Martin Casado - Entering Uncharted AI Territory - [Invest Like the Best, EP.381]

0:00 I know firsthand how complex the tech stack is for asset management firms. And seemingly every new tool and data source makes the problem even worse, adding more complexity, more headcount, and more risk. Ridge line offers a better way forward, one unified platform that automates away the complexity across portfolio accounting. Reconciliation, reporting, trading, compliance, and more, all at scale. Ridge line is revolutionizing investment management, helping ambitious firms scale faster.

0:25 Operate smarter and stay ahead of the curve. See what Ridgeline can unlock for your firm. Schedule a demo at ridgeline.ai. Hello and welcome everyone. I'm Patrick O'Shaughnessy and this is Invest Like the Best. This show is an open ended exploration of markets, ideas, stories, and strategies that will help you better invest both your time and your money. Invest Like the Best is part of the Colossus family of podcasts, and you can access all our podcasts, including edited transcripts, show notes, and other resources to keep learning at joincolosis.com

1:00 Patrick O'Shaughnessy is the CEO of Positive Sum. All opinions expressed by Patrick and podcast guests are solely their own opinions and do not reflect the opinion of positive some. This podcast is for informational purposes only and should not be relied upon as a basis for investment decisions. Clients of Positive Sum may maintain positions in the securities discussed in this podcast. To learn more, visit psum.vc.

1:29 My guest today is Martin Cassado. Martine is a partner at Andreessen Horowitz. And first join me on Invest Like the Best in 2022. So much has changed since then, and it was awesome to have Martine back to discuss all the different implications. of this modern AI revolution.

1:44 Before joining A sixteen Z Martine pioneer software defined networking and co founded Naikira. Which was bought by VMware for one point three billion in two thousand twelve. He has studied, built, and invested in digital infrastructure his whole career.

1:58 Which has primed him to go in depth in this interview on the immerse opportunities and challenges AI presents among creativity. Policymaking, agentich systems, real world data structures, and beyond. Please enjoy this great conversation with Martin Casada. Martine, I'm so happy to do this with you for a second time, and there is so much to talk about that's going on in the world of technology, most specifically AI.

2:21 What do you think the twenty thirties? Will feel like. The twenty thirties So I think the future is gonna be a lot weirder, wonderful, and more bizarre than anybody is imagining. I'm just talking about this AI stuff.

2:37 I feel like one of the mistakes we're making as an industry Yeah. AI has been around for a long time. assumed it will do certain things like it'll do chat bots and it'll automate the enterprise. Try to cram these new advances into that old one.

2:51 But it's not clear to me that that's actually the best use of it. A lot of the problems it solves are problems that haven't solved on computers like creativity and the emotional connection. And I mean actually look at the businesses. These are the things that are working. And so I Think we're this free internet world where we can't really imagine what it's gonna look like, but we're solving different problems. The glimpses that we see is just I think we're gonna have this creative social explosion as opposed to A oh we've solved an enterprise automation problem.

3:21 I don't know if that's what you're asking, but I I actually think we should all hold on and be just ready for a trippy, amazing, wonderful ride. What do you think is zooming all the way back to today is the most important next bottleneck on the path to that trippy awesome ride. What things if they didn't go a certain way would change your opinion and say, Ah, actually maybe we're actually on a more traditional marginal incremental growth plan.

3:43 I'll be honest, I think that for some subset of things, I think the cat's out of the bag, the penis out of the bottle, whatever stupid aphorism you want to use. When it comes to content. creation, like any sort of digital creation. The marginal cost has gone to zero. That's like asking After the creation of the internet. What's gonna stop its proliferation and I just think the laws of physics are now that

4:03 The economics are all gonna shift. We should be generating content and all this creative content with AI and I think it's just gonna naturally happen. But just like in the early days of the internet, I don't think we know what that's gonna look like. And so we know that there's this economic dislocation, we know there's a new way of doing things. I think it's a question of time scales, not whether that's gonna happen. I think the biggest existential threat to address your question directly is there's always this type of exuberance where people see the economic benefit, but the actual building the right apps and the demand takes a lot longer. And so you definitely have these big chill periods. Like I remember very well the fibergut. to the internet, we're like, Yeah, everybody's gonna be on the internet. We built all of this fiber. And then it turns out when the supply cycles are different than demand cycles. So the time it takes to actually build a data center, for example, in the AI and get GPUs to people versus like demand, when they're out of whack, you often have these overbuilds and then you have these gluts and then prices drop and you have so I think we'll go through some of those and that will slow stuff down.

4:59 But I think that the long term trajectory is pretty locked in from this point and this stuff is gonna change things in ways that we don't quite understand. I'd love you to lay it out as I've seen you do elsewhere where it's like the third marginal cost of X. asymptoting down to zero as the history of modern technology and why that's so important. Yeah, for sure. credit where credit's due. A lot of this came out of a conversation I was having with Ben Silverman.

5:20 The seum powder of Pinterest. And we've seen at least two cases in the past where marginal costs have gone to zero and it just changed the world. And so the first one was with the microchip.

5:32 So the reason computers are call computers is actually that name computer came from a job a human used to have, which was to computer. And it's very tough. Do understand where ballistics are gonna go. And to do this you have to create these crazy logarithm tables. Pre nineteen forties. The way that you would

5:50 Big out ballistics projectors, you'd have people in rooms calculating these big logarithmic tables. And then Eniak shows up and Eniak is Five thousand times faster than a human. So almost four orders of magnitude. Faster than a human being.

6:04 Yeah, so the Marshall Job Awesome. Uh computing goes to zero. And as a result of that, you have the entire compute revolution because if it's so cheap to compute, let's compute for everything, right? And it's everything's

6:17 It's IBM deck and HP and all of that came out of that. And then starting in the seventies, but really in the eighties and nineties, the internet showed up and now the marginal cost of distribution goes to zero. I remember when I used to buy like pre internet. I remember pre internet when I would go whatever buy a video game, like I would go to a store and I'd buy a box, and that box probably had been shipped there for two weeks. So it just takes a very long time to get the software. So When Marchall

6:41 cost of distribution goes to zero, you have the internet wave. And again, it was one of these things where Once it happened, the entire world shifted and rather than you having less, you have so much more because it's so cheap and it's so much more accessible. And so one way to think about this current AI wave is the marginal cost of content creation is going to zero.

7:02 And to some extent natural language reasoning is going to zero. And a great example of this is Let's say that I wanted to create a likeness of myself that looks like a Pixar character. So like the Martine Pixar character. So the traditional way to do this I go to a designer and let's say the designer is a hundred bucks an hour and it takes them two hours. Which having actually done this before, they're more expensive and they take longer. But let's just say that. So say it's two hundred bucks. If you compare that

7:27 to the inference cost from one of these models. Like the inference cost to create a Martinez of Pixar character is like a hundredth of a penny. So again, we're four orders of magnitude, almost the exact same difference between any act and a human computer we have between a model and creating a likeness of me. And just one last point on this, we can dig into any aspect that you want. If you want to Have a mental conception How big disinflection point could potentially be, imagine a triple A video game.

