EVERY SPOKEN WORD
70 min read · 14,468 words- 0:00 – 0:37
Introduction
- SPSpeaker
So polytheistic AGI, I think, is one very useful macro frame. Means every culture has their own AGI, and eventually every culture has their own social network and cryptocurrency and AI. You know, the AI is sort of like their oracle. This is the center of society, and they've got their deterministic law with cryptocurrency and their probabilistic guidance with AI, the social network that binds the whole thing together. So those three technologies are like social technologies that are almost like the reactor core of the, of the network state of a modern internet-first society.
- SPSpeaker
[on-hold music]
- 0:37 – 3:54
Personal Journeys in AI and Crypto
- SPSpeaker
Martin and I were talking off ca-- uh, offline about how amazing your thread was on AI and, uh, you know, n-normally or often you're a crypto guy, you're a network state guy, but you know, you're, you're, you're a technologist.
- SPSpeaker
Sure.
- SPSpeaker
And you've been thinking about AI for quite, quite some time. So w-why don't you trace us through your, your evolution a bit?
- SPSpeaker
Totally. Yeah. So, um, you know, actually, Martin and I are roughly contemporaries at Stanford. I got my PhD in '05, '06.
- SPSpeaker
I got mine in '07. Yeah.
- SPSpeaker
Yeah. I'm, I'm a little gray down over here, and Martin's a little gray up over here.
- SPSpeaker
Exactly. [laughs]
- SPSpeaker
And we're complementary, complementary grays. Um-
- SPSpeaker
We both have that, that ambiguous kind of Middle Eastern look. [laughs] Even though-
- SPSpeaker
Yes, exactly. That's exactly-
- SPSpeaker
Even though neither of us are Middle Eastern. [laughs]
- SPSpeaker
That's correct. That's exactly right. That's right.
- SPSpeaker
[laughs]
- SPSpeaker
So, um, so we're both men of a certain age. Uh, I, I think it's fair, as, as, as the phrase goes.
- SPSpeaker
[laughs]
- SPSpeaker
Uh, and the funny thing is, I'm actually, you know... I, I taught machine learning and computational statistics and so on in the context of genomics at Stanford for, uh, you know, the mid-2000s and founded a DNA sequencing company. And really, I would say, like, my original career expertise for ten years, that's all I was doing, was se- in a sense, ML full time. But then I got into crypto right in the early-mid-2010s, just as the deep learning revolution was getting underway with ImageNet and, um, you know, all, all, all that series of papers in the mid-2010s, early 2010s. And, um, so I've, I've-- So just, just my thought process, I have, I have a foundation in probability and stats and multivariable calculus and blah, blah, blah. Uh, and so, so I'm s- I'm conversant with the space. The thing I will say, and, and I'm sure Martin has thought on this, and I'll get to the specific tweet. In the 2010s, you know, we were all tracking diffusion models and language models, and so on and so forth, and it was improving. But, you know, like style transfer, for example, in the mid-2010s was working by that time, right? And, you know, GPT-2 and so on was interesting. It could, it could kind of blurt out like a sentence. But I admit that I never really thought it was gonna get past, like, Markov chain-y like stuff, you know. I was surprised at, uh, at how much better GPT... Like, DALL-E was a hint, you know, in, in ear- the early, early 2022. But I was surprised at how coherent ChatGPT was. I think everybody was. But it was like a huge jump up from what it was before in terms of sort of being Markov chain-y. And so I've been kind of observing over the last, you know, two years being very, uh, originally very deep in machine learning, then very deep in crypto. But, you know, you can't be deep in everything. You can't be deep in... Or maybe Elon can. Okay. Aside from Elon, it's pretty hard to be at the cutting edge of so many different fields at the same time 'cause they're deep fields whi-which have a lot going on. So I, I... You know, with, with respect to AI, I think, um, there's like several realizations I've had over the last few years in terms of the unarticulated limitations of the space, right?
- 3:54 – 7:53
Monotheistic vs. Polytheistic AGI: Competing Paradigms
- SPSpeaker
Some of these I think I, I sort of... I think I came up with relatively early in, in the history of modern AI, and others I think I've come to more recently. But let me kind of enumerate them in no particular order, okay? And Martin, jump in any time. So the first is that something that motivated, uh, you know, I think both Eliezer and Altman, a bunch of other folks who built OpenAI, was almost like the same kind of sentiment that motivated people to build, like, the Sistine Chapel, which was sort of like an implicit Abrahamic monotheism, which was like summoning God, right? Like, you know, communing with God. But in the Abrahamic God sense, right, of, like, also the vengeful God who would turn you into paperclips, like turning you into pillars of salt, right?
- SPSpeaker
[laughs]
- SPSpeaker
That was sort of... That was an implicit thing that was behind, right? So-- 'Cause when they, when they talk about AGI, they talk about AGI as implicitly a unitary thing. We will get to AGI, and then it will go to infinity and will be like, you know, rapture and singularity kind of thing, even though that's implicit. And, you know, the way I-- Because I've been thinking so much about crypto and other kinds of things, I was like, well, there's a different sort of implicit school of thought, which is polytheistic AGI, right? Do we have, rather than the vengeful God, do we have war of the gods? Do we have many superhuman intelligences, all from different cultural backgrounds that have the mores and the values imprinted on them? And very early on, I had, I had a tweet that said, "At a minimum, there's gonna be American AI and Chinese AI, and if we're lucky, there'll be decentralized, open source, you know, crypto style AI." Because at the time, the American AI was woke, highly woke, this is 2022, and I knew, I knew that China was gonna, like, stop at nothing to copy it, so I knew we were gonna get at least two. And I thought we might get N if we had... if we were lucky enough to get decentralized AI. And that wasn't obvious at that time because the cost of training models was so high and OpenAI was so far ahead, it took a while before other people caught up. But now it's very clear that we're gonna have lots and lots of high-quality, open source, decentralized models.That like a new one comes out almost every week, you know. And China's gonna work hard on this. It's a big thing, like all the DeepSeek models, et cetera. So polytheistic AGI, I think is one very useful macro frame, which takes away some of the sort of, um, AI apocalypse tones, I think, uh, because I don't think the image generators or, you know, text chatbots are going to cause the, you know, destruction that people had. People thought they were gonna like bust out and, you know, do systems programming, and Martine can speak about that. Martine and I kind of were talking about that. There are certain limita-- it's now more clear actually that they're worse at systems programming than they are at visuals. We'll come back to that. So, A, polytheistic AGI, and so what does that mean? That means every culture has their own AGI, and eventually every culture has their own social network and cryptocurrency and AI, which is, um, you know, the AI is sort of like their oracle that's at the center of society, and they've got their deterministic law with cryptocurrency and their probabilistic guidance with AI, and the social network that binds the whole thing together. So those three technologies are like social technologies that are almost like the reactor core of the, of the network state of a modern internet-first society, right? Um, it'll be customized for each different kind of group, and certain things will be disallowed and allowed, like image generation might not be allowed on subcultures or NSFW, whatever, all that kind of stuff would be, would be tweaked. And so okay, that's one concept. The second concept, you know, Eliezer Yudkowsky, who by the way, you know, even if I disagreed with a lot of his stuff, did do a lot to promote AI and people getting into it and so on and so forth. So even if I disagree with bombing the data centers and various other kinds of things, um, I give him partial, significant partial credit for getting people motivated to look into the space. He was definitely into it in that sense. So like directionally there, there was something. You know, I always try to see the right side
- 7:53 – 9:36
The Limits of AI: Chaos, Turbulence, and Predictability
- SPSpeaker
of things. But a second big concept that I think I really disagree with Eliezer on, that I think is being borne out, is the idea that AI could just cogitate for millions of years and figure things out, and it could, could outmaneuver you all the time. And we know that's not true because turbulence, chaos, cryptographic equations are not like that, right? You can come up with turbulent systems, chaotic systems where you simply with finite precision arithmetic cannot forecast out indefinitely. In fact, you get fracturing and breaking, and cryptographic hashes are set up in such a way as well to be hypersensitive to initial conditions where a small change of one character can get a totally different MD5 sum or, or something like that. And, um, so those are situations, and if you wanted to, you come up with a thought experiment where you inserted turbulence or chaos into your decision process, sort of like, you know, you shake a turbulent clock or something like that before throwing a pitch, and the AI wouldn't be able to predict your actions, right? And that's like a, just a simple experiment. In real life, just, you know, the flow of a fluid is turbulent or, you know. So, so that actually put bounds on what AI can predict, right? Quantitative physical and mathematical bounds on what an AI can, can predict. Right?
