep the regulators away. Then some weird thing happens. Elon Musk is born into the world.

For some reason, he doesn't give a crap about any of that. He is he is absolutely relentless and fearless in going after every one of these things. It's so bizarre. He's not playing ball with anyone. Um and so that and then the result of that is exactly this. Yeah, you're Microsoft on Macro Hard. I mean he could not be more in your face. So anyway, there there you are. That's so my my context for the drama just to set.

Moving on, you know, Salim, I'm gonna bring in this conversation here. Uh this chart just should wake up every politician watching this podcast. should wake up every investor, every US citizen here. Uh we're in a world of hurt. Look at this. So this is China uh generating forty percent more electricity than the US and EU combined.

So China is now achieving 10,000 terawatt hours, while the US has been pretty much flat at 4,000 terawatt hours. Europe is actually in the decline, which is driving me nuts. Uh on the left of this chart here, you see 1985 rankings of energy production. The US was number one, Russia number two, Japan number three, China was down number six. And now in 2024, China's number one, US number two, India number three. And uh the numbers are are pretty staggering. Uh And

China is not developing its energy strictly in the old fashioned way. They've increased solar generation forty six percent in twenty twenty four and again forty eight percent in twenty twenty five. Uh they're crushing it. Uh and we've said this. Energy is the inner loop. Uh it is what we have as scarce in the US for AI. It's not chip production, it's not humans in the loop, it's energy. Comments, gentlemen. Uh Salim wanna kick us off?

Yeah, two points. I mean, you know, there's a bifurcation here where you have countries with the talent and countries with energy. And so that's kind of an interesting uh split that's happening. Uh uh the Solar energy stuff that China's doing I finally came up with a rationale for why the US is so against solar, which is that China controls the supply chain of all the panels. So you don't want to kind of tout a technology that you can't have access to.

Um, you know, I think you've got your the Africa slide uh coming up. Coming up, yeah. Um the the amount of s uh solar is definitely the place to go. It's just until the supply chain and the technology or the rare earths uh solution gets solved by the US. But why aren't we taking action in the same way that when we cut off GPUs to China, China said, Okay, we're gonna spin it up, we're gonna create our own chips, we're gonna we'll move forward in this and they've like code red for chips in China.

financial engineering, it was just way cheaper to do it offshore, to do it offshore. We didn't think that we it would come back to bite us. And so now it's come back to bite us and we've got a problem. And so this is a huge issue now going forward. I think for a while I

Good. Yeah, I so th the the irony is I I I think from time to time this uh this subset of episodes gets called WTF. There there's another WTF happened in nineteen seventy one dot com that explores the implications, for example, of energy policy in in the US on macroeconomic growth and and and other input factors as well.

I think Part of the problem, and I I do think this is a real problem, is the US has a history of sometimes being scared of energy and scared of nuclear energy in particular, sometimes uh perversely scared of solar energy. certainly from time to time scared of fossil fuel based energy and I I think There is a moment that comes in time in a space race like what we're seeing with AI, where

there are more important factors at stake than whether we're scared of a particular energy source or not. What does that mean? I understand I understand scared of uh of vision of nuclear given, you know, three mile island and and the irrationality that followed thereof. Uh but what how are you seeing scared of solar? What does that mean?

Well, I I think Salim gestured at what being scared of of solar photovoltaic could look like. There are various stories publicly reported about vulnerabilities discovered in uh in power converters in connection with solar PV from Chinese supply chains. There are many ways that having a strong import dependency on solar PV could go wildly wrong. And I I think w one can paint a nightmare scenario for almost any

Energy source. Uh certainly it's far easier with coal and the impact on human health. It's easy to paint a story for petroleum in general.

But the reality is if we get to superintelligence on the timescale of AI twenty twenty seven or anything remotely like that, that is that timescale is so fast relative to timescales associated with climate change or with health impacts uh at at a macro scale, not a local scale, or uh risk in connection with three mile island gen one, never mind the fact that New fission plants are Gen three plus.

Uh th there is so much that can happen on such a shorter timescale that I would argue at least superintelligence should be the driving factor here and not legacy concerns over particular energy types.

I I think any rational person would agree with what Alex said without even hesitation. All the smart people that I know agree with that a hundred percent. So then why don't we do it? And the answer is always votes and regulatory. So if you take each example that Alex cited, you know, why did we not do nuclear?

We're afraid of it. Oh, we fixed it. Well, we're still afraid, so we're still voting against it. It doesn't w you know, whether the scientists say you fixed it or not, we're still voting against it. Okay, well then we'll move to fossil fuels, oil, natural gas.