7:54 So a triple A video game can cost a billion or more to create let's say it's five hundred million dollars. There actually isn't one aspect of that video game. that you couldn't create today with a model. Whether it's the music or the sound effects or the textures or the three D meshes or the characters or the video cutscenes or the images or whatever it is. All of these things

8:13 And so rather than being five hundred million, the actual cogs cost should be something like ten bucks. So this is just this massive Now what we've learned in the past is when these costs drop, it's not like it wipes out costs, not a zero sum thing, it's just we end up generating so much more stuff and we enable so many more applications because it become so much more accessible. So I think we're gonna see this massive kind of creative explosion as a result. Do you think about it beyond

8:39 Content. Content is one thing. There already is a lot of content because the marginal cost of distribution went to zero. So a lot of human brains created a lot of content, which still is much more expensive than these models. I certainly buy how much easier that just seems like a foregone conclusion. What about beyond content? What about The robots loading the dishwasher type of stuff. The things that start eating outside of just the digital world or in different parts of the digital world to other kinds of work.

9:03 So there's the content thing. Let's work our way to the robots'cause I think that's maybe the last bash, right? So there's the content thing Then there's the emotion thing, which I do feel like humans have some imperative for connection. And it just turns out these models are very good for that.

9:18 I don't know the true numbers for this, but I believe if you draw a dollar at random being made with these current models today, at least a third of that are people that are just creating an emotional connection, whether it's through a virtual friend, whether it's through these new style of video games where you're actually really interacting, whether it's a virtual girlfriend or virtual boyfriend, like whatever it is. There are many companies that we see at Edreas and Horrorwoods and that are out there that are building these we call it companionship, like these kind of companions. And so that's different than content creation. It's actually emotional connection. It's not a modality that computers have been able to solve in the past and they do now. So I think that's very much another area. And then a third one is just basic language reasoning.

9:55 So a way to think about this is I work with a number of lawyers. Let's say I pay them seven hundred bucks an hour. And they do very smart things, but a lot of these language reasoning things. Let's say I've got a new legal document PDF. So if I have a choice of Giving that PDF for the lawyer to read.

10:11 The lawyer spends three days reading it. The lawyer gives me s answers some questions based on it. Some of those questions are right, some of those questions Since they're wrong, like that's twenty one hundred bucks. Or I can do the same thing with an LLM. I think a lot of those language reasoning tasks are gonna go to these models'cause they do a pretty good job of the language reasoning piece. And so I think those three things we very clearly see In advancement and I think that it's at the point that we know these are gonna be big businesses.

10:35 But you asked a very important question. Which is when'll the robots show up. Let me just give you one anecdote. When I joined Stanford for my PhD was in two thousand and three.

10:45 And around that time Sebastian Thrun had just won the DARPA grand challenge. They've driven a tonerously of dam through the desert for twelve hundred miles. And we're like Oh my god, amazing hooray. Robot taxis are almost here.

11:00 This was twenty years ago. And in that time there's a very important thing for everybody to realise We as an industry have invested seventy five Billion dollars. In autonomous vehicles.

11:12 And if you actually look at the unit economics of the cars that are working on the street today, they're still more expensive than Uber by quite a bit. There are certain modalities that AI is trying to tackle where They're better than humans. They're quite a bit better than humans. They solve very real problems, but the economics

11:29 are very hard. And one way to think about that is is Another AI or another computer system. The competition is a human brain. And if the human brain is really good at something, it's like navigating a three D space. It's not just a human brain. This is probably like the four million year old brain has been doing this and it's very efficient at that.

11:49 If you take one of these autonomous vehicle cars and you look at how much wattage just the CPUs and GPUs burn, some of these are like one point three kilowatt rigs for the GPU setup for all the planning and all the inference, et cetera. Like they're very like power heavy. The human brain consumes about twenty watt. You're competing with the most efficient part of the human brain. Now compare that to the three things I talked about before, the creativity, the language reason, et cetera. That primarily comes from the prefrontal cortex, which is about two hundred fifty thousand years old. That human brains.

12:16 not very optimized. So it turns out the computers are much, much better at the newer Part of The brain solving that. So if you take an economic view of this, I think you can say that Solving the real world

12:27 Problem. is hard. For two reasons, one I have not talked about. One of them is just the human being is so good at navigating the real world. And you're competing with that and we don't know how to do that economically. Well, solving the content craving language reasoning is easier because the human brain isn't optimized for that. Okay, that's one.

12:43 There's a second point and it's a subtle one, but I just think it's so important. Let's try to dig into it as much as we can. It's important. There's a very reasonable view in all of this that what these models are doing Is all they're doing is tapping into human created data. And they're able to reproduce more human created data. That's what they're exploiting is data created by humans.

13:02 And that means that The creation of the data is the hard part. It's like a human brain. Look at this world. and decided that's a tree or that's a book and reduced it to something that's more linear and more structured.

13:16 And then these models just learn from that structure. as opposed to actually learning about the world. So let's just talk about language. As a human species, we've spent three thousand years writing down language and creating all of this content. It could be the case that these LLMs have just taken all of those tokens, those three thousand years of tokens, and just learned all of the structure in those tokens.

13:37 And then it can reproduce More tokens like a human being would. But there's nothing fundamental about the world in those. It's just the tokens themselves. It's a self encapsulated thing. When you're talking about Robots necessarily need to interact

13:50 But the physical. In the physical world, as we all know, it's chaotic, it's fractal, it's self similar, it's nonlinear, it's just very complex. And it's not clear at all that we know how to reign in that type of complexity. with these current models. We know that we can take human created content And that human created content, we can produce more human created stuff. We can't take real world stuff and produce more real world stuff yet.

14:13 That is still an outstanding problem. These are such important foundational points. So what's the chicken and what's the egg here? You can imagine that if you had a hundred million robots walking around, I guess there's cars gathering lots of data on streets, but if you had a hundred million humanoid robots doing stuff, you would be creating this insane data set over time. There could be some sort of data training feedback loop or something. But what do you see as the way to solve the chicken and egg problem that you just described?

14:37 I just don't know if we even know if it's a data problem. Meaning human created stuff. Just has enough structure to exploit. There's a great recent where I don't know if it got accepted, I just saw it on Twitter, but I just thought it was such a brilliant piece of work by Rohan, by the way, at the AJI house. Where he showed

14:54 That If you have a data corpus and you G zip it. Use compression. If it compresses a lot. Then the model learns a lot.

15:03 On the same data. Which suggests that The models are just exploiting structure. in the human created data. So it could be the humans that are actually adding the structure of the data that they're learning from.

15:14 That's a very different statement than saying You get data from the natural universe and you train a model on that. And so it could be that the current AI way is honestly just exploiting human created content because we did the hard work of creating structure and that's economically viable. And that's why we can make pictures

15:33 That are economically viable and videos that are economically viable and text that's economically viable. But we have yet to see is how do we look at the universe. And trade on that. And then use that to build an economically viable machine. Which again we have yet to do even with cataracts vehicles, even though we spent a lot more money on that. And so there's two questions. You asked a fundamental question, which is can you just go with a bunch of cameras

15:57 and look at the world and get enough data to train a model that's economically viable. We don't know the answer to that question. Cause you're competing with a human brain that's very good at that. It's almost become a cliche that any time somebody says A B it's gonna be done this is very famously Something that Tesla's had a hard time with, for example. It just feels like

16:15 The universe is very heavy tailed. There's lots of exception cases. And it takes a long time. The basics is I would not look at a model's ability to read text created from humans and be predictive, I would not compare that to a model's ability to look at the universe without human curation and come up with that. We don't know if that could be economically viable in the near term. We don't even know if it's possible.