- MCMartin Casado
Can I, can I, can I just, can I just add just a little bit-
- SPSpeaker
Go ahead. Yeah
- MCMartin Casado
... of color here? 'Cause I, I think this is great.
- SPSpeaker
Sure.
- MCMartin Casado
But I think we need to call out why, you know... So the, the way that you describe AI as like gods and monotheistic and polytheistic-
- SPSpeaker
Sure, sure, sure
- MCMartin Casado
... it's great at describing how human beings view AI, right?
- SPSpeaker
Sure.
- MCMartin Casado
But in reality, we're talking about software running on computers that are bound by those limitations,
- 9:36 – 14:10
Platonic Ideals and Real-World Systems
- MCMartin Casado
right? So I don't view them as gods personally. I'm a, you know, I got my PhD in systems-
- SPSpeaker
Sure, sure, sure
- MCMartin Casado
... so I view them as like system software. And I, I actually think the original sin in all of this AI, you know, anthropomorphic fallacy started with Bostrom, right? It was a, it was one of these kind of thought experiments where, you know... You know, when Nick Bostrom wrote "Superintelligence" was what? 2014? And, you know-
- SPSpeaker
Mm-hmm
- MCMartin Casado
... if you listen, he was talking about this platonic ideal of AI, and this platonic ideal of AI just happened to be able to recursively self-improve and just happened to like have these kind of like super physical things that like no AI today has. But it happened that the conversation around AI started then, and then it just coincidentally, we had these LLMs show up four years later. And so somehow our thought process on these two things dovetailed, right? And so people would take all of these kind of mental, um, ruminations, these thought experiments, and they would apply it to like actual systems. And I think the problem with doing that, the problem with taking some platonic ideal, whether it's Bostrom's or it's whatever Abraham, the, the, the Abrahamic's view of God or, or, or, or any kind of religious view, is that it is very blinkered to the limitations. And you've pointed out a great limitation. We've been doing computer simulation forever. We totally know the limits of simulating physical phenomenon, particularly chaotic systems.
- SPSpeaker
Right.
- MCMartin Casado
We have actually very hard bounds on these. Uh, it's, by the way, it's not just like the, the limits on, on the, on the size of like a, like a, a computer word or an integer or something like that. There are actually like very strong, you know, limits on time and the amount of compute necessary, et cetera, right? And so we know these. And so you can talk about these in two ways, and I think for this conversation we should do this. Like, Balaji, you're so good at talking about, like, what is this platonic ideal and how should we have a mental kind of model for this for like the non-computer specialist. And what I would love to do as we go through this conversation is talk about, like, actually these are still bound by computer systems. We know the limitations of computer systems, and so let's see how those bound it. And you just beautifully did both of those things in the same one. I just wanna make sure that we tease apart both of those things as we have this conversation.
- SPSpeaker
Totally. Totally. I feel I can kinda speak both languages here.
- MCMartin Casado
Totally.
- SPSpeaker
Where I underst- I understand where those guys are coming fromAnd because I kind of am like also a tech radical, you know what I mean, right? And but I'm also a tech pragmatist, so I like almost-- I think I straddle that boundary
- MCMartin Casado
The danger is, is if we don't say that this is a platonic ideal, people will map it to existing systems incorrectly, and that's exactly-
- SPSpeaker
Totally. Totally
- MCMartin Casado
...that's exactly what happened in twenty-twenty, twenty-twenty-one, which is they, they took-
- SPSpeaker
Yes
- MCMartin Casado
...this thought experiment that started with Bostrom. And it was a thought ex-- If you go back to the original book, you're like, "Listen, this has nothing to do with real systems," and they applied it to a real system. And I think that was kind of the original sin.
- SPSpeaker
Totally. Totally. I know, I know. Well, but, but let me poke on that a little bit and then de- both, both poke and then also defend your point. The Turing test was a thought experiment. That was a platonic ideal. No, no reference to neural networks, no reference to implementation details, and yet it served as something that went from a thought experiment to an applied thing with, you know, the Levshin-Gausbeck test. The CAPTCHA was like v zero of that-
- MCMartin Casado
Mm-hmm
- SPSpeaker
...you can argue.
- MCMartin Casado
Sure.
- SPSpeaker
It became commercially important. And now, obviously, AI's blown past the Turing test. It can, you know, right, uh, beat people in, in this. And then, uh, you know, the Chinese room, John Searle's thing-
- MCMartin Casado
Hundred percent
- SPSpeaker
...right, about machine translation. Another, another thing that was like a platonic ideal. And there's other things that are like that, like six degrees of separation, which social networks actually made real, right? So, so I'm not necessarily against platonic ideal-
- MCMartin Casado
Thought experiments are very, very important. I just think that we-
- SPSpeaker
Yeah
- MCMartin Casado
...we need to be very clear when we're having a conversation not to conflate them with, like, the actual systems.
- SPSpeaker
Yes. With what's-
- MCMartin Casado
And I think but that's-
- SPSpeaker
Right
- MCMartin Casado
...the thing that we've kind of fallen into in this conversat-- Not you and I.
- SPSpeaker
To-
- 14:10 – 25:45
Surprises in AI Progress: Language, Locomotion, and Double Descent
- SPSpeaker
AI rather than AGI. That alone, I think, kind of nukes a bunch of the concepts of we'll get to AGI and just win because, uh, it's kinda clear that there's just, like, a rapid onrush of new models, and it's, like, more of a continuous kind of thing. The fast takeoff scenario w- didn't happen, right? On the other hand, um, I think that it's not, um... L-like to, to, you know, can, can an AI write a sonnet? It can, right? It can do it better than most humans. Can it, can it write a screenplay? It can do that, again, better than most humans. There's a lot of things that w-we thought as maybe harder. We thought we-- maybe like locomotion or something would be easier to, to solve than what we think of as higher cognitive functions, but it's, but it's actually the locomotion that's still harder in some ways, right?
