Hivemind：伊朗停火、AI是泡沫？以及Drift遭利用损失2.8亿美元

1. State of The Market

All right, welcome back everybody to the hivemind. I'm Kevin Kelly, one of the co-founders of Delphi, a long-time listener but first-time host host of the show. Really excited to be here with the with the usual cast of the Delphi hivemind. Quick announcement before we get into things. We've got a bunch to get into. Uh this will actually be the last and final episode of the hivemind on the Blockworks Empire channel. Going forward, you can find the hivemind on Delphi's own media channels like the Delphi Digital YouTube, Spotify, Apple podcast feeds. We'll try and get all those links added to the to the show notes. But if you're fan of the show, subscribe to those channels. Uh do it now so you don't miss a single episode. We've got a bunch that's planned for this show ahead. Got a couple new ones that we're going to be announcing soon. So you're definitely not going to want to miss that. Want to give a quick shout-out and a huge thanks to Blockworks for partnering with us to really get this show off the ground over the last year. We're big fans of their whole team. Oh them a lot of gratitude for everything they've done and uh looking forward to hopefully maybe doing something with them in the future. Now on to the main event. As I mentioned, joined by the usual hivemind suspects here. We've got Yan, our managing partner of Delphi Ventures, Jose, who is our head of Delphi Labs, Saterus, head of research at Delphi, and Jason, head of markets. Yan, it's great to be here. Excited to do this first one with you and what a day to do it on. How's everybody feeling? It's ceasefire day, at least for now.

Yeah, we've had 12 hours of ceasefire or something. Everyone's kind of on edge, it feels like. I'm pretty long into it, I think. Positioning is favorable to be long right now.

I'm not super optimistic on this all playing out smoothly from here. I think my base case is basically that you have a breakdown in negotiations, maybe before the 14-day window or at the end of it, but it just seems like the requests are too far apart and basically diametrically opposed. There are some elements that are just completely mutually exclusive.

There's always the world where things get delayed even more, and I think that would be pretty favorable for markets. The big concern is obviously this backlog of fertilizer, oil, and other shipments. So the longer you have to undo the damage, the better off everyone is, I think. I think that's kind of the rationale for the negotiations and why it was pretty favorable for the US, at least, to do that. I think it gives them a bit more leverage if the economy isn't being squeezed as much, or the global economy isn't being squeezed as much, so they're not feeling as much pressure.

Obviously, the pressure isn't isolated to them. You hear whispers of China getting involved on the Iran side to push for a resolution. It's kind of hard to say what's really happening, but it generally seems like there's an opportunity for some short-term optimism around world trade resuming, obviously with tolls and however else these tankers are getting through. You're hearing rumors of BTC or yuan being used, but overall it seems like the market's in a good spot.

I think positioning is honestly the big part, where most people, rightfully so, were kind of sidelined. There's too much unexpected turbulence, and there's also just a contingent that's happy to wait until it resolves, which is totally fair as well. So it doesn't seem like there's that much long positioning. The question is basically, will you get enough follow-through to make these markets move and make the payoff in the short term worthwhile?

Yeah, I think Yan's base case is kind of where I'm at. Obviously, this is very nice in the short term. Nobody wanted to see Power Plant Day and Bridge Day happen. That would not be good for markets or just anything generally.

But I just don't see how this actually resolves anything because, to Yan's point, both lists of things people want are just opposites. It just doesn't seem like the outcome that the US is looking for. So to me, I don't think it really changes much. I think the outlook is still really precarious.

Unless you're actively trading and really good at it and have a track record of trading volatile environments well, I think it probably just pays to be cautious and either play defense or just not play the game. I think there's still a lot of downside risk. I actually like that kind of recession report that Scott wrote 2 to 3 weeks ago. Markets are pretty much back to that level right now, right? We're a few percent off all-time highs even though things are as bad as they've ever been generally over there. So I'm obviously happy that we get a little bit of a relief bounce, but I'm very cautious over the next 2 weeks. I think things kind of break down further before they get better, in my opinion. I'm very cautious.

Yeah, Santiago, I definitely want to get your thoughts. Maybe another question: how is everybody roughly positioned or allocated right now? Because I feel like that tells a lot about what sentiment looks like.

Yeah, I'm not a good person to ask this because I'm always just long assets. I just—I don't know. Outside of Bitcoin and HYPE, for my crypto holdings that I'm just holding long term, I'm very much in cash. I'm very tactical, very, very much in cash, just trying to buy washouts from headlines or fade things, just very short-term stuff. I'm not—

Is that it? Just BTC and HYPE in crypto?

Yeah, that's all I own.

Yeah, and Zcash, but it's not big enough for me to care yet. Of course, it's up 30% in the last day, so I'm just like, all right.

Yeah, pretty much. Equities, I'm not touching equities right now. I think equities probably come down, so I'm just very, very cashed up.

Yeah, I didn't change my portfolio that much. I'm mostly long equities. I did take out some hedges before all this happened, or at the beginning of Operation Epic Fury. I'm probably break-even on those, or maybe even slightly down at this point. I have to check. But yeah, I'm mostly long equities. I'm long some crypto—HYPE, Bitcoin, Zcash, ENA, and I might be missing something—and PUMP, basically. Something like that. I've been buying more Bitcoin and started nibbling on alts recently, but I've been mostly buying Bitcoin over the last month or so.

