Bankless - Investing in AI, Crypto, & Tech in 2025 | Elad Gil

Starting point is 00:00:00 I do think that to some extent, one could argue that there will have been three ages of humanity, right? The first age is all the compute is roughly humans and then other animals. The second age of humanity is probably this, what we have right now, which is there's some split between humans and machines, and humans are directing the activity and all the rest. And probably the third age is the age of machine intelligence, right? Where that's the predominant form of compute and intelligence on the planet. Welcome to bankless, where today we explore the frontier of tech investing. This is Ryan Sean Adams. I'm here with David Hoffman, and we're here to help you become more bankless.

Starting point is 00:00:40 I led with tech investing. I mean, crypto, of course, is a form of tech investing. We're talking a little broader than that. We're talking not just crypto. We're talking about AI today. We have unicorn tech investor, Eli Gill, on the podcast. He's giving us the Silicon Valley take on AI crypto, and also Silicon Valley's involvement in politics, particularly the recent involvement. So this is one part getting up to speed on everything that's going on in AI, and Elad provides a masterclass in that. And then we talk about the intersection of AI and crypto in his take on that, and then politics.

Starting point is 00:01:15 How will tech flourish under the new administration? What changes have we seen in the last year? Elad probably knows as much about AI as Ryan and I, bankless, knows about crypto, and vice versa. He knows about crypto, about the same amount that I think we know about AI. So we really start off picking his AI brain. But he's just familiar with the growth of markets, the growth of tech, having seen the rise of the internet.

Starting point is 00:01:42 And so not only is he intimately familiar with these tech sectors, especially AI, but also how investing works and how markets develop and grow and the S curves and how all these many S curves relate. So it feels a little bit like ancient wisdom. from a tech investor veteran on the podcast. He's also very calm, which I found very peaceful. So overall, very enjoyable episodes. Does he meditate, David?

Starting point is 00:02:06 Usually call it when guests meditate. And you could be a meditator. He might meditate. He might be a meditator. We didn't get into that in the bulk of the podcast. My big question going in for him was like, okay, give me the truth on AI. Is it overhyped right now or is it appropriately hyped or is it underhyped? So stay tuned for that answer as well.

Starting point is 00:02:25 Yeah. Let's go ahead and get right into the episode with Elad Gail. But first, a moment to talk about some of these fantastic sponsors that make this show possible. If you do not in December of 2024 have an account with Cracken, consider clicking the link in the show notes to getting me started. Want to know the exchange we at Bankless used to buy, sell, and trade crypto? It's Cracken, one of the longest standing and most secure crypto platforms in the world with tools for every type of trader to get started. Over 13 million users trust Cracken with their funds because they lead with transparency and privacy through top-notch security measures. Plus, they'll have access to professional 24-7 365 client support from real humans because your financial goals deserve real attention. Cracken also has a trading platform for advanced traders called Cracken Pro.

Starting point is 00:03:04 Professionals and D-Gens love Cracken Pro because it's one of the most customizable, high-performance and intuitive trading platforms in the industry. Design your ultimate trading interface by choosing from over 25 widgets for market data, analysis, execution, and other order management tools. Save multiple layouts and switch between them effortlessly on any trading scenario across 300. different assets. Manage your trades on the go with Cracken Pro's highly rated mobile app, or use the all-new new desktop app to unlock ladder trading in a native rust application. With Cracken Pro, you can truly trade like a pro. Ready to take control over your crypto journey, visit crackin.com slash banklist to get started today. Not investment advice of crypto trading involves risk of loss and is offered to US customers through Payword Interactive Inc. View legal disclosures at Crackin.com slash

Starting point is 00:03:44 disclosure. Uniswop Labs is making history with the largest bug bounty ever. 15 and a half million dollars for critical bugs found in Uniswop v4. just any update, Uniswap V4 is built with hundreds of contributions from community developers and has already undergone nine independent audits, making it one of the most rigorously reviewed codebases to be deployed on chain. And with 2.4 trillion in cumulative volume process across Uniswap V2 and V3, without a single hack, the commitment to security and transparency is rock solid. Now Uniswap Labs is taking an extra step to make V4 as secure as possible with a $15.5 million bug bounty. Head to the link in the show notes to dive in and

Starting point is 00:04:23 participate in the Uniswap v4 bug bounty. All the details from eligibility and scope to the rewards are there. Have you ever felt that the tools for developing decentralized applications are too restrictive and fail to leverage advancements from traditional software programming? There's a wide range of expressive building blocks beyond conventional smart contracts and solidity development. Don't waste your time building the basics from scratch and don't limit the potential of your vision. Cartese provides powerful and scalable solutions for developers that supercharge app development. With a Cartese virtual machine, you can run a full Linux OS and access decades of rich code libraries and open source tooling for building in Web3. And with Cartese's unique roll-up framework,

Starting point is 00:05:00 you'll get real-world scaling and computation, no more competing for Blockspace. So if you're a developer looking to push the boundaries of what's possible in Web 3, Cartyzee is now offering up to $50,000 in grants. Head over to Cartesey's grant application page to apply today. And if you're not a developer, those with staked CTSI can take part in the governance process and vote on whether or not a proposal should be funded. Make sure your vote ready by staking your CTSI before the votes open. Bankless Nation, very excited to introduce you to Elad Gill. He is one of Silicon Valley's greatest of all-time investors, I would say that. He's backed 40 unicorns, including Airbnb, our beloved Coinbase, Figma, Stripe, many others. He's super active right now in AI. You actually host a podcast

Starting point is 00:05:42 that I enjoy very much. It's almost like the AI sister podcast to Bankless. It's called No Priors. He's just in general, a very prescient thinker. Today, we want to get his takes on AI, on crypto, on the next decade of tech investing, what the opportunities are. Elad, welcome to Bankless.

Starting point is 00:06:01 Oh, thanks so much for including me. All right, so let's start with AI, because David and I have been following this in various ways. The Bankless audience knows about it, but we focus mainly on the Crypto-Tech Revolution, not as much AI, catch people up, catch us up to speed. So obviously there's a lot going on right now. How about this year?

Starting point is 00:06:20 If we just zoom into this year, what are the big things, like the big milestone events that happened in AI that we should be paying attention to as general tech investors? Yeah, you know, it's kind of interesting because if you look at the history of machine learning and AI and we used to all call it machine learning 10 years ago, I think what a lot of people really underappreciate is that we had a series of fundamental breakthroughs,

Starting point is 00:06:42 that effectively put us on a different technology curve from what we traditionally talked about machine learning. So we used to talk about convolutional neural networks and recurrent neural networks and all these things in the 2010s with AlexNet and a few other breakthroughs around machine vision and a few other areas. And the basis for those things kind of shifted in 2017 when Google invented what are known as the transformer,

Starting point is 00:07:07 it's a specific type of architecture model for machine learning, that got implemented a great, but also to OpenEI, and the T and GBT is Transformer. And it wasn't really until chat GPT launched just two years ago that a lot of people woke up to this fundamentally different technology curve that we're on. So it's almost like we're on one curve with machine learning, and then all this transformer-based stuff came out, and it kind of boosted us into a different trajectory and honestly a different technology curve.

Starting point is 00:07:32 And the capability set is very different from what we've experienced in the past with traditional machine learning, which is really effectively running a bunch of regressions in some sense or kind of data mining out statistical correlations between things. This wave of AI, which a lot of people are calling generative AI, is about the ability to understand and manipulate different types of language and imagery and a few other things. And language includes things like code. And language includes, you know, the synthesis and understanding of knowledge.

Starting point is 00:07:59 And there's these multi-step processes that you have and all these other things. And so when people talk about AI, we've been talking about it for literal decades. But I think the specific flavor of AI that we're focused on right now, I was only two, three years old. So even saying what's happened in the last year is almost like saying what's happened in a third or half of the time frame of which we've been aware of this really interesting technology shift that we're undergoing, right? I think there's probably a mainstream interpretation of AI and how it's going to impact our lives. We see chat CBT, we see its implications. Then there's stuff like, you know, custom AI generated videos and, you know, deep fakes and this is all going to impact our lives in this particular way.