- MCMartin Casado
I think that one now in retrospect is easy to explain via economic equilibrium. We just didn't have the right model.
- SPSpeaker
Why do you say that?
- MCMartin Casado
Well, so, for example, when you're competing with a human brain on 3D navigation in space, you're competing with, whatever, the four-million-year-old mammalian brain and a body that's been running away from predators and picking berries for four million years. It's incredibly, highly evolved. When you're competing with a prefrontal cortex, which is like, you know, the language learning and creativity, how old is it? Two hundred and fifty thousand years? And so if you take this question from an economic equilibrium standpoint, you're saying, "Well, listen, do you want to, like, compete with, like, the most evolved system that's solving the much more difficult problem, which is much higher dimensionality, has all of the chaotic... has to deal with like, you know, chaotic nonlinear systems like you said? Or do you wanna deal with the very new evolution that deals with kind of a much denser space [chuckles] that actually does a pretty good job with linear interpolation?" You know, the problems, the, the problems that actually works very well with linear interpolation, et cetera. Um, and so I think, I think you're totally right. It was, it was not obvious. I mean, listen, AI has all been, been solving what we thought was the easier problem, but the hour fallacy is, like, easier for humans 'cause we're really good at it, 'cause we've been doing it for a really long time. And it turns out [chuckles] the harder problems are just harder for us because we've only been doing them more recently. So this is a great case of where, uh, it did not, like, uh, like our intuition on the problems to solve were the wrong ones because of, of our own kind of anthropomorphic fallacy, right? Our own notions of-
- SPSpeaker
Well-
- MCMartin Casado
...of what problems are easy and hard.
- SPSpeaker
Yeah, I mean, okay, true. I would say, um, you know, one of the, one of the surprises I had, that's why it's interesting when you said economics, one of the surprises I had with GPT-3 and ChatGPT was how far you could get with language.
- MCMartin Casado
Yeah.
- SPSpeaker
Like, before the ChatGPT moment, it wasn't obvious to me that just... Then afterwards, you're like, "Oh, okay, language is sophisticated enough to encode almost any concept about the world," right?
- MCMartin Casado
Uh-
- SPSpeaker
Or, or rather-
- MCMartin Casado
Well, if I put you in a very dark room-
- SPSpeaker
Yeah
- MCMartin Casado
...and that's, that's an arbitrary construction, and I describe how you navigate and pick something up, I think you'd have a very tough time doing it.
- SPSpeaker
No, no, no, I know. And what I'm saying, though, is basically if you generalize language to be streams of symbols-
- MCMartin Casado
Yeah
- SPSpeaker
...right, you could send telemetry to some... Like, basically, I think I was surprised by how many concepts were encoded in language that could be learned, even to the point of, like, rough role models of, like, you know, the map of the Earth or the proximity of things. You can back out those kinds of things-
- MCMartin Casado
Right. Just the, this-
- SPSpeaker
Where-
- MCMartin Casado
The distinction is, is it could have been... I totally get it. The distinction is it could have been the human being, the human mind that looked at the world-Did the reasoning and created the world model, and that is cached in language. And then language is-
- SPSpeaker
Yeah, that's what I'm saying. So-
- MCMartin Casado
Okay. Okay, good.
- SPSpeaker
Right.
- MCMartin Casado
But, but, but the-
- SPSpeaker
Yeah
- MCMartin Casado
... but the, but the go from the world to the world model was the human, and then everything else, and then the lang-
- SPSpeaker
Yes.
- MCMartin Casado
Okay, great. I agree. Yeah.
- SPSpeaker
That's right. But, but it was, it was a little surprising to me that you could get that far with language models. As opposed to, like, spatial reasoning-
- MCMartin Casado
Hundred percent
- 25:45 – 29:18
Prompting, Verification, and the Age of the Phrase
- SPSpeaker
different. So another kind of angle I have on AI is prompts or tiny programs, which is more common today, but I think, uh, at least I attribute it, it went viral a while ago. And but they're programs in a hidden API. Because normally you have an API that is fully documented but very error intolerant, right? Prompts-- Prompting is the opposite. It's, it's completely undocumented, but it's highly error tolerant. It will usually do what you mean.
- MCMartin Casado
Yeah.
- SPSpeaker
But the better your vocabulary, the better you can prompt it. So now art history is an applied subject, right? Political sci-- like knowing a vocabulary of like Cezanne versus Picasso and so on and so forth, you can actually pull up the style that you want on demand. So the broader your vocabulary, the broader your subject knowledge, the more you can get out of it. We're in the age of the phrase, which is the prompt, the hundred and forty character tweet, and the twelve words for your crypto password, right? These, these phrases of power in AI, in social media, and in crypto just unlock everything. So the better your vocabulary, the more you can do, right? Um, so I think of prompts as tiny programs, and actually one of the things that I've gotten in the habit of doing is writing... The-- It's total opposite of search. You know, with search you, you learn to type things in keywordese, and you sort of figure out the, the word that has the most specific TFIDF, you know, on the page or whatever. I will write sometimes these long memos to an AI and then continuing, I know maybe you don't love this analogy, but I think it's funny. Continuing the polytheistic analogy, I'll give them to Brahma, Vishnu, and Shiva.
- MCMartin Casado
[laughs]
- SPSpeaker
Okay? So I'll give them to ChatGPT and Claude and now, you know, Sup- Grok and what have you, right? I'll consult all the gods, and then I'll make my decision on that basis, right? And, and then sometimes have them argue with each other, right? And, um, why do I say kind of half-jokingly gods? Because, you know, like, the Hindu kind of frame on that is not like the, you know, fearing God thing. It's not the same kind of thing. But in a sense, it is a superhuman intelligence that knows everything about your culture, and if you ask it the right question, it can tell you something that you didn't know. Uh, but in like the Hindu tradition, they're not infallible, right? Like, that is to say it's not the same as the all-knowing, all-seeing. It's like this interesting, it's like more like superhumans, or almost more like superheroes. People will argue with me on about that, but I think that's more-
- MCMartin Casado
No, I agree. Yeah.
- SPSpeaker
I think more true, right? Yeah. It's, it's like actually the Norse tradition is similar to the Hindu tradition in some ways, right? Where you have, you know, the gods were not infallible, but they were superhuman.
- MCMartin Casado
I think this is a, this is a great-
- SPSpeaker
Go.
- MCMartin Casado
And it's a useful framing. Just remember that when it comes to computer systems, we can put formal bounds on them. We can do this information theoretically. We can do this computationally. And like that's, that's gonna come-
- SPSpeaker
Totally.
- MCMartin Casado
And once that happens, we will understand these systems fully, and it'll be very hard to think of them as gods at that point.
