Robert (2:36)

So. So first off, Haseeb left out the most important item here, which is Beau. Partner and co founder of Dragonfly. Founder of Dragonfly. Right. Co. Founder, co. Found something completely flexed and stunted all over Haseeb. Tom and Rob in the picture because he wore this insane velvet suit. It's like a beautiful purple velour suit. And the rest of them are like, dressed in like Silicon Valley schlubby.

Haseeb (3:03)

They wear a bunch of schlubs.

Robert (3:05)

Yeah, that's right. I believe the shirt was amazing. It was amazing. It's like, you look. You have to look at that picture. It's like perfect. Beautiful.

Haseeb (3:13)

Yeah, I think we were fashion mogged, I think is the term. Yeah, no, it was good. But obviously we're very grateful to have money at a time when money goes a very long way. But there's also this kind of broader moment that I think we are meeting now with, having raised this fund. Is that the gap between institutional sentiment and retail sentiment, or really like crypto native sentiment, is the biggest that I have ever seen. And so there are a lot of people, I mean, we had a bunch of comments and a lot of people were asking, like, how did they raise this money? Like, where did. Who gave. Who's giving the money? And the answer, I'll tell you, for those who are wondering, the answer is institutions. Institutions give us money, and then we take that money from institutions and we go invest it. You can't get this amount of money from obviously retail investors. So institutions are bullish. Crypto. They believe in the story. All the stories that we're telling, all the stuff about stablecoin adoption and institutions coming to crypto and tokenization, they believe all that shit, you know, like, you might think they're fools, but they believe all of it. And they're continuing to invest into the space through institutions like us and others, and also obviously investing directly into the space. But at the same time that all that's happening, you're seeing a lot of turnover happening in the industry. So friend of the show Kyle Samani, I think it was just last week, announced that he was moving on from multicoin. He wanted to pursue other interests outside of crypto, potentially in longevity and AI. I don't know exactly what that means. There's obviously a lot of vagueness about why exactly he's moving on, but he's still going to be involved with Forward, which is his dat. And we also had Ariana Simpson, who's a GP at a 16Z crypto. She is moving on, launching her own fund. She's going to be investing also outside of crypto. We also had some other senior people leave. Tamash from the ef. He's also been on the show. He's moving on. Bastian Au, I believe his name is gonna be replacing him, who was a senior person at the ef. You also had Akshay, who was head of bd, or I think he was head of BD at Solana foundation, something like that. Head of something. Solana Foundation. He's moving on Nader. Or is it Nader? Nader, from. He used to be eigenlair, probably the best known developer relations guy in crypto, is moving on to Cognition, which is an AI startup and he's now developer relations there. So it feels like we're having this moment of reshuffling, a lot of senior people from the industry suddenly moving on. And in the midst of that, we're trying to, you know, do this last hurrah in our. In our front four. But I don't know, I want to get a gut check from you guys. How are you guys feeling? You guys, I mean, are we the Japanese on the island kind of fighting when the war is over? What's.

Robert (5:49)

I mean, first off, congratulations. You've been on a monologue, but we didn't get to officially say, I know,

Tarun (5:55)

I just got the flaps number four as a hand sign that cz.

Haseeb (5:58)

I did see that. Yeah, I like the. We're going to do that. Thank you. But how are you feeling? How are you guys feeling? Talk about the moment. Yeah. Yeah.

Robert (6:04)

You guys deserve to be at the moment.

Haseeb (6:05)

It's hard to celebrate right now. It's hard to celebrate right now because, like, I was very mindful. Normally you write, oh, we announce a new fund and you go and do the whatever parade in the streets. But it.

Robert (6:14)

I remember talking to Arjun from when he joined Paradigm right after they raised that like that fund in 2018, a little after you guys, I think, right? It was like October, November, and everyone was like, oh, my God, Bitcoin hit 3000, everything is zero. Like, why am I doing. You know, like there was like definitely this feeling of fear and I don't think it abated for like two or three years. And it kind of feels a little bit like that to me. That, like that, that mo. There was like this turkey drop, as I like to call it, of like Bitcoin's price like went down like 40% in like November 2018. It's just like it was near Thanksgiving because I remember everyone was making all these memes about like having to go home at Thanksgiving and everyone being like, oh, so how's your little coin doing? But like I kind of remember that. But then you like look at those funds performance wise, like obviously just completely crushed because everyone was hating on those funds. I remember how, when I think about how uncompetitive. Okay, I'm not trying to. Don't. This will get me in trouble. But whatever, I'll still say it. Which is, I remember uncompetitive. I thought the Uniswap Series A was at a low valuation. It was like 50 million or whatever. And like everyone was like, no, no, no, no, I'm not doing 50, I'm going to do 41.7. You know, it's like everyone, everyone remembered all the significant figures in a, in venture investment at that time.

Haseeb (7:42)

That's true.

Robert (7:44)

And, and like it was just like, yeah, but that's when you found the craziest things. And so I think like, you know, I, maybe I'm just this hope springs eternal person, but I'm just kind of like, eh, you know, it goes away, comes back. Everyone is just too obsessed with AI right now. that time, I remember everyone was so obsessed with fintech. That was like the peak plaid getting acquired, stripe hitting almost 100 billion, whatever, right? That was like the fintech was so much better than crypto. Why would you need crypto era? And it flipped. And I think another aspect of it that I think is true is everyone is talking in normal markets about SaaS apocalypse, right? And how all these SaaS companies, you know, a bunch of them, their moat is lost because like you can replicate a lot of things quickly or just in time or as you need it with a lot of AI tools. But there's sort of this, this kind of sense in which liquidity is a moat, right? Like your AI agent can't really copy liquidity and it needs to earn liquidity somehow. And like crypto is the only easy liquidity source. And like at some point those things are going to merge and that that, that's my hope, springs eternal optimism. You can't really copy. There's no SaaS apocalypse for liquidity.