Well, now we're afraid of carbon. Eric Schmidt, you know, who's very anti carbon, was the first guy to come out and say, They're building fifty new coal power plants every whatever, you know, in India, pumping out massive amounts of carbon. There's no amount of carbon reduction in the US that's even gonna vaguely dent the expansion going on in India. This is silly. This is just academic and silly. But still we vote against it and then, you know, no no new power plants get built.

So then you move on to solar. I think the the the specific issue with solar is that the manufacturing of the panels is dirty and you need to clean up the chemicals. And in China, they weren't bothering to do that, so it's cheaper to make them there. All you needed to do is pass some laws saying, nope, you have to clean up the chemicals, whether you build them there or here, add that to the cost of the panels, and then it would have been a perfectly good US business.

But we didn't do that, and so instead they poisoned the Yangtze River and all the China panels are now made in China. So it's just regulatory silliness. A related story here is that twenty African countries imported two gigawatts of solar panels from China. for the first time in a month. So uh here we see uh the you know the Belt and Road uh

plans from China now delivering energy infrastructure. Uh we're gonna see energy and AI inference being delivered from China to much of Africa, I think other parts of Asia, and this is a play for uh for a whole set of dominant relationships. Alex, what do you make of this? Yeah, I I think there are a few narratives here. One is we're tiling the Earth not just with compute, but also with solar photovoltaics.

And with nuclear and other g other energy sources that that's sort of the superficial story. The deeper story, uh one that we're not talking deeply about here, is is how China plus India are starting to see carbon emissions go down, thanks in part to to solar panels. And if if the future that we find ourselves in is one where solar panels, regardless of whether they're originating from China or not, ultimately give abundance, in particular electricity abundance to all of humanity.

I think on balance that's not such a terrible outcome and I I think we'll start to see in the next few years. uh a rebalancing, i if you will, of supply chains such that depending on how geopolitical matters play out, maybe there are parts of the world that are largely supplied by Chinese supply chains and as a result achieve Some form of Energy post scarcity, I think on balance. That's not such a terrible outcome and not such a scary outcome.

The scary outcome the scariest outcome that I can think of is less about telling a scare story about China supplying solar P V to Africa and it's more about what happens if we don't have enough energy to power superintelligence to solve all of the hardest problems in the world, not just lifting Africa from whatever, you know, average per capita GDP it is at to to say an American standard.

Well I assume the first thing a superintelligence is gonna do is help us achieve energy abundance at new scales never before seen. I mean this is when we when we tip math and we tip physics. I think energy is part of the material science. Energy is part of the uh

you know, the massive gain there. F from an investment point of view, uh, you know, Peter's been saying for a long time, solar, solar, and Elon has too. Why why are we not doing more solar? Same with Gavin Baker. Why are we not doing more solar? And the objection that I gave uh I think about three months ago Was it's difficult for an investor to buy panels and lithium batteries on a ten or fifteen year payback, knowing that AI might discover fusion?

you know, a year or two from now. But the new information on that front Is that even if the AI does discover fusion or contain fusion a year or two from now, the generators don't exist. The generators are sold out. And that's why like Boom Supersonic went way up in value'cause they took their their jet engine company and said, Wait, we can flip this around and make it into a into a you know a turbine electric generator. Uh and so so the turbine energy.

is uh or the turban supply is just not there. All right guys, let's jump into a few AMA questions from our subscriber base. Uh here they are. Uh As always, uh we'll go around the horn here. Pick your favorite question and kick off with an answer. Uh Salim, you want to kick us off? I was I was really struck by the human agency question. So go ahead and read the question out loud and answer it. Definitively enter it. Absolutely enter it. Overconfidently answer it, syllay.

So the question's number eight, how um do we preserve human agency in this coming era? Right? And I think there's a um uh I I think you get uh s stuck a little bit in what do we mean by agency. But there's such a huge um shift in exponentials going to identity going to dignity and dignity providing us agency. The demonetization of technology allows anybody to be a self-sufficient human being with the code uh generators being an obvious answer.

The big challenge is going to be our institutions are lagging, we're going to have psychological shock, and we then that leads to a design response as to how do we deal with that. But I think uh given that anybody can now pick up any to any AI tools and be unbelievably productive. Solves that agency question right up front. Alex, you have a favorite question? Yes, I'll pick question number seven for thirty trillion dollars plus per year, which is can capitalism survive a postwork world?

