Limitless: An AI Podcast - Delphi Digital: Why the 10,000x Crypto Fund is Pivoting to AI

Episode Date: July 21, 2025

In this episode, we dive in with cofounders and department heads of Delphi Digital, a transformative research, investment, and incubator firm whose crypto roots are expanding out into AI.Co-founders Anil, Yan, and Jose discuss their rapid rise to ten-figure growth, their unique research and incubation model, and insights on investment strategies in AI. They also introduce Delphi Intelligence, a new platform for democratizing AI insights. Don’t miss this quick dive into the future of tech and investment.------💫 LIMITLESS | SUBSCRIBE & FOLLOWhttps://limitless.bankless.com/https://x.com/LimitlessFT------TIMESTAMPS0:00 Intro0:45 Delphi Digital2:53 Transition from Crypto to AI5:49 Incubating AI Companies8:36 Building an Edge11:31 The Value of Research14:51 Founders17:57 AI's Market Dynamics and Future24:28 Finding an Edge in AI31:54 Structuring Opportunities in AI34:47 The Bull Case for ChatGPT Wrappers39:52 The Role of Customization in AI45:29 AI's Evolving User Experience52:38 Emerging Contrarian Trends in AI1:08:11 Introduction to Delphi Intelligence1:10:24 Conclusion and Future Insights------RESOURCESDelphi Digital:https://x.com/Delphi_DigitalYan Liberman:https://x.com/YanLibermanAnil Lulla:https://x.com/anildelphiJose Maria Macedo:https://x.com/ZeMariaMacedo------Not financial or tax advice. See our investment disclosures here:https://www.bankless.com/disclosures⁠

Okay, we have an incredibly special episode for you guys today. You're about to hear from one of the most well-researched strategic investors in frontier technologies. If you don't believe me, these guys raised $1 million, just one, back in early 2019 for their first fund, and turned that into over 10 figures in value in a year and a half. You want to know what's even crazier? The original $1 million that they raised was on credit card debt and loans. So we know that these guys go all in when they have conviction in something. What I'm really excited about is over the last two years,

they've been dialing in to all the stuff going on in AI, and I'm really excited to get into their heads about what trends they think are exciting and what they're excited about investing in. Anil, Jan, Jose, it's great to have you guys on. How are you guys doing today? Yeah, thanks for having us. Amazing intro. Yeah, I appreciate that.

Great to be here. Let's go. Okay, so from a lot of our listeners that tune into the show, they've probably never heard of you guys. And so maybe you guys can spend a few minutes painting a picture of who you are. And no, maybe you can kick herself. Yeah, for sure.

So, yeah, we're all co-founders of Delphi. The Delphi that everyone knows of today has like three main companies, right? Delphi research, Delphi Ventures, Delphi Labs. Jan heads up to Ventures. He's managing partner there. Jose heads up labs. And I mostly focus on research and ventures.

Essentially what Delphi does is we're very research focused, right? We started back in 2018 with research like embedded in our DNA. Jan and I and a couple of our other co-founders all met at her first job out of college at Bloomberg. We did, you know, a lot of like Tradify there with research and did leverage finance at Deutsche Bank. Essentially fell down the crypto rabbit hole when we realized that, you know, maybe like the future that Tradify promise wasn't all that great. And yeah, just kind of like fell in love with kind of the promise that crypto provided. We put our jobs in 2018, started to Delphi as a research firm.

And like, you know, the first few years, basically no one really paid us for research because like, There weren't that many fundamental investors in the space. You know, shout out like Multicoron and Hash. They were kind of like our first two, you know, real thing. But where we really bootshot was actually helping design and kind of consult and advice a lot of these protocols. A lot of what you saw at Defi and work with like protocols like Ave, Lido, very early on. And then especially with gaming as well with, you know, projects like Axi and Yale Guild.

Over the years, Delphi has kind of morphed into this like, you know, you know, three kind of like pronged layer where we build, we research, and we invest, right? And I think these three different perspectives work really well together because it lets us have, you know, as many like different hands on elephant as possible so we can really feel what crypto is and where it's going and, you know, have a really good pulse of it. It sounds like Delphi was extremely focused on the Web3 crypto world, right? And that's been your bread and butter since you guys have been incepted. And then over the last two years, you've been like dialing in very much on AI.

I'm curious like, like, what kind of like parallels run between the two technologies? Like, do you just see the AI stuff happening and thought like, huh, I just want to kind of like peek over the fence? And then did you get kind of like more involved in that? Like what made you more interested? Yeah, definitely. I'd say that like, you know, when we first fell down the crypto rabbit hole, it was almost, it wasn't even just obvious to us.

It was just like, you know, what else could we work on or spend our time doing other than this, right? It felt like nothing else mattered. And I think over the past two years, you know, shout out Tom, one of our other venture partners and co-founders. He really, you know, was early to, you know, the AI trend and everything like that. And we just, you know, a lot of people within the hive mind of Delphi got nerds snipped by AI and it felt we had that same feeling where it was like, you know, how could we not, you know, be infatuated and obsessed with this? And yeah, I think there are a bunch of parallels. I mean, obviously the speed of innovation, you know, just like when we entered crypto,

Just like so, you know, today, even with the team of almost 100, right, we have around 88 people across three companies of Delphi. There's just no way we can kind of, you know, keep up with every single thing happening in crypto. I think that's the same thing that we see in AI and why we think, you know, we'll get into this later, why we think it's really important to have a team focused on it and kind of like separate the signal from the noise. Yeah. No, in terms of parallels, I think just when you see something that looks so glaringly obvious, in terms of, you know, it's growth in application, but at the same time, there isn't really any widespread adoption of it or it's still, you know, orders of magnitude away from what it'll eventually be. Your eyes tend to light up because you start to think about all the possibilities

on the growth building and investing side. And so I think what we saw that with crypto, you tend to see here with AI. And then there definitely overlaps the two in terms of implementation and where they can be synergistic, but I think, you know, holistically you tend to, I think that positioning is really what gets you excited at first because, and, you know, for, and all the reasons you're bullish on it are, you know, fundamental reasons. And then you kind of put that against a backdrop of the fact that it's still barely permeated and there's still very minimal adoption of it is. And I think that position is what really excited us about it in the first place. Well, I think what's something that's really interesting is your focus on kind of crypto and web three for the initial fund,

Crypto is an incredibly fast-changing technology, right? And the whole point around it is it's meant to rebuild a ton of different sectors. Finance, media, you know, you name it. AI is exactly that as well. So I'm not, I can't say I'm exactly surprised that you guys are marrying both technologies together. You're doing a ton of stuff in this space. So you just mentioned a few arms, Anil. You're doing the research side, the investing side, and also the incubating side.

I saw that you guys are incubating a bunch of AI companies. Maybe you guys can speak more to that. Jose, maybe I can pass this to you. Yeah, very similar to these guys. I had my crypto-pilled moment in 2017 with Ethereum and pretty much had the same experience. Last year, actually, a bit later than I think some of the other people at Delphi when I read situational awareness.

Like, I'd been playing with Mid Journey, obviously, and chat chippedi. It was just so busy and kind of deep into crypto that I think I didn't realize just how momentous this thing was. And then, yeah, last year, once I read situational awareness, it really clicked into place. And we pretty soon decided with labs that we had to start doing some stuff in AI. So we put together our thesis on like crypto AI. I spent a lot of time on that, just figuring out where the good places for overlap was. And then ended up partnering up with NIR.