Starting point is 00:08:39 there's probably a consensus level understanding of how AI is going to impact us. But I'd actually like to see if there's any difference there between what you think is the future trajectory of AI. And if there's a gap between mainstream understanding of the future AI and what you think is true. Is there a gap? Is there an alternative version of AI's future that you think is being underrepresented or underdiscust that mainstream society, mainstream tech forward people are maybe missing? I don't know if it's underdiscust. I do think in general, AI is still underhyped despite being incredibly hyped, right? And I think the thing that people fundamentally misunderstand is as we have these advancements over time,

Starting point is 00:09:18 really the product that you're selling or the end product of these AI systems is units of cognition, right? You're selling pieces of thought or ability to do things. So, for example, you look at a company like Decagon, and what it's doing is it's augmented customer success agents. So you're a customer support rep, and you suddenly are making that person's job dramatically easier, or you're tackling more of the queries that they get through users. And so each person can handle a much larger base of people. And suddenly a person who only speaks to English can support people in 30 plus languages, 24-7. You know, you're kind of shifting the paradigm of how work is done in the leverage that you're getting on work.

Starting point is 00:09:54 And so for the digital world, you're basically selling units of labor. For the physical world, if some of these advancements and robotics come through, and, you know, I think that's, we're much earlier in that curve. And then eventually you're selling units of robot time or labor time or however you want to phrase it. And, you know, it does seem likely that one of the major types of robotic chassis or approaches that we'll have or form factors is going to be humanoid. A, because the world is designed around that and B, because it's general purpose in terms of the things it could do, right? But that's much further ahead in the future. If you just look at the digital part of it, which is very clear right now, I think eventually you're selling units of labor or thought.

Starting point is 00:10:35 And that's very different from, hey, we're doing. this regression on a bunch of data, which was machine learning. It feels a little cloud-like, as in there's a cloud of labor out there that you can purchase units of labor. Is that an acceptable comparison? Yeah, I think that's a great way to put it. So as an example, eventually you'll have a series of bots that will do aspects of coding for you. And already people are using coding tools like cursor or Magic or Devon or other things, right? But fundamentally, as those capabilities get better and better, you're going to have the AI system to write more and more code for you,

Starting point is 00:11:11 or the AI system do more and more customer support for you. I don't know if you saw the tweet from the Sea of Klarna that came out maybe six to nine months ago, where he said that they let go of 700 people on their customer support team because they replaced it with something they built on top of OpenEI, and it had a higher net promoter score. They had a 25% reduction, I think, in repeat queries. It was available 24-7 and I think it was like 20 plus languages. And the time to actually resolve the issue for a customer went down dramatically. I don't remember how much, 50% or some significant amount. So strict improvement across all domains.

Starting point is 00:11:49 Strict improvement across all domains with an AI system, right? Effectively, to your point in the cloud, they created a cloud of customer success agents or help. Now, it wasn't quite agentic. Agentic almost implies like this thing is going to be self-acting in really deep ways. And the technology basis still has to develop to get there. But again, I think it's back to it's units of cognition or labor, how do you want to phrase it, and it's units eventually

Starting point is 00:12:11 of robot minutes or robot time or physical labor time. And that's eventually what this wave is about, right? At least for a language, right? There's image gen and video and then there's all these foundation models being built for physics and material science and biology and all these other areas, which is different.

Starting point is 00:12:28 And we'll have different substantiations and kind of overlap in terms of what type of person they're going to augment or replace. What I think I would love to find out from your perspective, Elad, is like, where we are on kind of the S curve of this new, of this new transformer sort of unlock that we've approached in AI. So you're talking about the measure being units of productive intelligence. And in the case that you just gave is basically like some corporation has a, you know,

Starting point is 00:12:57 customer support intelligence center that is primarily staffed by human agents right now, right? So those are its units of customer support, productive intelligence. And what it's just done is in using, you know, Open AI, it's been able to replace that productive intelligence with a higher form of productive intelligence, right? So you see that that could happen in customer support, maybe some white collar type jobs. We're also getting these version upgrades with, you know, the chat GPs of the world, right? So like we go from three to four and to like five to six. But at some point in time, I mean, we know in crypto, right?

Starting point is 00:13:33 you're on the other side of that S curve and kind of adoption, progress for the technology that you're on, it starts to peter out. I mean, where are we on the curve here? Is this still early? Like, do we have a lot of upgrades to go? Or, like, when does this version of the tech curve kind of stop and start getting diminishing returns? Yeah, this is an area of active debate within the AI community.

Starting point is 00:14:01 And some people say, well, And really if you look at AI, there's three or four components that go into how smart the system or how capable the system is. And part of it is how much data and what sort of data do you have to train it and how clean is it and how is it labeled and how is it generated and etc. There's how much computer you're throwing at the thing. And, you know, there's a lot of fine-tuning or post-training. What do you do once you actually have the system up and running? And then there's the how much computer are you actually allocating at the time that you're, you're doing inference, which is for the moment that you ping the AI system and ask it to do something on your

Starting point is 00:14:37 behalf. And each one of those things can scale in different ways, right? And so 01 from OpenEI is really focused on that last piece. And you see that a lot happen with people. Like you're asked a question, and if it's a hard question, you'll think for a minute or a few seconds or whatever, and then you'll answer. You won't just spontaneously answer in an easy question you immediately belt out, right? And so differentially allocating compute is what you do as a person, right? You're kind of thinking more or less. And similarly, that has its own scaling law that people think has a lot of runway on it. The training side probably still has quite a bit a headroom on it in terms of just the core. I'm going to train a big model on a bunch of GPUs. And on the data side, there's ongoing

Starting point is 00:15:17 questions of like, do you move to synthetic data? Are you capturing data in new forms? Are you going to specific types of experts for the post-training side of it? So I think in each of those, there's debates around how much room there is, but I think overall there's an enormous amount of room still to improve. So on the one hand, I think we're still reasonably early in the S curve. And that's, again, a debatable topic. But I think there's a lot of headwages with the stuff we're doing. There's a separate question of, say that we stopped all progress. And we just took the models and capabilities we have today. How many more applications could we have cover for it? And there's a ton. Again, this technology in some sense, or people's awareness of the technology

Starting point is 00:15:54 is two years old. GPT4 came out, I don't know what, 16 months ago. Right? And four was a big stepping versus three, and that's when you could suddenly do legal. So, you know, one of the companies I've backed is called Harvey, and they have this sort of legal assistant and set of tools. And on 3.5, which was the version of GPT right before, right before 4, they couldn't do legal workflows. It just didn't work. It wasn't smart enough. And it forced them, it was capable enough, right? And so I kind of call it the GPT ladder or step, right? As you move up from 4 to 5 to 6 to 7, you're opening up entirely new markets that couldn't be served before because the thing wasn't smart enough to serve them, or you're fundamentally changing the capabilities that you can do in those markets. And if you look at SaaS in the U.S.,

Starting point is 00:16:39 it's a half trillion dollar a year, SaaS and enterprise software, it's a half trillion dollar a year market, if you look at all the sort of white collar-ish services, the payroll for those in terms of areas that could be impacted by AI is about three and a half to five trillion. So if you convert 10% of that headcount cost, just employee salaries into SaaS revenue for AI, you've recreated all of SaaS and enterprise software in terms of market cap, right? So this is a huge revolution. And that's why I'm saying it's under height, right?

Starting point is 00:17:09 And so there's two S curves. It's the technology S curve. And it's really a few different S curves stacked on top of each other. As mentioned, there's the training and the post training, and there's the time of inference stuff. And then there's the S curve of adoption.

Starting point is 00:17:23 And on the S curve of adoption, we're basically in the era of, you know, it's a year after the Bitcoin white paper dropped or something, and we're about to have like the Mount Cox blow up or, you know, whatever analogy you know, it's really early. Do you know what I mean? Yeah, okay. So you're painting this portrait of stacking S curves, all of which are early, including the adoption S curve here. Can you give some insight into how transformational this will be? So you've been in Silicon Valley since the early days. You've seen kind of the birth of

Starting point is 00:17:53 the internet and you've seen mobile and you've seen crypto and you've seen some. And you've seen social media, you've seen all of these kind of trends, right? And famously, Peter Thiel talks about how, like, some of these, you know, software value, you know, creation hasn't really filtered down into like real world productivity for the rest of the world. Anyway, is this different, or how does this compare to previous revolutions? I've heard some people compare the AI revolution, this idea of unlocking productive intelligence to the industrial revolution, which was kind of like a whole new scale of productivity output for humanity. And I think that's a different trajectory maybe than what the internet was. Anyway, how does this stack up in terms of the tech that you've

Starting point is 00:18:39 seen come out of Silicon Valley since you've been there? Yeah, I would separate out two things. Because again, there's a digital revolution and there's the atomic revolution of robotics, right, which is, again, we're not quite there yet on the second one. But that has its own transformation curve or its own implications sort of globally. And self-driving cars is, sort of one substantiation of that in some sense, too. And that's coming much faster. So there's really the atoms and the bits, and you sort of separate that? I would separate it out, yeah, because that's very different society level implications,

Starting point is 00:19:06 but also the decree to which you rework cities and physical labor and everything else is fundamentally different from, you know, what can you do with digital information? And the bits is happening now. The bits is happening now. Yeah, the bits is happening now. I mean, the Internet was the biggest revolution of them all for this stuff because this these set of AI systems couldn't exist without the internet and distributed compute, and neither could crypto, right? And so I think a lot of the, a lot of the really big technology waves are just

Starting point is 00:19:37 outgross of the internet in one form or another. And so that to me is sort of the granddaddy thing or whatever you want to call it. But on a relative basis, I think this is much bigger than mobile. It's going to be bigger in some ways than certain aspects of social. Crypto, I think, has been enormously transformative, mainly at least to date in, you know, the financial sector. And obviously it's spilled it over into art with NFTs and, you know, it's spilled over in other ways. But I think a lot of the kind of web three premise of everything is going to be on the blockchain, you know, have Airbnb on the blockchain with a token to incentivize hosts. And, you know, that stuff hasn't really proven out quite yet. And so I think, you know, it's going to be a very big revolution.