- SPSpeaker
Well, that's right. I mean, that's the thing is actually what's interesting is that the, um, the interpretability work that Anthropic and others have done and the work on like Groking or what have you, right? Um, that's actually really good stuff because, uh, the, um, you can pick apart neurons, you can pick up... You can find the Golden Gate neuron if you, you know, if you found, if you saw that kind of thing, right? You can dial that up, dial that down. You can start actually taking apart these AI brains in a way that hasn't happened before. A few other kinds of things, right? So, like, uh, the thread that you guys, that, that thread actually summarized, uh,
- 29:18 – 34:26
AI, Crypto, and the Grounding Problem
- SPSpeaker
several of my research. I should probably put this into a post so it's like kind of there for the record. So I'm just gonna do a, a, a bunch of these and maybe get your thoughts, right? Okay. So first concept in, in no, no particular order. AI doesn't do it end to end, it does it middle to middle. So the business spend, so basically you have to still prompt it, and then you have to verify it. And people talk about prompting, but they talk less about verifying. And Karpathy and I had this good conversation a few weeks ago where basically AI is gonna create massive numbers of jobs in proctoring and verification because it's so good at faking things. So a-one of my other kind of concepts is AI makes everything fake and crypto makes it real again because AI is a probabilistic technology and crypto is a deterministic technology. And so, like, crypto is in some sense what AI can't fake. It's like the hard cryptographic equations. It can't fake a, a Bitcoin private key. It can't fake even an on-chain NFT. That's what cr- AI cannot fake.And so that's, you know, like, like the hard b-barriers, right?
- MCMartin Casado
I mean, I generally agree. I don't... I mean, I don't think crypto solves the grounding problem, right? I mean, it's a, it's a, it's a mechanism you could use-
- SPSpeaker
Ah
- MCMartin Casado
... but it's a mechanism. Like, the grounding problem is-
- SPSpeaker
Well, so do you mean by ground-
- MCMartin Casado
The grounding problem is-
- SPSpeaker
Grounding in reality?
- MCMartin Casado
Yeah, yeah, yeah. The actual physical grounding problem.
- SPSpeaker
Oh, okay.
- MCMartin Casado
The data ingest problem, yeah.
- SPSpeaker
All right. I, so I disagree with you on that, and here's why, or let me give you a counterargument at least. Right now, let's say, just give a concrete example, let's say you ask Perplexity to summarize the FTX hack in twenty twenty-two. When it would do so, among the citations it would give you would be links to a block explorer, right, that would actually have on-chain data that you can cryptographically verify that this transfer of these funds happened at this time, and you... if you want to go even further, you can actually pull out the digital signatures and the hashes and the timestamps-
- MCMartin Casado
Sure
- SPSpeaker
... from that block explorer.
- MCMartin Casado
Sure. Yeah.
- SPSpeaker
Right? Okay. Now, here's my argument. My argument is that works for financial data, but what's happening now with Farcaster and other kinds of things is, with increasing block space, you could put more and more kinds of data on chain, and we're gonna have to because you're gonna need crypto instruments and you're gonna need cryptographically hashed posts and crypto IDs to know that it was posted by a human or to know the data wasn't tampered with. So more and more kinds of data are gonna go on chain, and then that will eventually mean that an AI's citations are to on-chain data, which is both financial data and social data. And so then at least it will map back in terms of grounding to an on-chain cryptographically provable assertion of some kind. And you might say, well, at least that'll be an assertion at the metadata level, like we can prove that this digital signature made this assertion at this timestamp with this probability.
- MCMartin Casado
Sure. I'm talking about real-world grounding. Like, I say something, I am a human being, I, you know, you have no idea whether I said it's true or not true. You know, there's a, there's a geographic place where there's a picture of the geographic place taken from a nineteen seventies photo. Was that doctored or not? I mean, like, like the actual physical world grounding just because you can't encode, you know, digital data yet at this point for the physical world. Now, over... It's a great mechanism to do that once we can solve the ingest problem, but this just-
- SPSpeaker
So, so let me, let me talk about something which is happening now that I've been kind of funding on the side. It's not a full solution, but it's, I think, a partial solution, which is, um, so, uh, crypto instruments, the idea would be that when you capture, like a frame of data, right? For example, sequencing machines, like DNA sequencing, when the data's coming off the machine, it's like TIFF files that are actually image data that gets processed into A, C, Gs and Ts.
- MCMartin Casado
Yeah.
- SPSpeaker
And many other kinds of instruments basically have a stream of data coming off the machine as you're capturing it. Cameras are like that, right? Okay. So you could, and, and there are things that do this already, take a hash of that and post it on chain at that time-
- MCMartin Casado
Totally
- SPSpeaker
... and do that, right?
- MCMartin Casado
Totally.
- SPSpeaker
And what that would at least say is that that frame of data existed at that time. And so if you had something that was like a scientific experiment, right, you know, like a, like a pre-registered double-blind trial or something like that, you could have not just a crypto instrument, you could also have other people with proof of humans there who have a sort of attestation ceremony, and now you have a number of different kinds of on-chain data that start to get harder to fake in a coordinated way.
- MCMartin Casado
Totally.
- SPSpeaker
Not impossible.
- MCMartin Casado
Totally. Yeah, yeah, I agree with all of it.
- SPSpeaker
But at least you have the set of-
- MCMartin Casado
As soon, as soon as you get it into the system-
- SPSpeaker
Yeah, yeah
- MCMartin Casado
... then cryptos are great mechanisms for ensuring kind of end-to-end guarantees. It's just the data ingest problem is a maybe. It's a long-standing problem in computer science.
- 34:26 – 37:19
Visual vs. Verbal: Where AI Excels and Struggles
- SPSpeaker
That's right. Okay, great. So, okay, so next kinda concept maybe to discuss, and I think this is a useful, uh, like division that, uh, that... This is a relatively recent point that I, that I made to myself that I thought was useful. AI is good for the visual and less good for the verbal. What do I mean by that? So when it's generating images, when it's generating video, when it's generating user interfaces like Vercel's v0 or Replit's user interfaces, the great thing about them is you can instantly see them, and with the GPUs that we have and hardware, you can verify cheaply whether they're good enough, right? 'Cause you can just instantly get the gestalt of it, right? Whereas when it's back-end code, when it's legalese, when it's like, you know, mathematical equations, you have to slow down and use system two thinking, not system one, right? And it's not just your gestalt impression. You have to actually go line by line and check whether it's right, and that is actually the expensive step, the verifying right over there. So that's actually... I think that's a non-obvious thing, where the more front-end and video and visual it is what you're doing, the easier it is to AI. And now the interesting concept is how much of it-
- MCMartin Casado
Can I just say I have one more thing, which is like, I th- for, for me, again, like I spend most of my time in like software and engineering, the big distinction is stateless versus stateful, right? So if you're, if you're-
- SPSpeaker
Yeah
- MCMartin Casado
... generating code that is gonna have semantics that evolve while you're running it, it's just impossible to spot check. Like some, some things are computationally irreducible. You actually have to run the computation to know the answer. So I to- like the image is the perfect example. It's visual, and it's basically stateless. Like all of the state is there. There is no kind of runtime semantic.
- SPSpeaker
Yeah.
- MCMartin Casado
Totally agree.
- SPSpeaker
That's right. And whereas even a relatively small snippet of back-end code could have a fairly complex finite state machine essentially underlying it-
- MCMartin Casado
So it just dynamically bounds and-
- SPSpeaker
... or even infinite state machine
- MCMartin Casado
... you know, like-
- SPSpeaker
Yeah. It could be, it could be very-That's right. And so simulating the time dynamics of that, you have to use something different, maybe formal verification if it's algebra, you know, it's-
- MCMartin Casado
Or you'd actually have to run it if it's computationally irreducible. Like, there's no way, there's no way to statically do it.
