Good Alexander - Why Hyperliquid Really Pumped
Good Alexander reads crypto’s rebound as a catch-up to the gold-versus-fixed-income debasement trade, with G10 governments increasingly behaving like emerging markets. Japan’s intervention, US deficits, and Scott Bessent’s talk of pushing rates toward “fundamental value” suggest potentially deeper long-end intervention; Trump’s stated ideal is a 2% 30-year yield versus above 5%. Crypto is roughly $2 trillion against gold above $28 trillion, leaving substantial room if “absolute degradation of the fiat system” becomes the dominant macro narrative.
Hyperliquid pumped as feared US enforcement appeared less likely just as tokenization gave crypto something economically real to compete for. Alexander says informed participants expected the CFTC to come after Hyperliquid and mentions “Jeff Yawn” in that rumor; Trump’s friendliness instead signaled a changed policy environment, whether or not the host’s Barron Trump explanation is right. The shift means more on-chain assets, fees, and liquidity: chains are finally “racing for something instead of racing for an indictment.”
Bitcoin’s setup turned positively reflexive after markets failed to take Michael Saylor off the chessboard and began lifting ETF holders back toward profitability. Alexander places STRC’s death-spiral threshold near 70, versus roughly 95 during the conversation; MSTR retained about a 1× mNAV while some Hyperliquid DATs traded near 0.6×. A sustained BTC break above $80,000 would move holders out of underwater positions, whereas a break below $60,000 could have made BlackRock’s launch “the most value-destructive ETF launch.”
The AI trade is rotating toward biotechnology because inference has yet to deliver the promised returns, in Alexander’s view. Consumer AI add-ons cost money without delighting customers, meaningful AI-driven workforce cuts have not arrived, and OpenAI’s economics still require expensive compute; “because we’re curing cancer” supplies a vast, unfalsifiable replacement story. His more cynical thesis is that biosecurity will eventually justify restrictions on cheaper Chinese open-source models, expanding Western labs’ margins by outlawing competition.
Alexander prefers OpenAI to Anthropic but thinks small-cap biotherapeutics with valuable data pipelines may offer the cleaner narrative trade. OpenAI can monetize hundreds of millions of users through ads, potentially hurting Google, while Anthropic remains exposed to enterprise customers trying to cut inference bills; he “will not be buying Anthropic’s IPO.” Memory suppliers retain three-year commitments, while his speculative equity expression is smaller biotech rather than Eli Lilly: “Fundamentally, dude, I have no idea.”
Tokenized equities are a jurisdiction-and-counterparty trade, not simply digital versions of identical stocks. The Robinhood product is a debt instrument; other wrappers may involve an Abu Dhabi regulator or debt to a Liechtenstein-based company, while a perp may reference an oracle. Parity is therefore not guaranteed. Alexander’s likely winners are Trade.xyz and Binance, with Abu Dhabi providing capital, regulatory clarity, and a safe physical hub for bearer assets: “They’re just gonna put the money behind it.”
The near-term scoreboard is unusually binary: real AI labor displacement and war resolution help equities, while crypto has separate upside catalysts. Without higher unemployment, Fed cuts risk a credibility crisis; a prolonged Iran conflict and continued Strait of Hormuz closure would pressure growth, energy, supply chains, and the AI-defense ROI story. The tail event is the CLARITY Act, which Alexander gives roughly a 25% chance: passage means “Valhalla” and a possible “true God candle” in XRP, HBAR, and Canton; a major USDT outflow is his clearest warning signal.
1. Crypto remains the ideological trade against institutional decay
Alexander’s response to founders abandoning tokens for AI labs is uncompromising: “There’s no such thing as halfway crooks.” Anyone drawn to crypto by distrust of government, fiat, COVID-era lockdowns, or institutional corruption cannot honestly conclude that the system has since become better run, less indebted, or less compromised.
His macro conviction is therefore not based on Solana transactions or speculative-user counts. It rests on the belief that society is “literally at the beginning of absolute degradation of the fiat system,” making a pivot from crypto back toward conventional fiat intellectually backwards.
