The Exchanges

Q and actually he states his kind of concern that AI will replicate autonomous driving in the way that we got so excited about 10 years ago. Everyone's gonna be unemployed, eight million truck drivers, but my question to you is, are we gonna see that similar plateauing Where for 10 years actually, kind of autonomous cars didn't feel like it was progressing. Do you, how do you think about that?

A I've never worked in self-driving, um, but if I, I'm gonna, I'm gonna apply a mental model, and you tell me, you, you tell me if it, if it lines up. I feel like in self-driving what happened was, there was an aha moment where, you know, you could get the thing to work at all, and then you're like, okay, well now it works 60% of the time. How do we get this to 99.99999% of the time? And every day you show up to work and you just play whack-a-mole on what's not working. Uh, and you just like hope and pray that this converges to that like 99.99999 thing. That's not true for AI right now. That's, I'm sorry, that's not true for, uh, specifically what I'm about to say is only applicable to building smarter and smarter models and agentic systems that ultimately help you do work. That's, that's the thing that I'm trying to talk about. Like, Um, for building that, that's not how, that's not how the, the underlying dynamics are right now. Like every day we go to work and there's like actually brand new scientific things we want to try that just dramatically improve the performance of the model. And, um, some of those bets don't work and some of those bets really work. Um, I think like the reasoning that we talked about earlier is an example of one. I think another example of one is like this, like universal multimodality that GPT-IV-O is. Like those, those breakthroughs are like visible.…

AI assessment note: “preventing this from just being a hype cycle that falls flat like AB”

Q You mentioned that the kind of co-pilot approach, I had the guests on the show say that the co-pilot and I think it was Miles Grimshaw at Benchmark or now Thrive said actually co-pilots is an incumbent strategy. It's leveraging existing distribution, and it's an incumbent strategy. Is that fair, or do you think actually it's not giving due credit to the co-pilot approach?

A I think both of these two things can be true. Like, I think co-pilots are a great incumbent strategy because it lets them morph their existing software business model to something that kind of looks the same while getting in on the AI thing. Um, but, but even separately from that, I just think, like, where are these systems going to be most useful? Like, I just feel like everybody in this field has this vision, right? That, like, AI is going to take all jobs. The, like, pricing by work thing is just a corollary of AI is going to take all jobs, right? Because then it's like, all right, maybe you price by work on invoices, and then next month you price by work on, like, consulting decks, and then before you know it, you price by work on, like, being AI CEO of, like, David Coe or something like that, right? Like, that's not, I don't think this is how this is going to play out. I think the way this is going to play out is, Is that what we're going to have as humans fundamentally be the drivers of these agentic systems that, um, like that, like basically give everybody tremendous amount of leverage on their own creativity. And like, how can that be built without a co-pilot style approach? It's like, it's like my question.

AI assessment note: “I think both of these two things can be true.”

Q I was quite shocked by Apple's partnership with OpenAI in terms of the looseness that they tied to it. They continuously said, oh, but we'll actually maintain relationships with others. And they very much left the door open to switching between different providers. I almost thought it was a net negative when I heard it. How did you, I'm just intrigued. How did you interpret that when you heard it?

A I am extremely impressed with OpenAI. I think in terms of their technical Their technical delivery. I think the degree to which GPT four O was like, I think relatively under hyped relative to what I think the true scientific improvements have been in that model, um, is, is, is this pretty big gap? Like, I think like, like we're, we're, we're moving towards a world where we're going to be training these like universal models that take any input in, right? Audio, text, video, uh, you name it, and then generate any Any output out, and all of humanity's knowledge will be encoded in one of these models, and GPT-IV-O is a much bigger step towards that than people realize. Um, so I think that Apple cutting that deal with OpenAI, I mean, of course, I'm not privy to what actually happened, but I think at least part of it is a recognition that I think OpenAI is on a different trajectory compared to others on actual model progress, but at the same time, it also really strongly hints at a commoditized future. Like to the same extent today, um, as a consumer, I no longer care whether my computer, my desktop at home is powered by an AMD or Intel CPU. I think trying to create a way in which Apple owns the interface and Apple owns the end customer, and then the like big brain LLM smarts is just like one hot swappable thing is brilliant for them.

AI assessment note: “it also really strongly hints at a commoditized future”

Q What does that do to the org structures of teams, David, do you think? Does this mean much, much smaller companies? Does this mean, how do you think that actually plays out?