- SPSpeaker
Right.
- MCMartin Casado
I mean, and so it does reduce to almost like the computation verification problem, which is this long-standing problem in computer science forever.
- SPSpeaker
Right. Now, now formal ve- formal verification, at least for a subset of programs, has become commercially viable for smart contracts-
- MCMartin Casado
Totally
- SPSpeaker
... because they're so high value and they're so small-
- MCMartin Casado
Totally
- SPSpeaker
... that they actually are-- it's worth doing that on, right?
- MCMartin Casado
Totally.
- SPSpeaker
And yeah, it's not gonna be... it's not gonna work for the general case, but you can do a constrained case. But okay, so, so that was one major division, visual versus verbal. Another when, when you're getting to like stateful a-and so on, is I think there's certain-- It's the limits of AI are the things I'm interested in, where you draw like a fine distinction of what it can do, are very crisp. And what I think AI is particularly bad at, that people are trying to make, trying to use it for, that
- 37:19 – 40:11
The Challenge of Markets, Politics, and Adversarial Systems
- SPSpeaker
they're gonna fail on, I-- in my view, is when they try to use it for markets or politics. And, and let me explain why I say that. Uh, for systems that are time invariant, you know, like the mapping of an image to the label cat, or the rules of a game like chess or checkers or even Go, or, um, you know, something where there's like a, like a sort of, like a static rule set or static mapping, right? Then you can do the train-test paradigm and train a model, and so it's different. However, when you have something which is time varying, especially rule varying and adversarial, like markets are or like politics are, then the same trade will eventually quickly start resulting in a loss. And by the way, the other guys are also using an AI on you, right? And so this is decentralized AI again, right? And so what that does is it actually argues that the CEO or the inf-- or the creator who is constantly sensing the market or sensing the political winds, and has a thesis on it based on human nature or other things or what have you, actually is the sensor that then prompts the AI. And that's a job that is hard for the AI to do at a, at a really deep level because it's time varying, rule varying, adversarial domains.
- MCMartin Casado
And it goes back to what you said at the very beginning, which is if you look at these type of equilibrium, they're complex differential equations which are nonlinear. And so in order to predict-
- SPSpeaker
Yeah
- MCMartin Casado
... what's happening, you'd have to like, do this nonlinear extrapolation, which we know that, like, these things are not very good at, right? That's what you use-
- SPSpeaker
In fact, actually what I... What's interesting is I wasn't even thinking of the stock market as complex different-
- MCMartin Casado
But it's a chaotic system. It's the kind of-
- SPSpeaker
But you're right. Even on the subset of time-
- MCMartin Casado
I mean, Mandelbrot wrote this great book-
- SPSpeaker
No, you're actually right
- MCMartin Casado
... on the fact that these things are, are these chao-- They're, they're super chaotic.
- SPSpeaker
Yeah, yeah. You're absolutely right that actually it would be a useful thing to show just with a toy example, a chaotic system that is time varying or another one that's adversarial, dynamic. Now, the thing about that though is, um, to argue against my point, uh, they've gotten AIs that are actually pretty good at StarCraft, right? Which are-- It starts to stretch the boundaries of what I was saying-
- MCMartin Casado
Yeah. Totally
- SPSpeaker
... 'cause it's definitely adversarial.
- MCMartin Casado
Totally.
- SPSpeaker
Right? Um, and it's like more time varying than chess, you know. It's, you could argue it's not rule varying, but it's, it's, it's time varying. But so, okay, let me, let me go to another point here, right? So, um, another concept is, I think-- So by the way, the, the commercial implication of that, that point on prompting and verifying is business spend moves towards prompting, proctoring, verifying, basically checking all the stuff that AI can generate. That's gonna be a huge,
- 40:11 – 43:37
Amplified Intelligence: AI as a Force Multiplier
- SPSpeaker
huge, huge thing. And that, that maps to KYC. That maps to, like, in a bad way, you know, the, the glass cases in Walmart, right? [chuckles] In a sense, a low-trust society is spending more and more and more on verification and, and, and proctoring and so on, right? Okay. Next. AI means amplified intelligence, not agentic intelligence, 'cause the smarter you are, the smarter the AI, as better writers or better prompters.
- MCMartin Casado
Yeah.
- SPSpeaker
What are your thoughts on that?
- MCMartin Casado
Yeah, I mean, I, I think to... It's interesting in the coding space that we actually start to have numbers on this now. Like, if you actually look-
- SPSpeaker
Oh, interesting.
- MCMartin Casado
Yeah. So if you actually look at, like, relative productivity gains, it just turns out if you're a more senior developer, you will have better productivity gains. And the ar-- I mean, the, the, the-
- SPSpeaker
Oh, I, I, I hadn't seen that graph. So you actually... It is something that makes the, the, the smart smarter, basically.
- MCMartin Casado
But also on a relative basis, which is really surprising, right?
- SPSpeaker
Yeah.
- MCMartin Casado
And, and, and, and then-
- SPSpeaker
Well, well-
- MCMartin Casado
But if you think about it-
- SPSpeaker
Yeah
- MCMartin Casado
... it's, it's actually not surprising. It's like, you know what the fundamental trade-offs are. You know what to ask. You know how to interpret the results. You know how to throw away bad stuff when it's bad. And so clearly, if you kind of know what you're doing, you can both verify, to your point, the output, but you can also be more specific of your ask. I think it's really important for all of us to realize that formal languages came out of natural languages, not the other way, right? Like, if you could explain all of this stuff in English to each other, we would, but it's just really inefficient, so we came up with, like, more efficient ways to explain trade-offs.
- SPSpeaker
Constrained.
- MCMartin Casado
Yeah, more constrained.
- SPSpeaker
Yeah, basically constrained languages that reduce the ambiguity.
- MCMartin Casado
This is literally, it's strictly an efficiency thing, right? And so, like, someone that knows how to speak these formal languages to the, to, to the models, they're gonna get-- is gonna articulate what they want better and is gonna be able to interpret the results better if the response is formal. And so i-it is kind of a nice codification of-
- SPSpeaker
I mean-
- MCMartin Casado
... of exactly what you're saying.
- SPSpeaker
Yeah, I mean, the thing about it is AI means everyone's a CEO because you kinda speak to the AI like you do to a... Like, to a great AI, you speak to them, like, in some ways to a great employee, where you give clear written instructions-
- MCMartin Casado
Yeah
- SPSpeaker
... and then you can verify the output. And then, you know, like so it actually kind of turns management into a skill that it hyper-deflates the cost of trying one's hand as a CEO or as a manager.Because you have to give those... And, and so the better you are in communicating what it should do, often the more people you can manage and, and so on and so forth. By the way, this gets to kind of the next point, which is AI doesn't really take your job, it takes the job of the previous AI, okay? And what I mean by that is, you now have a slot on your roster at every company for an AI image editor, an AI text, you know, a chatbot thing, an AI code, you know, IE thing, and so on and so forth. And each new release of Grok or Claude or whatever competes against ChatGPT and Grok and Claude, right? And so the AI takes the job of the previous AI because they're complementing. You kind of have a whole raft of AI augmenters that are augmenting your humans, but those AIs are competing in AI space to a large extent with the previous AI. Because once you've onboarded an image generator into your flow-
- MCMartin Casado
Yeah
- SPSpeaker
... then it just keeps improving, and you start using it in more places. But it's an AI taking the job of the previous AI. Let me know your thoughts.