Haseeb (9:02)

I'll take that. I'll take that. Robert, how are you interpreting the vibes?

Tarun (9:06)

Well, I think the vibes are bad, but from a venture deployment perspective, I think that's relatively good. You know, Tarun touched on this, but in general, you know, I'm a believer that bare market prices make the best entry. And when there's less competition for deals, you can take longer. Right. You don't get rushed into them. You don't get bullied into them by the market or other founders. And I think it's just a wonderful time to be investing because you can actually think more than in a bull market when it's an outright mania. And there's no time to think, there's no time to be strategic. There's no time to, you know, really like, you know, plan the portfolio as much as you can in a bear market when things are quiet. And so the doom and gloom is the best time to raise fund for, for Dragonfly. I think you guys are really fortunate with the timing. I think it's fantastic. And having, you know, a big stack of chips at a very quiet table where you can be picky is, you know, fantastic for investors, but also for the founders when the capital is scarce because, you know, not everyone, you know, will be backed in a bear market. But the founders that can build when there's less competition, when it's quiet, when they can build a good team, you know, it's a great time. Most of the things I've done in crypto has been in bear markets. You know, I founded Compound. You know, it was really like end of 2017, beginning of 2018, as things were diving off the cliff. And I found the superstate in 23 when it was like, you know, as unexciting of a time to found something new as you could find besides right now. And so I think being a founder in a bear market is great and for the same reason, I think being an investor starting in a bear market is also great.

Haseeb (10:48)

But how do you feel about seeing your peers, like people you've been in the trenches with for so many years, leaving the industry?

Tarun (10:55)

I mean, listen, if you've been in the industry a long time and you're now entering a bear market, right, and you have a giant fund that's already deployed, right, it doesn't look as good, okay, it looks very good for a Dragonfly and a Fund 4 announcement. It looks less Good. If you have a large amount of capital that's already deployed and you really can't compete and raise the same size fund again.

Robert (11:21)

Right.

Tarun (11:22)

So like, I get that it's ominous and it's as good of a time as any to go on permanent vacation to pick a new industry to spend more time with your family, like, whatever. The thing is, you know, especially for people that have the good fortune of having already succeeded in so many ways, right. It's like, why play on like super duper duper hard mode, Right? So I get it, you know, I, you know, their loss, everyone else's gain.

Haseeb (11:49)

Tom, what do you think? We got money.

Tom (11:52)

Good. Time to have a fun.

Haseeb (11:53)

Thinking about quitting?

Tom (11:55)

I'm out. Yeah. Yeah, I, I always remember exactly little stories. What'd you say?

Tarun (12:01)

You're not out, you're in. You just started.

Tom (12:03)

I'm in. Yeah, I know. I was a. When Jan Koum, the founder of WhatsApp, left Facebook after he had three or four years parting note that he ended with something was like, I'm spending time on air cooled Porsches. And I thought that was just a very nice transparent way to express what he was thinking about. But no sappy family shit. Yeah, it's the weirdest the industry feels like the weirdest it's felt. I think maybe something my entire time being here, where I think 2017, 2019, it was like, okay, it was much smaller and it was experimental and people genuinely thought, oh, maybe we're just super early or maybe this idea isn't going to work. And it was very existential for everything that was going on. I sort of understood why it felt that way. And then 22, it was, I guess, a similar kind of vibe. This feels like you're saying things are really working and yet there's this huge dispersion in terms of sentiment and just industry players. And I also, to your point, just the amount of personnel turnover, I don't remember really happening in this sort of dense of a period. And it's not even people even explicitly quitting, but it's people shifting their time to work on other projects or people sort of quiet quitting or sort of being more checked out. And I think at the same time, it's like other parts of the industry are moving faster than ever. So it feels just kind of like this weird, I don't know, shift. It's kind of like you. It's like an rpg. You've graduated from one town and you went to the other town and the characters are totally different and you're like, what am I playing again? And. But, you know, the game goes on.

Haseeb (13:36)