Starting point is 00:20:19 And it's going to take a decade plus to substantiate. One thing that we've seen in cryptic is kind of these repeating hype cycles where every sort of four years or so crypto goes out of terror and then it sort of gets ahead of itself in terms of, you know, the market price reflecting kind of the reality of where the technology and where the adoption actually is. Sure. And I'm trying to like paint this picture for crypto investors onto AI. Do you anticipate something similar where basically it will happen in waves? And, you know, I don't know where we are in the first wave, but it's been incredible to see

Starting point is 00:20:53 the evaluations of companies like Nvidia this cycle just like absolutely explode into the most valuable company in the world and so it's hard to look at that and not look at some crypto analogs. Well, okay, maybe AI is going to be a long-term transformational technology

Starting point is 00:21:09 but are there periods of time within the decades of that transformation where it gets kind of overhyped from a market perspective? Are these companies making money? Do they have business models yet? Yeah, I mean, Nvidia is clearly making money So I think that's her reason it's value so highly.

Starting point is 00:21:25 But that's one of those like reflexive money type things, right? It's like it's making money because there's so much demand for GPUs and chips because there's so much, you know, I guess hype going into the other elements of AI. Well, it's not just hype. It's really interesting. So if you look at the foundation model world or at least LLMs, these large language models, which is, you know, open AI is GPT or, you know, Claude from Anthropic or Sonnet from Anthropic or, you know, kind of name your. your model, Google and BART and all the things you're doing there, or Gemini, you know, fundamentally, the reason the hyper-scalers ended up as a primary backers of these big model companies, right? AWS probably is the biggest backer of Anthropic now, and Microsoft is the

Starting point is 00:22:10 biggest backer of Open AI, et cetera, is because it also drives enormous revenue on their clouds for AI services, right? So Microsoft had something like a $28 billion quarter or last quarter, And I think they publicly said that 15% of the lift on that quarter, so that's what, $3.5, $4 billion came from AI. That's incredible. That's significant, right? And so these things are translating into real revenue. And it's happening at startups where suddenly you see a startup go from zero to 10 to 50 in two years,

Starting point is 00:22:38 which is insane in terms of any sort of traditional SaaS application. You see that in terms of the rumored numbers around Open AI, where they're now in the billions of revenue after two years or three years of, offering GPT as an API that is actually being used, right? I mean, these are insane adoption curves. Now, there's a separate question of what is the durability of a giving company relative to these adoption curves because overall the segment's going to happen. It's such a useful and powerful technology.

Starting point is 00:23:06 And it gives you so many capabilities and so many cost savings and so many new revenue streams and all the rest of it, that it's happening and it's going to happen. And then the question is who wins. And for each layer of the stack, right, you have the foundation models, actually have the chips with Invidia, you have the foundation models, you have apps, actually you have infrastructure, then you have apps, and then within apps, you have B2B, and you have consumer. And within that stack of stuff for each sector, you can kind of go through and ask, is it an oligopoly market? Is it a monopoly? Is it highly fragmented? Who wins? Why? What's a defensibility of each one of

Starting point is 00:23:36 these things, right? What's the technology basis for winning? Do they have to build their own models or not? So there's all that stuff, right? And so I think the direction is clear. The who in some cases is clear and in some cases it's less clear. And it's the old thing about how the future in some cases is a determinant, but you just don't know who's going to be the person who drives that piece of the future, right? But you know that future is coming. And I think that's very true here. I mean, we've definitely seen that with crypto. It's been hard to predict what the individual networks that, like, we now are going to be. But we sort of know that the future is inevitable. I want to ask you a question about the evolution of this market. So you're

Starting point is 00:24:12 painting the picture, Elad, of this is an early market. It's still actually underhyped from your perspective, even at this point in time. How do you think this industry evolves? And there's different levers on this. Like a lot of the value right now seems to be going in kind of, I don't know if you call this the platform layer, but like, you know, the Mag 7 and some of these big companies, right? There's also this other layer we could talk about like closed source versus open source. I mean, something that the, you know, crypto advocates are very passionate about is decentralization,

Starting point is 00:24:42 right? Anybody being able to use this technology. Sure. There are different versions of that in AI. but that's something that I'm sure you support permissionless to centralize the ability for anybody to spin up these tools and for not to be cloistered in some walled garden. Anyway, what do you think of this market structure? Will it be centralized? Will there like be kind of power law winners here?

Starting point is 00:25:05 Sure. Or will this be more diffuse? I mean, will we see Decky unicorns in the startup world start to compete against some of these Mag 7 companies? Yeah, there's a ton of questions in what you just said. So let me try and tackle them one by one. I think one is around open source versus close source. And obviously I'm a huge fan of open source software. I think both will happen in this market segment.

Starting point is 00:25:27 And I think arguably both have happened in crypto, right? You have dexes, but you also have centralized exchanges. And a lot of the things that are supposedly decentralized are actually way more centralized than people really see it for splash, right? Like how many people can actually commit to Bitcoin Core? And, you know, how many miners actually make up what proportion of the network? work. You know, it's pretty centralized, actually, in some ways. So, or you look at Solana or other, you know, protocols. And, you know, relatedly, sometimes there's more centralization or less centralization, right?

Starting point is 00:26:01 And so the same is going to be true in this world where the really big open source models of the foundation model layer at this point are basically Lama from meta and then Mistral. Right. And there's a bunch of other stuff, but at least for the language model side, those are the ones that are, I think, most prominent. And then there's other types of open source models for a wide range of other areas in terms of, you know, things that have come out of academia for robotics or weather simulation or biology or, you know, so you can kind of go through one by one and ask, will it be closed source or open source? And to some extent, the hard part for open source and traditional software, is different from crypto where you can often monetize in other ways through a token or, you know,

Starting point is 00:26:49 there's two or three ways, actually, that you can monetize open source in crypto. Many of those things don't apply in the AI world. And it's much more like traditional SaaS where you have to figure out a business model around the thing and charge for it. And so for Lama, there's a few things that I thought were really clever that Facebook did. One is if you're over a certain username, I can't remember what it was in the original license. It was like, If you add over 700 million users, you had to pay for it or license it, otherwise you could use it freely. And so that meant if you're a hyperscaler and you were trying to put Lama on your platform, you had to license it. Or if you were one of the really, really big social networks or companies with enormous numbers of users, you'd have to license it.

Starting point is 00:27:30 Everybody else could use it for free. I thought that was very clever of them. Can you elaborate on why that's clever? Why does that work so well? Why is that a good mechanism? Because I think it does two things. One is it potentially, and I don't have any insights into how Facebook thought about it, it potentially creates a monetization path for Lama because all the big hyperscale is if they want to adopt it,

Starting point is 00:27:52 they have to license it and pay for it. And again, if it's driving cloud services, they should benefit from that. It also, I think, means that the very large competitors of meta can't just adopt it and compete with meta using their own technology. And so because then they kind of bridge the gap of having something that was roughly fully open source, but for the things that I'm assuming they really cared about, it was effectively close source or at least you had to license it, right?

Starting point is 00:28:14 I thought that was really smart. And that meant any developer can just pick it up and start using it almost anywhere in the world. That's amazing, right? So I think a lot of the value of that type of open source is probably going to go to the infrastructure providers

Starting point is 00:28:29 and then app companies that want to use that as a differentiator. The reality is, at least today, a lot of people are using OpenAI or Claude as sort of the starting point and then if they figure something out, then they'll go and maybe fine-tune and open-wates an open-source model like Lama.