- MCMartin Casado
Can I, can I actually... This is an adjacency to what you're just saying, but can I actually push on something you said previously?
- SPSpeaker
Yeah.
- MCMartin Casado
'Cause
- 43:37 – 48:17
The Polytheistic Counterargument: Convergence and Specialization
- MCMartin Casado
I actually agree with-
- SPSpeaker
Go ahead
- MCMartin Casado
... your polytheistic view of the world. I totally agree. But let me just let me just-
- SPSpeaker
Sure
- MCMartin Casado
... provide the counterargument for us to noodle on-
- SPSpeaker
Go ahead
- MCMartin Casado
... in this domain, which is, have you seen this kind of thing that all the AIs, if you ask to produce a random number, produce the same number? Have you seen this? Like four-
- SPSpeaker
Yes. It's like-
- MCMartin Casado
Forty-seven
- SPSpeaker
... it's like, uh, seven or something like that.
- MCMartin Casado
Seven or something.
- SPSpeaker
Yeah.
- MCMartin Casado
Right? Um, so one thing that is, to me, was non-intuitive but remarkable about these models is how easy they are to distill. Which is as soon as someone creates a, a leader-
- SPSpeaker
Mm-hmm
- MCMartin Casado
... everybody uses that leader and kind of sucks the life out of it, and then all the models kind of converge on it very quickly. Which you could argue that this is a counter to the polytheistic-
- SPSpeaker
Yeah, kernel intelligence-
- MCMartin Casado
Which is, which is-
- SPSpeaker
Which is basically the... Yeah, there's, there's like a core-
- MCMartin Casado
Maybe there are ten-- maybe there are a hundred AIs, but it just turns out they all have the same capability, so, like, is it, is it, is it just a s- [chuckles] like, a technicality that they're actually different and they've all learned from each other? Uh, a-and-
- SPSpeaker
Well, so it... So it's an interesting question, and, and, and my view is, and, and I've not called this a strong view yet, but my view is that's almost like the human body plan and spinal column, and then you differentiate on top of that core spinal column, maybe, you know. And so there's like, um-
- MCMartin Casado
I agree. Yeah
- SPSpeaker
... it's like you'd have some sort of... Just like every human to first order can see and speak and hear and so on and so forth, there's... But then some people have, you know, much better vision or they have much better speech or something like that, right? So there may be some distilled kind of thing... Oh, by the way, another interesting part on, on to what you're saying. In general, um, text on the internet is not emitted equally by every group. For example, liberals tend to write more text on the internet, and conservatives tend to be more visual than verbal.
- MCMartin Casado
Right.
- SPSpeaker
So it is actually, you know, you'll have an ideological skew that's actually hard to unskew because it's not a... Like, the people who are training AI disproportionately are, like, Redditors.
- MCMartin Casado
I-I-I-
- SPSpeaker
You know?
- MCMartin Casado
... I to- I t- I totally agree with, with what you're saying. I'm just saying the counterargument... By the way, I agree with what you're saying. But the counterargument would be, you just ask the AI model, you know, "Conservatives don't write so much, so give me the answer, you know, using the mediums that they use," or wha-- I mean, just all of that information would be in the model. I mean, here, here's how I would distill kind of my, my view on this, which is very much in line with what you're saying, which is I, I think the universe is very complex, and I don't think it gives up its secrets easily at all. And I think the universe is full of fundamental trade-offs. Like, you can't have both. You have to choose A or B.
- SPSpeaker
Mm-hmm. Yeah.
- MCMartin Casado
And so these models will, will align with those fundamental trade-offs, right? And maybe it's performance, maybe it's correctness, maybe, you know, whatever it ends up being. And so as soon as you want a specific solution for a given problem where it, it hits one of those trade-offs, you're just gonna need a different model because otherwise you just, you just can't end up having both. And, and again, because I know the coding space the best, we see this a lot, right? Which is, uh, uh, you know, a model that's very, very good for certain parts of code is just not gonna be generally good at other things because those are the trade-offs made when training it. And I, I think that this is kind of the, the, the future plurality of model where we're going.
- SPSpeaker
Well, is that true? I, you know, I, I thought somebody said something, I may be wrong about this, but I saw some counterintuitive result that said that making the AI specialize in one area makes it worse in other areas.
- 48:17 – 57:36
AI’s Impact on Jobs: Specialists, Generalists, and the Future of Work
- SPSpeaker
like, uh, i-in terms of what demands it's good at and so on and so forth, um, at least right now, I think, I-I'm not sure if you agree with this, AI doesn't really take your job, it allows you to do any job.Because you can get to like an okay level as like a user interface designer or sound effects or something like that, but you need a specialist for polish and that, um, you know, though, though I wonder maybe with enough RLHF from specialists, maybe that won't be as necessary. I don't know. What are your thoughts on that?
- MCMartin Casado
So here, so here's my current mental model. There, there's two personas. There's persona number one is the expert in the space, and there's persona number two is the non-expert in the space. So if the non-expert in the space is using AI, it's taking the place of the expert, right? And so, like, maybe you'll ask it, and it'll give you something. Let's say I want, I want a, a 3D asset for a video game, and I'm a programmer. Then I'm gonna ask it for like a nice 3D asset, and then it'll give me one, right? So I'm the non-expert. The expert user to our previous will actually know how to access, uh, ask it better [chuckles] and will likely get better results because it's an actually domain expert, and that expert is using it. So I, I think we see both. Actually, if you look in the market, you see both of these uses. Um, and I think both of them will persist, which is if I'm a programmer, I-
- SPSpeaker
It's like a, it's like a doctor talking to another doctor-
- MCMartin Casado
Exactly. Exactly
- SPSpeaker
... and they can instantly go to specialist language and so forth.
- MCMartin Casado
Exactly. That's exactly.
- SPSpeaker
So you're right. Even if there's specialist RLHF, you may not be able to-
- MCMartin Casado
Yeah, that's right. Like, why, why would I-
- SPSpeaker
You may not be able to access that
- MCMartin Casado
... why would I, why would, why would I want to learn the entire domain and make all of the trade-offs of, you know, 3D design when somebody else could have done all of that work for me, and they could talk to the model in the specialist way when I can just talk to my model using code, right? In order to very efficiently use a specialist model, I would have to become a specialist, would be the argument to our previous one. So I think, I think, listen, for casual use, I can use these models for whatever I want, but, like, to really use them very well, again, to our previous conversation, I'd have to become a specialist, and somebody else will have maybe already invested all of that time.
- SPSpeaker
Mm-hmm.
- MCMartin Casado
By the way, if you actually look at products-
- SPSpeaker
I think-
- MCMartin Casado
... these products actually have both these distinctions. Some products are very clearly for the casual user trying to replace, like, you know, like think about like Cursor versus Lovable, right? So Lovable is like, "I'm a casual user. I wanna create a website."
- SPSpeaker
Right.