Ilya is obviously an OG and AI. He's one of the original authors of Transformers paper. And yeah, we partnered with them to run our first accelerator in AI, which was really, really great. It had some insanely strong founders that applied just through Ilya's network. And then did a second one, finished about two months ago with the Cyberfund guys, which was also awesome. Yeah, we always think that the best way to, I mean, like Anil said, the old elephant groping metaphor that Anil likes, we like having a lot of hands on the elephant. And I think researching is awesome. and we're all kind of researches in, is at our core.

But building, you get a really unique perspective. And that happened for us in crypto too. Like there was things we learned by building protocols and being really deeply involved that we really couldn't have learned any other way. And it's been the same exact thing with AI. So it's just been really interesting. And also it's similar to crypto.

It's an entirely new paradigm. Crypto building is like very different from Web 2 building. Like you have these smart contracts that are immutable. Well, they used to be anyway. Nowadays, protocols take a slightly different approach, but still a lot of them are immutable. And so it ends up being more like hardware. Like you have to be really care. You have to spend a lot of time researching and then writing like it's less of an iterative

approach and more of a, you know, once this is out there, it's out there for anyone to exploit. And AI is like a different paradigm still where these things unlike like most of software before it aren't deterministic. They're probabilistic. And so it's really hard to ensure like a uniform. user experience and like they're not even standards for like unit test or anything like that. I really think the metaphor of it being a new kind of computer is great. So it's just been really useful diving in and learning like with our hands in the,

yeah, with just just getting stuck in and building. So there's a lot going on in AI right now. New frontier models are being released like every week at this point. Billions of dollars are being spent to train these things. There are numerous like consumer applications that are out there. And I can't help but think that this is like an incredibly expensive game to play. So I'm kind of curious, you know, what's your unique edge when it comes to investing in AI? You know, how do you view the market right now?

And where do you think you guys can make like the biggest impact with what you're doing? You obviously have like the whole Web 3 crypto background. And maybe it's something to do along with those kind of principles of investing that you had with that fund. But I'm curious whether there's anything new you guys are seeing in the market right now. I don't think we have an edge right now. I think we're sort of hoping to build our edge over time. Like we've definitely made a lot of investments in crypto AI. I think we have edge there.

We've made a couple of investments in AI. But I think we all sort of recognize that we're sort of paying tuition right now and getting to know the industry, getting to know as many founders as possible and kind of building our edge over time. That's the goal of like intelligence really. Same as when we started in crypto, you know, these guys didn't want to start a fund straight away, wanted to kind of build your edge, build your knowledge.

and then go for that. I think it's similar here, except now we have some capital behind us, so it makes sense to invest and start building that. So yeah, the hope is that we build the brand with Delphi Intelligence, get some really tough researchers on. And then we're also doing a couple of other things,

like we've been investing in young fund managers in AI, sort of looking to, like, when we started DelPie Ventures seven years ago, it was really hard to raise. And we know firsthand, both how hard it is to be a first-time fund manager raising and also how much edge you can have as a first-time manager.

And so we're kind of looking to find those people that were in the same position. We were in seven years ago in AI and back them and then kind of benefit from that deal flow and that learning. And I think the way we're thinking about it internally is we would like to aim to have edge and to really start accelerating our investment pace 12 to 24 months from now, something like that. So yeah, this whole thing is sort of us aiming to build that edge. I think like even when we first got started and we were writing reports, you know,

if we put out a report on, say, synthetics or something like that, people would always message us afterwards and say, damn, you guys really knew synthetics really well. That's why the report came out so great or anything like that. It's quite the opposite, right? Like we learn about, you know, whatever we're researching when we're putting together this report that we know is going to get like, you know, picked apart on places like crypto Twitter or by the team or by competitors, right?

So that's why we really do love having research embedded into our DNA, because it almost provides like this check and kind of this like, you know, high bar that anything we publish we know is going to be looked at by, you know, people either building in the space or other investors in this space, et cetera. So we want to make sure that the research is not just really good for us to use and build conviction, but also, you know, meets this bar where it won't get ripped apart. And that kind of like fear or intimidation is, I think, is like really powerful. Yeah. If I had to pick an edge, just to, you know, give you some answer to that. that question. I'd say it comes from a few areas, one just from, you know, investing for

however many years we've been doing it, right? And granted, that's an edge that's kind of consistent across anyone who's been doing it. So it's not necessarily a big one. I think we do have a decent variety of backgrounds and ways of thinking as well. And that's been an edge for us in crypto and should continue to be one here. And I think, you know, just being able to operate as a group is a big edge where we're able to take a variety of learnings that each of us are doing, bring them to the table, and get kind of immediate feedback and have just a variety of points of view. I think that's probably one of the bigger ones. And then patience, I think is another one that we've kind of learned over time in crypto in particular. And so here we realize we don't

really have an edge and we're trying to understand is where the best opportunity is, right? Is it early stage or does early stage really take too long to get a proper payback? Does it make sense to invest in some of these growth staged higher valuation, but lower risk type plays where you have a pretty kind of cemented path to becoming a large company? And so that's still something we're exploring. I don't think we really have an answer there yet. But I think it's just the patience and I think what's helped with crypto is that you go through so many cycles so quickly, and I think you can draw parallels to kind of other online experiences versus physical ones. So if you think about like, you know, online poker guys have seen an insane amount of hands,

right? And so they have a lot more experience than someone who plays live despite, you know, having a long-term career. So I think, you know, there is some benefit in terms of taking that from crypto and understanding those cycles and trying to draw parallels. parallels there. Yeah, I think we all agree, I definitely agree with Yon. I think being a venture investor is like a skill that's sort of generalizable across sectors, like a lot of it. Deeding founders understand it, but you kind of need to understand the sector to be able to properly do

diligence to founder and not get bamboozled by a high, by a charismatic, you know, sort of charlatan, I guess. And so, I think what we all agree with is that we all agree that this is going to be, I think the biggest bubble that, like, humanity has ever seen. I think just like all the ingredients are there. Isn't it like already a bubble? This was being said like last year and it's just been up only. I think what, Embedia crossed like $4 trillion in market cap this week. I feel like like how big do you think this bubble is going to go?

Because I agree with you like charismatic founders are super important. But I see a bunch of these like VC investors talk about like feces for decades, right? the next 30 years is going to look like this. AGI, we're going to achieve it in, whatever, 2027, or, you know, they're arguing about that. Like, how important is the founder when it comes to all of these kinds of things? I'm guessing quite a lot. Yeah. To me that, so we have different, I think, focuses even as investors.

To me, the founder is the most important thing, like, especially at the stage that we invest in, which is normally seed or pre-seed. Like, the idea is going to change a lot. And you're really betting on a founder that, and you want someone that, and you want someone that is just exceptional. It has a history. And exceptional people leave breadcrumbs. You can sort of put their path and see, you know,

you can be able to see some evidence of exceptional behavior before. And ideally you're looking for the things that are like he was insane at a video game or, you know, something in their youth, some like sporting thing. Those things are generally better because they're not as priced in as someone having, you know, done a successful startup and exited or whatever. And you're really looking for these kind of freaks, basically, that are, that her, like, insanely motivated, that are able to, like, go through walls to get, to get to achieve what they want.