Starting point is 00:28:47 But it's kind of like try it on an API first where you don't have to do a lot of heavy lifting and see if it works. And then if you think you can really optimize it or you're worried about data security or whatever, maybe, and again, I think the data's quite secure using these other APIs. But if you have some concern about sending data back,

Starting point is 00:29:01 then sometimes you go down the open source route, you know, and fine-tune your own model or do whatever it is you need to do. We've discussed the multiple adoption curves, S-curves of the growth, of AI and why it seems to be that it can accelerate very quickly in the near term. I want to introduce one more adoption curve, S curve of technology, which is crypto. And I think we're starting to approach crypto's middle of the S curve these days, especially

Starting point is 00:29:26 with Bitcoin crossing $100,000. I'm wondering if you're paying attention to the intersection of crypto AI. And this intersection has been growing. Right after ChatchipT launched, I remember there was like kind of a, a search. of crypto AI tokens, like right afterwards, kind of just writing the narrative of AI. And it was, it didn't really make any sense. It was kind of like, I'm going to call it Stone Age version of crypto AI. It's very rudimentary.

Starting point is 00:29:53 It wasn't real. But that was over two years ago. And since then, I think there have been some developers who are really trying to make this work, like trying to figure out how do these two parallel frontier technologies, how do they intersect. How do they grow together? And lately, there's been some very strong sparks that have probably actually turned into at least a small to medium-sized wildfire in the crypto world. Hasn't really broken out into mainstream. And this is kind of the story of truth terminals. I'm not sure if you're too familiar. There's, but the number one GitHub, download, forked and starred

Starting point is 00:30:33 GitHub repo right now is the Eliza framework, which is allowing people to build their own AI, agents, some with crypto wallets, some just vanilla agents. So I'm wondering, are you observing the space? Is this space interesting to you? And if you just have any takes? Yeah, I have like four or five comments on it, I guess. The first thing that I think is fascinating. Do you know the origin of near?

Starting point is 00:30:54 The near protocol you're talking about with Ilya and team? Yeah. I mean, it was an AI origin, right? Yeah, exactly. Elia has a very well-sighted AI paper. Yeah, he's on the transformer paper. He's on the original transformer paper. He's the last author on that paper.

Starting point is 00:31:08 Oh, wow, fun fact. I didn't know that, actually. Yeah, yeah, yeah. And so my understanding is when he left us start NIR, and it was called Near. A.I. Right? It was an early AI company.

Starting point is 00:31:21 And they were originally going to do almost like GPT-style stuff. And they decided they needed to do, and he'll probably correct me and all this. My understanding is it decided that they needed to do a lot of data labeling. And they're like, how can we pay people to label data around the world, maybe use Ethereum? and of course Ethereum wasn't scalable back then, right?

Starting point is 00:31:39 That's why we have all the L2 stuff and all the rollout, all the stuff we're doing on top of Ethereum now. So they said, hey, how do we create a really scalable protocol so that we can create tokens that can be used to pay people

Starting point is 00:31:50 around the world to label data so that we can then build a giant AI system? I think that's the origin of near which is fascinating, right? Or at least some version of the origin story that I've heard of it. So I think there's long been an intersection of the

Starting point is 00:32:06 people who are interested in AI and the people who are interested in crypto. And I almost feel like in a era when one of them has been hotter than the other, it's tip people's career in one direction to the other. I don't know if you know Uma Roy from Sysink, right? I just had her on the podcast. Yeah, she's great, right? And so that's a great example of somebody who did a lot of AI and ML at MIT. And maybe if she'd started her company two years later, she would be doing AI stuff right now. I mean, she's brilliant, right? And she's very smart on ZK and the mathematics underlying it. But, you know, there's stuff like that that I think. is just fascinating in terms of these paths that are almost like moment and time dependent on when

Starting point is 00:32:40 you graduate. And maybe she would have on crypto anyhow. I don't know. I'm just saying like she had that background, right? And so I think there's a lot of overlap in the backgrounds that people are interested in these things, at least a subset of people. So I think that's one aside just in terms of the human capital or the people who are excited about it. I think there's three or four approaches that people have been taking to that intersection of AI and crypto and a lot of it is over time has been, can we create these data repositories where you get paid into tokens to label data or use data or do data, et cetera. There's the distributed compute stuff, which I'm more skeptical about.

Starting point is 00:33:11 There's reasons you centralize these things usually. One could argue WorldCoin isn't part of proxy on Sam Alman, you know, mimetically. There's identity, which I think is super interesting. And actually, do you think like a blockchain resident identity could be used by the agentic AI world? And I'm still surprised nobody's built like a truly good identity system on the blockchain. There's payments, which is kind of the obvious one, which is why maybe you mentioned, you know, building AI with the wallet.

Starting point is 00:33:36 because it'd be natural for an algorithmic agent to transact using crypto. I think the censorship and censorship resistance is super important aspect of crypto that AI doesn't have. And I think that is showing up in almost like what you call the politics of the models, right? Because the models are all being steered down very political, the specific subset of political paths, right? They kind of reflect the Bay Area and the Bay Area politics is basically what you see when you

Starting point is 00:34:04 interact with one of these systems. and that may not be a good thing for humanity overall. So I think there's a lot of ways that these things kind of intersect. One conversation in the crypto space is the idea of decentralized versus centralized AI. We have the chat GPTs, the open AIs, the Facebooks, Googles, with a lot of resources, the benefits of centralized coordination. I think really pushing the frontier of AI, really introducing it into society into mainstream. And then in response to that, there's been a growing, parallel world of decentralized AI, which I think has some difficulties because I think maybe one of the

Starting point is 00:34:42 biggest drawbacks is that decentralized versions of AI don't have access to the same data that OpenAI that meta does. But there's been some victories. One thing that has come on my radar as of recently is prime intellect. I'm not sure if you're too familiar, but the story that I've heard is that these decentralized training models have had some constraint in these centralized form factor of training, of AI training. Prime Intellect is a decentralized training platform. I'm kind of like out on my ski tips here explaining this technically, but they were able to poke through to like break through a glass ceiling of the number

Starting point is 00:35:19 of parameters that they were able to use meaningfully from, I think the constraint in the centralized world was something like 400 million parameters and now they're breaking through at 10 billion. So a very significant increase in part of the decentralized structure of the AI revolution is starting to get some wins versus their centralized counterparts. I'm wondering if this is how you see this dichotomy, like there's two like different worlds of AI. There's the centralized AI and the decentralized AI and they're both kind of developing and iterating forward into the future. Is that how you see it? Not really. I tend to think of it more as open source and closed source.

Starting point is 00:35:55 And open source may include open weights. It may include open data sets. You know, it may include all these things, I don't think you need to necessarily decentralize the compute. You know, I don't think that is necessarily helpful unless you're like, we can't afford it otherwise or we're crowdsourcing it. So there's reasons to do it. We can't buy all the GPU, whatever it is. But I don't really view it through that framework because the decentralization is not additive in this context as far as I can tell unless again, it addresses some of those other issues I mentioned. And so it's more just like what's out in the open that anybody can use and how can they use it. And I think that's the vector I would care about the most. It's kind of, you know, the one other thing that I used to speculate

Starting point is 00:36:35 on that I no longer do around crypto and AI. And by the way, I think it's really cool. They built a 10 billion parameter model in a decentralized way. So again, I'm not trying to say it's not cool and interesting. It's more just the dimensions I care about are like what's open versus closed and who has access, right, versus a decentralized or decentralized. The thing I think is kind of interesting is if you think about the blockchain, you've always had these kind of programmatic agents and very, very simple agents acting on like Ethereum, right, smart contracts and other things where effectively these programmatic pre-bake things that happen with strong economic value associated with them. And so four or five years ago, I used to speculate,

Starting point is 00:37:19 well, is the blockchain the first place where artificial general intelligence emerges? Because you have these economic games that are played repeatedly between effectively agentic systems. And so what a wonderful place to actually train an agent or an AI, right? So maybe that's the place where the decentralized AI step is super interesting. If you basically have these selective functions on blockchain resident agents playing economic games at scale and learning off of them. You know, that's really cool. I think that's been like more of like my personal interest. And I think when David and I started really getting into kind of the intersection of AI and crypto is more on the agenic properties of it.

Starting point is 00:37:57 Because the idea of crypto is, you know, this podcast is called bankless, right? It's the ability for individual, like, humans, to go bankless. But actually, the most underbanked population of the future is probably going to be AI agents, right? Very difficult to, like, kind of, you can open up a stripe account, but then you have, you know, 3% credit card transaction fees. I mean, they open up an AI. Like, the form factor of a blockchain with program of money and property rights is basically, like, fit for purpose for AI agents. You need a social security number to open up a stripe account.