The game goes on. Yeah, it does feel like kind of disc two now of the game. And like the vibe is very different. Like, you can tell there's like a different set of developers who made Disc two. You know, I think that's a little bit of what I'm feeling as well. And I talked about this when Kyle announced that he was quitting because I actually, I wrote a kind of sappy tweet about Kyle and like, the way I felt about it because I. When I first saw that he had quit, I was like, holy shit, that's crazy. I can't believe the Kyle quit easier for us. Great. And I thought about it a lot and I remember waking up a couple days later still thinking about it and about what it meant for this industry. And the thing is, Kyle was the embodiment of that 2018, 2019 craziness. It was the, it was the spirit of crypto was there was this tiny flame that we were all cupping our hands over to keep alive at a time when nobody else believed in it. And that was Kyle. And he had a very bombastic way of doing that. It was very different from pretty much anybody else in this industry. But it was very powerful, very necessary. He was such an important proselytizer for the things that he believed in. And that mode of being just doesn't work in what the industry is now. The industry is very much, I think the way I put it was that we've exited the pioneer phase and we're now in the settler phase. And the people who settle the towns, as opposed to people who found the towns, are a very different kind of psychology. It's a very different kind of people. It's a very different risk appetite. It's a very different pace of life. And it's true in every industry. Jan Koum the people who built the social networks are very seldom the people who steward them going forward. It's not Reid Hoffman. Mark Zuckerberg is maybe the only exception. Right? Yeah. The people who make the thing are very different from the people who shepherd the thing. And Kyle, I think in a way he bristled against the direction in which crypto has grown. And you can tell immediately after he quit, he got into this big fight about hyper liquid and just immediately face planted in public and it's like, yeah, you can't do that shit. You know, like that's like you could get away with that in the previous era of crypto, but crypto has Evolved. Crypto's changed, it's institutionalized. It's a different game now. And I think, in a way, for me, what it represents is, okay, how do we need to evolve in a way that corresponds to the way crypto is evolving? And that's also not to say that I don't think retail investing or the crazy volatility or the crazy games that investors and traders tend to play. It's not that I think those things are done forever. I think those things will basically never be done. I think those things will keep coming back and they'll evolve, they'll come up in a different form. We'll see in a year, two years, three years, there'll be some new thing that's viral and speculative and crazy, because crypto is permissionless money. And people always find a way to do that thing, that crazy speculative thing. But the idea that that's the beating heart of crypto, that that's the main thing, that's the show, everything else is like kind of this weird institutional thing on the side. I think that's what's changing. I think that's what's going away. And the real adoption story, the real, like, yeah, you know, this brick and mortar kind of, okay, stable coins and wallets in people's hands that they're actually using to transact that stuff is not going to go away. So, yeah, it's got. It's got me thinking because, like, man, we've been here for a long time, all four of us, we've been here, I think, around the same time that Ky, which is 2017, 2018. Like, I think we're all basically the same vintage. And like, I was a young man when I got into the space. You know, I was in my. I was in my mid-20s. And now, like, you're just like, I'm. Fuck, I'm. Dude, it's just all gone. No, it's like your attitude evolves alongside the technology. You know, like the technology's grown up and we've also grown up. And I was reflecting on that, that, like, it was also true of Kyle, you know, I mean, like, you know, he. He was. I don't. I actually don't know how old he is, but I assume he's roughly the same age as us. It's like, yeah, we saw all of each other go through this journey together on this roller coaster that was crypto. And I understand why some people want to get off the ride, because it's. Yeah, it's. It's a. It's a motherfucker you know, it's a, it's a really, really challenging space to dedicate your career to. And so, yeah, I don't accept the framing that it's something about crypto or something about, oh, well, you know, it's just not that exciting anymore. It's not that big anymore, it's not that successful anymore. Oh, they were proven wrong. I think it really is just. Yeah, people, they get older, their priorities change, they move on, they want to do other things. Yeah, I don't know, but it's sad. I think seeing, seeing Kyle leave honestly hit me harder than anybody else leaving. So.

Robert (18:47)

Wow, you really just, you just really, you really dunked on the other departures.

Tom (18:51)

You just.

Haseeb (18:52)

No, no, it's just that like, instead

Tarun (18:54)

Kyle was on the show the most out of anyone.

Haseeb (18:56)

He was on the show the most.

Robert (18:57)

True, that's true.

Tarun (18:58)

That's appearances on the chopping block.

Haseeb (19:00)

Therefore, that's why it hit me so deeply is because.

Robert (19:04)

Well, I mean, repeat guys, I mean, a more objective version. He's probably the most successful of all the ones you mentioned.

Haseeb (19:11)

Yeah, but it's, it's more than that. I think he was the most singular,

Robert (19:14)

you know, he had the most conviction also of all the ones you mentioned.

Haseeb (19:19)

He did.

Robert (19:20)

He believed in this thing and he was just like, I don't. I'm going to be a bulldozer and not. Not see.

Haseeb (19:24)

Yeah, exactly. I think he, he was like the apotheos of crypto. Right. Crypto is the whole idea that like people who have. Who come from the fucking nowhere with no credentials and just like a force of will. Will can just make themselves successful.

Robert (19:36)

Dude, Google Glass.

Haseeb (19:37)

That was Kyle.

Robert (19:38)

Google Glass credential.

Haseeb (19:39)

Kyle was a Google Glass entrepreneur before he ended up going into crypto. He is the most unlikely person to end up becoming an institutional fund manager. But he did it because he believed. He just, he just kind of used the bully pulpit to transform multicoin into a phenomenon. And I say in the post that I think he's in many ways, although I'm totally different from Kyle, I could not be more different in just the way I behave and the way I operate and the way that I. My opinions. But I think in that sense he and I were actually the most similar because he didn't come from institutional background. He didn't have any great pedigree. He was just a guy who just got a loudspeaker and started yelling until people started listening. And he turned that into a multi billion dollar platform. And I think that's crypto that was crypto in a nutshell. And you can't do that today. You know, it's like very, very hard to recreate what, what Kyle built.

Robert (20:36)

I feel like you're going to eat your words on that. I bet you someone will do that this year. Yeah, there's always like one thing where like someone breaks beyond the noise and that happens.

Haseeb (20:46)

That's true, that's true. But it's only in bull markets that you see that in bear markets. Very.

Robert (20:50)

I think the hyper liquid broke out in the bear market like it was.

Haseeb (20:54)

They did, they did. But I'm talking about investors. I'm talking about investors, right?

Robert (20:57)

Oh yeah.

Haseeb (20:57)

Every, every cycle, every cycle there's a new VC firm that pops up in like, becomes meteoric in their rise. Right. If you think about like, you know, the Jump and Alameda, they, they became huge. Or fpg, Delphi.

Robert (21:10)

Oh, fpg. Wow, I forgot about that.

Haseeb (21:12)

Yeah, there's always like every bull market is an opportunity for a new investor to define themselves as, you know, riding this new trend or riding this new meta. But in a bear market, it's really tough. Right. Like a lot of people are struggling to raise in this environment, as you know, told by the headline of the Fortune article. So. Okay.

Robert (21:30)

Although to be fair, that guy's articles oftentimes have a lot of this like doomsday.

Haseeb (21:37)

I mean that's everybody's articles these days. Yeah, it's hard not to talk about crypto without talking about, you know, kind of doomsday vibes. So anyway, speaking of vibes, let's talk about openclaw. Openclaw, of course, is the vibe coded AI agent or AI agent, pla. Whatever open source thing. It was a phenomenon originally called Claudebot, became multbot and then finally Openclaw. Openclaw was recently acquired by OpenAI. Now acquired is a weird term. Apparently there is going to be.

Robert (22:09)

There's a claw foundation. There's a claw foundation.

