On The Brink with Castle Island - Weekly Roundup 12/05/25 (The trouble with Taproot, a prediction market paradox, HFSC OCP.2 report) (EP.689)

Episode Date: December 5, 2025

Matt and Nic are back with another week of news and deals. In this episode:  Quantum risk again Are banks filled with spooks? The trouble with Taproot addresses Microstrategy's dollar reserve Kalshi is going on chain Prediction market predicaments Are insiders monetizing Google search information on Polymarket? Are prediction markets insider trading machines by definition? DAT activist fights Vanguard adds Bitcoin ETFs House Financial Services Committee releases their final report on OCP2.0 What's going on with leverage in crypto? Are AI agents going to start stealing crypto? Is Eliezer Yudkowski to be trusted? AI is going to usher in the golden age of fraud

Matt Walsh and Nick Carter are partners at Castle Island Ventures. All of these expressed by them or the guests on this podcast are solely their opinions and do not reflect the opinions of Castle Island Ventures. Guests and host may maintain positions in the assets discussed in this podcast. You should not treat any opinion expressed by anyone on this podcast as a specific inducement to make a particular investment or follow a particular strategy, but only as an expression of their personal opinion. This podcast is for informational purposes only. Brought down by bad mortgage investments, Lehman, which has 25,000 employees will be liquidated.

The federal government loans American International Group, AI, IG $85 billion. This is a different kind of market, and the Fed is asleep. The federal government is stepping it to stabilize Fannie Mae and Freddie Mac, the two mortgage giants that have been threatened by the housing crisis. The Bank of England has pumped 75 billion pounds more into Britain's ailing economy with a new round of quantitative easing. You print a couple trillion dollars, and all of a sudden, people start to worry.

So out of this worry, we have something called the Bitcoin. Welcome to On the Brink. I'm Matt Walsh. And I'm Nick Carter. Recording this one old school style in person in a conference room. Live. We're doing it live. Busy week.

Yeah, so Wyatt sat down with Stepon Simkin of squads to discuss Stablecoin financial, stable coin powered financial services. And then you wrote, I don't think we talked about this last week, so you wrote another piece on Bitcoin and the Quantum issue. Yeah, it's called Bitcoin and the Quantum Problem Part 2, The Quantum Supremacy. That's a Bourne reference, Jason Bourne. Very good movie. The Bourne supremacy, the second one, is actually the worst of the three.

I've seen all of them. but I only really remember the first one. Yeah, the first of the most memorable. And then it's kind of, he just sort of plays the hits. You know, he does the same stunts many times. Like one of the thing he loves to do is he'll call up the CIA lady and he'll be looking at her from across the way. And you're like not meant to obviously be that vulnerable as the CIA person, right?

Right. And he'll be like, get some rest. You look tired. And they're like cut to music. And it's like, dude, You already did that in the series. You can't keep doing that. It's one of the coolest premises for a movie, though, don't you think?

That there's a spy with retrograde amnesia that doesn't know who he is. Yeah, it's just a very clever premise. He's just using his spy skills to try and figure out who he is. Right. Yeah. I mean, I find it very compelling. I guess we never really had a program like that, or maybe we did. Oh, maybe we did.

I feel like there's a current. It feels like there's a kernel of truth there. I mean, I don't understand why they know. needed to wipe all of his memories and psychologically program him for him to be a good spy. Well, because he was doing things that normal person would have ethical challenges with, I think. But that applies to, already applies to like those types of agents, wouldn't you say? Maybe, maybe. I don't have a lot of domain expertise in that field. Not to diminish, you know, the brave men and women in the CIA, but aren't they already fully, you know,

capable and willing to just kill people? I don't think so. I think there's CIA people that work everywhere. I think every bank in the world has a bunch of CIA people that work there. What do you mean? Covert. Convert? Of course. That's the worst posting ever. You have to go work in J.B. Morgan back office? No, it's probably you're making a couple million dollars a year working at a big bank and the government's paying for a bunch of your salary and it's a secret to all of your