And so that pattern of, like, we've seen a few with ventures over the years, and those have been our big winners. And you were just looking for more kind of an AI. And then on the bubble comment, I don't think so. I mean, I think when you, when you look at where, like, I look at 2000 as my mainly, like, maybe the biggest comp, like the price of earnings ratios of the, of the mag seven equivalent, we're still like, you know, two to three X what they are now. And then I think in the private markets, there's definitely a few bubbly things,

but there's also like insane growth and fundamentals, you know. Like CatchaPT is the fastest company ever to 100 billion in revenue, to a billion in revenue to $10 billion in revenue. Cursar, I think was the fastest actually company ever to half a billion in revenue. And you're seeing multiples of these, right? With DAU's like actual revenue, I do think there's some bubbly, behavior and some stuff that's that's kind of reminiscent of 2000 with these valuations. But I do think there's just a long way to go just because, like, first of all, you have the

most profitable companies in the history of the world that are stuck in like this game theoretic arms race where they're incentivized to spend every single dollar free cash flow into training better AI models because otherwise they might like miss AGI and have their company destroyed. And that's like a dynamic that's just going to be a constant tailwind to making these models better and every startup in the ecosystem benefits from from from better models so that there's that um and then i think there's just the fact that this stuff like the internet was kind of like um people got really excited in 2000 but there was all this all this infrastructure that still needed to be built for the killer apps that people imagined in 2000 to work right you needed you needed people

to have mobile phones to build uber you needed payment rails you needed um like GPS working you needed all these different enabling technologies. And with AI, it really feels like you don't. Like, everyone has a smartphone. Everyone has a, has a computer, fast internet. Like, there's nothing in the way of this thing just scaling. Like, it's really limited just by the quality of applications for people to use. And there's so much talent going into AI, there's so much compute going in, there's so much expending happening that I just think it's going to stay extremely, it's going to keep moving extremely fast. Yeah, so I don't think this is the bubble yet. Yeah, and on the bubble point, I think, you know, you can kind of think of it in multiple phases, right? So right now you have this

kind of scenario where the markets are really giving credit for just cap-back. So margins are coming down on some of these bigger players, and it doesn't matter because they need to spend and spend and spend and just get to this point where the like the next kind of wave is, you know, is proving out that the spend is actually valuable. And I think you're starting to see elements of that. But the market is kind of very forgiving right now. And so, you know, for the first time in a while, you have this technology that can improve efficiency by an order of magnitude.

And it just gets captured in so many ways, right? You'll have the big guys who leverage their distribution to just improve margins because they need to reduce headcount or just become more efficient. On the startup side, you have these smaller teams that can get to unicorn status without really needing these longer-term cash raises. And so I think the fact that it's kind of happening across multiple areas is what will give it legs for quite some time. But yeah, in the interim, you have basically this massive spend phase, and that doesn't

seem like it's going to be slowing down anytime soon once we're starting to see that there are actual improvements to be made to the base models, right? There was that concern up front where, okay, it was actually kind of solved. And then when there were these big breakthroughs, then everyone, you know, the CAPEX got turned back on again. And so it doesn't seem like that's really going to slow down anytime soon. But at the same time, you're having real efficiency gains at the early stage. And so, yeah, I think the trickiest part is probably the very late stage investing side in the world where they don't necessarily need to bring on that capital.

Yeah, the one thing I'd add here to is like, bubble has this very like negative connotation to it, right? I think like one reason we're really excited is because we actually do exactly what Jan said. We think they're going to be insane efficiency gains. We think there's going to be, you know, this huge period of abundance, right? Obviously with, you know, this new innovation. And I think like, you know, one thing that we think about and we were talking about just this past week at our founders' retreat is like, you know, there's this, like the turn rate of Forbes 500, the Fortune 500 company every decade has just been going up and up and up, right? So even if you use a turn rate from like the last decade, I think, you know, probably half of the companies would be kind of turned in the next like 10 years, right? We actually think, or, you know, this is at least my stance.

Like, I think it's going to, you know, that turn rate is going to increase exponentially because of AI. And I think, you know, you may even see 350 to 400 of the five, you know, top 500 companies get turned out in the next decade, which what does that mean? That just means there's immense value creation happening in other areas of the market. and capturing even a little bit of that upside, I think it's just going to be the craziest thing that you could have ever hoped for as an investor, right? So, yeah, I think we are excited for some of these big companies that already do exist, obviously, like the Max 7 thing,

to, you know, they're obviously fighting very hard to hold on in their spots, and there will be a lot of efficiency gains there, but I think more, more, you know, excitingly, and obviously going to be much harder to figure out are the companies that will, you know, go from zero to some of these top five-order companies, right? in areas all across, you know, all across the map.

So, yeah, honestly, we're just super excited. But yeah, I think it's going to be challenging, but that's why we're kind of, you know, pumped. Yeah. So one of these words that I keep hearing all three of you mentioned is the word edge. And it's like looking to find the edge. And what I want to ask, because I think this is what I'm personally interested in, and a lot of people who are listening, is what the process looks like in finding an edge

and what type of topics you guys are interested in pursuing where you can find that. because a lot of the times our episodes were interested in just exploring different frontiers, but there's a lot of different pillars in the world of AI. There's so many different industries and categories. Is there a particular spot you're excited about? And within that spot, how do you go about finding an edge and getting an advantage? It's honestly like a lot of trial and error and being very honest with yourself about where you sit. I think that's something crypto really gives you, like to survive and thrive in crypto, you need to be very honest about whether you have edge or not and

where you have edge. And in AI, I think for us, it's just been a pros. of, I think first of all, we started looking at, obviously we did crypto AI where we thought, you know, there's an overlap here with crypto. We have an existing brand. This sector is exciting. Here, I think it's pretty clear that we can have edge. Like, we're very early to it. And then we started trying to do more AI direct investments. And I think that's where we saw the bigger challenge. We were like, some of the stuff was hard for us to get our head around. But also, it was unclear to us, like whether we had edge. And that's always like a bad sign. Like you should, you should kind of know, I guess we know the feeling of having edge to some extent. And I think it's a mixture of,

there's like some reason, something that other people aren't seeing here, which I definitely think we're like more bullish on AI than the average person, but probably not than the average VC, right? So then we thought, okay, I think this direct investment, there's some negative selection happening here, like the deals that we're seeing are potentially not the, not the best ones. And so we started to look at, I mean, first of all, we started to look at fund managers, which I think was an interesting one where we saw, okay, there's these fund managers raising small funds, first-time fund managers. They're really struggling to because no one wants to back a first-time fund manager generally.

And the fund of funds are very risk-averse. And so, and we started seeing, wow, there's some guys here who are super plugged in, insanely well-networked and hungry, and really remind us of kind of ourselves seven years ago in AI. and this could be a way that we can have some edge. Like, not only will these guys perform, but also the deal flow that we get through them is going to be pre-vetted and give us some access that kind of overcomes that negative selection problem.

So we've been kind of digging into that now, and we think that there's edge there for us. We're also looking at China. Like, we've been looking at China for a while. Actually, both our members of the investment team spent a lot of their intelligence team, spend a lot of their time in China. I believe China is producing like over half of AI engineers.

And also the it's much, the rounds are much cheaper there because there's just less capital. Like the US investors aren't, aren't really able to invest in China, like institutionals. And there's obviously concerns like geopolitical concerns and stuff like that. So you've kind of been looking there and figuring out whether, whether there's a way for us to have edge there and to add some value in helping kind of these founders go global. So I think for me, I'm curious what the other guys think actually. And then we're also looking at kind of these secondaries of the big names, the anthropics, the Grox, the open AIs, and kind of figuring out, you know, whether we have edge there.