- MCMartin Casado
"I don't even have to know about code." And that's great. You can create amazing things. Cursor is, "I am a professional software developer. I have an IDE. I know an IDE." Now, now over time, maybe these things converge. That could be the case. But thus far, these are kind of very different user bases, right? There's professional coding versus basically casual coding.
- SPSpeaker
I mean, you know, part of it is, which is interesting and a little counterintuitive, it's like if you think about what a computer can do, it can do the job of an accountant. It can do the job of a physicist. It can do the job, you know, but then you clad it in something, and it's adding up numbers in Excel.
- MCMartin Casado
[chuckles]
- SPSpeaker
And you clad in something else, and it's doing simulations for MATLAB, right? And so we already know that when it came to logical system two thinking, that computers are actually really good at that. And now you have something similar where these, these models are actually very versatile, but you clad it in the power user interface, and you clad it in the, uh, the, the casual interface. And, you know, it's kind of implicit contextual prompting as well, probably as well as a system prompt that makes it do those things. I mean, the thing that's interesting to me about something like chain of thought is that-- and, and I think this is where people were freaking out in late 2022, and maybe they'll still be right to freak out, is computers have historically always been good at the logical style, much better, superhuman at that. Now they're also superhuman in a sense at the probabilistic style, at least of text generation-
- MCMartin Casado
Yeah
- SPSpeaker
... and so on and so forth.
- MCMartin Casado
Yeah.
- SPSpeaker
And so it's not inconceivable that someone could figure out a way to merge those two, you know, like, like a quantum gravity, you know, kind of theory-
- MCMartin Casado
Yeah, yeah
- SPSpeaker
... of, of things, right? Where you take the probabilistic and the deterministic and pull, pull them together.
- MCMartin Casado
Yeah. This is, this is-
- SPSpeaker
Kind of moving your thoughts on that
- MCMartin Casado
... this very old school, you know, systems part of me thinks that there's a fundamental trade-off here, right? Just like you can trade off-
- SPSpeaker
What's that?
- MCMartin Casado
Well, I, I feel like you can trade off... You can build a system for determinism, and you can, you know, build a system which basically cuts a bunch of corners. Um, but you can't build a system that does both, [chuckles] right? Like, if you, if-
- 57:36 – 1:03:41
Security, Drones, and Digital Borders
- SPSpeaker
when I, when I push people on this, what's interesting is some of the people who were, "Oh my God, we need to regulate everything," are now actually on the side of, "We need to build it before China." But the thing is, in both senses, first it was safety, then it was security, but it both comes to control. And you might argue that the security argument is a better argument. I think it's a more realistic argument in some ways. But, um, but the concept that killer AI is already here-
- MCMartin Casado
Yeah
- SPSpeaker
... is interesting 'cause they, they put so much stock in the like, "Oh, it's gonna persuade everybody to do things. It's a super persuader." But persuading is statistical, and, you know, drones are deterministic, or at least, you know, the, the guns on a drone-
- MCMartin Casado
Yeah
- SPSpeaker
... are deterministic, right?
- MCMartin Casado
Yeah, I mean, I, I-
- SPSpeaker
Um-
- MCMartin Casado
... I, I think that the interesting question around AI and, like, attack and defense is does it change the equilibrium? Because the internet did. The internet actually introduced the notion of asymmetry, which is the more that you rely on it, the more vulnerable you are. So to wit, the United States is more vulnerable than-
- SPSpeaker
So-
- MCMartin Casado
... like, you know, random, you know, you know-
- SPSpeaker
Yeah. I think, I think-
- MCMartin Casado
... random third-world country
- SPSpeaker
... there's-
- MCMartin Casado
And it's not clear to me-
- SPSpeaker
The most-
- MCMartin Casado
... you get the same thing with AI. Like, it could just be like it just enables everybody to have bigger weapons, but the equilibrium's the same.
- SPSpeaker
Well, I think that it actually has really huge impacts for borders, and unfortunately, I think China is well positioned here-
- MCMartin Casado
Yeah
- SPSpeaker
... for a very specific reason, which is China, their, their justification for the great firewall is they've just-justified it as digital borders. They say, "We can introduce physical packets. Why can't we introduce-
- MCMartin Casado
Yeah
- SPSpeaker
... digital packets?"
- MCMartin Casado
Yeah.
- SPSpeaker
Right? And now with the whole Ukraine controlling drones in your territory thing, that becomes more than simply a metaphor. It's a real thing.
- MCMartin Casado
Yeah.
- SPSpeaker
It's like controlling cloud space.
- MCMartin Casado
Yeah.
- SPSpeaker
Right? And cl- you know, if you can allow somebody to script drones or script humanoids in your jurisdiction, then they can blow things up, right? That's no longer a theoretical thing. That's, you know? And now the counter-counterargument is, well, maybe you just have them preprogrammed autonomous, so they don't even need an internet connection, and they can just do cameras or whatever. That's true. And in Ukraine, you know these things with the drones on cables, you know, this crazy thing they're doing there?
- MCMartin Casado
Yeah.
- SPSpeaker
Right? Like, you know those, like, big unwindy cable things that you sometimes see on, like, ships?
- MCMartin Casado
Yeah.
- 1:03:41 – 1:06:27
AI, Power, and the Balance of Control
- SPSpeaker
this saying, which is, "The mountains are high and the emperor is far away," right? And China always had a different conception of the balance of power between the government and the people than the West did.
- MCMartin Casado
Yeah. Sure.
- SPSpeaker
On the one hand, you know, and this is, this is a broad generalization over thousands of years of history, but, uh, very broadly, like on the one hand, because the state had all the weapons and the army and so on and so forth, like, like that chair could morph into an agent of the emperor or a CCP guy today if, if, if the government so desired, because there's no like limits truly-
- MCMartin Casado
Yeah. Yeah
- SPSpeaker
...you know, in the sense of it can just do whatever it wants, right? On the other hand, whatever the law has written down, the people just do what they want.
- MCMartin Casado
Yeah.
- SPSpeaker
Right? So they kind of in a very pragmatic way say the limit is really the limit of what people can enforce, right? And if the state has lots of power, then any written limit doesn't really matter. But the state can't find you since you're on the other side of the world, and the, then the written law doesn't matter either, right? Which is a different conception than the progressive versus libertarian-
- MCMartin Casado
Yeah, yeah, sure
- SPSpeaker
...within the West, where they'll always quote law against each other back and forth, what is written is what is permitted or what have you, right? But AI does change that balance because now the mountains are never high-
- MCMartin Casado
Yeah, the long arm
- SPSpeaker
...and the emperor is never far away.
- MCMartin Casado
The, the long arm is incredibly long.
- SPSpeaker
Because-
- MCMartin Casado
The long arm is infinite. Yeah.
- SPSpeaker
Yeah.
- MCMartin Casado
Yeah, yeah.
- SPSpeaker
They can, they can synthesize... You know, there was something, maybe you know, maybe you know this thing, Martin. I think it was called TIA, Total-
- MCMartin Casado
Yeah. Sure
- SPSpeaker
...Information Awareness in Iraq-
- MCMartin Casado
Oh, yeah
- SPSpeaker
...at a certain point.
- MCMartin Casado
Yeah, yeah.