colleagues that you're actually in the CIA and you're taking pictures of sensitive files at night. So you're telling me there's someone that's signed up for the CIA, they go through all the training, they get waterboarded or whatever it is that happens in training, you go through that, and then instead of getting posted to Baghdad or, you know, Santiago, you end up working in back office at J.P. Morgan. I don't think it's that. I think it's, hey, you just get promoted from analyst to director at some big bank and all of a sudden the CIA shows up at your door. And this isn't happening a lot, but you know, you're someone that might have sensitive PII information or you might know, be able to stitch together some foreign money flows. And the CIA shows up and says,

hey, it would be your patriotic duty to let us know what's going on on the inside of this big bank. Just be our eyes and ears. We might call you occasionally. It's not a heavy lift. And we're going to pay you some money. But can't they just demand that information anyway? Oh, they could. But you know, do you say they need a mole there too? Yeah, there's moles all over the place. This is completely unfounded. This is like Alex Jones Tears conspiracy. I mean do you ever watch Homeland? No. All right.

All right. So this article, there was a Jasonborn reference, but it was mostly about quantum and why it could potentially be really bad for Bitcoin. Yeah. But we got the solutions to it, right? Maybe. I mean, we have some exposures that are relevant, of course.

But I think the Bitcoin community just needs to take action. basically. I've been on the Bitcoin mailing list, which I hadn't been on in a while. I used to read that every morning. People are talking about it. People are going back and forth on email on the Linux list. And for those who don't know, all of the emails go back and forth, the Bitcoin Dev mailing list. People are talking about quantum. There is a BIP called BIP 360 by a gentleman named Hunter Beast. I don't know if that's his Christian name or. You don't think that's the Hunter that is affiliated with Castle Island.

It might be the R Hunter. Ethan Holman is involved, and Isabel Duke is the other co-author. So BIP 360 is a soft forked to introduce pay-to-tap script hash because we already have a paid-to-ta-tap route address format, but it's actually quantum vulnerable. Can you believe that?

Did not even know that. So we went through all of this work to put in a new type of address to Bitcoin and it ended up being quantum vulnerable. Why? Why? That was in 2021. That wasn't that long ago. Did we not have solutions back then? Quantum resistant? The other older addresses are more quantum resistant.

PTA script hash, good old P2SH. That's quantum resistant. That's my favorite type of script hash. With caveats is caveats, everything I'm saying. Pay to taproot is actually quantum vulnerable, which is insane. That is crazy. So BIP 360 fixes that and allows for a potential future quantum resistant signature scheme.

But I don't know. I'm not seeing the mobilization that I would expect to see. And I see a lot of denial is what I see. But Bitcoin devs that I know, many of them are in total denial. And they're not paying attention. They're not listening. It feels like the physics community has really accelerated in the cryptography community is maybe putting their head in the sand and saying this is 10 plus years away.

I would say the consensus is that it's about 10 years away. So that's the consensus. It's not 50 years away. It's not zero years away. So Eli Ben Sasson of Starkware, who is the co-inventor of Zcash, he did a sit down with Scott Aronson, who's basically the leading academic on quantum. And Scott's saying there is no fundamental discovery that needs to be made for quantum computing to scale from where it is now to actually breaking elliptic curves.

It's just an engineering challenge. So he basically thinks it's going to happen because we figured out the most important thing, which is basically error correction. That was the most important thing. And this guy is known as a skeptic historically. All right. Well, when's part three coming? I don't know. It took me a month to write the second one.

So it's pretty unusual for me, actually. It's good because you know, you go through an exercise like this. You probably feel like you could teach this to people now. Yeah, but if you ask me how sure Zellrhythm works, I honestly could not tell you. Shore is right down the street. He's still teaching. David Shore. All right.

Let's hop into the deals of the week. First one up is CalShe, the prediction market platform. There is $1 billion at an $11 billion valuation in a round that was led by paradigm with participation from Andrescent and Sequoia. Big round there. Wow. Yeah. Next up, Cracken has acquired backed finance.

