Marc Andreessen: The Original Tokenmaxxer — from Browsers to Claws

There's something about AI that causes the people in the field, I would say, to become both excessively utopian and excessively apocalyptic. Having said that, I think what's actually happened is an enormous amount of technical progress that built up over time. And for example, we now know that neural network is the correct architecture. And I will tell you, there was a 60-year run where that was controversial, or even 70 years. And so the way I think about what's happening is basically the period we're in right now is, I call it an 80-year overnight success, right? Which is like it's an overnight success because it's like bam, you know, ChatGPT hits and then o1 hits and then, you know, OpenClaw hits and like, you know, these are overnight, like radical overnight transformative successes, but they're drawing on an 80-year sort of wellspring backlog, you know, of ideas and thinking. It's not just that it's all brand new, it's that it's an unlock of all of these decades of like very serious hardcore research. If I were 18, like this is 100% what I would be spending all of my time on. This is like such an incredible conceptual breakthrough.

Yes. Yes. a16z, something like that.

We have some projects underway, but yeah, this is actually the original office. We're in where the whole thing started.

I mean, look, I've been doing AI since the late 80s. So I don't know, like as far as I'm concerned this stuff is all Johnny come lately.

There was a big AI boom at the time. And there was names like expert systems and the era of like Lisp and Lisp machines. I coded in Lisp. I was coding in Lisp in 1989.

Although that's really when the Nvidia phenomenon really, I would say it was in that period when it was very clear that at the time the vocabulary was more machine learning, but it was very clear at that time that machine learning was hitting some sort of takeoff point.

And then everything that followed. But look, machine learning, you know, there were, look, I've been working, you know, I've been working with one of my kind of projects working with Facebook since 2004, on the board since 2007, and of course they started using machine learning very early and have used it basically for like 20 years for content, you know, feed optimization and advertising optimization. And obviously many, you know, financial services, many, many companies, many different sectors have been doing this. So, it's like one of these things. It's not a single thing. It's like layers, right? And the layers arrive at different paces and but they kind of build up over time and then, yeah, and then look, in retrospect it was 2017 was kind of the key point with transformer and then as you guys know there was this really weird like four-year period where it's like the transformer existed and then it was just like let's go.

Yeah. Well, but between 2017 and 2021, I mean that was the era of which like companies like Google had internal chatbots but they weren't letting anybody use them.

Right. And then OpenAI developed ChatGPT or GPT-2 and then they told everybody this is way too dangerous to deploy. Right. You know we can't possibly let normal people use this thing.

Yeah.

But that dinner would have taken place in 2018.

Yeah. 2015. Yeah. 2015.

17, 18. So yeah. And then they didn't really, and then GPT-3 was what, 2020?

So, there's something about, say the following. There's something about AI that has led to this repeated pattern. And you guys know this, it's summer-winter, summer-winter, and it goes back 80 years. 80 years. So the original neural network paper was 1943, right? Which is amazing that it was that far back. And then there was, if you guys have ever talked about this on your show, but there was a big AGI conference at Dartmouth University in 1955.

And they got the grant, they got the 10 weeks and then, you know, 1955, no AGI. And like I said, I lived through the 80s version of this where there was a big boom and a crash. And so there is this thing, there is something about AI that causes the people in the field, I would say, to become both excessively utopian and excessively apocalyptic. And it's probably on both sides of the boom-bust cycle, you kind of see that play out. Having said that, I think what's actually happened is just, you know, we now know in retrospect an enormous amount of technical progress that built up over time. And for example, we now know that neural network is the correct architecture. And I will tell you, there was a 60-year run where that was controversial, or even 70 years.

And so we now, you know, everything we're building on today sort of derives from the original idea in 1943. And so in retrospect, we now know that these guys were right. You know, they would get the timing wrong and they thought capabilities would arrive faster or it could be turned into businesses sooner or whatever. But they were fundamentally, the scientists who worked on this over the course of decades were fundamentally correct about what they were doing and the payoff from all their work is happening now. And so the way I think about what's happening is basically the period we're in right now is, I call it an 80-year overnight success, right? Which is like it's an overnight success because it's like bam, you know, ChatGPT hits and then o1 hits and then, you know, OpenClaw hits and like, you know, these are overnight, like radical overnight transformative successes, but they're drawing on an 80-year sort of wellspring backlog, you know, of ideas and thinking. It's not just that it's all brand new. It's that it's an unlock of all of these decades of like very serious hardcore research and thinking. I mean look, there were AI researchers who spent their entire lives, they got their PhD, they worked for, they researched for 40 years, they retired, in a lot of cases, they passed away and they never actually saw it work.