In general, I think the macro situation is pretty good for Bitcoin. I've been saying this for a few episodes, but it's becoming clear now that the US isn't the world's dominant power. The world is more multipolar than ever, you know? I think Iran has just really clearly illustrated that. You kind of saw it with the Citrini piece, too, right? Which I know you want to jump into.

Before, everyone treated the US as someone you could rely on, right? They basically secured the world order. That's not the case anymore, right? Countries are basically flying out to Iran and making their own deals. China's still getting shipments through the Strait of Hormuz. I think a lot of countries are going to end up negotiating their own independent deals. Europe's making its own deals with China. And today we saw this headline that Iran's going to charge for the tolls in Bitcoin or yuan, which I think is just huge. To me, that's one of the biggest fundamental developments in Bitcoin in a while. I feel like we've all been waiting for this moment for—

I don't know if any real dollars are going to use Bitcoin.

Yeah, I think it's more of a headline thing.

Same. I think it's a really good headline, though. I just think this is what you need for Bitcoin to actually fulfill its purpose as a global, neutral store of value in this increasingly multipolar world.

2. Finding Opportunities In Crypto

Obviously, there are some quantum fears and some Michael Saylor fears as well. But overall, I'm pretty long still. I think when it comes to the crypto market, it honestly hasn't looked too bad over the past few weeks. And not just Bitcoin—Bitcoin has looked pretty good during the war, all things considered. Then there have been pockets of some alts doing well, right? TAO's done well. Monad recently, Zcash. Zcash is down a lot this year, but over the past month or so it's done well. HYPE obviously has done quite well this year. You've seen some real weakness in stuff like SOL. Obviously, the Drift hack didn't help that recently.

I also am wondering whether Kyle Samani leaving Multicoin is weighing on SOL. Multicoin may be getting a lot of LP redemptions, too, and SOL is such a big position for them that they're forced to sell it. SOL has been noticeably weak compared to other stuff.

There have also been a bunch of AI tokens, right? Zero has done pretty well this year. It's been healthier than it was during that period of just DCA down only for everything. People are starting to nibble at some things now, which I think is good. I'm pretty long. I have some BTC, and I'm deep in alts. You're unmuted, Jason.

I just laughed. I was like, “What are you deep in?” I'm curious.

Yeah, I started buying in late February. Those were decent buys. Then I bought some more in March, and I was up and then round-tripped a bunch. When Trump had that speech, I remember messaging Jason. I was like, “The market puked on it.” I was just like, “I don't think it's that bearish,” because the most notable part was that he gave a deadline, right?

I think that gives some element of certainty. He mentioned 2 to 3 weeks, and obviously there's no reason he might stick to it. But you don't throw that out with zero intention of trying to have some kind of approach and implementing or working around it. I thought that gave the market some certainty, so I loaded up a decent amount on that puke.

I also have some BTC. I think HYPE will do well as RWA becomes a bigger share of their L1. You see commodities, and I think there are a lot of positive headlines around price discovery and just the amount of volume they're having. Part of the idea is that these things will become everything exchanges, and I think this is a great step in that direction.

I bought some ENA, particularly after the announcement. I think it's insanely under-owned and super beaten down. USDe's supply has hit a floor, right? So now you have the potential for growth there. Do you want to jump in real quick?

Yeah, I was just saying, do you want to talk about what that announcement was?

Yeah, basically, they're looking to diversify outside of the carry trade when it comes to yield. They're already diversified outside of the carry trade, mostly, but they're trying to add additional sources of yield. Now it's prime brokerage and non-Treasury-yielding assets, so it could be other forms of debt.

I think it's a business development plus an ability to find creative opportunities. I think that team and Guy are really good at that, as evidenced by how quickly and deeply they became integrated into every exchange and custody provider.

On the AI front, those have been pretty interesting. You have to look at which teams are still building, right? There aren't many, and if you can focus on those, it becomes a bit more clear. I guess that's the benefit of bear markets that lead to attrition: the ones that stand out become clearer bets.

I like NEAR, Grass, and Venice. Venice is probably one of the more interesting ones, just in terms of its size relative to the amount of growth that it's had. You kind of have to piece together the growth based on Erik's tweets and other points, but it seems like it's been through the roof.

I think it'll pick up even more with Claude's adjustment away from allowing people to use OpenClaw with subscriptions. You're going to dig into more of these frontier models, because paying for APIs is pretty cost-prohibitive for a lot of people. For me, it's basically a combination of exchanges, ENA, and the AI coins.

Speaking of AI, Anthropic's announcement around Claude pushing people to the API instead of the subscription—it's hard to see how that isn't going to be extremely bullish. A lot of that was because it was very compute-based, right? They're basically getting strapped for compute capacity.

So how is that not potentially one of the next big legs? You have all this talk about the AI bubble, and one of the things I wanted to get your take on is the disconnect we're seeing between public and private markets. I know especially at Zions[?], where you guys spend a bunch of your time, it seems like some of this is Iran-war volatility in public markets, which just weighs down big tech, heavy names, the Mag 7.

But on the private side, you've still got crazy valuations and fundraising rounds that seem to only be accelerating. I don't know. This concept of an AI bubble, I've never really believed in it, or that we're anywhere close to the top of it. I'm curious to get your take on where you see this going forward.

Is the Claude-type news bullish? Does it reinforce that there's just so much demand for compute, that this infrastructure has yet to be built out and this has to continue? Any insights you're seeing on the private side? Underwriting AI right now is definitely a challenge.