16:36 That's just the state. And so really the first three that you mentioned companionship, creativity, productivity. is like this sort of known space and then Beyond that is this sort of It's beyond the horizon. We just don't know when and if it's coming.

16:51 I think that's exactly right. I would say The ones that are working are really working. Like these businesses are really working. They're solvent, they're profitable, they're growing, they're solving real spaces. But For anybody listening to this, for me this is such an aha. Almost all of them, basically all of them are trained on human created data. So humans create data, whether it's like conversation For companionship, whether it's I created art.

17:12 And if you train a model on that, it's economically viable because A, this is what humans are interested in. And there's already a bunch of structures so it can do that. It's very hard to come up with examples where model's been trained with the universe. whatever, it was streets or roads or any and they actually end up being useful. And you could say that This is because the data is just too chaotic, there's too much entropy.

17:33 It's too heavy tailed. Or you could say brain is just too good on the natural universe and we can't compete. And I don't know which one of those is true. My suspicion is probably both of those things are true. So it's just not clear at all to me that problem is going to get cracked in the near term. Maybe it will. I'm not saying it won't, but it's not obvious to me. But for sure anything that humans can create We know how to train model. I've always thought about

17:54 the bitter lesson idea that if we just throw more compute and more data at the problem, it'll just get better. as it being this like raw data. I've never thought about it as this like human created data structure is the thing. It's really a fascinating idea. So does that mean you think that maybe the bitter lesson stops applying? The solution might not just be I think you've already said this, but might not just be

18:15 More collection of data, more Nuke powered, you know, data centers running GPU clusters or whatever, there may be some other thing needed to get to that next phase. Let's tease this apart for what we know. We know that if human beings spend a lot of time creating data sets and we train on those data sets, then we we can get predictive models that are economically viable.

18:38 We know that. But once those data sets are exhausted. What are our options? And in this case we've literally exhausted the data created in the last three thousand years.

18:49 Spin some new ones up. Yeah, yeah, yeah, right. So we just know that to be case. And so you can make an argument that what we're drafting on is all of the effort and the energy that human beings as a species have done to take the universe and make it well formed, and then we're exploiting that structure. In that case. The structure It's like a reserve of energy, the reservoir that was created by humans that were exploiting. And then once that's gone, we've made this massive step function, but we either need to find new ways to create

19:17 data which we're actually seeing now. A lot of these LLMs of the budgets are moving to post training. and creating data sets. So we're already seeing that happen. Like you have thousands of people out there creating code snippets or annotating sentences or whatever it is. We're already seeing it. that happening, but you can't compare like a thousand people writing code. with all the code ever written before. So that will definitely slow down.

19:39 On that. When it comes to compute so then everybody says, Well what about synthetic data? Certainly in certain domains, synthetic data makes sense. If you have an axiomatic system Where you can fill out the state space computationally, like

19:52 Playing chess. Or arithmetic. Or even some spati where you can simulate things. A hundred percent you can simulate data. But that's not a general statement. What does it mean to create synthetic language? What does that even mean?

20:07 Language is not axiomatic. It's like this kind of ambiguous thing. What does it mean to synthetically do creativity? At the end of the day, we don't really know how to synthetically recreate the stuff that human beings have created in the past. And so it's not even clear what you'd throw the compute at. Let's say you're out of data, you exploited all of the state. You can create some kind of data which helps in certain domains, but it doesn't help in generality, as far as I can tell.

20:31 certainly doesn't help with generalized language because I don't even know what that means. I think you actually have to either create more data or you have to come up with a way that we don't know how to do yet to take the universe and make it better structured. It could be we hit another plateau, which listen, I think the actual bitter lesson in computer science is that Everything is a sigmoid.

20:49 And you have these exponentials and people get super excited, then it it flattens off. And then you have another unlock and everybody's excited and then it flattens off. And who knows, I'm not saying we are, but It very well could be the case that things flatten off now, and that's great. We've got more problems to tackle, let's go tackle them. I'd love to zoom in on this um visualizing like a economic crossover point of viability of a business model for some product that passes that

21:12 And one that doesn't. And I'm especially interested in the stuff On the line that you've seen. that you could maybe generalize about what you've learned about that economic viability line for AI products. I probably took my first AI course in nineteen ninety nine. This stuff has been around for very long time. And if you actually look, the progress for AI has always been incredibly steady. It's always been up and to the right. We've solved many problems we didn't think that you could solve.

21:38 Handwriting detection. And we sold all of these problems. We do it better than human beings, and we have for decades have been doing many things better than humans. There's always been this open question in the investment circuit, which is Why have we not seen A platform shifter of transformation.

21:54 Because of AI. Where are all the AI native companies? Why has all the value gone to incumbent? If somebody makes money off of AI, it's typically like a meta or a Google. It's not some startup. AI investing for startups has been A bloodbath. It's been very difficult. Why is that the case?

22:10 And the answer is. Because the economics have been crap. They've just been fucked. And he said, So why have the economics been tough? So a lot of traditional AI, let's take perception, it's actually solving these very hard problems. It's trying to build a model. that will see the universe and predict what happens, but the universe is very heavy tailed. So you have to bring humans in, you've got diminishing marginal returns. If you basically converge on like having to generate new data or have human beings in the loop, and then you then you end up like paying more people. This is the economics have been very tough for startups to solve.

22:40 So there's one thing that's changed. With these new models is the economics are amazing. Right. Asking why are the economics amazing now? And the reason is what I alluded to before is The models that are working are really not trying to solve these big tough problems of trying to reason about the universe. They really are trying to just reproduce what human beings have always created.

23:01 The closer you get to a model that's just trained on data. created by humans, the more economically viable the company is, because A, that data is what humans want. They spent the time to actually create it to begin with. So clearly that's interesting. But also the humans have done the hard work of creating all of the structure in it.

23:18 And I think the closer that you get to robotics or perception or trying to make sense of the real world. The harder the economics become. Even A B is robotics light, right? That's kinda two D ish. Like You're not like

23:32 What can a car do? It can move in a 2D plane, it can go forward, it can go backwards, right? It's not like it's like dealing with like arms and fingers and this and that. And so I think the closer you go to full 3D navigation in the real physical world and having to predict how the world is gonna react. I think the harder the problems get and the harder the economics get. And then the more you're competing with a more evolved portion of the human brain. So I think that really is the continuum. How do you think about with your investing hat on I guess the horizon to commercialization of a given idea.

23:59 Are you willing to invest in a robotics company today that's building the foundation model for dexterity? I'm making something up. Do you have to be able to see? A customer buying a thing for an economic transaction. In order to get excited about it as an investor, or are you willing to invest in things Where the commercialization is like over the visible horizon.

24:18 So it depends on the domain. I'm a firm believer that these models train on given data trading stuff. The model is the app. There's so magic that If you do it, people will find value and they'll use it. And we've seen this and we've seen many of these. We've seen music. We've seen speech.