Because I think there we're more just trying to capture the beta versus have a lot of edge. But yeah, for me, it's a trial, it's a trial and error process of like thinking through things, going in, doing some research and then figuring out being very honest with ourselves if we think we have edge or not. Yeah. No, I think the honesty is the important one. Edge comes in many forms, right? It's, it's selection edge, it's timing edge, it's some informational edge, and then kind of there's some that comes with experience in terms of bet sizing and everything else. And so for us, what we're in the process of doing now is basically trying to understand where we can have an edge. And I think even that on its own is very valuable, or it could even be considered Now we're like using this in a very nebulous way.

So you know, timing wise, it's it's on the early side for sure, right? So I think that's certainly one. Having the luxury to commit a lot of time to look at this without necessarily needing to generate a return immediately, I think is a huge benefit, right? Where to some extent other managers as part of their job, they're forced to deploy. right and so that I think comes with a disadvantage where you might be deploying in areas you don't necessarily want to so I think the patient itself is a huge benefit and should give us the opportunity to find those unique plays I think one of the biggest things and this is another

learning in crypto is so much of it comes down to bet sizing right and it's like it's really knowing what the opportunity is and and you know whether you're allocating one five 10 50 percent to a position is really what makes or breaks a lot of these, or what really drives, I think, the outperformance. How do you personally figure that out, though, Jan? I know you say that, and that's what all the investors say, but I want to get inside your head. Like, what, what's the difference between you being like, you know what, I'm going to give you around one to five million dollars, and then you're going, you know what, I'm going to pump in $20 million into your thing, which is not something you guys are unknown to, right? So what is that difference? Yon is a great. Yon is

a great person ask this question, to be honest. The big one is just risk. And so it's understanding, you know, how can this go wrong and realistically, what is my downside? And then, and then I think sometimes, you know, when things are going well, it's also knowing when to, like, on paper, you should be taking position down. But I think that's, there's an edge in understanding the position outside of, of, it relative to the rest of your portfolio, right? And saying, sure, by the book, I should probably

be downsizing, but it's more about how is this position relative to the rest of the market? Is everyone else under exposed? You know, will there be a lot of money coming in? And so I think that ends up really, it's like, it's understanding that your winners are winners and they should remain that way. And so you're either doubling down or leaving them as is. And so it's not often that you get really convicted and it's kind of in those scenarios where a lot of those edges line up, right? I happen to be down this rabbit hole

and I found this. Yeah. It's going to be a lot harder to get access to this in the future. I think it's derrisked more than people actually think. And so it's when the stars align in those scenarios that you really need to just

kind of have... Talking about optronic here? That's one of them. Yeah. And where you just have a lot of, and I think, you know, the risk tolerance is a big one too, where thankfully from crypto, you kind of get numb to the volatility. And I think that ends up being a huge edge as well, where you're just able to tolerate swings where if it goes wrong, it goes wrong, but ultimately, you know, more often than night, it will go right. And you really

want to be able to capitalize on those opportunities. Yeah, I think that the, Jan's really good at this. It's probably one of his biggest strengths. And we definitely have a lot of experience just from in Fund 1, we started with one position in the fund just by virtue of our size. And the rest of the cycle was us just selling that position to buy others. And so we just, you really, from that, like, understand deeply, like, the importance of bet sizing. And you also naturally have this, like, hurdle rate, right? Like, is this thing going to outperform door chain, which was our position at the time? But I think the sizing, that's, yeah, one of the biggest things is also one of the biggest things I look for in fund managers, like people who are going to be

concentrated and not afraid to take big swings. And it's also one of the biggest mistakes early fund managers make. They want to kind of, and like concentration just drives all the right behaviors. Like it forces you to think about whether this founder is going to be able to return the fund for you, whether this is someone you want to spend a lot of time with. It forces you to actually add value to the founder. It forces you away from like indexing and just following in to around because Sequoia is in or whatever. Um, so I, And then the other thing is just like conviction is it's like a feeling, right, that you build through research and speaking to someone and thinking about it. But when you have it, it's really important to recognize it because conviction, at least for me, it's not like you can have sort of 10x more conviction in something than you have on anything else.

And a lot of people will feel that and size them equally anyway, right? Or like, I have to have 10 positions or whatever. But actually, if you're conviction, if you have 10x more conviction in something, something else, you should size it appropriately because those things don't come along that often. You know, there's only probably three to five, if you're lucky, spots a year where you really find that kind of conviction where the stars line up. And when you find it, it's really important to size things correctly. And it's kind of the biggest difference, I think, in performance for people.

That's why we wanted to build this research team, build this conviction, right? is like we think we feel confident in our ability to see these opportunities. But if you don't have the conviction, you may not take the swing at the right size, right? And I think that's going to be really important for us. And then, you know, going back to Josh's question about, you know, obviously we've been using the word edge a lot. I'll say that like, you know, EJA started it. So the question was around it. So that's not totally our fault.

But the only thing I'd add to what these guys said is like for me, I think, you know, one of the biggest edges that we found it with Delphi is just different perspectives, right? And I think that's what we're going to seek out, you know, with Delphi intelligence as well. And I think like, you know, that's not even just within our team, which we really do like building those perspectives and insights within the team. But I think like more so just within our trusted network, right? You know, within crypto, we lean on our network all the time. And that really helps scale the amount and, you know, the speed at which we learn.

that's definitely going to be something we lean on, you know, within other areas that we're trying to explore and learn about. Yeah, so as you guys move into the world of AI, I'm curious of Delphi as a company, if you individually, you have a framework or a structure of how you think about these opportunities. Because AI is divided into a lot of big categories. I mean, on the show, we like to talk about it as a layer cake almost, where you have the chips layer, then you have foundation models, then you have dev tools and infrastructure, and then the tops the application layer. And there's all these different worlds that you could explore, I guess, to get that. edge. And I'm curious if any of you or if there's a company-wide kind of tooling or a way that you

explore these opportunities and find order in the chaos when you're evaluating everything. I definitely think we have different people have different perspectives on this. We've looked at things across the layer cake. I think personally I'm most interested in the top and the bottom. I just think that's like those are the places that tend to be the most defensible. So we've looked at a couple of, we haven't actually pulled the trigger on any, although I actually made a mistake on one of them, but we've looked at a bunch of chip startups and people doing new architectures and stuff, which have been really interesting. And then for me, I'm really bullish on the application layer. Like, I think, I think chat GPT rappers get, people use it as a sort of, you know, to throw shade. But I think chat chippity rappers are going to be insanely valuable. And you're kind of already seeing it with cursor, you know, and others like it. And to me, AI, the capabilities that it has already, it could do probably like 100x more than what people are using it for right now.

And that gap to me is the product opportunity of creating like verticalized applications with really clean products, with really smooth like context engineering and to solve like particular pain points. And I think you're going to have those across every single vertical and they're going to be, yeah, really big opportunities.