One bubbly component is that you have your SpaceX IPO later this year, and then you're going to have—it's this kind of ticking time bomb, right? The SpaceX one is just going to be monstrous, and there's only so much dry powder out there. You'll probably have SpaceX, Anthropic, and OpenAI going, and after that it becomes a bit trickier.

There are 2 ways it can go. You can have things continue to progress, revenue continues to ramp, and these things start doing really well. That enhances their ability to borrow cheaply and sell stock or raise debt, and the flywheel keeps going.

You also have this world where whatever happens in the broader economy causes a bit of a downturn, and then you have all of this new liquidity coming. There's going to be so much selling of this stock from insiders because it's all institutions that are just going to sell, and it makes sense. There's a world where some of this stuff trades down, and that hinders the ability to keep funding the growth.

I'm not sure how that plays out, but you can make an argument both ways.

Yeah, it's a race between capital markets being able to fund the build-out that is necessary and whether that capital is going to be there. These companies are probably going to launch, or IPO, with relatively low floats, because they're going to try to juice at least their early-day returns.

At this point in the private market, investors are looking for some type of exit, and these companies are too big to get acquired. Most of them are going to have to go public. You have to see how the public markets react, because IPOs still largely haven't done well.

A lot of the ones that have gone public in the last 12 to 18 months are down. You've even seen some big AI companies potentially looking at discounts on their latest rounds just to go public, just to give investors liquidity.

I'm not sure how that's going to pan out, but on the AI side, there's still a ton of stuff we're looking at on the private side that's interesting: your hyperscalers and your infrastructure side. And, Santi, I think you've become more bearish on the software application layer. Is that still the case?

Yeah, on the public markets, there's definitely a big difference between how private markets are pricing things and public markets. Obviously, the go-to example people use is Stripe and Adyen, but there are a bunch of other examples, like Databricks. Databricks and Snowflake are basically the best direct comps. They're both around $5 billion in ARR, but one is trading at around $150 billion and the other is trading at around $50 billion.

There are many examples like this on private markets, like Chime and SoFi and a bunch of these companies, where they're basically trading anywhere from half to lower than on public markets. It's a weird one. It's kind of like what happens in crypto, too, where the private marks are way higher than what can be sustained on public markets, and it just takes a while to adjust.

It's going to be interesting to see the SpaceX IPO because, like you said, there are a lot of people who've been holding that stock for a long time. They've raised a lot of money—I think $12 billion overall. It's hard to know how much insiders or how much investors own, but I think Elon owns 42% of the company still.

Employees have had a bunch of tender offers, but there's clearly a lot of investment capital in there. It's going to be hard to find—you need a lot of buyers, right?—a few hundred billion, really, of net buyers in public markets to offset that.

Obviously, SpaceX is an insane company, but it's doing about $20 billion in revenue. Yeah, I think it's like $15 billion or something.

Yeah, it's like Meta's trading at $1.5 trillion, right? And it does like $200 billion in revenue. It's kind of silly. It's literally trading at 100× revenue. I don't know. A lot of the private-market growth rounds are pricing in these multiples, but it's way more expensive than Anthropic or OpenAI on a revenue-multiple basis. So, yeah, I don't know. It's going to be interesting to see how it goes.

SpaceX, yeah. And on the application side—sorry, do you want to jump in?

I was just going to say, I agree. I think public markets are sometimes the ultimate arbiter. It seems like, when it comes to AI, public markets are very focused on asking questions around ROI and “show me the money today,” in a sense, whereas a lot of these private rounds are based on, to your point, future growth and growing into these valuations.

Again, if you looked at OpenAI 2 or 3 years ago, a lot of people would have said it never would have gotten to maybe even $20 billion or $25 billion in ARR, right? So it's a little bit more subjective, I guess you'd say. You can't really plug in a DCF; your models are only as good as your assumptions. Whereas once you go public, public-market investors and institutions have a very different mindset. It's almost like guilty until proven innocent right now.

Yeah, I think there's a high probability SpaceX trades below its recent valuations. The last round—you get a puke. Wow, that's good. The last round was $1 trillion, I think, so that would be—and they're going out at $1.7 trillion.

I guess below the IPO round, then. And—not, sorry, not below the last round; it trades down from the IPO for sure. If you get any kind of market downturn, that's going to be one of the earlier ones that gets puked, I think.

I mean, yeah, I just—the gall to IPO at a $2 trillion valuation is just... But the question is: Is this the new normal? Again, 5 years ago, Nvidia becoming the world's most valuable company on the planet at $4 trillion was just an insane, almost asinine idea, but here we are. Did you see that Nasdaq has been under some pressure to change the rules of index inclusion to basically allow companies to get listed—to get included—sooner? A lot of people are saying the idea is basically to get SpaceX included sooner so that passive index buyers can buy private investors' bags or whatever.

All right, so when it goes public on the NYSE too, the NYSE is incentivized to have this thing do well.

And then, I mean, there was a lot of—you were gesturing at that with the Fundrise Innovation Fund, right? The VCX ticker—that's not technically public because it's a fund, but it was basically a launch to the public. What was that, last week or 2 weeks ago? It traded—I mean, its NAV is like $19. It holds Anthropic, SpaceX, private shares of all these companies, but its NAV is like $19, and it wound up trading up to like $550, $560. People pointed to that as, like, here's so much pent-up retail demand to get access to these names, because they can't really get access to them in any other way.