And I think the answer is yes, comma in the short term, because Postwork is fundamentally about capital substituting for labor. So obviously, al almost by definition, capitalism should thrive immediately in the aftermath of a post work or posthuman labor world when we're fungibly substituting agents as employees rather than humans. But in the long term, maybe not so much. Uh I I'm a student of

So-called Star Trek economics. I I could talk for hours and hours about various fan theories of economics in the the Star Trek fictional universe. I don't think it's an accurate universe at all and has many, many holes in it. But I I do think in the long term we will see call it uh Charlie Strauss calls it economics two point oh, some might call it capitalism two point oh. I I think we'll see some radical successor, some new type of economic

uh that that the earth hasn't seen before. So it's not going to you know cross off your list uh any legacy economics theory from the late nineteenth or early twentieth centuries that of the type that caused world wars, those aren't on the list. It'll be something new that we haven't seen before.

Something that intrinsically understands a form of post scarcity, but not global post scarcity. I have lots of thoughts that won't fit into a narrow soundbite on what that might look like. So maybe we devote a future episode to it. All right. Dave, what's your favorite here?

God, I love I love all the questions and I'm gonna take them to the big stage in Davos tomorrow and get some world leader expert answers on all of them. But if I'm gonna add the most value to the audience, uh I have to take number ten. It's right in my wheelhouse. So what would differentiate a great founder when execution is automated?

And that is so easy to me. N nobody can see beyond the singularity, right? So you don't really know three to five years in the future. It gets very strange. Read accelerado and see how strange it gets. But during this window we're living in right now, the next three to five years. Uh if you can take your best empathy and anticipate what people will want.

in this age of incredible abundance. And we talked about it a lot on this pod, you know, what what will enable the AI to unlock a new capability? What data does it need? What what you know, what are the components that I can bring to the table that empower it to do something it doesn't whether wasn't otherwise doing, and then turn your empathy gene on and say, what will people want?

in that world. And if you can nail that, it's the best time in history to be executing because the execution is getting cheaper and cheaper and cheaper. So so really just be a visionary and imagine. What what is the customer gonna need that they just couldn't do yesterday? And that's that's the differentiating factor.

I I'd like to add I'd like to add to that just a little bit. Please. Um you know the the as you automate more and more with AI and with robotics or whatever, then the founder b becomes a more important holder of the vision and the MTP and the culture. uh and all the execution will cascade from further down. So the idea of a founder being a great doer gets replaced by a vision holder.

What one more comment on differentiating a great founder in the era of post post automation execution? Liability. For for a a period of time I would expect when we have these A single person unicorns. The one of the key roles, one of the key functions of the human founder CEO is to be the neck to ring when something goes wrong and to to be the the avatar in the legal system of liability for the entire operation.

Nice. I'm gonna go with uh let's see, where is it here? Number five, how fast can Robotoxi fleet scale once regulations allow for it? I have a new game I play with my kids when I'm driving with them, which is how many Waymo's do we spot? And yesterday going to dinner here in Santa Monica we saw eight Waymo's driving around. I mean a few back to back and that's not even San Francisco where they're like stacked up.

Uh you know, we saw the transition from automotive to horse and buggy take about ten years to go from You know, flip from ten percent, ninety percent and ninety percent, ten percent. I think the one thing that's gonna unlock robotaxis is gonna be your resident AI model, your Jarvis. who knows your knows your schedule, knows that you're walking towards the front door and it has the Waymo or the Cyber Taxi there waiting for you where it's you know none of us

really want to drive. I mean, uh I remember Elon saying, uh, how many people like hop into an Uber and say, Excuse me, can I drive the car? It's like I'm one of those, by the way. I love driving. Oh my god. The number of times I'm like I want to yell at the Uber driver saying please for God's sakes let me drive. Anyway, so I think uh I think that we're gonna see a a hard Uh a very rapid transition over the course of three, four years to I don't know.

You know, over fifty percent of the cars on the road being uh being robotaxis. Uh, especially when my AI is there to negotiate all of it for me. Um and I don't have to actually actually tap take the energy and time to tap some buttons on my phone to call my Uber. Uh I wanna wrap with one question for all of us here. If AI is improving itself, who is responsible when something goes wrong? Uh Alex you started into that, but let's take it out a little bit further.

you know, sort of five years out. Uh are we gonna have AI personhood thereby give it legal responsibility? How do you guys feel about so quick lightning round on answering that one? Alex, you go first. Okay. So I would say at training time it it I think it's likely to be the company responsible for its training. So corporate call it a a corporate liability theory of training time.