It is. It is sad. It is sad.

Yeah. Well, they were the guys, Alan Newell, I mean, there's tons of, John McCarthy, you know, John McCarthy was like one of the inventors of the field. He's one of the guys who organized the Dartmouth conference and, you know, he taught at Stanford for 40 years and passed away, I don't know, whatever, 10 years ago or something.

But like it is amazing in retrospect like these guys were incredibly smart and they worked really hard and they were correct. So anyway, so then it's like okay, you know, they say history doesn't repeat but it rhymes. It's like okay does that mean that there's going to be another like, you know, basically boom-bust cycle? And I will tell you, look, like in a sense like yes everything goes through cycles and people get overly enthusiastic and overly depressed and there's a timelessness to that. Having said that, there's just no question.

And so like I'll tell you what's different. Like now it's working. Like there's just no, I mean look, there's just no question. And by the way I'll just give you guys my take. Like LLMs, like from basically the ChatGPT moment through to spring of '25, I think you could still, I think well-intentioned well-informed skeptics could still say oh this is just pattern completion and oh these things don't really understand what they're doing and the hallucination rates are way too high and this is going to be great for creative writing and creating Shakespearean sonnets and rap lyrics or whatever, like it's going to be great at all that stuff but we're not going to be able to harness this to make this relevant in coding or in medicine or in areas that really, really matter. And I think basically it was the reasoning breakthrough, it was o1 and then R1 that basically answered that question and basically said oh no we're going to be able to actually turn this into something that's going to work in the real world. And then obviously the coding breakthrough over the holiday break was kind of the third step in that.

That's never happened before. And so now we know that it's going to sweep through coding. And then we know, you know, we know that if it's going to work in coding it's going to work in everything else, right? It's just then because that's like the hardest, in many ways that's the hardest example and now everything else is going to be a derivative of that.

And incredibly powerful. And then we just got the auto research, you know, the self-improvement, you know, we're now into the self-improvement breakthrough. And so the way I think about it is we've had four fundamental breakthroughs in functionality: LLMs, reasoning, agents, and then now RSI. And they're all actually working. And so I'm just, as you can tell, I'm jumping out of my shoes. Like this is it. Like this is the culmination of 80 years worth of work and this is the time it's becoming real.

It'll keep happening.

Yeah. So, a couple things. So, one is like look, Moore's Law was what we now call a scaling law. Like Moore's Law was a scaling law. And for your younger viewers, Moore's Law was chips either get twice as powerful or twice as cheap every 18 months. And that, and then, you know, that it's gotten more complicated in the last few years, but that was like the 50-year trajectory of the computer industry. And then by the way, and that's what took the mainframe computer from a $25 million current dollar thing into the phone in your pocket being, you know, a million times more powerful than that for 500 bucks. And so that was a scaling law. And then key to any scaling law including Moore's Law and the AI scaling laws is, you know, they're not really laws, right? They're predictions but when they work they become self-fulfilling predictions because they set a benchmark and then the entire industry, right, all the smart people in the industry kind of work to make sure that that actually happens. And so they kind of motivate the breakthroughs that are required to keep that going. And in chips that was a 50-year run, right? And it was amazing. And it's still happening in some areas of chips. I think the same thing is happening with the core scaling laws in AI. You know, they're not really laws, but they are basically predictions and then they're motivating catalysts for the research work that is required to be, and by the way, also the investment dollars are required to basically keep the curves going. And look, it's going to be complicated and it's going to be variable and there are going to be walls that are going to look like they're fast approaching and then engineers are going to get to work and they're going to figure out a way to punch through the walls. And obviously that's been happening a lot. And then look, there's going to be times when it looks like the laws have petered out and then they're going to pick up again and surge. And then it appears what's happened to AI is there's now multiple scaling laws. There's multiple areas of improvement and I think, you know, I don't know how many more there are yet to be discovered but there are probably some more that we don't know about yet. You know, like for example there's probably some scaling law around world models and robotics that we don't fully, you know, kind of acquisition of data at scale in the real world that we don't fully understand yet. So that one will probably kick in at some point here. There's a bunch of really smart people working on that. And so yeah, I think the expectation is that the scaling laws generally are going to continue. Yeah, the pace of improvement will continue to move really fast. To your question on like what to build. So I'm a complete believer that the scaling laws are going to continue. I'm a complete believer the capabilities are going to keep getting amazing leaps and bounds. The part where I kind of part ways a little bit with what I would describe as the AI purist, which is, which I would characterize as like the people who are in many ways the smartest people in the field but also the people who spend their entire life like in a lab and have, I would say, have very little experience in the outside world. The nuance I would offer is the outside world of 8 billion people and institutions and governments and companies and economic systems and social systems is really complicated. And 8 billion people making collective decisions on planet Earth is not a simple process of like just, like you see this happening now, it's like a bunch of the AI CEOs have this thing which is just like well there's just this, they just all have this kind of thing when they talk in public where they're just like well there's just these obvious set of things that society needs to do.