That's an asset tokenization platform. Then digital asset holdings, which is the developer. of the Canton network, the layer one blockchain, they raised $50 million from Bank of New York Mellon, I Capital, NASDAQ and S&P Global. BitStack. They're a Bitcoin brokerage company. There is 15 million from 13 books capital, Serena Capital, and Stillmark. 13 books.

What is that? Yeah, we talked about this last time. Do you remember? 13 books, yeah. We looked up what it's a reference to and we thought it was maybe biblical. Yeah, okay. But it wasn't. It was not, it wasn't like the number of books.

in like a series of unfortunate events either. We're just going to do this every time we say 13 books, are we? We need to now look it up again. Chat says it's a reference to Euclid's Elements. Euclid's Elements, that's what it is. The foundational mathematical treaties originally composed of 13 books. All right, so congrats to BitStack. 13 books.

We actually know these guys from a different name. What's that? This fund. What was it? Yeah, we like them before. I forget what it was. before. Okay. Antithesis, which is an Ethereum-focused simulation and testing company. They raised 105 million from Jane Street, Amplify, Spark Capital, and Hyperion. Well, what does that mean?

What do you mean? Simulation and testing for Ethereum? Yeah. That is the most ambiguous thing I've ever heard. Well, testing, making sure it doesn't break. Well, they have $100 million to do it now. Next up, Osteum, they're at decentralized exchange. There is 20 million from G. G-C, Jump, Coinbase, Wintermute, and GsR. Then you have Access, which is an on-chain yield protocol. They raised $5 million from Galaxy, OKX, CMS Holdings, GSR, CMT, and Maven 11. CMS Holdings, back in the league tables here. Yeah, shout out.

We haven't heard from them in a while. Yeah. Galaxy Digital acquired alluvial finance. That's a liquid staking platform. Then the last one is Finn, which is a stable coin payments platform. They raised $17 million from Pantera, Sequo. and Samsung next. Well, I want to start with micro strategy this week, Matt.

So they did something very different, which is they established a dollar reserve, not a Bitcoin reserve, a dollar reserve. Definitely an interesting move. Did you see Sailor's tweet last Sunday, kind of sent the market into a spiral? It was, what if we put a green dot on this chart? And everyone thought maybe he's going to sell Bitcoin and going to go in the

green is bad, green is dollar. in gold is Bickhol. Oh. Yeah. So he didn't sell. No, he didn't sell, but although could you make an argument that stacking dollars is kind of...

Is a form of selling. If you think about it like that. Yeah. But it's kind of a smart thing to do to have enough cash for at least 12 to 18 months of payments on the loans, isn't it? That seems prudent. Yeah.

So this is to cover, is their prepaying or they're covering the payments on their preferred. Yeah. They have many different now financial instruments, but I guess it's strife the main one, the perpetual one. There's like five or six of these, right? I mean, there's one school of thought that the price of Bitcoin will only go up always forever and that you don't have to do something like this. But if you have a bare market and the price of Bitcoin goes down, it's probably good to have a little cash to make your debt payments.

But if your constraint is you don't want to sell any of the Bitcoins, you still have to find a way to generate cash. right? Or are you saying you'd sell shares? I'm saying you'd maybe sell shares if you're trading at a premium, right? Yeah. So that was that. The Dats, I would say, I don't have that much to say about the Dats right now. They're generally beleaguered as an asset class. Yes. That's my summary of the Dats. I don't like punching down. So maybe we'll just skip the Dats this week. I want to talk about the prediction markets, though. So there was a lot going on there.

happening. Cal she's going on chain. Did you hear about this? Yeah. So what is Cal she is wrapping contracts in like SPL tokens or something? What exactly is this? So they're making some integration with the Salana blockchain where they're putting basically making a token for every contract. Is that right? That's really interesting because it was it last week that we said prediction markets didn't have anything to do with blockchains necessarily. Like they don't need blockchains anymore? Yeah. Because the, I mean, I mean, other than the UMA oracle process on Pollymarket, the only real blockchain angle there is getting cash into the system via blockchain rail. But it looks like Kalshi is leaning into it more from an underlying technology perspective here.