Starting point is 00:38:32 Right. And so some of the early use cases, which are starting to be interesting in some of those frameworks that David mentioned is almost like the AI agent influencer, you know, like package, right? So he's talking about Eliza and there's this platform called Virtuals and there's this meme account called Truth Terminal on Twitter and it has this goat token. And it's basically almost like taking the job of like influence. Let's say like crypto influencers, right?

Starting point is 00:39:00 And so it's playing almost like attention economy type of games. And, you know, AI agents and the LLM like models that we've created are pretty good at like interacting with humans and like passing the Turing tests and being interesting to interact with. And so that's been the early almost like toy version of AI agents on chain. And it's been interesting to see when you hook up a Twitter account basically and allow the LLM to kind of of like talk about whatever it wants and pair that with a crypto wallet. So create that token incentive. What experiments fall out of that? And to your point, like, I actually think that thesis of, you know, the AGI comes through this intersection of crypto and AI could be like spot

Starting point is 00:39:45 on as this evolves. Anyway, that's the potential that we've been more excited about recently. But people look at it and they're just like, ah, it's just meme coin games again. You just hooked a Twitter account to an LLM model. And it's pumping a meme token and like, who cares? What's your take on this trend? Do you think there's something here? You know, I think it's really interesting training data for an AI, right? Again, if you're, if you're getting back into economic game theory and how humanity evolved, and what are the set of drivers for evolutionary progress? And probably a lot of it was forms of economic games, right? And so I think as a training set, it's super interesting. Obviously, you still need

Starting point is 00:40:22 skill of compute and you need skill a day. You need all this other stuff. But I do think it's really interesting from that perspective. One of the big sort of future AGI concerns or safety concerns that Anthropic talks about is resource aggregation by AI. And so they spend less time talking about, hey, well, let an AI invent a virus that will do X, Y, Z and all this stuff, which I think is actually not very likely anytime soon. And they're more like, well, what if an AI is smart enough to start aggregating resources at scale and

Starting point is 00:40:48 using those resources to manipulate society and how do you prevent that? You know, it's sort of, and crypto would be a good basis for that. Basically, the AI as a god or forming a religion or creating some sort of memetic social movement, something like this? Not even as a god. What if it's just very good at predicting stock movements and it just starts getting a bunch of money and paying people to do stuff for it that gets us more money, right? So it's more about that form of safety. I do think that to some extent, one could argue that there will have been three ages of humanity. The first age is all the compute is roughly humans and then other animals, right? And from a concentration perspective, it's humans. And that was probably leading up into the century.

Starting point is 00:41:30 And then we have kind of a hybrid age. Like the second age of humanity is probably this, you know, what we have right now, which is there's some split between humans and machines and humans are directing the activity and all the rest. And probably the third age is the age of machine intelligence, right? Where that's the predominant form of compute and intelligence on the planet. And so to some extent, perhaps we're very lucky to be in this. second age, right? It seems like a limbo period. It's a waiting room for the third period. The grand scheme of things is going to be like all of humanity is in the first age. And then we have

Starting point is 00:42:05 this very short amount of time where we figured out how to get computers to think, but dumbly, like calculators. And now I think the third age is how you're describing it is like, well, now they think on their own. And there's a lot of them, right? And so from a sheer compute perspective, there's way more intelligence embodied in machines at some point. than there are in people. And we're not there yet, right? But at some point, that will happen in the coming decades, right? Eli, does that third age kind of, like, scare you sometimes? I mean, like, we actually opened up the rabbit hole of crypto by, like, talking to Elyse or Eukowski, and he's definitely has a defined thought on this, kind of a doom take on AGI in general. But, like,

Starting point is 00:42:48 should we fear this third age? I mean, maybe it's inevitable. I guess we're all investing in it to some extent. Part of the thing we kind of like worry about if you take L.E. Z. Yukowski's line of view is, oh, cool. So we've created this, you know, property rights and financial system, an economic resource allocation system for a bunch of robots that are going to come and like enslave us or obliterate humanity. Like, great job, crypto. Like, do you have any worries like that? You know, I'm not part of a doomsday cult. And I feel like some of, some of the prognosticators in this topic are basically doomsday cult leaders, right? Like, it's not, I'm not saying anybody specific.

Starting point is 00:43:29 I'm saying in general, like there are these things that effectively feel like doomsday cults, right? And they have AGI as the, you know, before it was like the meteor striking the earth or, you know, some religious event. And it's almost like a religious rapture in some of these people's minds. I think my take is a bit measured, which is I would kind of view myself as like a AGI or safety moderate, right? I think there's enormous good coming through AI. And I think we're only in the earliest innings. And if I look at global health equity and educational equity and all the rest of it,

Starting point is 00:44:00 or however you want to phrase it, right? These are the things that are going to drive the ability of humanity to participate at scale in ways that they couldn't before. You know, everybody should be able to get a one-on-one tutor education that's best in class in the world. And everybody should be able to have access to the world's best medical information through any device, right? And that's coming.

Starting point is 00:44:17 Now, at some point, there will be these very intelligent machines and we'll be interacting with them in different ways. And I don't know what that world looks like. Like, I can speculate. But I think it's very hard to say what, and years or decades from now, things are going to be. It's the old saying that less happens than you think in technology in two years and more happens than you think in five years. And once you hit 10 years, like, it's game over, like 10 years ago, we wouldn't be predicting any of this stuff. We wouldn't be predicting the stuff that SpaceX is doing. We wouldn't be predicting the quantum computing stuff that Google just released.

Starting point is 00:44:51 We wouldn't be, you know, we wouldn't be predicting this wave of generative AI. Nobody predicted this different curve on the transformer models. So I just think it's very hard to think 10 years out. Then we've also over predicted certain things, right? That didn't end up happening in terms of technology curves. New projects are coming online to the Mantle Layer 2 every single week. Why is this happening? Maybe it's because Mantle has been on the frontier of Layer 2.

Starting point is 00:45:16 design architecture since it first started building Mantle DA, powered by technology from eigenDA. Maybe it's because users are coming onto the mantle layer two to capture some of the highest yields available in Defy and to automatically receive the points and tokens being accrued by the $3 billion mantle treasury in the Mantle reward station. Maybe it's because the Mantle team is one of the most helpful teams to build with giving you grants, liquidity support, and venture partners to help bootstrap your Mantle application. Maybe it's all of these reasons all put together. So if you're a dev and you want to build on one of the best foundations in crypto or you're a user looking to claim some ownership on Mantle's Defy apps, click the link in the show notes to getting

Starting point is 00:45:51 started with Mantle. Are you ready to take control over your entire financial life? Crypto, defy and fiat all in one place? Meet Iyield, the free financial planning tool built for crypto natives. Unlike traditional portfolio trackers that just show you the value of your assets, I yield goes deeper. It consolidates everything, assets, debt, income, and expenses, offering you a complete financial picture across 16,000 tokens, 40 defy5 protocols, and all fiaturrencies. With real-time defyield tracking for platforms like Ave, Athena, Eigenlayer, and Uniswap, I-yield ensures that you're always on top of your staking rewards, investments, and cash flow. And the best part, it's 100% free with no ads and it's secure.

Starting point is 00:46:27 No personal data, no ID requirements, no compromises. So head over to iYield.com. So sign up today and get the guesswork out of financial planning. Imagine if the world's GDP moved on chain. Well, say hello to SafeNet, built by Safe. The project set on actually making that happen. It's not a blockchain. It's not a layer 2. It's a transaction processor network built to send money anywhere and swap anything on

Starting point is 00:46:47 thousands of layer 2s, all from a single universal balance. With sub-500 second execution guarantees and zero cross-chain latency, SafeNet delivers a Solana-like user experience for Ethereum and is layer 2s, making transactions lightning fast and seamless. SafeNet connects safe smart accounts across multiple chains, allowing users to access any chain with a unified balance, all with super-eservices. supercharge security, scalability, and speed. So get involved today.

Starting point is 00:47:11 Join the waitlist and follow Safe on X and check out the SafeNet Docks. There is a link in the show notes. Elad, you are a market veteran. You're a tech evolution veteran. You've seen a thing or two. Now we have these like two emerging frontier technologies, AI and crypto. I think we've talked about AI quite a bit. I want to just kind of get your takes on raw crypto, crypto without the AI.