And it's like how can society not, you know, whatever their theory is, how can society not see XYZ? And the answer is well society is, number one, there's no single society, it's like 8 billion and they all have a voice and they all have a vote at the end of the day of how they react to change. And then, you know, just like it's just human reality is just really complicated and messy. And so the specific answer to your question is like as usual it depends. You know, it depends. Look, there's no question people are going to, there's no question there are going to be companies, it's already happening, there are companies that think that they're building value on top of the models and they're just going to get blitzed by the next model. There's no question that's happening. But I think there's no question also that just the process of adaptation of any technology into the real messy world of humanity is just going to be messy and complicated. It's not going to be simple and straightforward. It's going to be messy and complicated and there are going to be a lot of companies and a lot of products and in fact entire industries that are going to get built that to basically actually help all of this technology actually reach real people.

Yeah. So, I should start by saying so I lived through the dot-com crash. And I can tell you stories for hours about the dot-com crash and it was horrible. No, it was awful. It was apocalyptic. By the way, a lot of the dot-com crash was actually at the time it was actually a telecom crash. It was a bandwidth crash. Like the thing that actually crashed that wiped out all the money was the telecom companies.

Global, yeah.

Actually, there was a scaling law in the dot-com era and it was literally the US Commerce Department put out a report in 1996 and they said internet traffic was doubling every quarter. And actually in 1995 and 1996 internet traffic actually did double every quarter. And so that became the scaling law. So what all these telecom entrepreneurs did was they went out and they raised money to build fiber anticipating that the demand for bandwidth was going to keep doubling every quarter. Doubling every quarter though is like, you know, grains of chess on the chessboard, like at some point the numbers become extremely large, right? And really what happened was the internet, by the way, continuously kept growing basically since inception, it's continuously grown, it's never shrunk.

But it wasn't doubling every quarter as of 1998, 1999. And so there was this gap in the expectation of what they thought was a scaling law versus reality and that's actually what caused the dot-com crash, which was they way overbuilt fiber, which is sort of the, by the way, fiber, telecom equipment, you know, so all the networking gear, and then by the way the actual physical data centers, like that was the beginning of the data center build and then data center overbuild. And so you had that but it was literally I think it was like $2 trillion got wiped out, right? It was like a big, and by the way the other subtlety in it was the internet companies themselves never really had any debt because tech companies generally don't run on debt but the telecom companies run on debt, physical infrastructure companies run on debt. And so the companies like Global Crossing not just raised a lot of equity, they also raised a lot of debt. So they're highly levered. And so then you just do the thing of just like, okay, you have a highly levered thing where you're overbuilding capacity. Demand is growing, but not as fast as you hoped. And then boom, bankrupt. Right. And then it's like they say about the hotel industry, which is it's always the third owner of a hotel that makes money, right? It has to go bankrupt twice, right? You have to wash out all of the overoptimistic exuberance before it gets to actually a stable state and then it makes money. So by the way all of those data centers and all of that fiber, it's all in use today but 25 years later. But it took, and actually the elapsed time was it took 15 years. It took 15 years from 2000 to 2015 to actually fill up all that capacity. The cautionary warning is the overbuild can happen. And you get into this thing where basically everybody who basically has any sort of institutional capital is like, 'Wow, it's just I don't know how to invest in these crazy software things, but for sure I can build data centers and for sure I can buy GPUs and I can deploy compute grids and all these things.' And so, you know, if you're a pessimist, you could look at this and you could say, 'Wow, this is like really set up to be able to basically replicate what we went through in 2000.' Obviously, that would be bad. The counterargument which is the one I agree with which is the counter on the other side is a couple things. One is the companies that are investing all the money are like the bluest chip of companies. And so back in the dot-com, like Global Crossing was like an entrepreneur, it's like a new venture. But the money that's being deployed now at scale is Microsoft and Amazon and Google and Facebook and Nvidia and, you know, these, and now OpenAI, Anthropic, which are now at really serious size as companies with very serious revenue.