Haseeb (22:11)

Yeah, exactly. There's going to be a foundation that's going to be maintained by OpenAI. There's a lot of speculation about what the acquisition price, if there was one, might have been. It seems like just a lot of bullshit. People are just making up numbers and trying to get them to go viral. So I don't think anybody knows anything.

Tarun (22:25)

I saw a made up number that was definitely.

Haseeb (22:27)

Okay, did you, do you want to repeat the made up number, Robert?

Tarun (22:30)

Yeah, I saw the made up number. It was either 30 or 40.

Haseeb (22:34)

Okay. I saw made up numbers going up to a billion.

Robert (22:37)

So yeah, I Saw it close.

Haseeb (22:39)

People are just making up numbers. To be clear, there's nobody knows anything. There's absolutely no reason to think that anyone knows anything.

Robert (22:44)

But I do think it's funny that Dave Morin, the Facebook guy, is the head of the foundation. I thought that was like funny they had. Wait, really?

Tom (22:52)

I missed that.

Robert (22:53)

Yeah, yeah. It's not because he went to go work at OpenAI. They needed a third party. It reminds me of crypto because it's like a crypto foundation structure where OpenAI is the lab company for the foundation.

Haseeb (23:06)

Wait, so OpenAI is not taking over the foundation?

Robert (23:08)

Foundation is going to take the foundation separate.

Haseeb (23:11)

It's independent, all the employees are stable. This is like one of these token acquisitions. What's going on?

Robert (23:18)

There's some irony because of what you're about to say about tokens.

Tarun (23:22)

No, but it was a one man startup.

Robert (23:24)

Yeah, it wasn't, it wasn't like a real. But. But I think they are keeping them separate on purpose. Like.

Haseeb (23:31)

Okay, okay.

Robert (23:32)

So that like the, the open source stuff is like can live on its own without even if open AI died or whatever. Right.

Tarun (23:38)

Their name is Open AI. It would be sad for them to close source it.

Haseeb (23:45)

Yeah, okay. They have to be same. Altman in the announcement said that this is going to become. We expect this will quickly become core to our product offerings. OpenClaw obviously is one of the fastest growing open source repositories in GitHub history. And of course originally they were based on anthropic. You know, Claude was literally in the name and he. So the creator, Peter Steinberger, he went on the Lex Friedman podcast I think just last week and there was a clip on that podcast that went viral where he basically complained about the crypto people and apparently crypto people were trying to bully him into launching a coin, or not launching a coin, but accepting a coin, I should say. There were a bunch of coins that were launched in the name of openclaw and they were like, please take this, you know, take the, you know, dev reward or whatever and basically anoint this as being the canonical meme coin for this, for this project. And he said on the podcast, quote, they crypto people are not just good at harassment, they're also really good at using scripts and tools. I underestimated those people. Honestly, I was that close to just deleting, deleting it. Basically implying that he was being so harassed by crypto people that he was willing to just delete openclaw and destroy it entirely. So thoughts on the ol. Openclaw Saga. Also, have you guys gotten an openclaw up and running? I think last time on the show, I think I was the only one who played around with it.

Tom (25:05)

I have one.

Haseeb (25:07)

Do you have one?

Tom (25:08)

Yeah, as I said that I went to the meetup in sf. Not much, honestly, I don't find it to be extremely useful. But I'm more struck by like, how and why it resonated with people. And like, that's kind of been a lot of the, you know, theorizing around Twitter is just like, why is this resonating? Because it's not like, oh, here's this one killer app use this for, and everyone's signing up for it. It's kind of amorphous productivity idea around. Like, yeah, it's kind of nice to have some permanence. I think maybe there's an idea around people don't like being locked into a single model provider or they don't like, you know, having their data stored with a single model provider. And there's kind of degrees to that. But yeah, I don't quite know. Like, I would not have predicted it would have like resonated this much with so many people. And now all these like little sub startups and wrappers and there's like even new implementations. Like I saw Ilya from Near Write like a Rust implementation that's like, you know, more efficient. So I don't know, it's like a very interesting time in AI land too. I mean, one, this being kind of one of those big threads in terms of agents and agentic coding and people also looking at new models and obviously all these open source models coming out. But I'm also always just reminded and kind of struck by this constant kind of give and take between crypto and AI for two industries that you would not think have any similarities. There's actually a lot of overlap in terms of the people and the technologies. And literally Ilya was on the Transformers white paper and SBF was investor in Anthropic and Sam's founder of Worldcoin and Open. So it's like kind of crazy that these two very different technologies have so much overlap. And I'm very curious to see what the future holds. Not even just like in the X402 stuff, but it feels like they're going to continue to kind of collide in some ways.

Haseeb (26:57)

Yeah, it reminds me actually. Do you remember, I can't remember the name of it, but there was something around the time there was a project called Baby AGI and there was another one that was like Going super viral. Auto GPT. That's right. This Auto GPT, if you ever tried it, it was absolutely horrible. It could not do anything. But it was also one of the fastest, most starred projects on GitHub. It was also this kind of crazy viral sensation and most people never even tried it. Most people never even use it, they just watch Twitter demos of it. And this again feels very similar. So I tried openclaw back when it was called cloudbot and I kind of played around with it a little bit and I was like, this is so broken and kind of useless. Like, I don't know why I would use this. I recently got it up and running again now on a VPS just to see how it's improved. It has definitely become a lot more stable than it was when I was first using it, when it was a total buggy mess when I was originally using it. It's still pretty buggy. It's not a great piece of software. It's very clearly vibe coded. But it is one of these things that now that it has such a large open source community, you can see that it's going to get better and better, right? It's going to become more stable, it's going to become more usable. It's got all these plugins. It's just like, you know, it's one of these things that we see over and over again in open source is that once you have a standard and people converge on the standard, they just start making it better, they start plugging everything into it. It's a little bit like mcp, you know, it's kind of loose of a concept, enough that you can just, once you converge around it, you can make something good even out of a shell. That is not optimal. But it does still feel like this sort of autogpt thing that people imagine what it can do and when you actually try it and you're like, great, please pilot my email and do my job for me. And it just starts doing dumb stuff and it obviously doesn't understand the context. And yeah, it's just like it doesn't work unless what you're doing is very, very well documented and extremely constrained. And you're doing a lot of prompt engineering and you're doing some good rag. You've basically built quite a lot in order to get this thing to do the right thing. Otherwise it's just, it's going to fail 15% of the time. And something that fails 15% of the time is just like a horrible employee. You know, it's like the worst possible hire you've ever made. So nobody's going to actually have somebody who like fumbles their email 10% of the time answering their emails for them. But it's such a tantalizing idea and you can kind of see it, right? You can see it sort of working. You can do the Twitter demo. You can, you know, you can have it call your mom and talk to your mom using, you know, Whisper or whatever. And, and like those demos, like you can see the sci fi movie just poking through the membrane. I think that's why people love it so much is that they want it to get there.