I don't know. I just think there's a very big uphill battle to climb, and it's almost like: The companies that go public first—are they the ones that benefit most because the dry powder that's there wants to get exposure to these names? Or does it make more sense to just wait and see if you're in the queue to go public? I don't know. You probably want to go public first, for sure.

I mean, no, for sure. There's just—it's the craziness roller coaster. Yeah, on the application side, you were going to say something. Is there anything interesting there?

No, I just—well, I've never changed my mind so often as I have recently with investment, in the last 1 or 2 years. I think I kind of said this last time. It just feels like if you change your mind, you have a chance to be right, but if you don't change your mind, you're almost definitely wrong, just based on how fast things are moving.

There was a moment when I was very bullish on verticalized software. I talked about it on this podcast as well. The idea was that AI is way more powerful than what people are using it for, and in order to unlock that power, what's really missing is context: being able to have data pipelines to input your specific context, and then a good UI to unlock that and make it easy for people. We would have those across all sectors.

I'm not sure anymore, basically. I don't even know if we have UIs if agents get good enough. Maybe you just say what you want and it happens. If you need a UI to see the data, it just gets generated for you.

One thing I'm pretty sure about is that there's more alpha than ever in public markets. I think for people who are paying attention in specific sectors and things like this, there's just never been more alpha, because for roughly 2 decades it was like nothing ever happens, right? The big just got bigger. The Magnificent 7 just carried the S&P, and passive investors did really well.

I think now there's so much happening, and it's so hard to play out all the consequences of it, that some of these active managers—because all the SaaS names got pretty much indiscriminately dumped 50% to 60%, or even more for some names—are definitely going to find value there. But I also think in energy, there are people working out the supply chain for AI chips and electrification. The people who got into the memory stocks early—I met a kid yesterday who thinks it's actually... Anyway, there are all sorts of different theses that people are playing with here. I think it's a more interesting game than it ever was before, and really hard to play part-time because things move and reprice so quickly.

I think this mindset and the increase in active investors is just going to make passive that much better. Because passive is basically coattailing on the amount of active-investor capital that's in the market. It's just going to make it even more beneficial. Passive investing has always been this free lunch. It's basically the only free lunch in the world, and I think it's only going to—

Coattailing? I think the opposite: active managers coattail on how much passive capital there is that's basically not making investment decisions. Passive just tails active. But passive tails flows more than active. Passive is just—there's money constantly bidding every 2 weeks, so I think that's what benefits passive.

On the other hand, if you look, the S&P hasn't really done that much, but there's been an insane amount of movement under the hood, which I think is more of an argument for active over passive.

Yeah, I think the S&P will do well because AI is, you know, net positive.

3. Is AI In a Bubble?

Yeah, I mean, most people should just own the index, right? If you're not doing it full-time, you're not going to be able to do it. Most people should just own the index. But I agree with you. I think we kind of mentioned that in The Year Ahead, that we thought, especially within crypto, but more broadly speaking, an active, stock-pickers' market would probably be the ideal way to treat the market going forward, given all the opportunities.

4. The Return of Active Investing

Obviously, we didn't have a lot of the things that have happened between then and now happen, but it only builds on that case. So, yeah, I fully agree. I just don't think a lot of people can play the active game well. That's the only thing. I think it's very hard to play it well.

Aren't you kind of contradicting yourself then?

No. I think the good active managers will do very well over the next 12 to 24 months relative to how they've done over the past 5 to 10 years. But I don't think there will be that many, because by definition, outperforming the market is hard and most people don't do it, right? So for most people, just own the index. If you're not doing it full-time. And if you are doing it full-time, it's still hard, but it's as good of a market as any to do it.

Yeah, I think the comparison would just be the spread between active hedge funds and passive 401(k) investors, basically. Whereas before, it wasn't—I’d say right now the spread is wider, where hedge funds can outperform just passive 401(k) peeps versus what it was before, when you just had this rising tide. Basically, if you missed a name, you really underperformed. And so, by being an index investor, you never missed the Apples of the world.

Yeah. And one way to frame what you said before is that there's probably going to be much bigger dispersion in the underlying makeup of the S&P 500. If before everything mostly went up, and maybe the Magnificent 7 went up more, I think now you're going to have some things get absolutely destroyed and other things come out of nowhere. I think that dispersion is where you can capture the big performances as a long-short manager right now in a way that you just haven't been able to in the last 2 decades.

So, yeah, I would take that bet. Buffett made that bet with whoever it was, an active manager, in 2000, right? The S&P would outperform this basket of active managers, and it did. I think now's a better time for active managers, basically.

Yeah, in particular, when these big guys IPO, there's going to be some fresh capital, but I think a lot of it is just going to be rotation, right? Which names are going to get decimated in that situation? I don't know what the ratio of fresh capital to rotation is, but I do imagine there's going to be quite a bit of it because it's kind of a reallocation of exposure from institutional managers saying, “I still want to be long this sector, this weight. In order to get exposure to that, I have to rebalance into something else.”

Then you layer in who the biggest losers from the growth of these other firms are, and you're going to get some really big volatility around those IPOs. By the way, on the topic of private investments and IPOing late and everything, did you guys see this yesterday? I don't know the type of liquidity that there is on this stuff, but I was pretty surprised. Someone was able to buy $500,000 of Anthropic on Solana, on-chain, a few months ago.

I've been pretty skeptical of these pre-stock things because a lot of times they're kind of just synthetics. They're like cash forwards and all this. Is this a spot thing?