And so this is institutional in a way that really wasn't at the time. And then the other is at least for now every dollar that's being put into anything that results in a running GPU is being turned into revenue right away. Like so, and you guys know this, like everybody's starved for capacity, everybody's starved for compute capacity and then, you know, all the associated things, memory and interconnect and everything else, data center space. And so every dollar right now that's being put in the ground is turning into revenue. And in fact I actually think there's an interesting thing happening which is because everybody's starved for capacity the models that we actually have that we can use today are inferior versions of what we would have if not for the supply constraints.

And so we're actually getting the sandbag version of the technology.

Yeah. But getting the good stuff is, it's just even if technical progress stops once there's like a much bigger build of GPU manufacturing capacity and memory, you know, all the things that have to happen in the course of the next 5 or 10 years, once it happens even the current technology is going to get much better. And then as you know, like there's just like a million ways to use this stuff, like there's just like a million use cases for this. Like this isn't just sending packets across a thing whatever and hoping that people find something to do with it. This is just like oh we apply intelligence into every domain of human activity and then it works incredibly well. Here's what I know. Here's what I know. In the next 3 or 4 years, it's like somewhere between 3 or 4 years out, basically everything is selling out. So, like the entire supply chain is sold out or selling out. And so there's no, like we're just going to have like chronic supply shortage for years to come. There's going to be a response from the market that's going to result in an enormous, you know, it's happening now, an enormous flood of investment in new fab capacity and everything else to be able to do that. At some point the supply chain constraints will unlock, you know, at least to some degree, that will be another accelerant to industry growth when that happens because the products will get better and everything will get cheaper. And so I know that's going to happen. I know the deployments, the actual use cases are really compelling. And then like I said, with reasoning and agents and so forth, I know they're just going to get much, much better from here. And so I know the capabilities are really real and serious. I also know that the technical progress is not going to stop. It is accelerating, like the breakthroughs are tremendous. I mean, even just month over month, the breakthroughs are really dramatic. And so, you know, I think if you were a cynic, and there are cynics, you can look at 2000, you can find echoes, but I can't even imagine betting that this is going to somehow disappoint in, you know, at least for years to come. I think it would be essentially suicidal to make that bet.

Because the pace of improvement of the software is faster than the depreciation cycle of the chip. And then my understanding is Google is running, I don't think, I don't know exactly what these are, rumors that I've heard or maybe it's public, but I think Google's running very old TPUs.

And very profitably. And so it actually turns out, as far as I can tell, it's actually the opposite of the Burry thesis. He was actually 180 degrees wrong. It's actually the old Nvidia chips are getting more valuable.

Yeah. Yeah. One of my, like an invitation to get your face ripped off.