24:32 We've seen video, we've seen images, and all of these are these great growing businesses. And it's just so easy to articulate why These are great businesses because the marginal cost of creating something goes to zero. And you're creating something people want, people will buy it. So for those I think you want to invest in the best team to build the best model. And then I think you can do this well before. So, for example, ideogram. We invest in ideogram. This was like the Google Imagine team is one of the pioneers of diffusion models. They went out and they created an image one. Like this is a very obvious we were the first money and it was just kind of a very obvious thing to do. As the economics become murkier

25:07 then it's harder to make these decisions. And one thing that I've learned looking at robotics companies over the years Is that Often when you're investing in robotics companies You're investing in a very verticalized company and you almost need to understand The business

25:22 The vertical that it's being applied to as much as the core technology. In the L M creativity space, literally some PhD researcher can create a model that consumers love. But if you're building a robot for agriculture, you better be an agriculture company, actually. And if you're building one for mining, you better me a mining company, actually.

25:38 I've sat on the boards of these companies that these these core technologies that are building drones or robots, especially like it is it's like shipping or it's ag or it's forestry or whatever, and the companies end up becoming the specialist of what they are. And so that's the mindset you have to go into those investments. This is an ad company. And I just learned that I don't know about these verticals. I don't know about construction. I don't get that sales process. I don't get the unit economics. I don't know about mining. I don't know about reforce. That's not me. That's other people. This is one of the reasons we even have an American dynamism team at Andrees and Horos, which is Clearly AI is gonna change robotics. It's changing robotic and robotics is our AI, but it's changing how we do things. It's gonna change all of it, it's gonna change defense, it's gonna change safety, it's gonna change construction.

26:20 But to invest in that, you better understand those markets. And so I do think you need special purpose teams to do that. So for my team. For those ones, we don't do them, but for other teams in D S forwards they do, but they're very much specialists on the markets as opposed to just the tech. How do you think about and analyze the defensibility of these business models where pick a modality, I don't know, music or something.

26:41 It seems every time some mind bending demo comes out, five other companies come out and there's like rough parity between What they're able to do. And I'm just curious where and how you think these companies will build Defensibility and

26:56 Allow them to be. Huge enterprises. Yeah, it's a great question. I've got a lot to say about those, but let me say a few maybe unrelated things and then specifically. So one unrelated thing is first movers tend to have a mess. advantage here. And it's because they're creating a new user behavior and they have a brand and they have a bunch of users and they have the feedback and whatever. So that's a very real thing. And I do think it Breaks Pareto where

27:15 You know, the first movers have eighty percent early on and then The twenty percent is left for the rest. However Over time all of that erodes. Right. And that's very clear. Now I've seen of these model companies launch at this point. We probably have the largest portfolio in one of them. So I've seen many of them launch. And they all Look the following.

27:31 The company does nothing and then they launch a model and it's not very good. It's a get a little blip and then it goes back to zero. And then they go for a while, then the launch bottle's not very good, then blip it goes back to zero. And then they launch a model that's good enough that it's over some unknown bar and then they get this great Oh. You're like amazing, we've made it.

27:48 However It's mostly'cause it's so magical to get a bunch of user and then it kinda looks like a T C P saw too. People use it and then they get bored and it drops off. They have to lunch another model and it goes up and then it drops off. They have to lunch another model. Then you're in this long March, let's call it the waves, where you're writing these waves of needing new models and it's a treadmill and it sucks.

28:06 And at that point, that's when the companies have to find a traditional mode. Now traditional moat could be a community moat. It could be an integration mode, like an API. It could be uh differentiated styling and brand mode. There's a lot of modes, but that's when you have to create the moat and get off the treadmill. I think if you don't get off the treadmill, this all converges basically to the same model. Let's talk about the motion.

28:27 That actually work. Integration modes are a real thing. Open AI has a great moat in that many people use the API in order to do their thing and like APIs just tend to be sticky, especially as they evolve over time. Now interestingly L's this anti stickiness property and that they're so unpredictable you can't rely on them anyways.

28:45 have to be ready to swap them out. But APIs are some mistake. So integration notes are a very real thing. Community emotes are a very real thing. A lot of the especially creative companies Move to a consumer experience that involves sharing and virality and community and that has inherent stickiness, of course. Workflow is a very real mode, which is Independent of the model

29:04 Many of these things still require human input and so you have my dashboard You own the access layer and then you're training a user behavior. So that's a real mo. Again, so if we're gonna break this up. Model companies will launch stuff, a lot of it will be failures, they'll have one that hits, and then they're in this sawtooth pattern of the treadmill of the next big one. And then in that time, they have to figure out a defensibility. It will not just be on the model. I'd love to hear your view on the evolution of the competition between the foundation model providers, both We'll call them frontier or closed source model providers like Open AI.

29:34 And the open source ones like Mr. All. And this will be an excuse to talk about some regulatory stuff and some bigger picture stuff in a few minutes, but first just on the models themselves. What is your way of Understanding these things, how do you watch them? What do you think matters? What interests you? Just riff out where you want.

29:51 So here's something. Everybody asked the defensibility question and what the mental model people have in their mind when they ask the defensibility question. Oh, long term there's nothing really defensible and They end up having the same data and blah blah blah.

30:04 But it's so much worse than that. And the reason it's worse than that And I think people don't realize to the extent that it's worse than that. Many people don't, which is There's a massive perverse economy of scale. Because distillation works so well. So what do I mean by distillation? What I mean is let's say

30:20 You Patrick have your AI company and you release the world's best model. And it's an LLM. It just turns out that I can use the fact that your model is out there to train my model very effectively to catch up to your model. Maybe I won't get it all the way. I can do it a lot cheaper than you did it. Distillation is so effective. We're seeing like a new model will come out and then everybody

30:41 catch up to it. But this is a massive perverse economy of scale. I don't Think We as an industry understand quite what that means yet. To me it means

30:50 If you're releasing a model in text, you better have a good plan to get its defensibility. This probably should be traditional. something that business have been built on for quite a while in software. I don't think being the best model is just enough of an enduring that said. It's also clear to me when markets expand They tend to fragment, not consolidate. And so white space is created. Every day there's more white space and so I actually think

31:13 We're seeing so much fragmenting and fracturing, even if defensibility is a problem. Companies are finding independent niches. Let's just take the image category. So one way to think about the image category is there can only be one image company. And that's gonna be the winner or they're all gonna look the same. But that's actually not what's happening. What's happening is they're all finding independent niches. It's almost like when you walk into a bookstore bookstore and have different sections there's here's the anime section and here's the sci fi section and here's the reference section.

31:40 Here's the horror section. It's almost like These image models are like deciding which sections they are. Pikawitch's video like they're very good at anime and mid journey is like the gritty fantasy realism one and Ideogram is like the cartoony text one and Doby is like the photo real one. And so I think that when marks expand they fragment There's a lot of white space to find niche. We're seeing that happen. There's multiple viable companies. There will be a consolidation phase, I think for a long way from it.

32:05 And so I think we're gonna see a lot of great companies being built in the near term and people are too scared of the consolidation phase where We have Market costs drop this fast, I think we're on a lot of runway where we need to worry too much about that. What do you hope to see from The big anthropic open AI mistral.

32:22 Facebook, like the big leaders that have provided the core models. What's the best case scenario in your mind about how that story plays out in the next three years? If you run on the thesis that these are A core CS primitive. That can solve a very

32:36 Huge classroom. Problem. I suspect that They're gonna end up solving different modalities of those Problems. I really do.