So they, yeah, they're converging. So polymarket started on the blockchain and went kind of tried to find Kalshys going the other direction. But what's the deal with some of these, I don't want to call them maybe insider trading, but what appears to maybe be someone with proprietary information trading on it and that it's getting really big on Twitter. Have you seen this? Yeah, so is this the account that is only trading things that pertain to Google searches? And the theory is that it's someone that works at Google

that actually has access to the data that can bet on all these markets with perfect information. Yeah, so there's an account called Alpha Raccoon. They joined Polymarket in 2024. They've made $1.2 million betting on Google search markets. So like, will Bianca Sensori be the number one search person on Google this year? They bet no. And they were right.

Wow. So it's a lot of trades like that. So does that necessarily mean that it's an insider doing that? Or could that just be someone that is? It doesn't. They might have just bought a dataset that is giving them some indication. But I do want to talk about this because. On the one hand, the prediction market people actually kind of like that this is a thing.

Like the official polymarket account will talk about instances where traders discover leaked information. And so they're pricing in information that the general public doesn't know. Through a leak or clever guessing the right URL, the press release, like stuff like that, The official accounts will actually excitedly post about this. Right. But at the same time, it's a very thin, it's like a razor's edge between that and, you know, exploiting proprietary information for financial gain, which it doesn't matter what the asset class is, insider trading is still a thing.

Right. If it's theft of proprietary information that doesn't belong to you and you're monetizing it, that can be insider trading. Yeah, you can insider trade a commodity. Yeah, 100%. It makes me uneasy. Yeah. You could see why you'd be uneasy about that. I mean, it also makes me wonder if, you know, junior people at these large hedge funds that are buying all these data sets just have tremendous informational advantage on polymarket. Yeah. And the thing that's weird about it is the reason prediction markets are good is because they surface, they create a financial incentive to surface information that's not broadly known yet. However, that information gets surfaced.

Right. So the reason they're good is also a thing that encourages people to commit crime. Yeah. And then the libertarian idealistic view is like, hey, it's just good to get all this information out there and surface it. But, you know, people are getting hurt in these financial transactions, I guess would be your argument. So I have two issues with that. One is the thing I've said before, which is that, okay, so look at these markets.

Google Search is Bianca Sensori going to be the number one most. The person on the other side of that trade is getting shellacked because they're not an insider. So they're getting farmed by the insiders. That's probably kind of bad in a way. Right. Whatever. The real problem I have is this perverse incentive all across society for people to basically steal proprietary information from their employers or just from anyone and monetize it through a prediction market. That didn't really exist before unless you were trying to do options trades on it.

merger information or something, but that was very easy for the SEC to apprehend. This is much harder. I don't think it's good that that gets normalized. Yeah. It's going to be really interesting to see how this gets regulated over time. That sort of dovetails into something else that happened this week in the prediction market game. Connecticut's Department of Consumer Protection issued cease and desist orders to Calci, Robin Hood, and Crypto.com, alleging that those platforms are conducting unlicensed

online gambling in that state. I don't know if sports betting is legal in Connecticut. I assume if it is, it's through the regulated channels. And the argument here seems to be Calci, Robin Hood, and Crypto.com don't have a right to offer sports wager in the state. Yeah, I think you're going to see these fights all across the U.S. It seems like it. It kind of reminds me a lot of ways of kind of what's going on the AI market, where you have

state and federal preemption arguments. I think you're just going to have to have some. cohesive framework here. Ideally, probably at a federal level, I'd imagine. So elsewhere in prediction markets, Fanatics is launching a prediction market in 24 states. You see a bunch of these new entrants, huh? Fanatics, uh, interesting entrant to the game. And I assume the 24 states or the 24 that you feel like you could go into, probably not Connecticut. Connecticut has weird, um, crypto bit license-ish type of dynamics around their MSBs. It's always been a tough state for crypto companies to operate in.

Really? Yeah, there's a lot of... Even despite gray scale, but DCG is there. Yeah, DCG is there, but this is more for the on-and-off-ramp businesses have historically had a really tough time getting approval for their MTRs in Connecticut. Elsewhere in News binomial, those Chicago-based derivatives exchange, they're rolling out the first spot cryptocurrency trading platform overseen by the CFTC, believe it or not.