Well, so just to start, yeah. So I think open source is very important for a bunch of reasons. I think edge inference is very important for a bunch of reasons. I think just practically speaking if we're just going to have fundamental supply crunches for the next, I mean you guys know if you just project forward demand over the next three years relative to supply, one of the dismaying predictions you can do is what's going to happen to the cost of inference in the core over the next three years and like it may rise dramatically. Right? So what is, and then as you know, the big model companies are subsidizing heavily right now. And so what's the, what will be the average person's, you know, per day, per month token cost, you know, three years from now to do all the things that they want to do? And I don't know what it's going to make. I mean, I have, you guys probably have friends, I have friends today who are paying $1,000 a day for OpenClaw tokens to run OpenClaw, right? And so, okay, $30,000 a month, right? And by the way, those friends have like a thousand more ideas of the things that they want their claw to do, right? And so, you could imagine there's like latent demand of up to, I don't know, five or $10,000 a day of tokens for a fully deployed personal agent. And obviously, consumers can't pay that, right? And so, but it gives you a sense of the future scope of demand, right? And so even if there's a 10x improvement in price performance that still goes to $100 a day which is still way beyond what people can pay. So there's just going to be like ferocious demand. By the way the agent thing, the other interesting thing is I think the agent thing, so up until now a lot of the constraints have been GPU...

I think the agent thing now also translates into CPU constraints.

CPU and memory. Right. And so the entire chip ecosystem is just going to get...

It's all bottlenecked potentially for years. And so I think that Brad and I think it's actually possible. I mean generally inference costs are going to keep coming down but I think the rate of decline I think may level out here for a bit because of these supply constraints and then at some point maybe the lab stops subsidizing so much and that again will be an issue. And so there's just going to be so much more demand for inference than can be satisfied with the centralized model. And then you guys know this but like all the dramatic innovations that have happened in the Apple silicon to be able to do inference is quite amazing. The level of effort being put, like the open source guys are putting incredible effort into getting this recurring pattern where the big model will never run on a PC and then 6 months later it runs on a PC, right? It's like amazing and there's very smart people working on that. So there's all that and then look there's also other motivators which is just like okay how much trust are the big centralized model providers building in the market versus how much are, at least in certain cases with some people for certain use cases, people being like well I'm not willing to just turn everything over. So there's all the trust issues. By the way, there's also just straight up price optimization. There's many uses of AI where you don't need Einstein in the cloud. You just need a smart local model. There's also performance issues where you want your doorknob to have an AI model in it to be able to do access control. Obviously, like everything with a chip is going to have an AI model in it. And a lot of those are going to be local. And so yeah, I think you're going to have, and then by the way, also wearable devices, you don't want to do a complete round trip. You want your whatever your smart devices are, you want it to be super low latency.

Yeah, we'll see. We'll see. Look, number one, I do think we care. I do think we care who makes it. I would say this, the previous presidential administration wanted to kill it in the US. Like they wanted to drown it in the bathtub. And so they wanted to kill it. So at least we have a government now that actually wants it to happen.

Yes. And the new and the past. Yeah. So this administration for whatever other political issues people have which are many, this administration has I think a very enlightened view and in particular an enlightened view on AI and in particular on open source AI. And so they're very supportive. My read is the Chinese have a very specific reason to do open source which is that they don't fundamentally think they can sell commercial AI outside of China right now or at least specifically not in the US for a combination of reasons. And so they kind of view open source AI as a bit of a loss leader against basically domestic paid services and then kind of ancillary products. They're very excited about it. By the way I think it's great that they're doing it. I think DeepSeek was like a gift to the world. The great thing about open source is the impact is felt two ways. One is you get the software for free but the other is you get to learn how it works, right? And so like the paper and the code. And so for example I thought this was amazing. So OpenAI comes out with o1 and it's an amazing technical breakthrough and it's just absolutely fantastic but of course they don't explain how it works in detail and then of course they hide the reasoning traces. And then everybody's like okay this is great but who's going to be able to replicate this? Are other people going to be able to do this? Is there secret sauce in there? And then R1 comes out and it's just like there's the code and there's the paper and now the whole world knows how to do it. And then 3 months later, every other AI model is adding reasoning. And so you get this kind of double, like even if the Chinese models themselves are not the models that get used, the education that's taken place to the rest of the world, the information diffusion, is incredibly powerful. So that happens and then I don't know, we'll see, there are a bunch of American open source AI model companies. I mean, look, there's going to be tremendous, there already is, there's going to be tremendous competition among the primary model companies. There's, depending on how you count, there's like four or five big model companies now that are kind of neck and neck in different ways. And then obviously both OpenAI and then Meta are involved, both have huge attempts to kind of leapfrog underway. And then you've got a whole fleet of startups, new companies, including a whole bunch that we're backing that are trying to come out with different approaches. And then you've got whatever it is, I don't know how many mainline foundation model companies are there in China at this point. It's probably six.