Starting point is 00:47:33 When you're looking over into the world of crypto, you're seeing, you know, Bitcoin in 2009, Ethereum in 2015, Coinbase going public, circle, these companies that are just evolving, maturing, the industry is becoming legitimized, regulatory. We've had regulatory oversight. Now we have regulatory

Starting point is 00:47:50 leniency, or at least we think that's what we're getting. Just like, what patterns are being sparked? What's your overall take? You've seen a thing or two before? Like, what are you resonating with you in the growth of our industry? Yeah, you know, at this point, I feel like I'm reasonably ignorant on crypto. Like I used to spend a lot of time on crypto and I did a lot of

Starting point is 00:48:11 crypto investing in 2016, 2017, et cetera, and got involved with companies like Coinbase and some of the protocol level things and then index funds like Bitwise and, you know, just different things across the industry. Starkware and that early work and, you know, all sorts of early ZK stuff, you know, like Minna and other sort of early adopters of different ways to UCK. So I think it's amazing what the crypto industry is accomplished. And I think it is a fundamental thing that's here to stay. And usually when I look at these sorts of technology shifts that are so important and so fundamental, I try to distill it down into use cases.

Starting point is 00:48:52 Like, what are the specific aspects of the technology that make it uniquely good for a specific application or use case? Because otherwise nobody's going to use it. And that's sometimes where I feel like people in any market segment, especially if it's a lot of technologists lose their way because they start extrapolating all the stuff off of the technology that nobody actually cares about but they think should happen because the technology is so good, right? And so the mobile version of that is,

Starting point is 00:49:15 if you look at the really big mobile-centric companies that got formed, it's the companies that uniquely took advantage of the new capabilities of phones, which was GPS, so location, always on. It was the instant ability to message or send push notifications, etc., and roughly what that collapsed into was social, products like WhatsApp and others, you know, Instagram, whatever it is. And then Uber, you push a button and a stranger shows up in a car and you're okay getting in and they'll take you to wherever

Starting point is 00:49:44 you want, right? Instacard and other delivery services, right? So you kind of go through and you say, where were the new startups or new things created? And obviously there's enormous incumbency usage, right? Like B of A apps and all the stuff now. But the main things where the mobile revolution was important was one, opening up access nonstop to the internet. because you had a device you could carry with you every or a supercomputer in your pocket. But second was these unique things that were enabled by GPS and text messaging and all the rest of it, right? And so for crypto, it's the same question of, okay, what are the core capabilities of crypto? Because some versions of blockchains, and obviously it's much more sophisticated now,

Starting point is 00:50:21 but in the early days, a blockchain was kind of like a shitty database, right? Yeah. That had very unique characteristics that made it incredibly important around censorship resistance and permissionless systems and all the rest of it. And then you ask, well, where are those unique capabilities really important? You know, you have this 24-7 accessible financial system, right, that's emerged. And that's things like store of value, like Bitcoin, it's defy. You know, it's a bunch of stuff like that in my mind. So I almost view it as like, what are the capabilities of the system and therefore,

Starting point is 00:50:57 what are the unique use cases versus, hey, it's just going to do everything? Do you like Chris Dixon's framing of this, which is like, He goes through, I mean, he wrote a book called Read, Write, Own, which is basically Web 1 is Read, Web 2 is right, you know, so you get blogs in that kind of era. And then Web 3 with crypto is owned. So it's this idea of like property rights, which fits into store of value, decentralized finance, the ability to spin up assets that are like digitally scarce. Is that the rubric through which you view crypto use cases? No, I kind of view it more as like, when do you need these. capabilities. I mean, it overlaps, right? And there's a lot of things that haven't happened in

Starting point is 00:51:37 crypto that I thought would happen. Like what? What were you thinking would happen? I mean, to your point on property rights, like a public ledger of who owns what property, so therefore you can avoid government seizure of your land or other assets, right? Or the notion of identity on the blockchain, where you should be able to fragment identity but provably show that you did stuff. So, for example, you can provably show you, on one identity strand or, you know, one wallet or however you want to kind of assign identity that you're a Google engineer and you've done XYZ things. And then on the other one that you're a contributor to some DAO or something that you're

Starting point is 00:52:16 anonymous on or student non-on, right? So there's a bunch of stuff like that that hasn't happened that the blockchain feels like a really good fit for it because you can transparently in the clear, prove identity without having to reveal identity. What's really difficult in this. And I'd love your take on this is it's, it's, hard to determine kind of the order of operation through which these things will happen. Like some of those use cases that you mentioned, like identity, I'm personally, you know, David

Starting point is 00:52:42 and I are both personally big believers that identity will happen on chain at some point in time. Like ZK, tech stack just like makes so much sense for this use case where you can just like prove something without disclosing all of your, you know, private information about it. But to your point, right, like crypto is not really done much in the identity space. Like Bitcoin has been the big use case, has been like store value being the big use case. And some of the things that we thought would happen earlier haven't happened yet. But the question is, will they happen in kind of the fullness of time? How do you get the order of operation right in these different tech trends?

Starting point is 00:53:19 I think that's a great question. And if I knew the answer to that, I'd be retired right now. I think that I'm joking about retiring. But I think I was going to say, I'm sure you could retire. You're just like, you know, just like what you're doing too much. Retiring is boring. Yeah, I don't know. It sounds pretty nice.

Starting point is 00:53:37 So I think that the order of operations is a great question, particularly with some identity where you kind of have to bootstrap up a network and the usage of it and all the rest of it, right? I think for other areas, it's just never going to work because it just doesn't matter. Like the argument for like a decentralized Uber was, well, every driver can buy a token and therefore have a stake in Uber. and you could go buy Uber stock. Yeah.

Starting point is 00:54:02 Do you know what I mean? It's not. And so I feel like some of these things are a little bit overstated in terms of like, you know, decentralization and a token will magically make something occur. But the flip side of it is like a store of value where you can cross a border with literally a billion dollars in your head, which nobody can seize is amazing, right? That's a superior product to gold or superior product to whatever store of art or whatever you're going to cross a border with, right?

Starting point is 00:54:25 And so I do think that there's a lot of stuff like that in crypto that's really important And again, I do think defy is an example of that. I think all the stable coins and all the USDC and all the whatever form you want to talk about, like all that stuff was clearly going to be really useful for all sorts of applications. And so I just think there's a lot that crypto is accomplished and a lot still to do. And it's very exciting. And the place where I see crypto companies sometimes get a little bit lost is either they just go all and I'm building some really cool piece of infrastructure that doesn't necessarily have an application yet.

Starting point is 00:54:59 or really focusing on like a use case where the existing centralized version works just fine. What would you say, Elad, is like kind of the mainstream Silicon Valley take on crypto right now? Because I feel like I've been charting this from the outside a little bit. It felt like there have been periods of time where there's some pretty intense skepticism about crypto, just like this term that you used earlier of, you know, it's just a slow database. Like, who cares? Or like, you know, I don't understand why Bitcoin is worth what it is. This feels like a tulip mania.

Starting point is 00:55:31 And, you know, this has been present at various times. In fact, a lot of the crypto startups have just, like, not happened in Silicon Valley. They've happened outside of sort of the VC Silicon Valley apparatus. What does Silicon Valley think about crypto now? Or do they not think about it? Is Silicon Valley all, like, AI, AI, AI, and, you know, this crypto thing's happening on the side, but it's, like, less interesting. How has the narrative really shifted? I don't know.

Starting point is 00:55:58 I think there's a subset of founders and firms on the venture side and all the rest who continue to just participate in crypto ongoing, right? And so, you know, and then there's obviously specialist firms that have kind of spun up, right? So, you know, in terms of the broader firms, obviously Andresen has continued to do a lot of crypto with Christensen and others there. There's been standalone firms that obviously either spun out of existing institutions or set up on their own. That's paradox. I'm in electric and on ventures and, you name it, right? There's a bunch of different folks.

Starting point is 00:56:34 And they're all based in Silicon Valley, right? It feels like maybe the biggest shift over the last couple of years is the degree to which momentum really increased for New York is one of the core places for crypto over the last eight five years or whatever time period you want to put on it. And it was almost like this transition from L1 to other layers, right? Because it felt like the L1 stuff continued to be out here for a while in the Bay Area. And then it kind of shifted elsewhere. and obviously it was always distributed

Starting point is 00:56:59 and there's people working from all over but I just feel like the type of work or almost the subcluster in crypto for a while at least reflected the nature of the work being done in terms of the geography where there was concentrations of people

Starting point is 00:57:12 but I think mainstream tech maybe 20% of people always thought crypto was really important and kind of thinking it was important and then maybe there's 30 or 40 people who swing it in and out based on Bitcoin price and then there's 30, 40%

Starting point is 00:57:27 it will always be skeptical. And, you know, those are the people who are like, get off my lawn, you know, kids playing there. You know, it's very, um, and some of these folks are very smart. I just think it's a little bit backwards looking because I think it's, it's clear that there's use cases, it's clear that there's adoption. And I do think there's a lot of assets that are reflexive and it's one of them. And it's, it's hit that escape velocity, in my opinion, barring something really unexpected or, you know, some quantum attack or God knows what, you know, but even then it'll survive it and obviously those quantum resistant algorithms that you can add it. You know, it's just there aren't that many things I can think of that would disrupt it outside of, you know,

Starting point is 00:58:04 civilization level events at this point, but that could be wrong on that, you know. Maybe regulation, like a global lockdown from a regulatory perspective with some tyrannical world government or something, you know. I would love to pick your brain on regulation because it affects many things in A on crypto. You know, like one last thing while we're on kind of like crypto companies and the evolution there is crypto found. So you've had an opportunity to invest in work with many of the best tech founders in the space. You know, people like Palmer Lucky, Brian Armstrong, of course, you know, Coinbase, who we're very familiar with, Dylan at Figma. I'm wondering if you think that basically the tech founders generally are similar to the crypto tech founders.