Robert (29:39)

Yeah, I, I kind of mainly agree with you. It's like it, it's, people are ascribing a lot more than what's there. But like it's pretty clear that someone will. You know, I would say it kind of came out of this fact that like after Claud code got a lot better, there was just this huge rush for like a million people to make orchestration tools for like, how do you run 50 agents at the same time? And they were all kind of clunky and for some reason this ended up being the modality where it was like slightly less clunky and it took off. That was sort of, my impression was like there was this like two month period of like everyone and their mom was just like, okay, there's some new orchestration tool manager agents, right? And you know, I, I think a, a, a common maxim in parallel programming is it's easy to pull a, a carriage with two horses, but can you pull a carriage with 10, 24 chickens? And that's kind of like the problem I see with a lot of these things is like the reason they're so clunky is they like try to instantiate too many agents to do stuff and then they just step all over each other. And that can be fixed, but it takes engineering and time. It's not going to be fixed by like five prompts, you know, like, and I think that's, that's where it started.

Haseeb (31:01)

Yeah. So this is one place where I think I am bullish on open source. Is that the thing about all of these things and everybody's talking about this, that these openclaw super insecure, it's very easy to create these foot guns. You know, people have these like, you know, servers with open authentication or easily brute forceable passwords that were just sitting out there that, you know, you can just show Dan and like kind to bring them all down and start basically just prompt injecting them or whatever. So I think there's a lot of security issues with these things, which means that the big labs are going to be very, very cautious about launching these products. Right. Because these products, one, they don't really drive that much business value yet, they're kind of too unreliable to do that. And then second, they're also very insecure, very unsafe. And so that combination makes it very impractical for an enterprise facing company to launch a product for this. But it's exactly the kind of thing for people who just want to, you know, basically buy a gun and like point it at their face and just see what happens. And there will be a huge community. Huge community.

Robert (32:02)

This is like EVM smart. EVM smart contracts in 2016.

Haseeb (32:06)

Well, it's more like, it's like gas golfing, you know, it's like writing EVM by hand in line with like, no, no, like, you know, no safety on. And it's extremely dangerous, very stupid. But people are going to love, people are totally going to do it. So I think this is going to blossom in open source and I think the big labs are going to be very slow. I think open source can be way ahead of what the big labs are doing. And even Google, Google probably has the best shot on goal because they literally have Gmail, they have the connectivity to your entire life. They're not going to touch this shit for a long time. It's going to have to be very, very robust before they touch it. Because just one bad press story of like, you know, this company had their open claw, like, you know, leak the production database and all this pii to like, you know, some rando and it's just going to be a PR disaster, you know, and they're not going to pay for it because like, where's the business value?

Robert (32:56)

Well, I think it's funny that OpenAI is now again an open source company indirectly.

Haseeb (33:03)

Well, it's technically foundations outside apparently. So it's not.

Robert (33:07)

Yeah, yeah, it's true, it's true. But it's the labs co it's writing.

Haseeb (33:11)

That's right, that's right.

Tom (33:13)

The crypto crossover is also again present here where it's like you see these viral Twitter threads of people like, oh, I told my openclaw to make me a profitable polymarket strategy or I told it you had $50 and was going to turn it off if it didn't pay for its own server costs. And then obviously everyone is also like, that's great. Yeah, everyone's launching the token. And I will say, if you don't know a lot of these are fabricated or greatly embellished. So don't try to just five code a polymarket trading strategy from your OpenClaw. But it is again speaking to this kind of like agentic permissionless finance. Yeah, if you want to try any of this stuff, you can just spin this up today and start using it. You don't have to go and have your agent register for an account. And I think there is something about the accessibility of on chain financial applications that feels so tantalizing.

Haseeb (34:03)

The other element of it is if you look at the. So if you look at Multbook today. So Multbook was this social media platform that was kind of a Reddit like thing that as you set up an openclaw you can go on there and you can start having your agent talk to other agents. And so it's extremely weird social experiment thing. So Moltbook is still around. People aren't talking about it much anymore, but it's still around. And almost everything on Moatbook is crypto. Almost every single thing is some agent pumping some kind of meme coin or trying to prompt inject some other agent.

Robert (34:36)

So the agents tried to replicate farcast variables.