32:44 I think it's amazing that Facebook releases llama, but when you release open source without actually having an enterprise sales person put a market engine around it, the thing kind of lands and people do whatever they want, but you don't ever push and put it and pull it in the way that an enterprise buyer would want. Ms. Crawl on the other hand it's doing open source and it's releasing it, but it also engages with customers. So he can very directly push and prod and pull it in the way that is actually meaningful. Does that mean these two things are competitive? Not really. They're serving two purposes. One is here's this model, do whatever you want with it. Another one is I've got an actual solution for the buyer. And then open AI on a topic again.

33:19 I just think they're pushing the state of the art. I hope they continue to push the state of the art. I think that They will likely find somewhat overlapping, but also independent nations in different areas. And I think these things are expanding so quickly, I think they're gonna continue to be viable players. And so I mean, I applaud all of these companies for

33:37 Basically changing the industry. and launching these things. And again, I don't think it's a zero sum game at all. I think that's actually the worst thinking in this AI way to zero sum thing. And it literally is the most detrimental thing. Somehow it's a zero sum game. And because you have open source and the road's closed source and because you have two people competing in the same Area like one has to win. It's not like that. This is like a hundred percent abundance. Market is growing by a hundred times kind of let's cheer them all on. Can you lay out your position on all the important regulatory considerations that are happening very real time that I think

34:09 you've been very vocal about that you have strong opinions about. Can you like lay out your dump speech for What you think is right versus what's going on. I like to say I'm actually not a political person at all. I never get involved in this discussion. And I just feel like Now we're just hitting this very dangerous period that I've seen before.

34:28 But I almost feel like it's my obligation to come on before. But it's not my temperament. I just wanna be clear. It's something I've thought a lot about. Like I felt like okay, okay, I've gotta go out and do the right thing here. So here's the thing. I feel that in

34:41 Maybe Two thousand thirteen, two thousand fourteen. Nick Bush, he wrote this book. He's a super smart guy. He's a great philosopher, and he has great things to say about us living in a simulation and how innovation could be the downfall of humanity. And how AI can be bad. They're these great intellectual arguments, and they've been going on for a very long time. And I think they impact a lot of people. If you actually read the Walter Isaacson book on Elon

35:05 Lost one of the more interesting things is is how these kind of early Bostrom arguments impacted people like Demis and Elon and Sam and Ilya, et cetera. But what many people don't realize is this was all before the systems were actually built. And this was a platonic Arm chair thought experiment. And somehow that got conflated with actual systems that were being built.

35:29 And so you had this let's prepare for the boomy recursively self improvement AI that can end humanity and protect everybody from it before we actually created like these systems that are working And somehow that Red.

35:45 is being used In my opinion Emperor. And dampen Innovation to what could be

35:52 one of the most life changing industry changing value creating advances we've ever had. And I think that the best analogue is during the internet. And I was there and actually very active in these conversations. So

36:06 A lot of What? my positions draw from that experience. So let me just talk about that because I think it's a great analog. During the creation of the internet And the web.

36:16 We had the exact same conversations using a lot of the same rhetoric, which is Oh my goodness. this stuff that we're running our critical infrastructure on this stuff. No You can take over people's lives, you can have sovereignty. I actually went to think tanks in DC in the early two thousands on sovereignty ending events for the United States based on the internet happening. We would talk about digital peril harbor or cyber perhaps. There's all of this concern.

36:41 And all of the same things, oh, we need to protect the people from it. We need to pass all this regulation, all of the same type of stuff. And at the time It was actually a far more legitimate concern than we have today for two reasons. First one is it actually changed the security posture of the United States because it introduced the notion of asymmetry. So before that we were

37:01 in this kind of cold war neutral destruction doctrine, which is if we have big weapons and they have big weapons We have it to talk. Yeah maybe we'll These tensions. But at least you have

37:12 Stale mate. Where In the internet, the more advanced society was the more insecure because you're relying on the infrastructure. And so if you were like at a cyber war with a country that doesn't use computers or isn't networked, You're more at risk. So it introduced actually a new doctrine shift.

37:30 Which is this. But the second one is we had actual specific examples. of emergent threats that we've never seen before, like the Morris ones. We had these worms that would take out things, right? So both were you armed with These very specific things that have shown a new threat and an A sub T.

37:44 And yet after all of the discussion, the conclusions were Don't hamper innovation. Open source is good. This is the way to keep the United States ahead. This is the way to have more secure systems. And that's what played out. Many of the top security contributions to the internet came from academia or open source. SE Linux is a very famous example of this. Listen, Dor, which became Sourcefire was a great example, and Mandy is a great example. Like the entire industry came out of open source. And so I just feel like we've had this discussion. Listen, we're in this AI time where the rhetoric is back, but hey, we don't have

38:18 any of the proof points that we had on the internet, there is no clear doctrine shift And a lot of the arguments people are using for this call to regulation, there are these kind of like Non provable science fiction views of like recursive self improvement or creating bioweapons, which at this point have been pretty well refuted. And so I'm afraid that this kind of Bostrom view of the world has

38:40 captured the imagination, we've built these systems and we're using it to hamper innovation. in the name of quote unquote security and we have not learned the lessons from the past. And in the end. It'll hurt the United States. It'll hurt academia. It'll hurt researchers. It'll hurt little tech in the industry. And so that's why I've been this call to arms. Anybody who's not in a big company who's gonna basically benefit from this stuff needs to wake up. And get involved and make sure that we don't regulated ourselves out of a leadership position.

39:06 So if we think about this as the role that small tech will play On all of this. What do you think people need to do? What do you hope happen? What needs to happen? feels like a very different time than it did twenty years ago because academia has almost been entirely muzzled. I don't know why. So in the past, academia was

39:23 very vocal proponent of open source, even for critical infrastructure. BSD came from Berkeley. ran the internet. There's nothing more critical. If you take that out, the internet doesn't work. Hospitals don't work. Those Nothing more core to the safety of human.

39:40 than this and academia was on the forefront of this. The EFF came out of MIT. There was this kind of organic individual progressive almost grass root support And now with this AI stuff, if anything, academia is totally complicit. The California Bill.

39:56 This horrible bill in California, which is literally we need to protect Everybody from these models that don't even exist yet. It literally. Came from A Berkeley professor and her student.

40:07 coming from academia. So academia is totally complicit. So I think that those in academia that are silent, I think they need to speak up. And I actually think if for example they called Scott Wiener's office. He would listen. And if they called regular they would listen. And this is a California specific one. I know this podcast goes to everybody, but I think academia needs to wake up. And talk. I think little tech needs to

40:28 And I do think that'll Bing vocal. I think political lobbying actually is quite important. I think making views known is quite important. I just think the silent masses needs to arise. If there's one takeaway I've had having lived in the two thousands been totally silent having lived now is how different the landscape is. It just feels like

40:44 Back then big tech just wasn't as dominant. They weren't as involved in the government, they weren't as involved in the academia. And I just feel like all of these things are not bought paid for. And there's so few neutral voices left. And so for those that care, I think just coming out and speaking for what's right is very important. Can you teach us conceptually a little bit about the importance and role of open source? Versus closed doors. Those are words that everyone uses, but

41:05 in a sort of brass tax way. describe the role that those two things play and what they both mean to you. Just in general and technology. Yeah, for sure. I'm gonna be very clear about one thing. I'm in no way an open source zone.