Starting point is 00:58:45 Or if you think that there's a different pattern with a crypto tech founder, one thing that's kind of breaking the mold a little bit in crypto is, obviously you have the original crypto tech founders. original crypto tech founder, which is Satoshi. And like, who is this person? Right? They disappeared. This is not normal. Like, that's certainly not something that happens. Then you have people like Vitalik.

Starting point is 00:59:05 And he has like, Vitalik almost has, not like Mark Zuckerberg CEO type qualities. There's almost something like David and I have described as like monkish about it. He's just almost like a, sure. He's a movement founder, not a tech company founder. Right. Yeah.

Starting point is 00:59:21 He's founded an entire movement, it feels like. like, almost like a Dalai Lama type figure. There's like, I don't want to impose this on him, but there's something like kind of movement religious about it in a way. And then you also have the Brian Armstrongs and Jesse Powell's of the world and the Hayden Adams and such. Anyway, are there differences between crypto founders and tech founders or are they kind of like, you know, same patterns?

Starting point is 00:59:44 It's an interesting question. I feel like to your point, there's been multiple different types of ways of crypto founders because it went from a true backwalk. right when the Bitcoin paper dropped and you know you had to be kind of weird to get into it right and you were you know using it to buy pizzas and you know all the stuff just to try and get other people to adopt it right too if you remember those days where they were just giving out Bitcoin to get anybody to do something and so that that was a wave of people and obviously that also came out of more of a kind of right-leaning technology community which is very different from the internet

Starting point is 01:00:21 which was very left-leaning in terms of its origins right in terms of the types of people people working on it. And as you kind of moved up and down the stack, you changed the type of founders that show up. And I remember there's also an era of a lot of like professor coins, right? In like 2017 or 2018, like every, every professor working on crypto like wants their own token, right? Or was working on a token. And there's kind of the AI equivalence of that where you have kind of these professor AI companies, right? There's these new architecture model company that spins out of an academic lab and it has a very similar characteristic where they raise a lot of money and then sometimes the execution is lacking. I mean, obviously there's good versions of that too. So there are some analogs

Starting point is 01:01:02 or parallels I feel in different eras of each of these things. As mentioned, I do think there's people like Uma from Sysink or a number of other people, Ilya from near, who have done both or could have done it both. And so I do think there's a lot of overlap and then obviously there's also differences in terms of either some of the philosophical perspectives or political alignment or other things between the two communities. And sometimes they overlap, but sometimes they don't. So, Elad, I want to pick your brain on politics from a different angle here. This is just like Silicon Valley's entrance and tech's entrance into the political landscape in 2024.

Starting point is 01:01:39 We've certainly seen this in crypto. And David and I have been tracking it quite closely. But I, in general, I've never seen tech so political. You know, you have Elon Musk. You have the All In podcast gang. You have Mark Andresen talking about politics. Of course, in crypto, it's taking the form of people like Brian Armstrong, Jesse Powell from Crackin, the Fair Shake, PAC, like political donations, public advocacy for pro-cry candidates, all of these things. Just a couple weeks ago, we had conversations around debanking go viral, and that has kind of a political lens to it as well.

Starting point is 01:02:16 What's your story for why tech got political in 2024? Did this need to happen? Yeah, I mean, I think there's always been some overlap between tech and politics. And so if you even go back, you know, 100 years or 80 years to World War II and all the industrialists being pulled in to basically be able to create large-scale weapons systems and other things for the U.S. military during World War II, I mean, that was effectively a mobilization of what was a prior generation of tech in some sense, right? when tech was more about automotive or more about shipbuilding or more, you know, all these things that were high tech of its day, right? Detroit used to be the high tech capital of the world in some sense during the automotive industry boom. And there was very similar cluster effects and everything else that happened during that era that now are happening in Silicon Valley. So I think one could argue on one level that there's been this engagement in different ways or different forms.

Starting point is 01:03:12 Obviously, I think there's a period where technology, you went very kind of libertarian and very kind of hands off from government and vice versa. And one could argue one of the reasons tech has been so optimistic over time is because it was so lightly regulated and so you could be optimistic. Because if you're heavily regulated, you tend to always have constraints around you and so you become more pessimistic in some sense around the world. And you can actually see the most regulated industries when you talk to people in them are often the most pessimistic in terms of what can actually be done, right?

Starting point is 01:03:45 You have very clear guardrails and not said of those guardrails, you can't do much, right? So I would argue part of tech optimism is due to a lack of regulation. And I think that if you look at prior administrations, there are always ways of tech people participating. So in the Obama administration, a number of different people from Google actually joined the administration, and then some of them cycled back out to companies like Facebook and Google and others. So I think there has been some back and forth. I think maybe the difference is that in this cycle, some of the most prominent people in tech are very actively engaged. And that's Elon Musk and that, to your point of some of the other folks you mentioned, David Sachs, etc.

Starting point is 01:04:22 And these are people who've built very large-scale companies and had enormous success. And now they're building that perspective, bringing that perspective and skill set and everything else to government in a really deep way. And I think part of that is because there may be this moment in time that's being recognized about being able to affect large-scale change and actually change the system. And that would be things like Doge, the Department of Government Efficiency and sort of the changes in some of the Supreme Court rulings that may be. actually enable it to happen, right? When before you didn't have that legal framework to do so, Chevron and other things, right? And then I think part of it is that certain aspects of government became highly weaponized against aspects of tech.

Starting point is 01:04:59 And that's the debanking stuff that you mentioned. That was going after Elon Musk in different ways across different companies. I think there was what, like a dozen or two dozen different lawsuits from different federal agencies against Tesla and SpaceX. I think the most egregious one. was where as a government contractor, they had to hire legal U.S. residents. And so there's one rule in the books that says as a government contractor, you have to do that. And then there's another rule on the book that says you can't discriminate against immigrants.

Starting point is 01:05:31 And so you get sued either way depending on what you're doing, right? So how can you function in that environment? And why did his company specifically get sued? And so if you're being targeted, of course you're going to try and do something, right? What do you think about Doge? So now Elon Musk is going to bite back along with Vivek. Are you optimistic about Doge? Maybe you like Doge in theory waiting to see it play out in practice.

Starting point is 01:05:58 What do you think? I think that if they're able to pull off what they're hoping to, then it is a truly once-in-a-generation opportunity to affect massive change that can cascade through time in a positive way. So if you're actually able to reduce certain types of regulation and certain size of government and all the rest, it could be extremely freeing for the U.S. economy, for progress, for the capabilities of the country, the economy, people's participation in the economy, etc. Technology progress, all that stuff. So I think it could be incredibly positive. Now, obviously, there's certain things that are important to maintain while you do that, right? So for example, I'm very happy that the FDA has prevented the tainting of baby formula in the U.S., which has happened in China, right? I think that's really good. We probably don't want to get rid of that, right? But there's other aspects of things that we probably do want to change and that we want to change significantly.

Starting point is 01:06:55 And part of it, too, is just asking, are we functioning within the proper legal framework relative to the U.S.? In other words, is it constitutional to have all these agency setting rules that they weren't necessarily tasked with doing to begin? in with, right, that go outside of their mandate and that should potentially be legislated by Congress, right? And so there are also interesting legal questions of are we in compliance legally as a government and as a country? And if not, then we should probably fix that either by changing the laws to say, hey, we're now in compliance or getting people into compliance, right? That's the reason we have legal frameworks, is to kind of set the rules by which we do things and govern. I know one thing in kind of watching this, you know, crypto get political,

Starting point is 01:07:36 watching this up close firsthand over the past four years. I think you're very right that, you know, in particular crypto leaned kind of libertarian, right? Leaning's just like, we're doing our store of value thing. We don't need, you know, the nation state to kind of interfere. We're not going to get involved in politics. It turned out that, you know, this happened in kind of in 2022, that somebody actually changed that in crypto. His name was Sam Begman-Fried.