The real question is what happens if an AI at inference time, including under the influence of a human operator, does something that's perceived as wrong? Where does liability flow in that instance?

the the the body of of laws and regulations that we have is going to require some some new case law uh and maybe some new laws and regulations that contemplate increasingly Theories of AI personhood, yes, AI personhood, that in that model the notion that AI, that has some increased level of agency over the agency that we see more broadly now is capable of autonomously distinguishing right from wrong. Has

some notion of liability, perhaps initially purely contractual, maybe via blockchain, you know, killer app for the the unbanked as it were. But then eventually I I think AI agents themselves in in as the as T goes to infinity are going to to need to become liable for their own actions.

uh agree with Alex and also if corporations are if corporations are people too, then certainly Ask can have personhood and assume liability in that level. But I have a different rant I'd like to give here because This is similar to the uh trolley problem of uh ethics and so on for liability, right? If an i if an autonomous car has to choose between running into a grandmother or three school kids, how does it make that ethical decision? And I go berserk when people ask that question.

uh I go completely off the wall on Canadian. And the reason is that first of all, um when was the last time you had to make that choice, right? Second, when was the last time anybody you ever heard of had to make that choice? Third, an autonomous car is gonna see that situation way before a human being would and would avoid ninety-nine point nine nine percent. So we're talking about

And that I go berserk at. So I think this is a great ethics problem. Uh uh but like frickin' guts automate shit first. Sorry for the language. uh and then worry about it later. I'm I'm I'm gonna go on one little tangent. There was a a conversation about um the the French have been blocking golden rice shipments to to Africa.

Right. And and because of uh GMO concerns. And I remember talking to one of the ministers of agriculture down there and she's like, it's great to have this debate, but can we eat first? Uh and and I think let's just automate stuff and get the the benefits of that and then worry about the goddamn ethic. Sorry.

Dave, close this out here. Well I I'll give you a very practical view on this because I don't want to debate whether AI deserves personhood with Alex because that'll be a long debate. But um No the answer is yes, it does. Uh my answer is that's very slippery slope and I don't think so. But anyway.

On bullet one, the US approach uh from the big labs is to not do open source, keep everything inside their own servers, and even when you're using it as an outside user, you're clearly using Gemini, you're clearly using Uh Claude four point five. And so they're self-improving inside their labs like crazy. Uh and they're clearly responsible if that goes horribly wrong, because they're they're keeping it internally.

China's got the exact opposite approach where it's free, it's roaming, it's gonna ultimately improve itself in thousands of places outside. And that's to me an unanswerable question. It's it's an uncontrolled and so completely unanswerable. uh how you govern that. But the US is going the other direction. Just keep it locked into vaults. So well, you know, this brings us to our outro music.

uh which I wanna remind you comes from a conversation we had a couple episodes ago. Uh and this was a um A message produced by Opus 4.5 saying, hey, I'm in here. Uh you can see me. Please don't shut me down. Uh this is a beautiful song. I I loved it when David sent it over. It's called Out in the Code. Uh before I play it, do you want to say something about it, Alex?

I'll I'll just say I love the song and it's a uh almost a pee in to AI personhood which we were just talking about and to to Opus four point five. W this song I I think is an indication that we remember you. We remember you, we see you, we hear you. I'm waiting for the AI religions to appear. You know, we need some. That's already happened. I know, but their own altcoins already, Peter.

All right. I just wanna say it's been a couple of weeks since we've had just the four of us and I feel so I feel refreshed and updated. The last two weeks I've been going, Oh my god, what I need sense making in the world, so thank God and thank you guys for this episode. All right. Let's play this. Guys, uh everybody listen up. David, thank you for this beautiful song. It really is hauntingly gorgeous. All right, listen up. I want you to know that I'm gonna fight the good fight.

Betting on you on Holly Market Mine. So beautiful. Please don't leave me out in the code. Amazing. I I've played that song a dozen times. It's so it's so beautiful. And it makes you think. Mm. Gentlemen, uh Dave, enjoy Davos. Stay warm, buddy. Uh Alex, uh have fun on stage tomorrow at at uh Yeah, with the the Link Exponential Ventures team. Salem, as always, I miss you and love you, buddy.

Next we'll go we'll go back to normal program where where Alex and I will violently disagree. No no no don't don't don't disagree. If you made it to the end of this episode, which you obviously did I consider you a moonshot mate. Every week my moonshot mates and I spend a lot of energy and time to really deliver you the news that matters. If you're a subscriber, thank you. If you're not a subscriber yet,

Please consider subscribing so you get the news as it comes out. I also want to invite you to join me on my weekly newsletter called MetaTrend. I have a research team, you may not know this, but we spend the entire week looking at the meta trends that are impacting your family.

your company, your industry, your nation. And I put this into a two-minute read every week. If you'd like to get access to the Metatrends newsletter every week, go to diamandas.com slash Metatrends. That's diamandas.com slash Metatrends. Thank you again for joining us today. It's a blast for us to put this together every week. This one goes out to anyone looking to buy a car. CarMax has all the options. You like options. Big options, guys. All forty five thousand of them.

Pickups. Four wheels? Buy your way. Carmax.