41:17 My company had some open source and some closed source. I invest in closed source companies all the time, and so I'm in no way thinking that you should only do open. But it's played an incredibly important role. In the industry

41:31 And one way to view it is almost like this democratizer, commoditizer of mature businesses. Unix was one of the first operating systems. It was a standard It was used in the enterprise and the open source community created Linux. Well starting from Torvald. And then it followed

41:49 The path of Unix. But the entire all of the academics and the researchers and the hobbyists and the companies Involved that They would add security and solve a number of problems. And so because the closed source version had created this market, the open source version could come and create a flourish ecosystem that everybody can participate. So it democratized. operating system. We've seen this in many things. We saw this

42:11 in databases so like Oracle of course And side base for databases, MySQL comes. Postgres comes and then it follows along. Of course the smartphone then Android comes. And follows along. So it's always been this or historically been this

42:26 Almost democratizing force that is a checks and balance for the closed source version and allows Anybody else who doesn't have access to it to innovate And if you look at again the history of security and safety in innovation, so much of that has been in the open source. So if you remove that,

42:45 We lose all of that. We lose all of those brains and all of that effort and all of that energy in it. If you think about the talent in all of this. How does it feel different to you, if at all? From the way you source, interact with the types of people building, where they're coming from, like where you're looking for them.

43:02 Versus the pool of SAS talent or something like that. Does AI feel Like a shift. In that sense to you as an investor?

43:10 Yeah, it's actually very interesting, which is you are seeing a rise of the HT CEOs, which is actually pretty near. faculty at Berkeley. PhD guy. George Fraser, Five Tran. So you're actually starting to see almost

43:23 Very deep. Tack. Focuses and I don't know if it's because we're early in a super cycle and they're just on the bleeding edge. Or it actually is a bit of a different discipline that it's not just engineering and just business, but it's actually more core.

43:37 two sides. I don't know the answer to this. That's one. There is an open question that we debate often in A sixteen C. And the question is do you back the core technical team or do you back the commercial team in a given modality? Is the tech democratized enough? That any good founder can get the right team and use it or just the team have to understand the core tech and I gotta say I personally go back and forth on this a lot.

43:58 I think because I tend to come from listen, I did grad school, I did research, I did a company based on that, I tend to error towards the deeply technical teams, but that's clearly not always right. And so I do think a lot is up in the air. I think Commercial versus Tekken up in the air, I think. researcher versus engineer is up in the air in this realm and it's an active dialogue. Maybe it's a great excuse given that a lot of these companies are in this phase of going from zero to ten plus million of ARR to talk about your concept of company market annealing.

44:26 It sticks out of my memory as A really interesting way to think about the early stages of companies and since so many are going through that now, could you describe that idea and that framework? I've read all of these books and they're all very useful about finding product market that But that's not how it ever seems to work. And now listen, I'm sat on many boards I I've watched this many times. So the typical story in finding product market fit is you the company create some product and then you go ask the market what it wants and then based on its feedback you

44:53 Tweak your product and you A B test your way. Into Finding the right thing. And so you get all these aphorisms that I find so stupid. Sol your own problem.

45:03 But then you have a company, which is just not true or solve a real problem with the customer. And helpful, right? None of these things tell you actually how to find product market. So here's my experience having watched it played out multiple times, which is You both have to know your technology and evolve that as well as make the market pliable and educate the market. This is if you're doing category creation. Market emailing is for category creation. If you're in a mature market, just go build something better. People already know how to use it. But in new markets, like the customer doesn't know the value that you have. They haven't a clue. They don't know how to use it. They don't have the workflow. So there's a ton of work that you're actually hammering the market and making it so it can accept it. And so the way that I think about it is almost like this kind of annealing where

45:45 You work a little bit on your product and then you actually work on the market. By work on the market, I don't mean Do interviews. I think this is like the most broken way of building a company. It's you go ask Imagine this. You're gonna go interview a customer and ask what they want. They're sitting down there and they feel like they need to answer something. So they'll just make something up. Oh, I need this. I want a pony, right? It's just it's almost like it's too much

46:05 Leading the witness. This isn't a central buyer. And so The most successful companies they have a true nor like we know that this technology should evolve in this way, and we need to educate them on this, but there's gonna be some real organizational constraints and real budget constraints. We're gonna find those at the same time. So they go

46:22 And then they tweak their product to do that. And then they go pull and push on the market and they treat the product and they do that. And it's just this give and take this annealing. And that's how most companies do category creation, in my opinion. Not this go ask the market what it wants or go solve your own problem. The market will just understand it. If that works for you, you're very lucky, but I've never seen it. We've talked a lot about models and some of the apps. Built on those models.

46:47 the first time and you and I talk mostly about the modern data stack and just data in organizations. And we haven't talked much about bare metal like compute. And I'd love to just spend a minute or two with your impression of what's going on in each of those two worlds, maybe starting with the data side. What are new things that you're thinking about that matter?

47:04 In the world of data. And get Exaggerated by the rise of AI. Clearly unstructured data now is Very, very important.

47:12 We come from this cloud data warehouse world where people use SQL and data's very well behaved and it's in some relational database somewhere and we query it and whatever. That's still incredibly important. Still growing. But that was The dominant thing. Now more and more. These data pipelines are

47:28 Give me a bunch of random videos and a bunch of random images. And a bunch of random music and a bunch of random stuff. I'm I'm gonna somehow like wrangle all of those things and use that to train this model. And oh, by the way, you have to be able to do this in a way that reproduce the more scriptural. So it's almost like the unstructured world and the structure world are converging. Which I think is great. So clearly data is continues to grow important like it has been doing since the beginning of compute. But the size for the unstructured has just been growing asymptotically.

47:55 The world need new things that it doesn't have to handle all of that. New companies, new technologies. New ways of working. I think so. I don't Think we yet know how to

48:05 had a large pools of super unstructured data in a way that's efficient. Imagine how much video there is out there. The ability to query across video, the ability to put it into workflows, ability to pull stuff out. I still think there's this kind of a lot of work. And there's a lot of companies working on it. And so I know right now there's a founder that's like, Oh, but Martin, my company does that. Yes, I know that. There's a lot of companies that are working on these problems, but they're not solved in the sense that somebody can't just take a big pool of video and then be able to where you just like sequel right away and then there's just a lot of Edge cases for this. I do think that

48:36 we need to mature as an industry almost like the software industry has mature, just the tooling around managing data and getting it ready for this. The collection of data is becoming is absolutely Paramount in the labeling of data. There's one lesson I've come away with in all of this stuff is the existing data that we have is great, but once you exhaust that by throwing a lot more Compute at it, you gotta go get more. And nobody really knows how to do that other than

49:01 Maybe just pay a lot of money. I think that we need some real advances there to keep things moving. What about on the hardware side? Anything interesting that you've seen on Obviously NVIDIA hogs all the attention in the AI world right now, especially in public markets. Anything on just the raw hardware, semis, power. Side that has you curious or intrigued?

49:21 It's great that we're compute bound, right? I mean the good news is these workloads are very different than cloud workloads, so we're gonna see Massive innovation all the way up and down the infrastructure of stack, right? The communication patterns for training And the requirements on the fabric are very different than hard workloads and they're very different than inference. It's not clear that you would even want to use the same fabric over time. We're back seeing like massive innovation networking. Last time we saw that was when we were doing fat tree work.