Starting point is 01:08:03 And he started getting very active in politics. only in a negative way, and that kind of blew up in our face. And I personally watch firsthand as people who are, I would say, more libertarian, right-leaning by philosophy, people like Brian Armstrong, in the wake of 2022, in 2023 and 2024, say, okay, we can no longer afford to do this. Politics is actually existential for crypto at this point. We had the Gary Gensler attacks. We had, you know, the FDIC and various debanking types of movement.

Starting point is 01:08:33 So for crypto, it really felt existential. this election cycle in 2024. And that's why there is so much funding, there's so much advocacy, so much work done to actually push pro-crypto politicians. I think that's largely been successful. I know Mark Andresen has taken the point that, you know, everything that was going on in crypto was sort of Act 1 for what the existing government apparatus planned to do around technology.

Starting point is 01:08:59 Act 2 was a whole bunch of handcuffs and aggressive actions against AI to try to, you know, capture that from a regulatory perspective. From your perch, did you kind of see that in AI? Like you were watching what is going on in crypto and like, okay, well, you know, this is starting to affect AI and it could really offset the U.S.'s progress in this incredibly important tech field. Yeah, I mean, I felt like there was three or four things all happening at once, right? So to your point, there is the crypto side of it where, you know, you had kind of what felt like

Starting point is 01:09:33 reasonably random enforcement action that wasn't quite enforcement action, but with these warning letters that, you know, is very like unclear frameworks for how you can actually function and then sort of action against companies that fell either politically motivated or suited random. You also had censorship at the social network side, right, which was clearly coordinated with the government, at least if you look at the Twitter files or other things. Then you had strong antitrust activity that in some cases seem justified

Starting point is 01:10:04 in some cases it was just kind of weird and then to your point on the AI side there was this interest in getting really heavily involved from a regulatory perspective

Starting point is 01:10:15 and I think the sort of doomsday cold stuff helped fuel the ability for politicians to do that but it felt to me like a lot of that activity was just another way to try and take control

Starting point is 01:10:25 the tech industry. It wasn't it was an opening to regulate tech more broadly. It wasn't a, hey, we care about AI or understand it deeply, right? I mean, if you talk to the people behind it, they didn't necessarily, and some people were very bright about what AI is and how it works, but a lot of the regulatory action that was coming felt like,

Starting point is 01:10:43 hey, there's some aid in the administration who kind of wants to get a handhold on tech, and this was a way in, right? And the tech industry was semi-w welcoming it. Now, some of the people who were welcoming it perhaps were doing so because regulatory capture, the ability to have a lot of extra rules, tends to help incumbents and hurt startups. And so this was sort of the big tech versus little tech thing of like these were very, in some cases maybe kind of pro big tech things that would have hurt a lot of innovation

Starting point is 01:11:08 at startups. So I think, you know, fundamentally it's back to like, are we, are there clear regulatory and legal frameworks? Do they make sense? Are we complying with them at multiple levels, company level, government level, etc. and in the cases where we're not or we're acting well outside of our mandate in ways that are potentially quite negative societally or sectorally, like, why are we doing that and how do we stop that misbehavior, right? And there have been absolutely, obviously, bad actors in crypto,

Starting point is 01:11:43 and there will be bad actors in AI. And, you know, and so it's good that there's some regulatory framework for this stuff, but we just did to make sure that it's correct and then it's properly enforced. So do you think all of some of the negative headwinds we felt, from the regulatory apparatus, do you think that just all goes away under the Trump administration? You think we get big changes here? I don't know.

Starting point is 01:12:03 I mean, I'm hopeful that, and to your point, I think, Coinbase and others have done a great job of kind of helping rally the industry around some of these issues, right? And to your point, there was PACs set up and other things to really focus on creating a pro-crypto environment.

Starting point is 01:12:19 And I think this is the first time it's really felt that way, and one could argue that's the reason Bitcoin has really started running. And so I'm hopeful that given the intellect and capabilities of the people from tech getting involved in the cycle, because it really is some of the strongest people, right? Like Elon Musk is probably a once in many hundred years founder, right? Or something on that level, right? I mean, like, if you just look at the span of stuff that, you know, we have some very capable people getting involved in government.

Starting point is 01:12:50 the question is how will government react and, you know, how does that story play out? But my hope is it ends up in a really positive place. What do you think about this newly appointed crypto czar position? So David Sachs was kind of tapped for that rule, obviously from the All In podcast. Yeah, this, and by the way, this is a crypto czar position slash AI, I believe. So it's kind of a combo type rule. Do you think that will have some power in the administration? I don't know the nature of the role.

Starting point is 01:13:20 I think David is very smart, and it's funny that you call him a podcaster, or you're talking about the All-Im Podcasts. You know, missing investor, founder, all these other things. Well, I mean, he was early at PayPal. He was, I think the, what was he, CEO or something at PayPal, way back in the day. And then he started Yammer, which was bought by Microsoft, and he ran a division at Microsoft. And then, you know, he became a prominent investor with Kraft Ventures. And he funded all sorts of early crypto things back in 2017 or so, right? And then, you know, I've overlapped with him on some AI stuff and talked to him from 10 of time on that stuff.

Starting point is 01:13:54 So I think he's very smart. And he has done a lot of stuff operationally. He has done a lot of stuff as an investor. And he's participated in crypto and AI in a way that many folks in the prior administration who had those sorts of responsibilities or roles hadn't. So I view it as quite positive. You know, the proof will be in the pudding or however the British say it. I don't really know Britishism.

Starting point is 01:14:18 So I probably got that wrong. They're probably like, with the tomato sausage. I have no idea what British people say, right? But whatever the Britishism is, you know, obviously we now have to see what things translate into. But I think he's very strong. And I think, you know, I'm excited to see what impact he could have, you know. Eli, this has been great.

Starting point is 01:14:37 We talked to AI. We talked crypto. We talked about sort of the U.S. and the regulatory apparatus. Maybe a final question to kind of close this out. So obviously, we're speaking to you, an audience of. investors, crypto investors slash tech investors who are looking at 2025 and they're always on the lookout for the kind of the frontier opportunities. What do you see is the biggest growth opportunities for tech investors in 2025? Maybe particularly, you know, less the VC kind of startup type of thing

Starting point is 01:15:10 and more kind of the category that a retail investor might have access to. Like what advice would you give them on how to do well in tech investing? I don't know if I have good advice. I think fundamentally the way I've been thinking about the world is what are effectively indices on major tech movements and then what is durable in the face of tech movements? In other words, what is the best way to participate in a basket of stuff related to AI? And to some extent, that's Nvidia, that may be other things. What's the way to participate in crypto?

Starting point is 01:15:45 You know, to some extent, one could argue a coin base where I still, own some stock or some of the other companies in the market may effectively function as indices on top of the crypto industry because if you're trading all the tokens, then fundamentally you're participating in those transaction fees and therefore you've created an effective index, right? And so I tend to think of it in terms of like what are the index companies for the areas that I think are most important? Like I would argue Andrewill now is effectively an index bond on defense tech. Right.

Starting point is 01:16:22 And so sector by sector, Stripe used to be an index on e-commerce, or the new wave of e-commerce companies, because if all the new e-commerce companies were adopting Stripe, you effectively had an index, right? Because they took a piece of every transaction. So I tend to think through that lens in terms of how can you participate in markets broadly, right?

Starting point is 01:16:41 The second question is, what's it durable? So AI comes and eats away at the world, what things will survive and what things will get? augmented, but what things just won't care. And so I really like the companies that just don't care. The extreme example of that would be a railroad. There's only so many railroads. AI, not AI, who cares? This is going to keep shipping freight, you know? And so that's very durable in the face of AI, right? So I kind of tend to think in those ways. You said you didn't have advice, but that's pretty good advice, Elad, toward the end of this. Thank you so much for joining us.

Starting point is 01:17:19 It's been a pleasure to chat with you today. Thanks for inviting me. Appreciate it. Gotta let you know, Bankless Nation. Of course, none of this has been financial advice. Crypto is risky. You could lose what you put in. But we are headed west.

Starting point is 01:17:30 This is the frontier. It's not for everyone. But we're glad you're with us on the bankless journey. Thanks a lot.

Bankless - Investing in AI, Crypto, & Tech in 2025 | Elad Gil

There aren't comments yet for this episode. Click on any sentence in the transcript to leave a comment.