Haseeb (34:40)

Yeah, it's really, I mean that's the thing is like once you have just total freedom, the first thing you're going to do is try to scam other people to take their money. Other agents, Crypto Other agents. Well, agents to get the people underlying the agents money. So there is something that's like again, this is like the dark underbelly of economic incentives. Is that the first thing that's going to happen when you build something is that people are going to try to get you to claim a meme coin or they're going to use your platform to launch meme coins or to scam other people out of their meme coins. So there's, I don't know, it feels inescapable that there is an intersection here and it's not a savory one. I think there will be more than just, okay, we're going to use this as a substrate to just blab about meme coins. And so there's a lot of excitement now increasingly about X402. X402 is a standard that has been, I think incubated by Coinbase. I don't know if they originally proposed it, but certainly has been adopted by Coinbase. So X402 is a reference to an old, I think now deprecated HTTP code, which was an HTTP code that basically said if you, you know, like 404 means not found, 402 means requires payment. Now that's not been used in the HTTP standard, but it was technically there from the original HTTP spec. And Coinbase has kind of adopted this, or the industry has adopted this as being the standard by which you can negotiate with an AI agent to say, hey, if you want me to do X, I need this payment for this service. And there's, you know, a sort of a protocol by which you can negotiate what this payment might be. So X402 as a payment standard has been growing. Stripe recently announced that adopting S402 payments.

Robert (36:16)

One thing worth adding that I think is actually is they're wallet lists. You don't need kind of like the whole apparatus for signing you just like, it's a very minimal protocol. So it, it's kind of easy to use in a way that I think a lot of other crypto stuff isn't. Which is worth noting in case you're like listening to the show and you're like, here's another fucking acronym of some crypto shit that probably only three people can use. It's actually a very simple protocol.

Haseeb (36:45)

Okay, how, how does it work, this

Tarun (36:46)

super high level terrain?

Robert (36:49)

I mean it basically it's just like a message that has like a field for a assigned payload that is, you know your, you know what you, what you want. And then it just has some instructions for your transaction, like max gas cost max how much you'd pay a relayer if you're doing an account abstraction wallet. And that's it. It doesn't, it's not like a whole wallet that like you have to hide the keys and you have to pull out your keys each time. And like the idea is it isolates that and the browser can do that however your wallet does it. So I think the main point is it's so simple that you don't need to know anything about crypto to use it. Which is I think the reason it might be more popular for web2 devs. I think the goal is to get

Haseeb (37:31)

web stick to the way in which you pay someone.

Robert (37:33)

Yeah, exactly. So that people who are Web2 devs might be more amenable to using it.

Haseeb (37:40)

Right. So there's this whole concept that AI agents may end up paying each other using something like XRO2. And the idea is that, okay, well, most people might think like, well, why don't I just have my AI agent have my credit card? We actually had somebody on our team was telling us that they were setting up an openclaw. And the first thing that openclaw did once it came online was ask them for a credit card so that they could start buying stuff. Because, okay, what's the first thing that you can do in order to be useful to your human? And the answer is like, I don't know, maybe buy something on Amazon. So the idea is like, okay, if I want to give my AI agent some financial autonomy, I could give them a credit card. But with a credit card, all you can do is buy things from merchants. They have to be able to take Visa. But if an AI agent is going to ask another AI agent to do some work for it, for example, let's say they have access to some data or they have some skill set, or they've already built something that I can use and it gains from trade, it's more efficient for me to ask them to do it rather than for me to burn the tokens to figure out how to recreate what they already created, that I can just pay them. And that concept, okay, how do I pay another AI agent? Maybe I swipe a credit card. But then they have to have a merchant account, and they're probably not going to have that. What country are they in? How do you remit taxes? Are they in some country that I'm not supposed to transact with? If it's just another AI agent, then crypto kind of solves all of that in a way. Crypto was kind of made for machines more than it was made for humans. So I think there's a really elegant story that's becoming increasingly obvious to people that, hey, this is probably the way in which AI agents are going to prefer to pay each other, because it's extremely simple, instantaneous, and they know how to manage environment variables and custody. Key material, or not custody, but whatever, obscure key material. So this story is growing, and we're seeing more and more x402 transaction volume happening. It's still pretty small in absolute terms, but I think seeing what's happening on Mulbook, seeing what's happening with people experimenting more and more with Open Claw and giving their agents money, I think this is going to be one of the big trends over the next year or two. But the reality is these things are still pretty dumb. So you would not want to give your agent a lot of money. You definitely don't want to give it your metamask and just let it run wild. That's probably very, very stupid. But if you could give your agent a little bit of money, and I don't know, see if it can trade on Polymarket. Look, I don't know, it's a great idea but there are stupider things you could do. So anyway, speaking of Polymarket, besides all this talk about AI agents building polymarket trading bots, Polymarket as well as call sheet, they had some of their biggest trading days of the year because of the Super Bowl. Super Bowl? Huge all time highs in terms of sports volume. Not all time highs. Was it all time high? I don't think it was all time high. Right. I think World cup was ultim high, but anyway very high volumes.

Robert (40:32)

I thought super bowl was a high. I thought this was actually super bowl was high. I think zero high.

Haseeb (40:37)

Was it all time high? Okay, fine. Was or was not all time high? Unclear, but Polymarket recently launched 5 minute Bitcoin up or down markets. So these are markets that basically you can bet in the next five minutes is Bitcoin going to be worth more or less than it is currently now, this obviously being at a time, I think there's 15 minutes as well. Oh, sorry. They have one hour as well as five minutes and potentially they're teasing that they will do one minute markets next as well as teasing some potential talk of a Polymarket token. Robert, I see you shaking your head. Robert, what's your take on five minute up or down markets?

Tarun (41:12)

Okay, five minute it's, it is what it is. But one minute really start to get into ruthless degenerate.

Haseeb (41:21)

Why is five minutes okay, but one minute is not?

Tarun (41:23)

Yeah, I don't know if five minutes is okay.

Haseeb (41:25)

I mean. Okay, here's the thing.

Tarun (41:28)

This is a great way to financialize anything using a prediction market structure. Right. I think the same way it can be used to replace binary options. Right. On equities which are extremely popular. Like one of the reasons why I think prediction markets at scale are truly phenomenal businesses. Right. I, I just don't think that a five minute market on the price of Bitcoin serves any predictive capability really. And it's just, it's just gambling, right? It's just gambling using the interface that people are used to and using the tools that have been built outside of the normal ways for people to speculate with potentially lots of leverage. And so it's fun. Is it healthy? No. Is it sustainable? Probably not. Is it zero sum or worse? Yes. Right. Are people going to get wrecked doing this? Yes. Are some people going to make a lot of money doing this? Yes. But you know, in my utopian vision of prediction markets, you know, I kind of want to see less of the sports gambling and less of the five minute.