Yeah, that's interesting. It doesn't seem like that would have been the first one, because you kind of look at this and you're like, hold on, aren't there already a bunch that are overseen by the CFTC? But of course, no, because everything else was overseen by the states. Actually, we did forget to talk about this one dat thing that's happening, which is that are now activist fights happening at some of the dots. Oh, my gosh, I forgot about this one this week.

This was a juicy SEC filing. So CZ YZ Labs, they launched a proxy fight to alter the board and the management team of BNC, which is, is the DAP focus on BNB, which I think is trading at a huge discount to Nav. It is. And this reads like one of these lawsuits almost. Did you read it? No.

It has kind of a blow by below of how upset they are with management. It has just bullet points of dates that they were in contact with management and how management didn't take the proper actions. On this date, we had dinner with the CEO and outlined all of our concerns. On this date, we followed up. On this date, we followed up again. On this date, we found out that the CEO is also involved in another dad and not paying full attention to our dad. It's a juicy read there.

And the other one where there was a fight was Pompstatt, right? Oh, was there a fight there? Yeah. Pomp changes compensation model in response to an activist threat at I think ProCAP, Tigger Burr. I saw Laura Shin's article, but I didn't actually see that they had changed that. So Pomp went live and said he's, he's. I think he's taking away all of his guaranteed compensation. It's all incentive based.

That seems like a reasonable thing to do. Yeah. Elsewhere in Bydance news, did you know that Binance has a co-founder? I didn't know that. It is Yee He. I didn't realize that she was a co-founder. Is Yehi the one who's married to Cizzi? I don't know.

But she is now the co-CEO of Bydance. Okay. I'd never come across Yee He before my life. Oh, yeah. I've heard of Yee. Oh, okay. I think she was the chief operating officer. Well, now she's the co-ceo.

All right. Well, this was a big one this week. Vanguard, which is the second largest asset management company in the world. They have reversed their prior stance that they will never have digital asset ETFs on their platform. They will now allow ETFs in mutual funds that hold cryptocurrencies to be traded on their platform.

I think that is what you call reacting to your customers. See, you know, a lot of people complain about this. But they're actually sort of doing people a favor because they protected their clients from buying Bitcoin at $126,000. Yeah, they also protected them from buying Bitcoin at, you know, $40,000. That's true. But if they're going to do it, you might as well do it during a dip. Well, I think last year they probably had some very upset customers. Fortuitous timing.

Well done, Vanguard. Welcome to the party. And was there another major bank that said that their client? could. Yeah, there was a, I didn't put it in our newsletter, but Bank of America was rumored to have issued some guidance around, I think it was up to 4% allocation in wealth management portfolios for Bitcoin, but I couldn't actually find the source. That was maybe just a total rumor.

All these crypto publications were reporting it, but, you know, if that's true, then that's good news, too. So there was an interesting report from HFSC, HFSC, House Financial Services Committee. they had a report on one of our favorite topics, Operation Shugpoint 2.0. Yeah. What did you think? I mean, this has been your issue for the longest time. I guess this is the final report into the investigations,

at least from the House Financial Services Committee. I'm glad they did it. This is, I guess, out of Chairman Hill's office, spearheading the effort. It was really nice to see that they basically agreed with what we in the industry had been saying for years. two years on. I mean, we first started making a fuss about this in early 23. Now we're at the end of 25. And it's nice to see these things put in pen to paper and having Congress basically endorse

our interpretation of events, which is, yeah, we were wrongfully harmed by the Biden admin in a manner that, you know, it wasn't really strictly legal. It was very political and very arbitrary. So thank you to Chairman Hill and staff for writing this. I'm glad it's all out in the open now. It's kind of hard to read through this and not just get so angered by what happened to all these startups that lost access to their backingouts. Yeah. And, you know, the version that ultimately made it to print is not the full story. No.