A I think the main way it plays out is, um, actually, this is something I'm gonna steal from, um, our angel investor, Scott Belsky, who, uh, has just thought about this so much. Um, he always calls it like this collapsing the talent stack thing, and the idea is basically that, um, uh, or this is my interpretation on it, like, projects and teams where the same person is simultaneously the PM and the designer and or the engineer or the go-to-market person or the marketer or whatever, Like, the more that, like, those different skill sets are smushed in the same person, the faster that thing moves and the more effective the thing becomes. So I think what it's gonna do is it's gonna make humans at work much more like generalists, and it's gonna have, like, causes to create larger and larger, oh, sorry, like, uh, uh, or giving people sort of, like, larger and larger scope over various different, like, areas that are different functions today, while they ultimately supervise, like, a cohort of, like, AI co-pilots that are the specialists.

AI assessment note: “collapsing the talent stack... make humans at work much more like generalists”

Q How do you think about the variation of agent requirements based on a power industry basis? Do you know what I mean? That it's so varying. It's completely different.

A Yes. And that's what our advantages. That's what our advantages is like, is like, you know, we get this question all the time, right? Where the depth trying to build, build something that lets anybody like, like we want to be the system of record for workflows and enterprises, like any employee at any large company should be able to teach adapt. Hey, like, here's how I do this particular thing, right? Like, here's how I handle, um, Here's, here's how I handle fetching all the data for an insurance claim, right? And this is really show adept that, and then adept should be able to do it for them. And like that generalization, all of those edge cases and variability is why the only way to solve that is to have vertical integration of model with use case. And it's also why I think we'll do better than companies that are just focused on a vertical, like a particular narrow problem, because every Like, uh, I was talking to, um, Parag, who used to be the CEO of Twitter, we were just hanging out the other day, and he's like, dude, every enterprise workflow is an edge case, and he's absolutely right, and that's why you need to control the same thing.

AI assessment note: “all of those edge cases and variability is why the only way to solve that”

Q How do you think about that? Is that true?

A So let me put it a different way. I think that, um, I think, uh, I am not a lawyer, but, but even so I will say it depends. Uh, and so I think the, the way, the better way to go think about it in my view is like, there's two parts to model scaling with compute. Um, one part to it is you simply make the model bigger and then you throw more data and more GPUs at it. And if we go look at CPUs and data centers, right? For a long time, we had Moore's law, right? Every year, the, every year, chips would get better at some predictable pace, and everybody's, ah, Moore's law is gonna die, you know, we're at three nanometers or whatever, there's, like, no more nanometers left for, but what actually happened is you go look at the amount of compute available, um, uh, even, like, for, for, for chips, it's actually, um, continued to trend up, because now what we do is we build systems that have multiple chips in them, so we have, like, both the scale up of a single chip and the scale out, and as a result, every year, Humanity has more and more compute available to it, right? It's the same thing with, um, with giant model scaling. Like, the base model itself, um, even if the base model itself stops scaling at some point as you throw more compute at it, there's a whole new way to go make models smarter that is just being tapped right now. And that whole new way of making models smarter is not …

AI assessment note: “the better way to go think about it in my view is like, there's two parts”

Q I was quite shocked by Apple's partnership with OpenAI in terms of the looseness that they tied to it. They continuously said, oh, but we'll actually maintain relationships with others. And they very much left the door open to switching between different providers. I almost thought it was a net negative when I heard it. How did you, I'm just intrigued. How did you interpret that when you heard it?

A I am extremely impressed with OpenAI. I think in terms of their technical Their technical delivery. I think the degree to which GPT four O was like, I think relatively under hyped relative to what I think the true scientific improvements have been in that model, um, is, is, is this pretty big gap? Like, I think like, like we're, we're, we're moving towards a world where we're going to be training these like universal models that take any input in, right? Audio, text, video, uh, you name it, and then generate any Any output out, and all of humanity's knowledge will be encoded in one of these models, and GPT-IV-O is a much bigger step towards that than people realize. Um, so I think that Apple cutting that deal with OpenAI, I mean, of course, I'm not privy to what actually happened, but I think at least part of it is a recognition that I think OpenAI is on a different trajectory compared to others on actual model progress, but at the same time, it also really strongly hints at a commoditized future. Like to the same extent today, um, as a consumer, I no longer care whether my computer, my desktop at home is powered by an AMD or Intel CPU. I think trying to create a way in which Apple owns the interface and Apple owns the end customer, and then the like big brain LLM smarts is just like one hot swappable thing is brilliant for them.