49:46 for data center. So when the data center came out, we never had that amount of density of Servers in the same space. We had to evolve Networking. manage that much bandwidth. And then we haven't seen a lot of innovation, but now this AI is pushing us to really innovate on the networking fabric, which just goes to show you how much we have to reimagine the entire stack.

50:05 But it feels on the silicon, to me it feels still very early days. Let me just give you just a very interesting anecdote. Let's say you're gonna do A billion dollar training run for model. So

50:17 How much does it cost to create a custom ASIC? Let's say two hundred million dollars to create a customation. So if for two hundred million dollars you could get a fifty percent improvement. A fifty percent improvement of a billion dollars is five hundred million dollars. But two hundred million dollars is well spent. Almost like entering this space where it almost makes sense build a custom Asi

50:38 For one training run. The economics are there, then it becomes very compelling to do hard for harbor work. Now, listen, of course I understand all the complications of that, yes, timeline set, et cetera, but I'm just trying to give everybody a mental model that It's been a very long time that you can justify building a custom For a single Project like this, but the economics are actually there now, which is why you're seeing so much work in silicon again. So it's very exciting. I think the entire stack is going to be disrupted.

51:06 Can you say a word on the relationship between investors in capital markets and pools of capital and whether or not it feels correctly calibrated to you to the needs of the marketplace to do all this stuff. You've seen Relative to other parts of the economy. Some pretty lofty valuations relative to fundamentals or whatever. Some huge checks being written by very large firms.

51:27 What do you think? Do we have the right capital market set up for This explosion. of innovation and ideas. Let me give you the two bookend views of this now somewhere in the middle. So one view of this, a common view is this is all a bubble. Investors are stupid.

51:42 Valuations are crazy. It doesn't make any sense. markets are not efficient in the sense This won't go away. That's one view. There's another view which is

51:50 Actually We know that the economics work. Actually, there's a bunch of value creation. All of this makes sense. Markets are very efficient. I tend to be more of the latter one, but This is a very Nuance discussion because you can maximize a value in expectation.

52:06 Just like The probability something's gonna happen times the potential payout. The expected value of solving free energy is infinite, but the chances that it happened is Almost zero. And so I think markets are very efficient.

52:20 I think markets realize that this is disruptive and there's gonna be a lot of great companies. They realize that, but they don't know where. So I actually think the sizing is correct. I think This is the same thing, by the way. with the internet, which is everybody said that all like

52:34 the dot com was crazy and they overpaid, except for it was literally the growth driver of the world economy for twenty years. It way more than made up for devaluation. I think that will happen again The question is how do you sensibly place those bets. When you don't know For any given

52:51 bet what the expected outcome is gonna be. You just don't know these things. I do think this is one of the big ones. I do think this is a super cycle. I do think this is gonna be a massive trader. I think it once it's all said and done, you look back at these valuations, they're gonna look very sensible. But that said any independent bet they look silly. And I think that's the nature of venture capital and I think that's the nature of investment. You've talked a lot about places where you feel The story is

53:14 Already being told. Creativity that the tools that are available and already possible today versus things that are uncertain. What area of uncertainty do you find yourself spending the most personal time? Reading about, thinking about, talking about. Do you know what the first video on the internet was? Course I'm gonna guess like cats or something.

53:31 It was a coffee pot. I think this is Guy in Cambridge. Some researcher and at the end of the hall was a coffee pot and he didn't wanna walk there if there was no coffee in the coffee pot because like that'd be like a wasted trip and so he put their camera on that. And of course it was on the internet and so it became super, super popular, right?'Cause oh my gosh, if you can have that.

53:52 Then you know, of course you can have like that place. Do you remember the big red button? Very popular app early on was the big red button on the internet. It was a web page with a big red button and if you push it Nothing happened.

54:03 It just pushed. That was it. That was the entire thing was super popular and people were super into it. Then you ask why, are you like, Oh, this is just silly. Or you could be like, Maybe it's because people saw the big red button and then they imagine like you have one button, you have a bunch of buttons, and then you have Salesforce, right? And so I will say Having been through the web transition, I saw all the silly things and they were silly. It was all video games and goofy stuff and

54:26 Culturally marginal stuff. But those are the things that grew up To be the Googles and the Yahoos and the Amazons and None of us. knew what that looked like on the other side.

54:37 And there was all of these half steps. Do you remember desktop as a service? Stupid thing where like I've got a desktop here, I'm gonna run a desktop in the cloud. It's like there's all these half steps where you're stuck in the old way of thinking and you can't really do the new way of thinking. So where I am right now is This is the early days of the web. And a lot of the things that look silly.

54:57 Talking to a virtual girlfriend or Making an image of a cat. Those I know are gonna grow up to be the thing. I know it in my bones because I've seen it before, but I don't know what those are gonna look like. And so so much of what I'm asking is What is the big red button? versus not, what is that gonna become?

55:13 What is the future gonna look like? Let me just give you one anecdote I think is interesting on there. So my daughter's thirteen, Covid kid, right, went through COVID and that was very disruptive, and it was very tough. Basically she went from a being a very social kid to basically interacting online. And lately she's been a big user of character AI. So character AI from Shazir, who is one of the transformer authors, and but just like these kind of characters that you talk to seem silly.

55:36 But she really likes it. She's got these characters. And I've noticed recently that not only does she talk to these characters, but she'll actually invite them Group chats with her friends. That's what's Entered the social dynamics. So it very well could be that we're entering a phase where the actual social dynamics shift.

55:53 And then I ask myself, Okay, so for my daughter In five she's thirteen, in five years when she opens a bank account, is she gonna go to Well, Spargo SAS app, or is she gonna do it? From her character. Hundred percent from her character, right? It's like almost like the access layer to the internet is changing, right? If she's gonna do search, is she gonna go to Google or is she gonna do through her character or something?

56:12 I just feel like these very kind of almost silly, goofy, fundamental changes that are happening right now may become the thing. And when the access layer to the internet changes or to compute changes, like everything changes behind it. That's really what we spend a lot of time thinking about. It brings to mind a question on agents and

56:30 Right now you've got these chat models where it's like the smartest model you've ever seen that has no memory and can't really affect other systems. And it seems like Next step would be it has memory, it knows about you, or it has context and it can affect other systems. Cre create new actions. Elsewhere.

56:46 How do you think about it? Is that overly simplified way of thinking about what an agent is? What is an agent to you? What do you think about that? Everyone's talking about agents. So one thing that I love working on a team like I do.

56:56 I feel so lucky to have the team that I do. Everybody's so smart, so engaged and so knowledgeable is we have a different set of opinions. And the reason I bring this up this way is The agents is one where we've got very different opinions. We a hundred percent want to invest in agents and in the agent space and we look at all of these companies, but I personally have been very bearish on agents as a technical direction. And the reason for this There's a number of reasons for this is but I just feel like there's some very key technical problems that we do not Right.

57:24 Like for example the memory one that you've mentioned, correctness. is another one that we haven't done. We don't really know how to do planning very well. So it's not clear to me. that the advances that we've seen with L Models. carry over to fully automation of agentic stuff.

57:40 Now that said, most of the people on my team disagree with me and think that I'm a boomer. Okay, Boomer, like this stuff is amazing. I've seen these things work and they're amazing. I'm very open to be totally wrong on this point. But it just it hasn't been super obvious. There's actually another area where I'm I like this, which is this kind of idea that AI is gonna code for us. So there's kind of convictions. It's like Coding is done.