The Blackprint Manifesto distilled his positioning into two questions: after society “invented awesome amounts of money,” did everything become awesome; and will the institutions that elevated Jeffrey Epstein really democratize immortality through AI-generated biotherapeutics? “I know my answers, and I act accordingly.”
2. Biotechnology gives AI an unfalsifiable second act
Alexander sees a structural problem for AI labs: inference consumes enormous compute while revenue may decelerate and margins disappoint. Meanwhile SK Hynix, Micron, and SanDisk secured multi-year commitments because, in his telling, roughly 70% of AI capital expenditure has to go to Korean companies.
That mismatch creates the incentive to graduate from assistants into medicine. Alibaba described inference as a step toward proprietary products; Dario Amodei invokes curing cancer; Demis Hassabis pursued genomics ventures. The resulting promise—AI might cure all human disease—makes current revenue, bookings, and gross-margin objections feel temporarily beside the point.
His corrective on Moderna is important: its personalized-medicine work reflects 2018-era artificial intelligence, not an LLM being asked to invent a vaccine. The rally and surrounding AI-bio narrative helped make biotechnology the market’s new “SpaceX moment”—a story whose theoretical endpoint is so large that near-term falsification becomes difficult.
Alexander’s most categorical, speculative conclusion is that biosecurity becomes the route to banning Chinese open-source models. Western labs can promise cancer cures, warn that unauthorized models might enable bioweapons or another pandemic, and then “expand their margins by outlawing their competition by citing national security reasons.”
3. Longevity may be a private club before it becomes a public market
Capital is already moving from AI toward bioinformatics, Alexander says, producing everything from heavy demand in South San Francisco to shortages of laboratory monkeys. The constraint matters because personalized treatments require physical testing; monkeys and human trials do not scale like compute.
His best example comes from clinics in Panama and elsewhere: once a treatment produces credible human evidence, price-insensitive family offices can buy the facility or company outright. That helps explain why public longevity funds may underperform even when science advances—the valuable breakthrough disappears into private ownership before listed investors capture it.
Alexander calls this the “Altered Carbon tape”: wealthy networks obtain privileged access to life-extension technology while the public receives a universal-benefit pitch. Citing Epstein emails and secondary private-market information around Peter Thiel, Eric Schmidt’s family office, and AI-bio circles as his mosaic, he concludes: “It’s a club. You’re not gonna be allowed to be in it.”
4. The investable bio trade is data-rich acquisition bait
Ginkgo Bioworks, down roughly 98% from its IPO in Alexander’s account, anchors his skepticism toward public longevity investing. He jokes that Bryan Johnson might not be selling olive oil had Ginkgo risen 1,000%, using the example to separate compelling immortality narratives from shareholder returns.
Thread Guy asks whether investors can simply own Eli Lilly and ignore the rest. Alexander says no: Moderna’s roughly 140% daily move changed short-seller psychology, making smaller biotechnology names more dangerous to fade and more responsive to acquisition narratives than established mega-caps.
His preferred expression is “small-cap biotherapeutics that have a data pipeline.” Tempus AI, once a potential short in his framework, becomes a possible long below a roughly $20 billion market capitalization because an AI company could acquire it and use the data-and-biotech story to support a trillion-dollar valuation. His hedge is explicit: “Fundamentally, dude, I have no idea.”
5. AI adoption is bifurcating between unwanted features and cost control
The failed category, according to Alexander, is bolting assistants onto existing products: HubSpot tools and dealership chatbots cost suppliers money while customers respond, “Dude, I just wanna talk to a person.” The successful Palantir/Atlassian pitch is different—control sensitive data, host models privately, and reduce inference spending.
His base case, absent regulation or a new frontier model, is migration toward self-hosted DeepSeek and Qwen systems that are “radically cheaper” than Anthropic and OpenAI. He interprets both Western labs’ new privacy products as defensive responses to that enterprise demand.