AI assessment note: “one hot swappable thing is brilliant for them”

Q How do you think about the variation of agent requirements based on a power industry basis? Do you know what I mean? That it's so varying. It's completely different.

A Yes. And that's what our advantages. That's what our advantages is like, is like, you know, we get this question all the time, right? Where the depth trying to build, build something that lets anybody like, like we want to be the system of record for workflows and enterprises, like any employee at any large company should be able to teach adapt. Hey, like, here's how I do this particular thing, right? Like, here's how I handle, um, Here's, here's how I handle fetching all the data for an insurance claim, right? And this is really show adept that, and then adept should be able to do it for them. And like that generalization, all of those edge cases and variability is why the only way to solve that is to have vertical integration of model with use case. And it's also why I think we'll do better than companies that are just focused on a vertical, like a particular narrow problem, because every Like, uh, I was talking to, um, Parag, who used to be the CEO of Twitter, we were just hanging out the other day, and he's like, dude, every enterprise workflow is an edge case, and he's absolutely right, and that's why you need to control the same thing.

AI assessment note: “all of those edge cases and variability is why the only way to solve that”

Q How do you think about that? Is that true?

A So let me put it a different way. I think that, um, I think, uh, I am not a lawyer, but, but even so I will say it depends. Uh, and so I think the, the way, the better way to go think about it in my view is like, there's two parts to model scaling with compute. Um, one part to it is you simply make the model bigger and then you throw more data and more GPUs at it. And if we go look at CPUs and data centers, right? For a long time, we had Moore's law, right? Every year, the, every year, chips would get better at some predictable pace, and everybody's, ah, Moore's law is gonna die, you know, we're at three nanometers or whatever, there's, like, no more nanometers left for, but what actually happened is you go look at the amount of compute available, um, uh, even, like, for, for, for chips, it's actually, um, continued to trend up, because now what we do is we build systems that have multiple chips in them, so we have, like, both the scale up of a single chip and the scale out, and as a result, every year, Humanity has more and more compute available to it, right? It's the same thing with, um, with giant model scaling. Like, the base model itself, um, even if the base model itself stops scaling at some point as you throw more compute at it, there's a whole new way to go make models smarter that is just being tapped right now. And that whole new way of making models smarter is not …

AI assessment note: “I will say it depends... there's two parts to model scaling with compute.”

Q kind of you. I really do appreciate that, man, but you've been at some incredible companies as a training ground, so to speak, one of which was Google Brain, and I just wanted to start there. When you think about your biggest takeaways from your time with Google Brain, what would you say ones, two are, and how do you think that shaped how you think about building ADAPT today?

A Google Brain, um, uh, uh, was, and also now as part of DeepMind is a really magical place. It's, uh, it's the, I think during the, uh, peak days of AI AI progress on the research side, right, where every day there was a new paper that came out that just changed the world. That, like, 2012 to 20 18 or so era, Google Brain was just, like, incredibly dominant. They did an amazing job picking talent. Like, the people who invented Transformer, the people who invented the diffusion model, people who did, um, all of these new optimization techniques that we all take for granted today, they were all a brand at the same time. Like, truly the Bell Labs of, of the era. And I think I learned a lot about how to make, um, how to make pure bottom-up, like, to see what good pure bottom-up basic research looks like at Google Brain.

AI assessment note: “I learned a lot about how to make... see what good pure bottom-up basic research looks like”

Q You mentioned kind of AGI being kind of infinity there in people's minds. You said before to me that the last step is human computer interaction, and that's the last ingredient to AGI. Before we do a quick fire, what did you mean by that? I didn't get that one either.