Yeah. Yeah, it's by PreStocks. I need to look into them more because I don't understand how this guy was able to buy $500,000. It must all be direct with the issuer, because I don't understand how—

Not direct with the issuer. No, no, not with Anthropic—with PreStocks or whoever's tokenizing these, right? I spoke to someone who's got an interesting design for this. It's actually based on the original FTX design from back in the day.

FTX—I don't know if you remember—listed Coinbase and Robinhood back in the day. Apparently, it was this Swiss law firm that they ended up acquiring that had this super interesting design. It's very simple. It's like a trust. You put the asset in it, and then you can sort of—it's a trust that's cash-settled on IPO or exit. You put the asset in it, and then you can issue—

Yeah, I guess I was just kind of shocked. I knew these things existed over the last few months, but I didn't realize that because most of them are like, “Oh, you can buy Tesla xStock,” and you buy $1,000 and move the price 2%, right? I don't know. It's an interesting space to watch, especially as there's more and more tokenization over the next few years. I was pretty shocked he was able to get that size.

Because even moving some of these, it's messy, right? These vehicles have carry fees, broker fees, all of these elements, and so it makes even secondary-market transactions pretty brutal. I'm impressed that that worked.

Yeah. I got to look into it more, but I saw some people whose opinions I respect tweeting that it was legitimate and stuff. I definitely think this is going to happen, though. There's going to be some way to bring private stocks—private markets—on-chain. I think it's kind of a perfect fit. I just don't know what the winning design is going to be.

I mean, just look at the explosion in secondary sales. That's the market right now.

Yeah, that plus if some of these IPOs go south, everyone who's in late-stage privates is like, “Holy—how do we run for the door?”

That's the thing, right? The private market is just so big right now. It's also going to be interesting how they get around this, because for certain companies and certain private investments, there's a good amount of secondary supply on basically any of these platforms. But then there are others that are arguably the most highly coveted, where the teams themselves put restrictions on secondary sales or want to know who's on the cap table.

I don't know how you get around that if you don't have the team's sign-off or if the documents are structured that way.

SPVs.

Right. So you're on the cap table because you're approved, and then you're basically just tokenizing or creating a new—

The cap table, right? And I'm just selling that thing around. Then somebody else sells to somebody else.

Or your SPV is on the cap table of an SPV or something—the cap table of another SPV, which hopefully is on the cap table of the underlying. The Russian dolls.

Which hopefully the underlying is somewhere in one of those.

Yeah, exactly. That's going to be really interesting with SpaceX, too. I think a lot of people are going to realize that these things did not have access to the underlying, or that you were paying so many layers of fees that you were basically making way less money than you thought on some of these things.

Yeah, I know a buddy who was in Anthropic at, I think, the $18 million round, and he recently sold. After all the fees, it was like an 8X.

Well, what would it—well, and, yeah, wow. $18 million. There was the big round that came in, and there was 50% dilution. The round after that was the big one. I forgot who led that, but that was basically a 50% dilution round. So there's that, and then there are the fees. These things are super capital-intensive, so you're getting diluted pretty heavily each round.

5. Prediction Markets

Yeah, it's a tough world when you're getting disappointed by an 8X, too—on what could have been.

I also saw an interesting chart, kind of going back to disconnected markets, around prediction markets. We can get into some of the Polymarket stuff because I know they had a couple of announcements this week. It's just been interesting to see the rise of prediction markets.

I saw a great chart that overlaid the volumes of prediction markets with the volumes in crypto, and they were just the inverse of one another, right? Prediction markets were just stealing a bunch of attention. I still feel like the crypto market suffers from that. We talked about some names that have done decently well, especially in the last 3 or 4 weeks, but I don't know. I feel like the attention is still shifted elsewhere.

The attention from prediction markets?

The attention just in general within crypto. Bitcoin has its own narrative around it, but I don't know. I think prediction markets are going to be an interesting one we can dig into because I know they've had some news this week.

Polymarket itself seems like it's had a week. It took down some of the markets that were tracking the fate of U.S. pilots who were downed. They had one that resolved “Yes” to “boots on the ground,” which, Santi, I know you've had some thoughts around, or some debate around, the usefulness of these things.

Yeah, I think prediction markets have a lot of work to do to be useful. A lot of times, you get these markets that resolve around technicalities and wording. I'm stealing some of this from a Matt Levine column the other day, but I totally agreed with it.

The “boots on the ground” market—the intention of that market is, “Is the U.S. going to send troops to Iran for an operation, as an attack?” The intention of that market is not that they're going to go rescue a pilot. That's some outlier technicality.

If you're really good at prediction markets right now, they're in this phase where you've had so many markets resolve around these weird nuances that get away from the point the market is really supposed to be about. If you're somebody who spends a lot of time figuring out these nuances, you can make a lot of money in this, but it's not useful, right?

If the whole selling feature of prediction markets is that they make information more accurate and surface information, then you do want markets that are measured by their true intention.

I'll take the other side of that, though. I think you need to remove subjectivity; otherwise, it can get very messy. If you take every market literally and treat that as a feature, not a bug, then it becomes more about the prompts and the conditions for the market, so you don't repeat previous mistakes.

The next one will say, “It doesn't matter if somebody gets shot down; it's not boots on the ground.” You become more specific about what triggers the parameters.

And so, there's kind of a learning curve there, but I think you want to remove subjectivity. At the same time, these guys just rely on UMA, which has been less than honest sometimes with manipulation and so on.