They weren't able to get in touch with everyone that was affected, but it was virtually everyone was affected. Correct. So there was a real chilling effect. And okay, yeah, some firms were debanked, but it was just delays and impediments and cost. So it was more than just the immediate fallout from debanking. It was just this general malaise that we felt for years. Yeah, and then they, you know,

would have been great if they could have given immunity from prosecution to some of these bank executives that were, told certain things on the phone, but told that it was confidential supervisory information and that they could not actually relay the contents of those conversations. We don't really have any of those disclosures. Yeah, I think also the agencies were hard to penetrate. You know, the Fed clearly wasn't very cooperative here. FGIC was sort of. They released some documents. That was nice. So I don't know if we're ever going to get the full story, you know? Well, I think when enough time goes by, someone will write a book. That's kind of my thinking on this. We're kind of overdue for a big

crypto book, don't you feel like? Yeah, I haven't read any good ones recently. Now, I mean, you had that wave of Ethereum ones, and then, of course, you had the wave of FTX ones, of which, I guess, Zeke Fox, he probably wrote the best one. Michael Lewis wrote the worst one. I read the Michael Lewis one. That was terrible. I like that he talked about the tungson cube. The tungsten cube ending

was maybe the only good part about that whole thing. I wonder if they ever got that tungsten cube. But that was an awful book. Yeah, somewhere in a shipping container in the Bahamas, there's a 14-inch tungsten cube.

They should try to get that back. And like 2,000 years from now archaeologists are going to dig that up and be like, wow. Our predecessors really worshipped this one specific metal. Tungsten doesn't rust, does it? I think it's pretty durable.

Yeah. Well, anyways, someone's got to write a book about kind of the four years of the Biden admin and the impact that it had on the crypto space. There was a good research piece from our friends of galaxy research on the state of crypto leverage. And I know we told you we would monitor the leverage in the system and tell you when it was getting out of control. What do you make of it? Because it definitely went up a lot from the second to the third.

quarter. Yeah, I mean, it's definitely in early 2021 territory, I would say. But, you know, the problem here, I guess, when you try to comp this to 21 is that you have more established players in this market. And so just the overall market is a lot bigger. But, you know, it's creeping up. I'd say it's been creeping up for the past eight quarters based on this data. Yeah. So in absolute terms, the C-Fi lending is below. was in 2021. But it's a totally different cast of characters. It's the defy ones too, right?

So now Tether is the biggest C-Fi lender. Did you know that? I did not know that. And if you add up D-Fi plus C-Fi, we're actually above where we were in 2021, but that's because D-Fi lending is so much bigger. That's true. But D-Fi-Lending doesn't scare me, really,

because it's sort of over-collateralized. It's over-collateralized. I think, you know, what does maybe scare me a little bit about some of the D-Fi stuff is just are there hidden participants there that could just blow up and, you know, do they have a CFI arm or something like that? But it's not clear that there's anything systemic right now. Yeah, at one point we thought this kind of pendle points looping thing would bring down the whole market, but I don't think it did.

No, we went really deep on that pendle thing. That was, that kind of broke my brain. Our associates did. Yeah. We just, they just talked out us about it and we tried to understand. We tried to understand that, yeah. But yeah, you know what? I would characterize my current state as not worried about the amount of leverage in the system,

unless there's hidden leverage that we're not seeing. Now, honestly, I'd be more worried about like a counterparty blowing up on 10-10 and just being on the ropes right now. Or an exchange that, you know, I think this theory that some of these exchanges internalized a bunch of losses on their perpetual futures contracts, that would worry me more than the leverage here. So there's one thing that we haven't talked about yet. What did you make this anthropic blog about how they just told their agents to go make money and they found $4.6 million laying around that they could steal from the blockchain?

I think that this reminds me of this book that I just read called, what is this? If we build it, we will, if anyone builds it, everyone will. Even reading, you'd count. Skoski Slop, Matt? Yeah, it's why superhuman AI would kill us all. It's not a very pleasant thing. You know, I have a hard time. I finished the book.