The physical economy sharpens his objection. Walmart was guiding toward unusually weak consumer growth while mortgages and gasoline remained expensive; in Puerto Rico, Alexander experienced rolling power and water outages even as PWR rose roughly 600% through AI power deals. “Hospitals are on generators…while we’re just cranking ChatGPT tokens.”
The missing proof is broad labor automation. Alexander once expected Accenture to eliminate 20% of its workforce and produce explosive margin expansion; Thread Guy thought Jack Dorsey’s Block layoffs were that moment. The predicted generalized follow-through did not arrive: layoffs occurred, but not a broad replacement of white-collar labor by AI.
6. OpenAI has monetization options that Anthropic lacks
Alexander dislikes Anthropic bringing its IPO forward: his dating analogy is that someone first available in two weeks appears in demand, then suddenly being free tomorrow creates doubt. Enterprise inference is precisely where customers are cutting bills, and he believes Codex is a better product than Claude Code.
OpenAI, by contrast, can turn hundreds of millions of daily ChatGPT users into an advertising business. Because its data-center commitments with Oracle need funding, Alexander expects it to push hard into ads, converting search from Google’s one-player market into a multi-player market. His concise equity pair is “long a bunch of biotherapeutic stocks and short Google.”
He will not short memory suppliers because contracts extend roughly three years and vendors say 2027 pricing is tighter than 2026. The bear case analysts can identify is extreme: customers go bankrupt and cancel. That leaves old-line memory businesses with better visible economics than the labs consuming their products.
Alexander also sees ultra-fast inference as a coming product catalyst: GLM-5.2 served at high speed impressed him, 5.3 Spark disappointed, Groq 4.6 is “good,” and he thinks 5.6 Soul fast on Cerebras could transform Codex’s user experience. That potential churn further weakens his appetite for Anthropic’s IPO.
7. Bessent’s words matter more than the first intervention
Alexander dates the macro move to an early-August BOJ joint intervention. Japan owns substantial US Treasuries; defending the yen can therefore pressure the Treasury market precisely as American wartime deficits expand. The broader signal is that G10 balance sheets and currencies increasingly exhibit behavior once associated with emerging markets.
His emblematic duration loss is Austrian 100-year debt issued in 2021 and trading near three cents on the dollar. The unifying trade is “gold versus fixed income”: gold exceeds $28 trillion while crypto sits near $2 trillion, and ETH around $2,000 badly trails the roughly $8,000 level he would once have expected under this macro backdrop.
Bessent’s phrase “fundamental value” sounds Trumpian to Alexander because Trump argues that the world’s best credit should pay the lowest rates. With the 30-year above 5% versus Trump’s imagined 2%, merely rebalancing government purchases toward long-duration bonds—Operation Twist—could become explicit yield-curve control if the objective is 3% rather than roughly 7% mortgages.
Thread Guy asks whether the policy itself is unprecedented. Alexander says the opening move is calm compared with 2020, when Powell invoked crossing “red lines,” an SPV with BlackRock received emergency authorization to buy corporate bonds, and the Fed had to support low-quality credit. Gold printed a 5% candle. What matters now is Bessent signaling willingness to go further; even a small purchase produced an almost “COVID-like” gold response.
8. Crypto survived its test and gained a real asset universe
Alexander concedes why crypto lagged: the Ethereum Foundation sold tokens, Ripple’s equity structure pressured XRP, and governance failures repeatedly destroyed capital. Yet hacking has produced a Darwinian culling, with teams forced to use AI to code-review their work: “You didn’t do your governance DAO correctly? Guess what? You’re down $200 million”—leaving survivors such as Zcash technically stronger.
The decisive “test,” borrowing Soros’s framing, was the attempt to take Michael Saylor off the chessboard. Alexander believes STRC near 70 could have triggered a death spiral; around 95, with MSTR still near 1× mNAV, the market had chosen not to kill him. Survival means Saylor can continue selling equity and buying Bitcoin.