Well, yeah, minus the UMA thing, right? I mean, there really isn't a solution other than what Jason said, right? Just make a better market that's more explicit about what the resolution criteria are.

Well, no, that's what I think they should do. I think they've been sloppy with the way they've talked about designing and measuring these markets. In a lot of markets, you look at the measuring criterion, and it's kind of weak. They have a lot of work to do in that respect.

You're going to have this come up a lot until they get better at it, because there's obviously money to be made here in figuring out what the flaw in the market is, right? What do you think of the thing they dropped? Polymarket did the exchange upgrade thing with their own Polymarket USD. What do you think about that, Santi?

I don't remember that. What was it? Didn't they announce something like that?

Exchange upgrade, I think.

Yeah, they're going to have a native stablecoin, Polymarket USD, that kind of thing. Is it yield-bearing? Is that part of it, so you don't have the opportunity cost?

The other way around: they internalize the yield.

Yeah. Oh, really? [Laughter] When you don't charge fees, you have to make some money. But it kind of means everything is now directly controlled by Polymarket.

Everybody, I mean, every successful app—I know Ethena has done this with MegaETH, right? A lot of these people come to the conclusion that the good, solid fundamental revenue in crypto is the treasury yields behind the stablecoins.

If you're an application and you're thinking, “Look at all these stablecoins I have on my platform. I should get paid for that.” Solana, for instance, has $15 billion, and I don't know, half or more than half of that is USDC. They don't even freeze $200 million when there's an exploit, and they get nothing for that. Why? What is that relationship getting Solana? They have an issuer who shares no economics and leaves their DeFi ecosystem out to dry when they have the ability to stop the damage.

I think this is a point that a lot of blockchains and applications in general are just going to get into more and more.

Yeah, it'll be interesting to see what the end game is on that. I think we talked about this Nic Carter post a while ago, but I thought it was really interesting. He said that it used to be that everyone thought a few giant stablecoins would win because of liquidity and network effects, but his argument was that maybe the end game is actually a bunch of different stablecoins.

His arguments were that the float is too valuable to give away, and everyone's going to want to launch their own stablecoin—every app, like we've seen with Polymarket and others. Launching your own stablecoin has never been easier, either. Before, it was super hard, but now you have all these white-label providers that let you do it very easily.

Users might not care what stablecoin is underneath because the front ends will abstract it away as just USD, right? It doesn't really matter. Swaps and interoperability are getting so cheap that the network effects aren't as big as they used to be.

With yield sharing, there's a lot more incentive for these things to be adopted, and we're kind of seeing it. Ethena did really well with MegaETH and a bunch of others. It's an interesting argument, I think. I still think there's a lot of—

Now, I'm curious what you think, Jason, actually.

Yeah, I was going to say, when your app doesn't really need the network effects of your stablecoin, when all the activity is contained to it, and you're able to do it on the back end—“Deposit your USDC or USDT, and we'll just convert it on the back end”—then there are really few trade-offs for the app to use an existing one. They might as well just go with their own.

Yeah, I'm very much in the same boat. I think the alternative is basically when you need the network effects, which is becoming a smaller surface area. That's when you'd rationalize not doing it.

So, yeah, it's kind of bearish for Tether and USDC, right? Maybe there's not really a way to express this bet other than on the applications themselves, or if Ethena or another white-label provider ends up being very dominant here.

Tether might be insulated because it kind of exists as this monster way to transact. You're never going to see Polymarket USD or any of these other specific stablecoins take over Tether's primary utility, for example. But USDC relies more on these integrations than Tether does, at least, and I do think that definitely hurts its business.

From a user standpoint, though, wouldn't you still make the argument that network effects matter if interoperability and swaps become essentially frictionless on the back end? You're going to default to the stablecoin where you can get the highest yield if it's just sitting idle, even if it's within an application or an ecosystem.

I don't know. I'm trying to square that with the dispersion of many, many stablecoins versus just a small handful.

Yeah, most stablecoins don't share yield. You bucket them into stablecoins that share yield and stablecoins that are just lubricant for the app. If they're lubricant for the app, then you can internalize it because they're not there for the yield. If it's specifically a yield product, then I think that's where the demand for the stablecoin comes from.

Yeah, it's a good way to put it. It's a really good way to put it. I do think being the highest-yielding one—as much as possible, you want to be sitting on one of the vertices, right? You want to be the most liquid, have the highest yield, or have some other differentiator, because everyone in the middle is just going to get eaten and internalized by the applications.

I do think the changes they made are important there. I don't know if you guys noticed, but Sky's USDS supply recently—has anyone checked the DAI supply? It's $12 billion.

What's that?

$12 billion. I've dug into the loop, and I just haven't put enough time into fully understanding how it works, but it seems like they can just choose—I think it was $21 billion or some arbitrary amount—that they can mint. Then they have this kind of tiered security apparatus behind it.

With Maker?

Yeah, and with Spark and all of that. I think it's just been up since then. It's sort of absorbed most of Ethena because they had better bear-market yield sources. They'd already integrated with Apollo's credit funds and all the things they had already integrated with.

Their USDS base yield is 3.75%, but then they've got higher yields if you stake it and stuff. They have three different yield sources. That was enough to absorb what seems like a lot of the USDe supply. It's kind of what USDS did to DAI initially, right? When it went live and had a 60% yield.

It's also interesting that USDe seems to have this floor at around $5 billion, even though the yield is lower than what you could get with USDS and even Treasuries. Stablecoins that have been in the market for a while and have had a lot of holders seem to have this huge advantage: There just seems to be forgotten supply that's willing to accept lower yield.