I finished the book. I found that he took a lot of logical leaps. There was a lot of like, AI is going to do this with no explanation of how AI is going to do it. But one of the things he pointed out was that they would use cryptocurrency. Well, Meg Yude is not the only guy to point that out, but he's right. I mean, you know how he got his start as an author, right? Elizer?

Yeah. No. He was a writer of Harry Potter fan fiction. Oh, so this is the guy we're trusting with the AI Dumer stuff? Yeah, which is why I'm saying we should not trust him. He's not to be trusted. All right. Well, I fell victim for that.

So this is fan fiction, which you're reading. This is further fan fiction. So just ensconce it within that universe. You know, started with Harry Potter. Now it's AI, but it's all fan fiction. You know, the conclusion here is basically that we shouldn't have AI to some level. I don't know. That's not actually the conclusion, but I didn't find it to be a very useful book, but I like to get all sides of an argument.

Yeah. Well, I don't agree with him, obviously, as you can tell. But I do agree that AIs will use crypto and certainly misaligned or malicious AI will, and they're actually going to steal the crypto. Oh, I think this smart contract vulnerability thing, I'm sure Lazarus is all over this already. but Anthropic putting that piece out basically means that every teenager who wants to go hack a bunch of smart contracts is going to start using LLMs for that. So did you know that there's already a benchmarking system where they tell the AI go steal crypto and then it steals as much as it can?

No. And it's been increasing exponentially every year the amount that it's able to steal. That doesn't make me feel good. So starting with DeepSeek in late 2024, early 2025, Sonnet 03. keeps going up and up. So as the AI gets better, they're able to extract more vulnerable cash from crypto. What's the solution here? Just build safer, smart contracts? We have to use AI defense. We need AI defense and AI attack. We could have Tezos, the formally verified blockchain. I mean, Cardana was meant to be formally verified, right? But some vibe coder brought down that

whole blockchain. Did you see that? That was wild. So, but while Charles Hotskinskin got very upset about this, and he, they community noted my tweet about it. Oh, really? I didn't see that. And he said that the Cardana blockchain was not brought down, but in my opinion, it was. It didn't seem like transactions were going through. So what would you call it? Yeah, and there was a chain split.

There was a chain split, a meaningful one. So in my opinion, that's downtime. But I think, so I don't think formal verification actually protects you. I think it's we have to use AI. in defense, you know? Use AI to do penetration testing on your smart contracts and find the bugs before they get exploited. I think that's

probably right. It seems like these smart contract AI tools will actually be much better than humans at finding these vulnerabilities. But I do think it's true that once AI gets really good and AI gets in the hands of like

malicious entities, that's what they're going to do with it if they tell their AI to go make money. Yeah. It's not going to go get an honest job. No, I mean, it's just the low-hanging fruit is to go steal vulnerable cryptic fish. Fishing campaign. Super targeted, specialized fishing. Right.

Or literally spark contract exploitation. That's going to be massive. There's going to be an explosion of this stuff. I agree. You know, once people use open source AI and it's really good and agentic, the world's going to become an extremely scary place. Yeah. Switching gears, I heard. a really bad rumor that Elizabeth Warren is going to run for president again.

Oh, well, I don't think she's got much. I mean, it's not like she's very popular. I don't think so. But can you imagine how much content we're going to get out of that on the podcast if that turns out to be true? I mean, it's not like her previous presidential runs went that well, did they? No, I mean, she got to run all of financial services under Joe Biden. I'd say she parlayed that presidential run into a big role.

She was the most important senator in the whole country. Yeah, that's an even bigger win than RFK running and ending up in charge of... HSS, yeah. She got to be... The position kind of only existed in the ether, but it was head of financial services for the whole country. It is interesting that that's a way to obtain power is to have a failed presidential run.

Yeah, you run for president, and you say, I'll give you my endorsement in exchange for X. Smart. It is smart. She is smart. She still doesn't like crypto. No, and she's not going to like this anthropic thing. No, I don't think she's...

She likes anything at what's going on here. No. All right. I think that is it for the week. We will leave it there. Everybody have a safe and healthy weekend and we will see you on Monday.