- SPSpeaker
Where, where the idea was they had satellites covering Iraq, and so every time some guy was put down, putting down like an IED or something like that, they would rewind the satellite to find who the guy was that did that and where he came from, and then put a bomb through his window or what have you, right? And in a sense, it was like tracking someone for like their whole life, you know, because you were sewing together the trace of them through all of these cameras, right? And that's totally possible for China to do now. And the difference is that AI makes it possible to-- For a long time, that data was ingested, but it couldn't really be parsed or queried because it was too difficult to look through five, you know, thousand hours of video on one person or whatever. That's increasingly becoming queryable, right, and ingestible and summarizable in a way that it never was. And so I think that the real check on something like that is going to be, gonna have to be cryptography, exit, you know, and so on and so forth, which is ultimately get out of the jurisdiction, you know, have, have property that they cannot actually seize. Like you go back again to the limits of power and what have you, right? Um, anyway, let me pause there. Just so-some thoughts on balance of power since you talked about that.
- MCMartin Casado
Oh, that's great.
- SPSpeaker
Okay. Last one. I think that there's going to
- 1:06:27 – 1:09:10
The Coming Anti-AI Backlash
- SPSpeaker
be... There already is an anti-AI backlash that's like the anti-crypto backlash and will be part of the anti-tech backlash, because a lot of people are not using AI for what we're using it for. They're using it for like therapy, or they're using it like, you know, forLike as a companion or something like that, you know? And go ahead.
- MCMartin Casado
It's the top of the pyramid of needs. It's kind of funny, like if you actually look, it's like self-actualization, spirituality, therapy. It's like finally computers are addressing kind of like further on up. Exactly. Yeah.
- SPSpeaker
The top of that. That's right. And there's another aspect to it that I think hasn't gotten as much press, but that's interesting to understand is that just like, you know, the tariffs are meant to kind of ward off Chinese competition, I'm not sure if they'll work. I d- in fact, I'm skeptical. There's a similar, much less publicized thing that's happening at many media corporations where they're unionizing to try to ward off AI competition. Like they have union contracts that say the editors, owners cannot use AI, right? So they're making their organization very brittle, where they think they own the market, but they're not allowing themselves to use AI, right? And then eventually they're gonna be beaten by AI-enabled competitors that pull all of their followers and views and so on away from them 'cause they're just more efficient. So I think that's gonna result in an anti-AI backlash, and I think that, um, it's already kinda here where, you know, in so- on some like artist forums or whatever, they'll say, "Are you an AI supporter?" Have you heard that?
- MCMartin Casado
You know, you know that, you know, like AI artists spend as much time on building things as traditional artists. It's just a different tool set, that's my view.
- SPSpeaker
I know. But yes, yes, yes, that's true. But basically, they feel that it's, it's kind of similar to the reaction by master craftsmen-
- MCMartin Casado
Yeah, yeah, yeah
- SPSpeaker
... in the 1800s, right? When, when mass production started taking over what they were doing in the physical world, this is now happening in the digital world, right?
- MCMartin Casado
Totally. Yeah.
- SPSpeaker
And so, uh, a- and the other aspect of this is I don't think people have thought about the international aspect where, um, if you've got, let's say, a lawyer who's making 200K a year in the US or a doctor in, in the West, and then you've got somebody from the Philippines or India or anywhere in the world who's making currently $2,000 a year, maybe the converged wage with AI plus their IQ or what have you, AI plus human convergence, is like 20K a year.
- MCMartin Casado
Yeah.
- SPSpeaker
Which is a 10X-
- MCMartin Casado
Yeah
- SPSpeaker
... for the person abroad, but a one-tenth for the person in the West, and it radically increases consumer surplus-
- MCMartin Casado
Yeah
- SPSpeaker
... and so and so forth. But I, I do think that that's gonna be something that's gonna be, uh, it's gonna be a big deal in the, in the years to come.
- 1:09:10 – 1:11:40
Global Implications: Labor, Politics, and the Future
- SPSpeaker
And so we'll have to figure out how to mitigate that.
- MCMartin Casado
To your previous point, Balaji, I just think this is so important. Like I agree, there's gonna be a huge backlash, and I think some of it's gonna be rooted in like the experience of individual people, like, you know, my job is shifting and, and that I, I am very sympathetic and I think we should address.
- SPSpeaker
Yeah.
- MCMartin Casado
But I think there's something more pernicious going on, which is, and this m- maybe this is my cynicism, but more and more I kind of view politics as like you've got, you know, pretty sophisticated people, and they have clientele classes. And then what, what they s-
- SPSpeaker
Yes. Patron, patron client. Yes.
- MCMartin Casado
Yeah. What they say is, is basically what will move the clientele class the most. Like, that's what... They just look for soundbites that'll move the clientele class. And actually, you know, the, like the, the, the patron are sophisticated people that can actually hold nuance in their head. They know their complex topics. But that's the-
- SPSpeaker
Right, right, right. But, but they dumb it down on purpose.
- MCMartin Casado
But they dumb it down on purpose, right? And like what better talking point than AI? This goes back to the Promethean legend. I mean, we're terrified of technology. You can anthropomorphize it. You can talk about it as gods. I mean, it is the perfect tool to mobilize, and we're seeing this on both the right and the left, right? So this is not, uh, in any way beholden to one party. So I think this is like the ultimate, you know, political, you know, tool for any purpose, and we're seeing it for that. And I think for that, even more than crypto, by the way. I think, I think the AI strikes to the heart of people's insecurities more than crypto ever could. And so I think that this is the big battle.
- SPSpeaker
Well-
- MCMartin Casado
And I think it's gonna be bigger
- SPSpeaker
... it's interesting. It's, it's all of the above, right? Because AI is disrupting media, crypto is taking power over money, robots are taking power over manufacturing, and drones are taking power of the military. So all of these... And by the way, there's a crypto angle to at least three of them, 'cause obviously there's a crypto angle to money, there's a crypto angle to AI in terms of constraints. There's a crypto angle to the drones because you're gonna want the control plane for the drones to be on-chain, since that's the part that can't get hacked, whereas the Pentagon can get hacked. So I do think, yeah, there, there's-- This is something where it's, it's going after quite a few power centers. Go ahead.
- MCMartin Casado
I just don't think people-- I just, I just don't think 300 years ago if you said, "Well, listen, you know, we're gonna do this new crypto thing," people would be like... [gasps] But like if you said, "Listen, we're gonna create AI, these artificial, you know, intelligences that have unbound power," I think-
- SPSpeaker
No, you're right. You're right
- MCMartin Casado
... I think you're really getting at a, a core-
- SPSpeaker
Yeah, yeah
- MCMartin Casado
... human insecurity that we've seen in myths and legends for 3,000 years and probably longer.
- SPSpeaker
That's true. That's true. I, I think, well, we'll see what happens with, with, you know, currencies and so on. But I think you're right. It's both.
- MCMartin Casado
It's both for sure. [chuckles]
- SPSpeaker
Perfect. Balaji Srinivasan, thank you very much. [upbeat music]
Episode duration: 1:11:52
Install uListen for AI-powered chat & search across the full episode — Get Full Transcript
Transcript of episode LM7snohbu4k
Get more out of YouTube videos.
High quality summaries for YouTube videos. Accurate transcripts to search & find moments. Powered by ChatGPT & Claude AI.
Add to Chrome