I'm one of them, honestly. There are people who just won't get rid of their USDe for whatever reason—for laziness or something like that. But it's a huge advantage going forward because it means the yield for stakers is inherently higher. It's inherently leveraged.

Yeah, that's why I think if the changes at Maker are enough to raise the yield by 100 basis points, it'll be enough to steal back a bunch of supply from Sky.

Yeah, I saw another element where they were talking about how there's a decent amount of DAI that's lost. That's effectively free buffer for the protocol, since that collateral will never be redeemed.

Yeah, I haven't even looked at DAI supply. Wow, there's still $4.4 billion of DAI, too.

So I guess cumulatively, it's like—assuming I'm right—it's 16 billion or something. Which, yeah, I think that's the biggest decentralized stablecoin ever at this point. Or not? I don't know if you can call Maker decentralized anymore.

6. The Drift Exploit Fallout

All right, so we've obviously covered a ton of stuff. I want to end this with a final question for all you guys. We've talked about a bunch, but I'm always curious because I chat with you guys about this stuff a bunch: for everybody out there, is there anything we haven't talked about that you really want to hit on? Or I'm curious what has got all of you nerd-sniped right now?

I think I'm not nerd-sniped, but I'm reassessing risk after the Drift thing. I'm kind of nerd-sniped, I guess, by these hacks—the way that they're doing them now. The Drift exploit was a 6-month, in-person operation where North Korea hired an external team. They were quant traders, and they became close with the Drift team over 6 months. They made a vault on Drift that people could deposit in and eventually made a TestFlight app, got the Drift team to download it, and did all this other stuff. It was super sophisticated, right?

You're seeing these social-engineering attacks happen way more than smart-contract hacks. In the past, hacks were always smart-contract hacks—just some shitty code, right? But the code has honestly gotten pretty good for a lot of DeFi because it's been around so long and it's been so battle-tested. This hack, the Bybit hack, and the Radiant hack—all of these were social engineering, right?

My main takeaway, first of all, was that when this happened, I withdrew everything from any Solana DeFi protocol because it's an admin-key thing, right? It got me thinking: do I really want to have money in these protocols with admin keys, where we've kind of just seen this complete rug? One thing I think we need to do better as an industry is—I don't want to say—I mean, the Drift team needed to be better in this.

If you're on a multisig, the Drift team's mistake was that they thought they had cold-signing keys for the multisig. They used hardware wallets, but they were signing on their main computers, right? If you're using a hardware wallet on your main computer, that's not a cold key anymore because your machine is hot, right? Individuals should be using separate machines to sign transactions.

That might be overkill for a lot of individuals, but if you're in a multisig on a 200-million-plus TVL protocol and you're not using a separate machine to do the signing, where you don't download anything on it, to me that's just lazy. I think there's a lot of laziness when it comes to this. We've seen 3 insane hacks now—with Bybit, Radiant, and Drift—that are all of a similar variety, dealing with admin keys.

And then what? Anthropic released a model yesterday that can just find all these vulnerabilities, too. So that goes back to smart contracts. Maybe the smart contracts aren't as safe as they were, right? I don't know. The day after the exploit, I wrote in the Delphi internal chat—and this is a chat with people who've been around forever. We all know how to use DeFi, and we're all pretty sophisticated with everything.

The first 10 respondents—I said, "Do you have 25% or more of your crypto portfolio in DeFi?" Maybe that number was too high, but the first 10 respondents were all no, right? These are people who work in the industry, and we don't even have a quarter of our capital in DeFi protocols. Some people came in afterward and said yes, but with all the points programs that have been enticing people for years and nothing's come out of them, you're taking all this risk. I don't know. Yan, you want to say something?

No, no, I was going to say I fully agree with what you're saying. The admin-key stuff is one of those things where, the second it happens to one of these, everyone should have their ears perk up and make some adjustments. If this happens to a second one, it's even more nuts that it happens. I guess the main hope is just that adjustments are made on a go-forward basis.

On the share in DeFi, I wonder how much of that is risk versus just a lack of compelling things to do at the moment.

I also think DeFi has kind of been hijacked. The term DeFi has been hijacked—I don't know who tweeted this—but something with an admin key, or even a multisig, is explicitly not DeFi, you know? Or at least it's not the way DeFi is supposed to be.

Yeah, it's on-chain.

It's on-chain. It's supposed to be noncustodial, permissionless, et cetera. I just think that—and no hate to the Drift team, because I actually think Cindy crushes it—the risk profile for these CeDeFi things, where you're using a DeFi protocol with an admin key, is actually the worst of both worlds.

You have smart-contract exploits and the slowness of being able to react on-chain with a multisig, but you also have the risks of key management for admin keys and most of the risks you'd have with centralized platforms. I really think it's the worst of both worlds. It's a problem that DeFi has been hijacked so much, because I wouldn't consider either of these hacks DeFi hacks. Both of these things were basically custodial, right? Resolv and Drift. So, yeah, it's unfortunate that it ends up spilling over to the rest of the DeFi brand.

The last DeFi one was Balancer, right?

Yeah. What was that? I don't even remember what the hack was, actually. I don't know, but it was in their—I think it might have been their V2 contracts, which have been around for so long.

That's the scariest stuff, right? It's just this honeypot that's been there. Then that ties into Claude's models getting released—or, I mean, the release of what it can do and what that means for a lot of these long-standing, large honeypot setups.