ByteDance would be like the next tier. They weren't as prominent. They won't have a... Yeah. But at least you know Seedance is very inspiring and presumably they have more stuff coming and Tencent probably has more stuff coming and so forth. And so look here would be a thing you could anticipate which is there are not going to be, between the US and China right now there's like a dozen primary foundation model companies that are at scale at some level of critical mass. It's not going to be a dozen in three years, right? Like just because these industries don't bear a dozen. It's going to be three, there's going to be three or four big winners or maybe one or two big winners. And so there's going to be a whole bunch of those guys that are going to have to figure out alternate strategies. And I think open source is one of those strategies. And so I think you could see a whole, I think the question like who's going to do open source I think that could change really fast. I think that's a very dynamic thing. I think it's very hard to predict what happens and I think it's very important.

Well, I was gonna say well exactly and then you got Nvidia and then just again there's an old thing in business strategy which is called commoditize the complement and it's right. And so if you're Jensen it's just kind of obvious of course you want to commoditize the software and to his enormous credit he's putting enormous resources behind that. And so maybe it's literally Nvidia and I think that would be great.

Yeah. So I think the combination of the two of them I think is one of the 10 most important software... OpenClaw got all the attention but talk about Pi. Pi is kind of the architectural breakthrough. For those of us who are older, there was this whole thing that was very important in the world of software basically from like 1970 to I don't know it still is very important but from 1970 through to basically the creation of Linux which was basically this thing we used to call the Unix mindset. Because there were all these different operating systems and mainframes and then all these Windows and Mac and all these things and then there was this but kind of behind it all was this idea of the Unix mindset. And the Unix mindset was this thing where basically you don't have these like, in the old days like the operating system that made the computer industry really work in the 1960s was this thing called OS/360 which was this big operating system that IBM developed that was supposed to basically run everything and it was this giant monolithic architecture in the sky. It was like a giant castle of software. And by the way it worked really well and they were very successful with it but it was this huge castle in the sky but it was this thing it was almost unapproachable which is like you had to be kind of inside IBM or very close to IBM and you had to really understand every aspect how the system worked. And then the Unix guys originally out of AT&T and then out of Berkeley came out and they said no let's have a completely different architecture. And the way the architecture is going to work is we're going to have a prompt and a shell and then all the functionality is going to be in the form of these discrete modules and then you're going to be able to chain the modules together. And so it's almost like the operating system itself is going to be a programming language. And then that led to the centrality of the shell. And then that led to basically chaining together Unix tools and then that led to the emergence of these scripting languages like Perl where you could basically very easily do this. And then the shells got more sophisticated and then look, that worked and that was the world I grew up in. I was a Unix guy from call it 1988 all the way through my work and it worked really well. It's in the background. Normal people didn't need to necessarily know about it, but if you were doing system architecture, application development, you knew all about it. And then it's been in the background ever since. And look, your Mac still has a Unix shell in there and your iPhone still has a Unix shell kind of buried in there somewhere. So they're kind of in there. And then the Windows shell is kind of a weird derivative of that. But look, the internet runs on Unix. And then smartphones, actually both iOS and Android are Unix derivatives. And so Unix did end up winning but anyway and then we just started taking that for granted. And so basically the way I think about what happened with Pi and then with OpenClaw is basically what those guys figured out is I always say the great breakthroughs are obvious in retrospect, right? Which is...

The best kind. They weren't obvious at the time or somebody else would have done them already. And so there is a real conceptual leap but then you look at it sort of the backwards looking and you're just like oh of course.