ETF reflexivity reinforces that turn. Below $60,000, many IBIT holders could become underwater and feel trapped in a disastrous launch; sustained trade through $80,000 reverses the psychology. With Saylor alive, sidelined capital available, and investors tiring of AI, Alexander sees “a great positive inflection.”
Tokenization improves the substance beneath that reflexivity. Crypto credit historically meant dubious counterparties and episodes such as Stream Finance; real-world assets create a broader tradable universe. Trump-era friendliness toward Hyperliquid therefore means more on-chain fees and better assets for Hyperliquid, Ethereum, Solana, and other competing chains.
9. Hyperliquid’s rerating is regulatory, while the next move is political
Alexander says Hyperliquid had carried rumors that the CFTC would come after it; the transcript mentions “Jeff Yawn” in that rumor, but the name is unclear. Trump’s positive signal made that feared outcome look false; the host’s joking inference is that Barron told his father Hyperliquid was “the jam.” Whatever the channel, the policy reversal explains the violent rerating better than ordinary four-year-cycle mechanics.
Family-office demand can magnify it. Alexander estimates the top 0.1% owns roughly 40% of assets after offshore wealth; their gatekeeping questions are whether crypto is legally investable and whether a hotter alternative exists. With gold expensive and an Anthropic IPO no longer an obvious full-portfolio trade, crypto becomes the credible catch-up.
Tokenized stocks will not converge automatically. Alexander compares them with ADRs: Yandex’s Dutch wrapper went to zero after Russia invaded Ukraine, while today’s products may be Abu Dhabi-regulated claims, debt to a Liechtenstein-based company, or oracle-priced perps. He expects jurisdictional winners—likely Trade.xyz and Binance, backed by Abu Dhabi’s capital, clarity, and secure infrastructure.
His final triggers are stark. Equities need genuine AI-driven unemployment and resolution of the Iran war; otherwise stocks may grind lower while coins outperform. A CLARITY Act passage, assigned roughly 25% odds, could create “Valhalla” across XRP, HBAR, and Canton; failure favors non-US operating bases such as Hyperliquid, Binance, and Ethereum, while a large USDT outflow would make him “freak out.”
Full transcript
Mr. Alex, good. How are you, man?
I'm doing well, Thread Guy. How are you?
I'm good, man. It's good to see you. The last time you were on, which was, I don't know, part five maybe, was a couple weeks, 10 days before the SpaceX IPO.
It was. Man, we need to stop doing this.
The fact that everything's gonna be fine. Once we stop doing these episodes, society will just be on the right track. Every time we do one of these interviews, it's just a sign that things have deteriorated progressively, right?
I was cracking up. I took a screenshot of this and sent it to ChatGPT. Their current, quote, “doom index”: 68 out of 100. Doom is the base case. I'm like, “All right.” There's nothing better.
But look, a lot has happened since you last came on, and I think the last 72 hours, if you are in crypto, have been the most eventful 72 hours in a long time. So we have a lot to go over, a lot to cover.
And I want to give you some flowers. I think you're an interesting guest right now because you're a “know what he's on” kind of guy for me on Twitter, and I think that you, of all people, have been a real crypto bull over the last 6 months. But not because you're looking at Solana transactional volume, FOMO users, or pump.fun activity, but for ideological shifts in society, in markets, and in the importance of Bitcoin. I think you have a little bit of a different bull spin than a lot of people, and you've been pretty vocal about it, definitely over the last quarter, of everyone on my timeline.
You know, I have a lot of conversations with people. A lot of VCs have been pressuring crypto projects to pivot out of crypto. People will come to me and they'll say, “So-and-so top-name VC thinks crypto is toxic. They want this and that. They don't want a token anymore. People want an AI research lab.”
I was talking to guys who, last year, were pitching everyone on decentralization, and then they're like, “Oh, what are you doing now?” You're like, “I'm doing an AI research lab.” I'm like, “Yeah, okay.” I mean, how do you not believe in the government? How do you get into crypto and not believe in the government, see all the stuff with Jeffrey Epstein, live through a COVID lockdown, get into this space, and then be like, “Yeah, you know what? Never mind”?