Dude, the doomsday scenario is a bug in the actual L1, right? You imagine undercover North Korea operating state-of-the-art Ledger this whole time.

Claude will find the bugs.

On a more positive note, something I'm not nerd-sniped by, but clearly HYPE being one of my only positions is something I'm clearly invested in. Obviously, everybody's talking about prediction markets a lot—clearly one of the winning use cases of the cycle.

HIP-4 is in testnet, which is effectively them launching prediction markets, unified spot-margin accounts, and everything like that through HIP-4. It's their first real movement beyond perps—perp trading and just trading as we know it. I'm curious to see how it plays out, whether there's actual traction there, and whether they steal away any market share, especially on the crypto-native side of things, from Polymarket or Kalshi.

To bring the vibe back down, how do you feel as a HYPE holder about the key-management stuff? I think HYPE is one of the things that I wouldn't really consider DeFi, honestly.

No, no, I agree with you. That's always been the big risk, right? Even since HYPE was super new and before it had all of this growth and success, it's always been the big risk.

Hopefully my HYPE isn't all in HYPE, right? I have some of it in my wallet, but if they steal a bunch of money or something happens, clearly it's going to go down a lot. I just have to trust that they're doing the right things and hopefully learning from all of these mistakes. I believe Tay, who was super critical initially, has been working with them over the last year, so maybe that's good. I don't know. I have no insight into it.

I saw something on Twitter that they have some migration going on, but I don't know how true that is. I need to look into it more. They might be updating the bridge a little bit to get away from the current structure that they have. But, yeah, Jason, that's clearly the biggest risk. Everything else is going great, right? That would be by far the biggest risk.

How do you guys feel about LIT, or Lighter, here? I got shilled and ended up buying a bag while I was in China, but I'm curious how you guys are feeling about it.

I'm bullish. I think you're going to have—I mean, it's trading a little cheaper than HYPE, even with the FDV adjustments for investors.

A bit cheaper. It depends on how you adjust the FDV. If you don't adjust it, then, yeah. But if you adjust HYPE's and then do all that—

What are you adjusting HYPE for—the supply and treasury, or—

To normalize the two, basically.

HYPE—you don't have the investors. There's an ecosystem; there are no investors who are going to sell. So you kind of do apples to apples.

Yeah, the rest. And just what share of overall supply is investor versus community, kind of thing. Stuff that's been pretty broadly talked about. So the gap converges. It's still there. It seems like they're going down the licensing route and just deeper U.S. integrations, which should be bullish. I think there's a lot of open space on the RWA front, and I think you're seeing growth on both of those. So, I mean, I'm buying the number 2.

When has buying the number 2 in crypto that looks cheap ever worked out? That was my main worry. I do think it's differentiated in a way that not many people talk about. The way they built it with ZK, I think it actually has an eject button, although I don't know if it's working right now. I think if it was, they'd probably make more of a big deal about it. It also allows them to do cross-collateral more easily, apparently. But it sounds like HIP-4 is going to bring that out anyway. I don't know. On the AI front, did you guys see that viral tweet that went around when they basically listed every protocol that had North Korean involvement? Like, everything.

[Laughter.]

Well, did you see yesterday? There's this protocol on Solana called STABLE—not “stable.” It's spelled S-T-A-B-L-E. They were tweeting yesterday, “Everybody withdraw your liquidity ASAP.” They were quote-tweeting a ZachXBT tweet saying that this guy was a North Korean developer who had worked at protocols. They were tweeting, “Withdraw. This guy worked here.”

But the thing is, it's so hard in this industry: that tweet itself could have actually been a hack. They could have hijacked the front end of STABLE, gotten everybody to panic-withdraw from the protocol, and then that was the hack, right? That's why this is so hard. Did you see the other viral tweet of this interview where they're like, “We found a way to detect if the candidate is from North Korea”? They basically ask him to publicly defame Kim, and he says he's from North Korea.

No, because if people think—listen, that's the long game that they're going to play. There's no way you can trust that process, man. They're just going to be like, “You can denounce him,” and then people will get a false sense of security that they're not from North Korea. There's no way that's going to work.

A recording of them denouncing him—hack North Korea and play it on the TVs there. Boom. Checkmate. Yeah, and do that for every single one. [Laughter.] You’ve got to fight fire with fire, bro. Come on. This—I’m going to share my screen. It was so ridiculous. I was like, “What is going on?”

Emergency, guys. Please temporarily withdraw your liquidity instantly. Better safe than sorry. The new STABLE team. They spelled “temporarily” the way they spell “STABLE.”

[Laughter.]

But then, basically, yeah. This is the worst crisis comms I've ever seen.

Basically, this is what they retweeted.

[Laughter.]

Panic now. You need to panic. They quote-tweeted this: “Better safe than sorry.” I don't know. I was just like, it's so funny—the team itself saying that: “Better safe than sorry.” It seems we had one a year ago. We have a new team that took over 4 weeks ago. So that's good: a new team. Our values: “Better safe than sorry.”

[Laughter.]

So, I don't know. Anyways,

Funny. Good times. Always entertaining. Listen, crypto. Yeah, dude. Never a down week. Well, 2 weeks from when we do the next one of these, it'll be right after the ceasefire ends. So that'll be—I mean, we're going to have fireworks along the way, I think.

Global peace. Yeah, we'll have global peace the next time we all hop on, which is super exciting. Something to look forward to. I really appreciate you guys, as always, jumping on and doing this.