Yeah, exactly. But like it right. But she's even saying it wasn't obvious until somebody actually did it, right? And so the conceptual breakthrough is real and deep and powerful and very important. And so the way I think about Pi and OpenClaw is it's basically marrying the language model mindset to the Unix basically shell prompt mindset. And so it's basically this idea that what is an agent, right? And as you know like many smart people have been trying to figure out what an agent is for decades and they've had many architectures to build agents and the whole thing. And it turns out what is an agent? So it turns out what we now know is an agent is the following. It's a language model. And then above that it's a bash shell. So it's a Unix shell and then the agent has access to the shell, hopefully in a sandbox. So it's the model, it's the shell, and then it's a file system. And then the state is stored in files and then there's the markdown format for the files themselves and then there's basically what in Unix is called a cron job. There's a loop and then there's a heartbeat and the thing basically wakes up. So it's basically LLM plus shell plus file system plus markdown plus cron and it turns out that's an agent. And every part of that other than the model is something that we already completely know and understand. And in fact it turns out that the latent power of the Unix shell is extraordinary because there's just enormous latent power in the shell. There's enormous numbers of Unix commands. There's enormous number of command line interfaces into all kinds of things already. I mean your entire computer runs in a shell if you're running a Mac or a phone, your computer is running on a shell already. And so the full power of your computer is available at the command line level. And then it turns out it's really easy to expose other functions as a command line interface. And so this whole idea where we need like MCP and these fancy protocols, whatever, it's like no we don't, we just need a command line thing. So that's the architecture. And then it turns out what is your agent? Your agent is a bunch of files stored in a file system. And then there's the thing that just completely blew my mind when I wrapped my head around it as a result of this, which is like, okay, this means your agent is now actually independent of the model that it's running on because you can actually swap out a different LLM underneath your agent and your agent will change personality somewhat because the model is different, but all of the state stored in the files will be retained.

Right? Exactly. And it's like, right, swapping out a chip and recompiling, but it's still your agent with all of its memories and with all of its capabilities. And then by the way you can also swap out the shell. So you can move it to a different execution environment that is also a bash shell. By the way you can also switch out the file system and you can swap out the heartbeat, the cron framework, the loop, the agent framework itself. And so your agent basically at the end of the day it's just its files. And then of course yeah it's basically just the files. And then by the way as a consequence of that the agent itself it turns out a couple important things. So one is it can migrate itself, right? And so you can instruct your agent migrate yourself to a different runtime environment, migrate yourself to a different file system, migrate yourself to a different, you swap out the language model. Your agent will do all that stuff for you. And then there's the final thing which is just amazing which is the agent actually has full introspection. It actually knows about its own files and it can rewrite its own files.

And then that leads you to the capability that just completely blew my mind when I wrapped my head around it which is you can tell the agent to add new functions and features to itself and it can do that.

Right, extend yourself, give yourself a new capability, right? And so literally it's like you run into somebody at a party and they're like, 'Oh, I have my OpenClaw do whatever. Connect to my Eight Sleep bed and it gives me better advice and sleep.' And you go home at night and you tell your Claw, or if they're at the party, by the way, you tell your Claw, 'Oh, add this capability to yourself.' And your Claw will say, 'Oh, okay, no problem.' And it'll go out on the internet and it'll figure out whatever it needs and then it'll go out to Claude Code or whatever. It'll write whatever it needs and then the next thing you know, it has this new capability. And so you don't even have to, you can have it upgrade itself without even having to do anything other than tell it that you want it to do that. And so anyway, so the combination of all this is just, I mean this is just a massive incredible, I mean it's just incredible. Like if I were 18 this is 100% what I would be spending all of my time on. This is like such an incredible conceptual breakthrough. And again people are going to look at it and they already get this response. People are going to look at it and they're going to say oh where's the breakthrough because all of these components were already known before. But this is the key, the key to the breakthrough was by using all these components that were known before you get all of the underlying capability that's buried in there. And so for example, computer use all of a sudden just kind of falls trivial. Of course, it's going to be able to use your computer. It has full access to the shell, right? And then you give it access to a browser and then you've got the computer and the browser and off and away it goes. And then you've got all the abilities of the browser also. And so the capability unlock here is profound. My friends who are deepest into this are having their Claw do literally like a thousand things in their lives. They have new ideas every day. They're just constantly throwing new challenges at the thing. And by the way, it's early and these are prototypes and there's security issues and so there's a bunch of stuff to be ironed out, but the unlock of capability is just incredible. And I have absolutely no doubt that everybody in the world is going to have at least an agent like this, if not an entire family of agents, and we're going to be living in a world where...