A I personally find a world in which sort of increasingly generally intelligent systems run around with their own agency and goals, uh, and, uh, and, um, not involve, um, Not involve what, what, what humans most care about to be not a world that I really want to live in. And this goes back to the, what you were saying about like selling AI by work versus as a, as a, as a software tool, right? Like I, I would much rather live in a world where we have sort of these like, like AI teammates, um, and assistants that we interact with instead. And then I think the question becomes, um, the, then the question becomes, how do you find the right interface Between smarter and smarter AI systems and people and how that interface is defined actually changes a lot about what training data you collect. How can humans align these systems towards the preferences of what humans want? Also, ultimately, like how these models are even built and what their architectures are. And so in a weird way, like the way the field is moving is let's make models smarter and then let's make use cases starter and then let's go put them in people's hands and then let's figure out what this means for people. Like it's kind of this waterfall sequential method. Which I don't think is a very good way to develop the technology. I think we should start back from ultimately how did, how should humans use these things and t…

AI assessment note: “we should start back from ultimately how did, how should humans use these things”

Q Is that why we haven't seen agent progression in the way that we wanted to or hoped we would? Because a lot of the tasks that people do are not actually codified in data. They're codified in conversations in rooms, in whiteboards, But not in data, but not in data.

A Yeah. I mean, I think that's a key. That's a key insight. Um, it's like, like, I kind of think that chat bots like chat GPT and stuff and, and agents, um, are kind of becoming different species of technology a little bit, right? Like, um, like I think they'll be useful in very different ways and, um, and what they need to be used for super different, right? Like just one concrete example is, um, is, is the hallucination problem. Having hallucinations in chatbots and in, like, image generators is, like, a really good thing, right? Because it gives you, it gives you, um, like, a starter tool for, like, getting to, like, solve the blank page problem, right? Like, and, like, gives you, like, little bits of novelty and creativity. But, um, but they, but agents, on the other hand, like, if you want something to go, like, consistently, I don't know, like, do your taxes for you or, uh, handle all of your shipping containers or something like that, you do not want that thing to go, Randomly hallucinate and, like, make up stuff along the way, right? And so, like, these things are speciating in an interesting way right now.

AI assessment note: “Yeah. I mean, I think that's a key. That's a key insight.”

Q Why has no one been able to solve memory? People often talk about this, and respectfully, it seems a confusing one to me, because it's like, computers have memory anyway. Why, why in AI is memory such a challenge?

A You can kind of think about memory as being two different things, right? You kind of have short-term working memory, and then you have long-term memory. I think people have made really good progress on short-term working memory, right? Like, um, if you could look at Gemini, Gemini's context length is like a million, it might even be more now, I actually don't quite remember, like a, like a million tokens long, which is so cool, as you can feed it, like, Giant snippets of video and be like, hey, like, write me a step-by-step of, like, every, everything the person cooking on this, uh, on this, in this particular video did, and it'll do it. Like, that stuff is insane. Like, that's making good progress, and the reason that's been hard is for computational reasons, um, but I think this, this sort of longer-term memory problem, this goes back to, like, another thing that I believe, and that, that's why I'm excited, I'm, I'm, I'm slightly less excited about model building, slightly more excited about application developers, because The underlying thing that everyone's realizing now is that LLMs themselves are not a product. Like, an actual product is this entire software system that uses LLMs in it. So for example, like, um, uh, what we should be doing is we should be finding ways in which, uh, in which, like, end application builders can be themselves responsible for, uh, for how to …

AI assessment note: “You can kind of think about memory as being two different things”

Q Is that why we haven't seen agent progression in the way that we wanted to or hoped we would? Because a lot of the tasks that people do are not actually codified in data. They're codified in conversations in rooms, in whiteboards, But not in data, but not in data.

A Yeah. I mean, I think that's a key. That's a key insight. Um, it's like, like, I kind of think that chat bots like chat GPT and stuff and, and agents, um, are kind of becoming different species of technology a little bit, right? Like, um, like I think they'll be useful in very different ways and, um, and what they need to be used for super different, right? Like just one concrete example is, um, is, is the hallucination problem. Having hallucinations in chatbots and in, like, image generators is, like, a really good thing, right? Because it gives you, it gives you, um, like, a starter tool for, like, getting to, like, solve the blank page problem, right? Like, and, like, gives you, like, little bits of novelty and creativity. But, um, but they, but agents, on the other hand, like, if you want something to go, like, consistently, I don't know, like, do your taxes for you or, uh, handle all of your shipping containers or something like that, you do not want that thing to go, Randomly hallucinate and, like, make up stuff along the way, right? And so, like, these things are speciating in an interesting way right now.

AI assessment note: “I think that's a key insight. Um, it's like... I kind of think that chat bots”