So that's one I'm really excited about. But yeah, we look at stuff all across the stack, I think, just at this point, just to kind of build knowledge. I mean, actually, in the crypto AI area, we did look at a lot of data stuff too. We kind of had an intuition that that would be somewhere that crypto would have a particular advantage. Like being able to, it's always been kind of a crypto thesis, right? And initially it was this idea of Web3 Social where everyone would own their own data and you get paid for it. But I think the idea of coordinating a bunch of humans to provide valuable data to train AI's always was like an obvious or seemed like an obvious crypto AI idea. So we did make a lot of bets there too.

I think we're a little bit more cautious now just given where things are going with synthetic data and just RL and like we're being a bit more cautious there. But yeah, those are two that kind of came to mind. So you mentioned that you're bullish chat GPT rappers. Can you just give us the bull case for them? Because I like you, have seen so many people shit on them, basically. Yeah. Why are you so bullish? The sort of preconditioned for me being bullish on a chat GPT wrapper is the founders,

or the app gets better as the models improve, right? So it actually becomes more useful as the models get better. And there's a lot of examples where that's the case. There was a lot of examples initially where you're just building some scaffolding on chat GPT to do code. or therapy or something. And that's not interesting. Like all that stuff will get picked off by the models. What is interesting is just like verticalized applications

which improve as the models get better. And like some of them, like I think even the more interesting ones or the most interesting ones are the ones which actually don't work right now. They're actually just betting on the models improving enough that one day they'll work well. And there was a bunch of examples of that initially. But I think there's there's some interesting ones now too. But to me the bull case is just,

Yeah, kind of what I said what I said before that to get the most out of models is actually hard work. You need pretty good system prompts for whatever vertical you're using it for, right? Like if you're using a model for therapy, it needs to not be so

agreeable. It needs to actually tell you hard truths and stuff like this. Whereas if you're using the model to write you I don't know, a Twitter or show post or something, then maybe you want it to be persuasive and stuff like this. If you're using a model for investment due diligence, you need it to have access all your investment notes.

You needed to know what the thesis is behind your firm. So there's all this, like, people call it prompt engineering. I like context engineering, which is a combination of metaproprone and context. And that stuff is actually really hard. Like it's hard to get the most out of a model. And there's going to be applications that, like, optimize that process for a specific vertical and just give users, like, really refined experiences for it. Curse is a great example, I think.

but also Anthropic just released Claude Code recently, right? And so I'm curious about your thought around how much of the application layer you think the model makers can actually kind of take, right? So I'll give you another example. XAI just launched GROC4, and they have this huge distribution network, right, which is X. And granted, Elon is a very unique case

because he's just buying everything. He's probably going to be implement. the chip sector at some point as well. He's putting chips into our brain, blah, blah, blah. And he's building up a massive competitor in terms of like data centers. What edge do you think application builders

that either you're investing in right now or that you're looking for right now have over what model producers can just kind of like replicate themselves? Is it in the context engineering that you're talking about, Jose? Like, is it the fact that these founders can basically and intuitively describe how an app should behave. Because a lot of this is just around social behavior, right?

Like the thing that makes an app successful is if you go on it and a bunch of people like it and really vibe with it, right? That's it. Like OpenAI just launched their agent yesterday. And the number one bit of feedback I've seen was, this is cool, but like, what am I going to use it for? And if you have your potential target market saying, what am I going to use it for? you haven't nailed the application there.

So I'm wondering whether like there is like, you know, maybe just a list of items that you think separates kind of like founders that are building applications in AI versus like model producers that are just going to like steal their stuff eventually. I think it's a great question. It's kind of the golden question if you're investing in AI applications. Like is this something that the models can do? I think coding is an interesting one where like if,

I think if Clod turns out to be the best, best coding model for everything, it's going to be hard for for Cursor to, to win, right? Right. If it's just literally a clod wrapper, although there's still like cool stuff that the cursor's built, like the rules, you know, which are, which I think is a really interesting primitive. I don't know if you guys have used Cursor much, but it's, it's like, it's a very interesting, like U.X framework that they've built and there's other stuff like that. And I think there's definitely advantages to being the only, to being like, to being like,

are focused on just pretty much user experience and not having to build your own models. It's hard to answer in the abstract and in the general. I think you have to go kind of like application by application. Yeah, user experience is a big one. In the sense, I think one parallel is looking at Gemini, right,

and how underutilized it is because it's just the UX is tough, right? And so it's kind of clunky. It doesn't really, it's not as widely used as you'd expect it to be considering how many people are using Gmail and all of that. And so I do think, you know, the UX is a big component. And so it depends on how much of the value is just in the raw processing ability of the model

versus how much of the value in the product is in building out everything else around it and making the experience fluid. There's a lot, for instance, Harvey is an interesting one where they've just built a lot of scaffolding, as I understand it, a lot of scaffolding to make the document creation for lawyers extremely fast and seamless. So Harvey, AI, just for context of the listeners, is like a chat GBT for lawyers. Is that right, Josie? Yeah, basically for creating memos and stuff like this. And you want to be able to have your firm's standard boilerplate stuff and like whatever the style is that your firm writes in, the key documents. And you want to go document by document because this is like very

high, fake stuff that you don't want to get wrong. And I think that's going to be the case for almost every vertical who's going to have this. Because reliability is also a huge thing. Kind of talked about that before. But these models are not, they're getting more and more, but they still have hallucinations. They're not super consistent. That's another thing that the kind of verticalized applications can help fix with really good scaffolding and system prompts and stuff. But yeah, I think Harvey and Kursor, probably the two biggest examples of ones so far that I think have built cool stuff on top of like a basic wrapper. Nice. Yeah, I do also think customization is going to be a big key. And I wanted to jump in

after y'all, too, because I think like, this is something I go back and forth on a lot is a lot of these model creators obviously have a lot of data on, you know, who is paying for compute, how much they're paying. And, you know, can very quickly figure out why, you know, if this person is paying, they're obviously building something that is valuable. Let's go copy and paste that. And yeah, to each as his point, obviously a lot of, you know, these guys are all going towards its agent space towards like creating something that is scalable to, you know, the masses. I think, you know, the last decade was very much about, you know, there's an app for that. And I think this upcoming decade will be very much like there's an app for you, right?

So very like custom app. Maybe Jose, like, I don't know if you want to leak or share some of the conversations we were having this week about like something labs is building for Delphi itself. I don't know if you want to go into that. But like, I think that's a great example of like something that, you know, yes, we know a lot of these model creators will have something that will probably accomplish 70 to 80 percent, if not maybe even more, you know, in the future for us. But it's something that, you know, I think labs wanted to roll up their sleeves, get their hands dirty and build something custom fit for us. That would be, you know, fulfill basically 100% of our needs. There's like Delphi we operate. We like to call it like the

hive mind, right? It's also the name of our of our, of our pod. And it's, it really operates that way, where there's a bunch of people in different divisions, some doing research, some building stuff, some investing that are having a bunch of interesting calls. And right now, it's, it's, it's, the sort of bandwidth between surfacing the interesting conversations for the whole firm to benefit from is really slow. Like, we have to schedule these like biweekly calls and then by the time that's happened, people have forgotten about it. And so I think the initial sort of vision, is for it to be sort of an organizational knowledge base, or like we call it DelphiOS or High Mind OS,

which can just, first of all, like, have all the conversations that people are having across the firms in a retrievable and, like, queryable format. And then building, like, intelligence on top of that, so this thing can, for instance, generate IC memos really easily. Like, I have a bunch of calls of the project, and then it has RIC memo format. Maybe I can put in podcasts that the founder's done.