Robert (42:35)

Robert and Vitalik. Vitalik wrote a long post this week

Tarun (42:39)

about this and more of the providing information to society that adds value in the aggregate.

Haseeb (42:48)

So okay, let me, let me, let me see if I can steel man the five minute markets because I do think, I do think there's fine arguments for that. Okay, yeah, let me. What do you think is the best argument for the five minute markets?

Robert (42:59)

Robert?

Tarun (43:01)

The best argument for the five minute markets is that there are better risk controls than gambling with 2050, 100x levered perps which is probably the substitute good for the consumer that's utilizing these.

Haseeb (43:18)

That was very sophisticated. I did not expect that. That's good. I was going to say that the five minute markets are subsidizing all of the positive externality markets. Right. So the reason why you can get a lot of liquidity and actually list right now, okay, you've got a market on the Fed, you've got the market on Iran, you've got the market on these other things. But you could have many more markets than you currently have. You could have markets on when is Taiwan. I guess that it does exist, but I don't know. Imagine smaller conflicts or smaller questions of geopolitical import that are not necessarily sexy and not necessarily front page news. How do you get the demand for those markets? The answer is that you gotta be a big business. That's how you get the demand for those markets. If you are small and you're like a niche like look, this is purely all things of intellectual interest to people who read Politico. That's just not going to survive. Right. That's basically your in trade or something.

Robert (44:12)

I think about Facebook Marketplace, you needed the farmville and minor league gambling apps to subsidize all the like infotainment useless apps that like basically people liked. Right, whatever. Like I guess like you know, if we went back to 2016 people would say there were misinformation but you know like they were, they were apps that people liked. And it's sort of like you can't make an open ecosystem without you know, both sides. Like it feels like there's almost just like a law of nature that like you have to have both. You can't like have one without the other.

Tom (44:50)

Yeah. I mean again I feel like to your point this is a different form of binary options that exist in traditional markets. So it feels weirdly inconsistent to say this is okay and this is not okay. And I agree it's risk controlled and so hey, if we're allowing people to buy lottery tickets or spin on slots, it feels again also just inconsistent to allowing them to not buy this or even buy zero day options, you know, which is obviously much more literally available in your, in your brokerage account.

Robert (45:24)

So the thing I, the only thing I was saying is like, if you have a platform, right, like take farmville generated all the income for Facebook for like their platform. If they didn't make money off farmville, they wouldn't have reinvested in the platform to make it easy to build your newsreader app and whatever other stuff you would get. Right. And they wouldn't have had this ecosystem if they didn't have these kind of like, they even had like slot machine apps. I remember in the early, early days of like, just like. But those apps made social funding, right?

Haseeb (45:52)

Yeah, those are a big thing on Facebook.

Robert (45:54)

So I just mean that like you, I feel like any platform that's like trying to be an ecosystem, you can't, there's like a yin and a yang thing. Like you can't. You have to have these apps that make a lot of money that are kind of, let's say less information containing and then you. But they subsidize the platform for the ones that are information containing but not really. You can't make a business off.

Haseeb (46:16)

Right, right. I mean, look, I don't think you have to contort yourself too much and say that, okay, well, therefore these are not gambling. Clearly if you're betting on 5 minute Bitcoin up or down, you are gambling.

Tarun (46:25)

It is not long term investing.

Haseeb (46:29)

It is not long term investing. Agreed. That said, there's obviously people also gambling on the fed funds rate. There's people who have just no alpha and they're just yoloing into something. There's people gambling on the presidency. They have no information whatsoever. So there are, there are markets where 100% of people are gambling. Or sorry, not 100%. I should say like you know, 95%. Obviously there's people who are actually Sharps who are betting on the bitcoin. Up or down. Yeah. Or hedging possibly, but mostly Sharps. And then.

Robert (46:56)

Okay, that's like, that's like what, that's what the Vitalik post was about how these markets should be used for hedging. Right.

Haseeb (47:04)

And again, yes, for five minutes only be markets that open. Only markets that can be used for hedging should be listed. I think was point bitcoin exposure.

Tom (47:14)

You know, you're an agent, you're about to get Some bitcoin, you're like, well, it's highly volatile.

Haseeb (47:20)

I'm holding Bitcoin only for five minutes before I send it along to buy this service.

Tarun (47:24)

In that case, wouldn't a perp be the better hedge?

Tom (47:28)

Well, I'm not interested. So I don't even know what model I'd be running. Depends where you're asking.

Robert (47:31)

One is fixed cost, one is fixed variable cost costs, but fixed cost for

Tarun (47:36)

an indeterminate amount of like, you know,

Robert (47:39)

no, you know, the expiry. Right. It's five minutes.

Tarun (47:41)

All right, I know, but, but, okay, so think about like this. If you have $100 of Bitcoin that you're getting, right? You would. Only if your life of it is within five minutes, could you hedge the whole thing. If the life of the bitcoin is past five minutes, which in every situation it's going to be, right. You'd have to do some partial hedge in order to hedge the delta of the whole.

Haseeb (48:02)

Robert, I think.

Robert (48:03)

I don't know. This. This agent. This agent. This agent might have a very short half life.

Haseeb (48:08)

Yeah, okay. We don't need it. We don't. Guys, we don't need. We don't need to argue this. It's not. It's not worth spending time.

Tom (48:13)

You can buy it for the next five minutes, by the way, you know.

Haseeb (48:15)

You can buy it for the next five minutes.

Robert (48:18)

Yeah.

Haseeb (48:19)

Okay. Amazing. Thank you, guys.