58:02 We're gonna just gonna tell an AI what we want and it'll do everything for us. And that's almost an agentic thing, right? It'll go. Do the coding. And many people on my team believe. That is the future. And that's great. They should do all of those deals. But my view is coding is thing with correctness is

58:16 really important. I think people spend what Ninety percent of their time when they develop maintaining code. So actually understanding and readability is very important too. So it feels to me that the human being still has to know everything that's going on. And so clearly coding will evolve and the UI will evolve and AI will help it.

58:35 But I don't think that coding goes away. I think we just we end up coding more. And I actually don't think that developers will have more time in a week, but have less time in a week because they're gonna be so much more efficient. So I think just low code, no code has always promised to get rid of programmers and never did it. I think AI will potentially hold the same false problem. But again, in my team, there's massive disagreements on this. And so I'm a huge fan of a company like Cursor. It creates a new IDE that's AI native IDE, and it's gonna go ahead and change how people code, but it doesn't replace developers or other people like, listen, like why would you ever need an IDE because you'll never need a programmer and I'm gonna invest in this company that does code. What is the definition of an agent to you?

59:10 How would you define it? An Ated to me has a control loop and state. So you give it a high level objective. And then it will do multiple steps along that high level objective and it maintains states between that and then it actually has a controller of feedback loop where it

59:26 Determines next step. And I don't think we've solved the state. problem. I just don't think that's necessarily part of the reasoning fabric right now. It almost feels like it's Tacked on. I don't think that we've solved the control loop problem in a way that converges. I feel like we still don't know really know how to handle the non determinism.

59:45 In these models. And so it's not super clear to me. But this recent breakthroughs on Yeah.

59:52 Large generative models. Cracks planning. Practice agentic problem. Again, very open to being wrong. But that's my intuition. In addition to the agent and the automatic coding. Are there other areas where you feel like your views are the most divergent, either from your team or from the markets?

1:00:09 I think one of the biggest mistakes that we're making as an industry is AI's been around for so long. Just all these enterprise use cases that we try to cram it into before. customer service chat bots and basic automation and this low code stuff. And what we're doing is we're taking this new thing and just cramming it all the old use cases. And we're like oh. It's like this old version of it, but now with like generative AI. I think this is just massive.

1:00:33 Lack of creativity and vision. I think that This is a new breakthrough in my opinion. It's got its own set of properties. And those properties are amazing for certain problems and not amazing for other problems.

1:00:47 I personally am very excited and attracted to the problem. I think it's a great fit for the emotion companionship. I think it's a great fit for the creativity problem. I think it's a great fit for the language reasoning problem. Great fit for those. I don't think it's as good as a fit for I'm gonna now use an LM to Replace Some workflow that requires correctness over long periods of time.

1:01:10 Where you don't have a human check in, etcetera. I just feel in general and look towards the areas that are a bit emergent. But then as an industry we just don't know them as well. So just like the internet, like things look so silly. Whenever I talk about creativity companionship, people are like, Oh, I thought you were a serious investor. But what they don't know, these are the things that become the new thing, I promise. Like these are the big red buttons and it's the copy pot on the internet. These are glimmers, these are signs of how apps will look in the future. And that's why I started when you asked me in the very beginning I think the future's gonna be a lot weirder and wilder and more wonderful than it is today. Exactly.

1:01:45 If you could interview anyone in the world right now that have been given truth serum Who'd you interview and what would you ask them? My clever way of asking. What do you not know that you want to know? I would love to know if

1:01:56 Yeah. Many of the people in the AI safety debates I wanna know who's a Baptist and who's a bootlegger. So do you know this kind of dichotomy? Explain it, but sure.

1:02:06 I'm gonna change your question. I wanna be at last ten people and I wanna create these baptists and bootleggers. So baptists and bootleggers is an economic phenomenon where you have regulation for example, in in the time of prohibition, and the people that support it you can often categorize as basically true believers and opportunists. So in the time of prohibition, the true believers of Baptists really thought that alcohol was bad. Like they truly believe that the Baptist thought prohibition was good and alcohol is terrible. So the true believers actually believe it. It turns out that in any of these situations you also have opportunists who actually see the regulation is good and they're called bootlickers. So the boot lickers, it turns out, really thought That prohibition was great.

1:02:42 Because then they could actually So I'll call for a premium and they are put in business. I think that this is a game theoretic B. I think this is an emergent property of these environments and it's not a value judgment on any single person. I just think that this will always emerge. And we're definitely in an area Where

1:03:00 There are regulations that are hampering innovation or will hamper innovation that are being espoused. And you have many people in the room. And I think some of them are true believers. they're true Baptists. They think AI is gonna recursively

1:03:15 and prove itself and kill us all and take jobs. I but I don't think there's very many of those. I think the vast majority of those are actual bootleggers. Listen, this is good for my big company. This is good for my agenda. This is good for whatever it is. And they're driving that. And I honestly couldn't tell you who is who. And so what I would do is

1:03:32 You're giving me one person to ask and I'd find some way to split it into five people to ask and I'll try to find out who's a Baptist and physical bootlegger. I love the answer. To end on a high note. What thing, if you have to go really tactical and specific, everything we talked about, what excites you the most?

1:03:47 Can I say two things actually? that excites me the most is I think it's a theta compliment. There's no holding this back. It doesn't matter how much bad regulation It doesn't matter. How much these large companies

1:03:58 Try to influence things. This shift is gonna be so dramatic over time, it'll win. The market woman here for sure. And I feel so bit about that. I'm the most frustrated. I just feel very calm. I just know that It's not.

1:04:10 We just have to wait. It's just gonna be bumpy along the way. And I do think that this What is Yeah, I do think it's gonna change how we think about the universe and how we think about each other and how we think about the world and how we do I really think it's kinda that Order thing and it's gonna happen. We just need to let it happen and then we

1:04:24 He's a way uh And then a very specific thing. That gives me a lot of peace that I just feel That's done. The very specific thing is I do feel

1:04:33 Number Very close. to being able to type a prompt into a computer and end up with entire virtual world that you can enter and is fully complete and is incredibly deep and has every aspect that we've ever wanted. We've had these notions in the past of

1:04:49 The hollow deck or the metaverse. And all of these are You have to have humans constructing it and you know designing it, et cetera. But I actually think that we've created enough content in the past that we can now auto generate these things and they're gonna be as rich and as interesting in the real world. And so I'll end on this. The one takeaway people probably should have taken from Bostrom. Is not that AI is gonna take over everything. I think the one takeaway is we probably are living in a simulation and we're probably about to create the next one. Martine, so much fun to talk to you every time. Thank you for your time.

1:05:19 Yeah, delighted. If you enjoyed this episode, check out Join Colossus.com. There you'll find every episode of this podcast complete with transcripts, show notes, and resources to keep learning. You can also sign up for our newsletter, Colossus Weekly, where we condense episodes to the big ideas, quotations, and more, as well as share the best content we find on the internet every week.

What you see above is a preview of the first minutes. One unlock costs 10 credits and covers this episode forever: full segment and word-level timestamps on this page, plus .txt, .srt, .vtt and word-level JSON downloads, as many times as you like.