There were a whole bunch but I'll give you a couple. So one is, and to be clear this was totally different. We didn't have the capabilities we have today because we didn't have the language models underneath this but we did have this idea that human readability actually mattered a great deal. And specifically in those days it was not so much English language but there was a design decision to be made between binary protocols and text protocols. And basically every old school systems architect that had grown up between like the 1960s and the 1990s basically said you know the internet is starved for bandwidth. You have these very narrow straws. Look, when we did the work on Mosaic, people who had the internet at home had a 14 kilobit modem, right? So you're trying to hyper optimize every bit of data that travels over the network. And so obviously if you're going to design a protocol like HTTP, you're going to want it to be binary, highly compressed binary protocol for maximum efficiency. And you're going to want to have it be a single connection that persists and the last thing you're going to want to do is bring up and tear down new connections. And you definitely are not going to want a text protocol. And so of course we said no, we actually want to go completely the other direction. It's obviously we only want text protocols. By the way, same thing in HTML itself. We want HTML to be relatively verbose. We want the tags to actually be human readable. We want to use...

Yeah. We want to do the inefficient things.

Yeah. Exactly. Yeah. Basically...

Well, actually this was the conscious thing which basically says just assume a future of infinite bandwidth, build for that. And then basically what it was is it was a bet that if the latent capabilities of the system were powerful enough and that was obvious enough to people that would create the demand for the bandwidth that would cause the supply of bandwidth to get built that would actually make the whole thing work. And then specifically what we wanted was we wanted everything to be human readable because we at the engineering level we wanted people to be able to read the protocol coming over the wire and be able to understand it with their bare eyes without having to disassemble it or whatever, right? Have it converted out of binary, right? And so all the protocols, HTTP and everything else were always text protocols. And the same thing with HTML and in many ways some people say that the key breakthrough in the browser was the view source option which is every web page you go to you could view source which means you could see how it worked which means you could teach yourself how to build new web pages. There was that human readability. And again human readability in those days still meant technical specs, now it means English language but there's an incredible latent power in giving everybody who uses the system the option to be able to drop down and actually understand how it's working and that worked really well for the web and I think it's working really well for AI. That was one. What was the other? A big part of the idea of web servers was to actually surface the underlying latent capability of the operating system and to be able to surface also the underlying latent capability of the database because basically what was a web server? What is a web server fundamentally architecturally? It's the operating system. So it's the operating system's ability to, it's running on top of an OS. So it's the OS's ability to manage the file system and do everything else that you want to do and process everything. And then of course a lot of early websites are front ends to databases. And so you wanted to unleash the underlying latent power of whether it was an Oracle database or some other Postgres or whatever it was. And so a lot of the function of the web server was to just bridge from that internet connection coming in to be able to unlock the underlying power of the OS and the database. And again people looked at it at the time and they were like well does this really matter? Like is this important because we've had databases forever and we've always had user interfaces for databases and this is just another user interface for a database. It's like okay yeah fair enough. But on the other side of that, it's just like this is now a much better interface to databases and one that eight billion people are going to use and is going to be far easier to use and far more flexible and you're not just going to have old databases. Now you have a system where people can actually understand why they want to build a million times more database apps than they had in the past. And then the number of databases in the world exploded. And so again, this goes to this thing of building in layers. Some of the smartest people in the industry look at any new challenge and they're like, 'Okay, I need to build a new kind of application. So the first thing I need to do is build a new programming language, right? And then the next thing I need to do is build a new operating system, right? And then the next thing I need to do is I need to build a new chip, right?' And they kind of want to reinvent everything. And I've always had maybe it's just a pragmatic mentality or something or maybe an engineering over science mentality, but it's more like no, you have all of this latent power in the existing systems and you don't want to be held back by their constraints, but what you want to do is you want to liberate that power and open it up.

And so I think the web did that for those reasons and I think it's the same thing now that's happening.

I mean I think this is going to change a lot because I don't think the models care what language they program in. And I think they're going to be good at programming in every language. And I think they're going to be good at translating from any language to any other language. Like, okay, so this gets into the coding side of things. I think we're going through a really fundamental change. And look, I grew up hand coding. Everything I did was actually written in C. I was...