And then I can answer some questions to the AI and then it can just generate and I see memo format, you know, something that takes me kind of hours to do. You might have the same with research. Or for instance, if we want to have a kind of CRM of all the companies that we've ever spoken to, we can see all the conversations people have had with people at this company and also all the conversations people have had about this company, right? We can sort of search this and see, oh, this founder actually leaked to a mouthful like these guys are not performing well. they ended up using a different service provider or whatever. And we want to have, and I think every company will basically have this in the future. Like, all the knowledge of the company will feed into this central memory, knowledge base, whatever you want to call it.

And then there will be various kinds of agents you can run on it that both help the company operate better and just automate and augment its people to be able to do more. You know, you could kind of see this getting kind of crazier as time goes on, right? Like recently we just had this big founders retreat and we always like to like kind of like share a book that we all read and stuff like that. And this book for this last week was essentialism by Greg McCohen, right? And you could see us using all this, you know, all this data that this knowledge base like fills. And then in our chat, add an agent that is based off Greg McCohen who like kind of follows our, as follows our calls and then kind of shits on us whenever we're drifting away from, you know, what the thesis of his book is. So it's not like us holding each other accountable, but this agent almost holding us accountable.

accountable for the decisions we're making at an org level. So yeah, I think we're super excited to play around with it. And I think it will be super useful for other companies. And at the same time, to answer your question, do I think this is something that like the big models like OpenAI, Anthropic, et cetera, you know, crop are not going to build in? No, of course. They're, they're obviously building it right now, as we've seen with all these recent announcements. I think the customization is something that's really special. I think will be like, you know, again, what I said earlier, an app for everyone rather than here's an app for, you know, No, you.

Yeah, this is a super interesting point, right? It's because you were able to build DelphiOS using AI, and that would previously have been something that you'd have to go to a larger company or use a lot of resources in-house to develop. It's become much easier. And then you mentioned that, well, Grock is probably going to integrate this. ChatGAPT will probably see these types of tools in. I'm curious where you see the most forming,

because a lot of the new innovations tend to become commoditized fairly quickly. And I think one of the most that we've seen perform the best, at least in the consumer world, which is what a lot of the people who are listening are involved in, is chatypte's memory function. And memory is amazing because it includes all the context of previous conversations you've had, and it really locks people into that platform.

But outside of memory, I haven't really seen many other moats that make me want to use a model. So I'm curious what your takes on moats are, if they're possible to capture a large amount of a user base, or is it just going to be commoditized software all the way up? All the models get better. They all kind of copy everyone's features. Is there any moats that you guys are excited about?

One funny one is there's a big moat to the brand and what kind of gets normalized, right? So as we kind of all agreed on earlier, we are using a very, very small fraction of the potential of these, right? And so if you think about the earliest adopters of this tech, which, you know, chat GPT has an insane amount of users, but the penetration is still pretty low. And that's why it's so valuable. And so the first cohort is going to be kind of the most diligent about figuring out, okay, this one is better for this. This one is better for this.

But each incremental on-border is going to be less particular. And at the same time, all of the models will keep getting better. So what that basically means is each one will continue to use less and less of the potential of this thing. And they're all going to be relatively commoditized for their use case. And so what it'll boil down to is what gets normalized, you know. Going back to, you know, use Xerox for copying, then Google, everywhere you Google it. And then now, like, Chad GPT has won that so far, right? That's just kind of the one that comes into mind for anyone who's looking to start dabbling in this.

And I think, you know, that as an onboarding tool and as a customer acquisition tool can't really be slept on. In general, AI stuff has less of a network effect than the Web2 giants did, right? like social media and ad-based stuff has way bigger network effect where it's just much harder to disrupt. But I think the moat in AI, there's some things that have a data moat, right? Someone like Tesla that has like so many hours of driving data and there's other like robotics companies that we've looked at where that's a moat. I think open AI in itself, like the amount of chats that they have and like the ability to use that for training and things like this is also somewhat of a moat. But I do think in AI

the main moat is just going to be like ux and and speed like the team that is the best at constantly shifting to where the meta is and and building the next thing ideally you don't want your memory to sit with with chat chipiti or whoever and this is like i think pretty visceral for people when they're sharing like i've shared some pretty personal stuff with with chat chipit like in in in i think we all have yeah like more personal than i ever thought i would have so i think ideally remember we would actually sit and we have a project that we're incubating that's actually building this like ideally you would have private memory built on T or ideally you know FHE once once once that works and then you would like give uh in sort of a cursor like UX you'd be able to choose which model you want

to give permission to access certain parts of that context to answer a query right I mean the ideal ideal would just be you have a model that runs locally but I think that's going to be super tough. So I think that's like one interesting area, but I agree in general, like the moats and that's why we've also been looking at deep tech stuff. I do think the modes sort of end up also moving to like hardware, to IP, to just things that in the past were seen as not sexy. You know, like it's not software. It's too hard. But I think those things will actually have like some of the most persistent modes in an era of, uh, of AI and just insanely deflated cost of software. Yeah, I'd say that on the memory front, I really hope that's not a moat.

Right? Like, if memory is emote, that just means that you're kind of like stuck into one of these ecosystems and you're really relying on that one builder to build every, you know, the best app of everything. Whereas like, you know, yeah, so, you know, to Jose's point, yeah, we are incubating a project that is, you know, based off this thesis, that memory won't be locked in in one place and won't be this mode. I feel like this whole memory term is just like another term to just. describe data, right? And that's what all the top social media technology platforms have nailed

so far, right? They just aggregate the most amount of data. I mean, Jose, you just mentioned that you use so much personal stuff where you say so much personal stuff to chat GBT. I am talking to this thing for hours on end, right? So at this point, I'm just like naturally inclined to use chat GBT even though there's like another model that comes out. I really hope the portability gets figured out, to your point. I just don't know what the incentives would be. for some of the bigger model producers. It's sort of different from social media, though, because in social media, it's not just the data,

it's the fact that all your friends are on there. So you don't just have to pour it over your data. You have to get all your friends to sign up to whatever new thing, whatever web through social media thing you're using. Whereas here, you know, portability, you actually have access to all your chatchipati chats, right? Like you could, and it's not that heavy.

Like, no matter how much you've talked to it, it's text data, you know, it fits on any computer. So I do think, If someone builds a great user experience here, it's something where it has, it can actually win because it's a better product like fundamentally. It just has to work really well. I also think I guess you put out this tweet. I don't even know if it was today or yesterday. It's been a long week.

But like, you know, you talk about the different personalities to these models, right? I think that's an interesting way to think about emot as well, right? The conversations I have with Rock are way different than the conversations I have with like 03, right? Right. So yeah, I think that's an interesting way to think about it as well. No, that's a good point. For those of you who are wondering what this tweet said, I basically described all the top models as having different personalities. So I said, Grock was kind of like, whatever, naughty and rude and extremely horny, just to be frank. And then Chad GBT was like kind of this incredibly agreeable personality.