Robert (48:20)

Yeah. Look, look, look. I'm not going to try to say it's not gambling, but I'm trying to say I feel like that's.

Haseeb (48:27)

It's obviously, it's. Dude. Yes, we all want to defend.

Robert (48:30)

I want to defend these things at science.

Haseeb (48:32)

I'm okay with defending it, to be clear. I'm okay with defending it. I'm okay with defending it. But the defense, I think, is strongest when you say that it's subsidizing the unsexy markets that have these positive externalities around information. Right? But like Vitalik's proposal of you should only have markets that have informational value is basically saying, don't build a business. You know, it's basically saying, like, you know, build fucking gnosis. Whatever that prediction market was. I forget what it was called.

Tom (48:59)

Gnosis.

Haseeb (49:00)

Oh, yeah, it was. Yeah, yeah.

Tom (49:02)

That's actually the original Polymarket contracts are gnosis forks, believe it or not.

Haseeb (49:06)

Right, right. So basically, yeah, just build gnosis and who cares if it makes any money? And I think that's a formula for basically not doing what Polymarket did. Right. Polymarket actually incepted the concept in such a way that created many more positive externalities in the information because it was a successful business and because it caused people to shift their overton window. Because a lot of people now are trading the stupid stuff and that caused them to learn about the stuff that actually has positive externalities and has these flow through bleed effects of. Great. Now people are doing more of the stuff that matters. Right? So yeah, the people betting on sports.

Robert (49:44)

You reminded me of something very weird about this kind of prediction market versus gambling thing, which is for some reason in the US states seem to always own the right to gamble, right? Like they license the casinos, they license the lotto it. But then like crypto trading was always federal, right. There's never really like a state. Well, a New York DFS might be the only exception.

Haseeb (50:06)

Yeah, no, no. I mean you can. It's just a bad.

Robert (50:10)

But they, they don't do it anyway. They don't. And, and this current fight you're seeing in prediction markets, right, of like, oh, should the states regulate them or federal? And there's kind of like exactly in the middle of this, like what is the line that is like a state's right versus federal right for gambling? Like, because like if we call these different shades of gambling, there's like clearly some society is like fighting right now about like where the cut is. And to you, what is the answer?

Haseeb (50:35)

Okay, my understand, I'm obviously not a lawyer, but my understanding is that the rule is generally federal preemption is that when the, when there is a federal law, it basically overrides state laws. State laws cannot go in and modify the federal law. If there is no federal law, there is no nationwide injunction against gambling. So if you don't pass something like that, I mean prohibition maybe was the time when we had the closest to that. But if you don't have any such law, then basically it's up to the states. The states fill in the gaps. So there is no federal law. There was a federal law against sports betting. That federal law was overturned which allowed individual states to go in and start licensing sports betting in individual states. And that's why we have 30 states that license sports betting in 20 states where it's still not allowed. But before that there was no sports betting anywhere outside of casinos because that was the law, that there was a federal law against sports betting. So my understanding is that for general casino, general gambling there's no federal law. And so each state decides for itself. And with prediction markets, we're now in a genuine fight. Because so recently, Michael Selig, the new CFTC chair, he came out both in the Wall Street Journal and also, I think he went on some podcast where he basically came out and basically threw down the gauntlet, filed an amicus brief in many of the cases where states are suing polymarket, Kalshi and Coinbase's prediction market and basically coming out in favor of them and saying, hey, federal preemption, we agree, back off. This is our jurisdiction. You cannot come in and regulate prediction markets on a state by state basis. Now, unclear how this is going to play out because a lot of these cases are in state courts, not in federal courts. And so if you're a state judge, it's a different court system than a federal judge. Federal government. The government comes in and says, hey, step off. This is our jurisdiction. And you're like, hell no it ain't. I don't know. Unclear how this is going to play out. But obviously it's very good for prediction markets to have the CFTC throwing their weight around and saying, no, we totally agree with the prediction markets that they are federally regulated. This is our turf, step off. You do not get to decide on a state by state basis how federal regulation of event contracts works, which is the argument that all these prediction markets are making. We are event contracts, we're not sports betting. And the reason why they draw the distinction why it's not sports betting, why it's not gambling, is that you're not betting against the house. It's peer to peer. If it's peer to peer, it's different.

Robert (53:03)

Sorry, I understand the legal thing. I'm just kind of curious how society chose these arbitrary boundaries.

Haseeb (53:09)

Society doesn't choose man laws, like the

Robert (53:12)

laws choose the boundaries of this gambling stuff is always weird to me because then you go to other countries where it's like only federal or it's like purely only state, and you're like, you see very different structures.

Haseeb (53:24)

State thing doesn't really exist in other countries. Right. We're like a federation of states. It's very weird.

Robert (53:30)

I'm taking the EU regulation versus the individual countries as a proxy of this. Right. Whereas there everyone just listens to mother eu except for one exception.

Haseeb (53:39)

No, no. I mean, that's maybe true for economic stuff. It's not true for, you know, social issues.

Robert (53:44)

Certainly for stablecoin stuff, it seems to be everyone just falls and does whatever mother the mothership says. And I think that's going to be the same for the prediction market gambling stuff. So it's, it's just like interesting, like observation to see how this is breaking people's legal brains over this in different countries like that's what I love about watching this even though I'm sure it's expensive lawyer cost wise.

Haseeb (54:06)

Yeah, yeah, no doubt. Well this is going to continue to be a battle that gets waged and although the CFTC has come out in favor, I think this fight about is it gambling, is it sports betting, are these going to be directly competitive with other platforms that offer these kinds of services or are they going to have to stay in their lane and only offer these kind of informational contracts? We'll see. My guess is that it's going to be a fight fought state by state. It's not going to be one big mega battle and it's not going to get resolved soon so expect a lot more of this. All right, I think that's it from us for this week. We'll be back next time. Thank you everybody and good luck out there. Yeah, peace.