Claude was kind of like, yeah, there you go, there you go. It's much more human, right? It's much more intuitive. Anyway, they have a bunch of different personalities. and it kind of like attracts a certain type of audience or it kind of like secretly molds you into being some one type of a user, right? You end up saying information to one model that you went to another

and it just kind of like creates this weird kind of socio dynamics that I think are interesting. But kind of moving on, guys, I remember when you first started your fund in 2019, the stuff that you guys were investing in, I thought you guys were insane. And this is coming from someone that like worked in the space, right? And then, of course, years past,

And it turns out that you guys nailed it. So my natural question now that you're focusing on AI and investing so much in AI is, and I'm going to put each of you in the hot seat, so prep your answer, is what is one emerging contrarian trend in AI right now that you think everyone is missing, but they should 100% focus on

because it's going to become a big thing over the next couple of years? I guess one thing I'd say is I don't think you necessarily have to be contrarian in venture, actually. I think you have to be right, but not necessarily contrarian. Although it helps, for sure. Like, it definitely is helpful when you're looking at something

and you're really bullish on it and no one else happens to be. But I do think just, yeah, I mean, the area I'd say is the one I already spoke about, which is like GPT rappers. I think a lot of people are sleeping on them, and I think they're going to be absolutely, like,

giant kind of businesses. What's a GPT wrapper that isn't like a coding wrapper that you think people should focus on or pay attention to? I mean, this application that we're building internally and there's a couple of teams that we've spoken to that are building it, one of them is Den. It's like a YC company. So it's kind of like think about it as cursor for work, right? It ingests like all your work data, your emails, your memos, your calls.

and then you're able to use any model to like run on that data. They also built a Slack clone, which I think is really interesting, because the idea being that you're chatting with these models anyway, and actually in the future, and so you can open these chat groups with the model and your team in them, and you can all chat to the model together in these groups and have different models in the different chats, which I think is really interesting,

like the idea that you're chatting already, why not have a chat app where you can have group chats with the models, and they can be on calls and stuff like this. I think various versions of those, I think you'll have a new Slack. I think all the company CRM stuff that Salesforce does right now is going to be rebuilt around AI.

I have to think of some more examples of good rap. Those are the ones I've mainly been focused on. But I think in hiring, for instance, you're definitely going to have something like that that's just going to know exactly, you know, what kind of person you're looking for, it can do the interviews for you, sort candidates for you.

Like, there's in every vertical you can think of, AI is going to have, like, you're going to want AI to do a huge percentage of the work. And there's going to be like an app that facilitates that workflow, I think. But you're giving more high-level ideas, though. I feel like he just wanted specific. Yeah. I want specific company. Yeah.

Yeah. Yeah. And the crazy or the better, honestly. Yeah. The crazy of the better. Just lean in. Mine aren't going to be crazy. And I hope Jan and Jose will make up for that.

like going off of what Jose said, which is like, you know, the contrarian part, I think is like over indexed. And I think, you know, in Crypto Venture, it definitely worked out really well for us. But also I think nowadays in crypto, there's not much stuff that is contra, like every, you know,

every conversation you have, people are bullish, you know, hype or pump or something like that. But I think, you know, when it comes to, you know, just generally AI, I think for us, we thought the contrarian thing was,

we thought even the most bullish people were going to be under exposed, right? So for us, we just want to be, And, you know, the thing that I go back and forth on, you know, to Jan's point is like, I think finding alpha here is going to be extremely difficult. Obviously, we're up for the challenge, but I think it's going to be extremely difficult. So for me, what I've been kind of pushing internally, and I think, you know, this is open to kind of like anyone inside or outside of Delphi is, you know,

capture a lot of this beta exposure. I think sometimes like investors and people just like to work very hard to feel like they're smart. But I think almost like, you know, you can capture, you know, a nice index of, you know, Open AI, Anthropic, like Anderol, Neurrelink, all this stuff. And capture a lot of this beta upside in a lot of these like sectors that you think are going to be massive, right? Even in the public equities, I think like companies like Google, you know, maybe Tesla and stuff like that, I think are worth like looking at, you know, I'm super bullish Google, for example, even though people maybe, you know, are dancing on their graves because they're thinking that, you know, their big search is going to be like cannibalized by AI, right?

Or open AI is launching this browser, which is going to like kill Chrome or something like that. So yeah, I think, you know, again, definitely not contrarian, right? I'm literally fucking talking about Google and, you know, open AI and stuff. But I do think that people will mid-curve it and say, you know, that's too easy or, oh, these things have run away, like maybe the 10x is behind me, 100x is behind. me or something like that. So let me try and find that 100x and then probably invest in things that go to zero instead. So that's my kind of answer. And then more in Jose's vein of giving broad ideas and not specific names. I think like, you know, one idea that I think

will be massive in the next, I don't know, 12 to 18 months is I think if you're using Twitter nowadays, you kind of get really annoyed at all these bots, right? And these agents that are like the in your replies, they're really bad. And so a lot of people are kind of like looking at for a social network that is like, you know, people only, right? Maybe you do this sort of world corner or whatever the fuck. I think actually the opposite is even more interesting where it's like a one on, you know, one where it's like you entering a social network where it's all agents, right? And you basically can kind of like get these agents to have a conversation about whatever

you want based on personalities that you actually do follow, right? Instead of, you know, people listen to the All In podcast and you're waiting for, you know, the topics that they're talking about, hoping to talk about, a topic that may be relevant to you, you can kind of create your own podcast of those personalities you do want to follow talking about the exact topic you want to talk about. So I think something like that will be really cool and I think will kind of exist in the next like 12, 18 months. I don't know if the company exists yet, but that's something that I think like meta is, that part of their strategy is just to kind of create a bunch of AI companions. GROC is launching

them as well. And I wonder, I wish I could somehow track how much time each human user, spends with some of these AI agents and companions as they go live. I bet you like it's going to be incredibly sticky. And what's really interesting about that, Anil, is that it's basically going to be a reflection of the person to an extent, right? And it depends on how much you dial up the sycophancy trait or if you dial it down and it becomes kind of like your mentor that kind of like abuses you every now and then.

It says like, no, you need to work harder or whatever that might be. All right, Jan, you're up next. So one area I've spent a decent amount of time looking into and I'm super excited about is the humanoid space. So I think, you know, us speaking to a bunch of emerging managers and early stage investors, it seems as if as in as most of them are kind of fading it to some degree or they think it'll be more of a application specific form factor that makes more sense from. a cost perspective from a utility perspective. Part of it is them talking their book naturally because it's a building out the humanoid component is very difficult and so and expensive. And if you're doing early stage investing, it makes more sense to do these targeted use cases that can get to market a lot more quickly and start to generate revenue. And so I think there's a massive world

where those make a lot of sense, right? The unit economics can be very predictable because most of the tech already exists, and I agree, there's a huge market for those. But I think fading the humanoid side doesn't make much sense. And the way to think about it is, the market for the humanoid form factor is insanely huge. I'm very aligned with the idea that there will be billions of these in probably two decades, just because of the amount of time it takes to build them, but I think there will be a massive just supply crunch for them within the next three to five years realistically. The human form factor makes a lot of sense because it can easily slot into everyday life now.

I think the cost component is starting to really get close to achievable. So the humanoid form factor business model usually fell off in the transition from protocols. type to scalable model. And that makes a lot of sense, right? You have these insanely expensive robots that can breakdance, but that's not really valuable from a business perspective. Ultimately, what you want is reliability. So you're paying for hours worked, right? That's kind of what really drives the value prop here. And so I don't think there's a winner take all in this market because the demand, I think, is nearly infinite, right? And as they get better, the surface area for deployment and implementation only grows. They all kind of gather within, you know, they all learn

together, which is, I think, something that isn't really appreciated enough where whatever it's learning in one factory, it gets to apply everywhere else. And so, and then, and you also, I think one of the things that that gets faded on the humanoid side is the fact that people think there will be kind of a societal uprising, right? They're taking our jobs. But for, yeah, exactly. But for the foreseeable future, it just kind of amplifies productivity, right? If you zoom out and think about demographics in terms of the population that wants to do

some of these roles, that's only going to decrease. So cost of labor will increase. On the other hand, you have electricity costs will come down, production costs will come down, reliability of these things will come down. and these businesses become pretty profitable pretty quickly, especially when you think about their creative kind of forms of financing. So I think that space isn't really as appreciated. And so realistically in the U.S., there are basically three major players for it, right?

You have Tesla as the leader with Optimus. Figure is second in line. They did a raise at $40 billion that's kind of getting wrapped up. And then I think Eptronic is the clear third. and that's they're trying to do another race soon and that's the one we're really excited about internally because we see a lot of value there. We think what they excel in is the actuator side, which is basically the joint of the robot. And that's something they've been building for quite some time. And I think there is a moat in that because of how that contributes to the dollar

spend per hours worked formula and in terms of what it does for reliability. And then on the other hand, their partner they're partnering with Google and plugging in Gemini right and so you have the physical humanoid and then the model and the two needs to work in tandem and so you can try and build the model from scratch which is what what figure is doing after their kind of separation from open AI but I think partnering with someone and focusing on your strength makes a lot of sense and so yeah it turned into an electronic show that point around the actuator Jan is such a crazy thing to think about can Can you imagine, like, in the Industrial Revolution when humans were, like, just working at factories, that they were each graded by, you know, their ability to move their kind of, like, elbow or whatever at a 90-degree angle.

That's just insane. The fact that you can kind of, like, program economics into these things is crazy. And I think you're right, like being able to kind of picture and visualize these robots as actual, you know, not some otherworldly creature, but just functioning humans. And then monetizing that is just, it's just a new model to kind of. wrap yourself around. It's just insane. I think human noise is a really good one because you can kind of, like, I think being in crypto so long, you can kind of identify what things cause a bubble.

And I think obviously the thing has to have very strong narrative potential, right? Like humanoid robots, replacing all physical labor has that. And then you also have to have a lot of hate. Like you kind of need, because it both forces people to talk about it and also creates like these really hated rallies. And I think human-eyed robots actually has a decent amount of hate from like smart people who just think that specialized robots are going to win out. So it's a very, I think, good contestant for that. I'd give you two names that I think are interesting, maybe contrarian. I think Anthropic is really valuable. It's like the least valuable of the model companies.

I think you could get it at like 60 bill when I last looked a month or two ago versus three to 400 billion for OpenAI and 150 billion or so for Gruk or for XAI now. And they're clearly the winners in encoding. Like they have been over and over again. I think they have a lot of market share encoding. Like every dev and any dev you speak to is using quote code. And I think that's insanely valuable. Like if you think software has eaten the world, is going to continue to eat the world, and you are literally the world's software factory, right, where everyone is going to produce software. I think it's insanely valuable. It's also one of the things that's easiest to train on because you have these easy kind of RL loops that you can do. It's formally verifiable and stuff.

So I think they're actually like in a really strong position. And it's tough because they don't have their own users. I think a lot of people use it via API. And that's generally not a not a great place to be. But I think if they win coding, that's like I think tens of trillions of dollars like use case. I think it's only going to get bigger. And then the other one, the one we're speaking about at dinner is just, it's in a hated sector. It's not to do with AI, but it's epic games. So those guys, they're doing like $6 billion in revenue and I haven't found supply for it yet, but it trades at something like $15 billion, which, you know, it's a very depressed multiple.

And just because gaming is not hard at all right now, gaming is in kind of a secular decline for the last two years, sort of the time people have spent not just crypto gaming, but time people have spent gaming has gone down for two years straight, which no one really thought was possible. No one knows the reason either. A lot of people speculate it's literally just TikTok eating your leisure time, like the people used to be spending gaming. And people talked a lot about the metaverse in crypto. Like, Fortnite has actually built the metaverse.

It's not VR like most people expected, but they have the closest thing to a metaverse in terms of, yeah, in terms of just different worlds that are player created, all the different maps that are player created, like 500 million users, they're having concurrent players, maps with, like, thousands of players, and just a really thoughtful CEO. And I think, like, everything is going to be leveraged by eye, and I think they will be too, just in the speed of what they can do. I think it's an interesting one that, like, it's always interesting to look at sectors, that people aren't excited about it all. And I think gaming is one of them right now.

Awesome. Before we round up, guys, you made a big announcement this week around something called Delphi Intelligence, and you gave Josh and I access to the platform beforehand, and we have to say, like, we were super impressed. Maybe you could tell us a little more about what this is and why it's important towards what you guys are doing. Yeah, definitely.

Yeah, so obviously we've talked about this a lot on the pot already, but like research is just at the heart of everything we do. And to be honest, like any decision we make, we kind of want to go in with conviction and as much like insight and knowledge as possible. So we know we're not only making the right decision, but when we are making that decision, can size it properly, right? And I think for us, you know, right,

basically, you know, Jose right after he kind of like passed around this situation in his paper, which he actually read on a, you know, week off, which is like probably when we get the most work done, it's like our weeks off, to actually like read and think about, you know, the future of Delphi and everything like that. I think that's when we really, you know, probably nine, ten months ago at this point, realized that, you know, this was like a no, not an option for us, right?

We think to be the best investors, builders, researchers in crypto and honestly any area, you kind of need to start building expertise in AI. So that's when we really started, you know, rolling up rest of leaves and doing the hard work of building out a team and building out kind of like an MO, which is just public, a lot of great work in areas that we're interested about so we can kind of build conviction and build expertise in this area to help us make these decisions. So that's what Delphi Intelligence is. It's a research platform free to access for all so you can go on Delphi Intelligence.io right now, put your email in and you'll get all of our research basically biweekly free. We already

have two reports out, one on just like AI in the era of entertainment, and then one on video generation models. Both are like great. We have another one coming out next week on AI powered browsers, which I think is going to be like really top of mind for a lot of people. And essentially like, you know, it's us open sourcing our learning to the world. And what's cool about it too is it's not just going to be our team. We're going to be curating a lot of great reads from within our network and people we respect, you know, including some of the fund managers that, you know, Jose brought up. So yeah, I mean, if you're interested, please subscribe, you know, follow us on Twitter and everything like that, but we're really excited about it.

Awesome. Well, thank you all for spending time with Josh and I and kind of going through your thoughts on the AI market. As you can imagine, like, there's just so much going on. And, you know, our Twitter feeds or rather our X feeds are off the hook. We are talking to like five different AI models for various different things a day. And it's just not easy to think strategically and long term and have conviction around investments, right? is such a hard thing to kind of nail. So, you know, hearing your perspectives has been hugely informative for us and I'm sure for our audience as well. For the limitless listeners, thank you so much for joining us for another episode. As you know, Josh and I are trying out something new, which is

just put out loads of content as and when it comes live, as and when the topic is trending. So we appreciate you and your feedback. The main bit of feedback that we've got so far is that you love the guest episodes and we want to get more interesting guests on. We hope you see this as one of those pushes towards that. And again, if you have any friends or colleagues or whatever that might be interested in this thing, we appreciate you sharing, liking and subscribing. Thanks, folks, and we'll see you on the next one. See you guys. Thanks.