It's like one of these things where I'm like, there's no such thing as halfway crooks. Once you get into this space, you can't really intellectually, honestly leave, right? You either think that fiat is a thing or you don't think that it's a thing. And then you don't randomly start—
I don't look at the government right now and be like, “You know what? Everything is so well run. It's just so much better than I expected. It's not corrupt anymore. We're on track to get rid of the deficit.” I'm like, “No, it's way worse than it ever has been before. We're losing a war with a third-world country. Oil is ripping. We're pumping Hyperliquid because Trump's son is probably ripping oil contracts.” We're in a banana republic, and I'm like, “Dude, why would you pivot out of crypto and into fiat when we're literally at the beginning of absolute degradation of the fiat system?”
Good Alexander is like, “Think of the worst-case scenario and then make it 10 times worse.” But you just made this tweet, and it was a really good tweet. It was long. I'm not gonna read it back to you. I'm just gonna read the end, and it's 3 lines:
Question 1: We invented awesome amounts of money. Is everything awesome as a result?
Question 2: Do you think the system that elevated Jeffrey Epstein to the pinnacle of society is gonna let you become an immortal because of AI-generated biotherapeutics?
I know my answers, and I act accordingly. Do you?
I mean, do you? I was asking the question. I have my answers.
How do you act accordingly in this situation?
Well, I wrote a manifesto back in 2023. It was called the Blackprint Manifesto. The theory was that AI and biotechnology are going to structurally intersect.
I don't know if you've seen, for example, the chart of ARKG, the Genomics Index.
No, no, no.
It's just up-only technology. Citrini has done a great job pushing things like Twist—
And what is this?
The ARKG Genomics ETF. It's just the most momo, “we're gonna cure cancer”—
Got it. Got it, got it, got it.
And my theory has always been that AI and biotechnology are going to structurally intersect. In Puerto Rico, I live on Ponce de León, and Ponce de León is the famous Spanish explorer—
Yeah, yeah—
—who's looking for the Fountain of Youth. And of course, this is a—if you go back throughout the history of money, rich people get super rich, and they always try to become immortal, right?
Yeah.
They always fail.
Yes.
It's like German eugenics or—it's like— The most recent permutation of this is Bryan Johnson—
Yes.
—and trying to basically synthesize treatments for yourself.
And then you really get into this, you look into this, and you realize that these AI companies have a very, very strong structural incentive to pivot out of the inference story, right? So, for example, Oracle CDS is at all-time highs. Alibaba just came out with earnings, and they said that inference is just a step towards something bigger, where they're gonna be developing their own products.
If you're an AI research lab, you're sitting there being like, “Okay, I need to IPO. Maybe my revenue's decelerating.” OpenAI, for example, disclosed that their revenue was below their expectations, and their gross margins were negative, right? Because they're just lighting compute on fire.
Then you look at the memory guys, and they're just crushing it, right? Because they're like, “Yeah, we've got 3-year deals. We forced everyone into them. And 70% of the AI CapEx has to go to Korean people.” SK Hynix is crushing it, Micron's crushing it, SanDisk—they just have the moat, right? They're these old businesses.
And so you have fixed contractual commitments. The inference business doesn't look like it's gonna come through, and you're like, “What do we do?” Right? And so Dario Amodei shows up on Twitter, he's like, “Well, let me tell you, we're gonna cure cancer.” We've got Claude within it. Claude has cured disease, man. Hallelujah.
And so now the entire biotech sector is absolutely on fire because Moderna came out, which—and by the way, this has nothing to do with Claude. Moderna's personalized medicine is 2018 artificial intelligence. It's not large-language-model shit. It wasn't like they told Claude to invent a vaccine or something like that.
Furthermore, if you look at the history of Moderna, it is the scam company. I don't know if you remember Joe Biden—
Joseph Biden.
Joseph Biden, yeah. He had this debate where, classic Biden, he went off script. He's like, “I'm pretty confident the Moderna vaccine...” And then it was before the Moderna vaccine was even authorized, right? So he knew it was part of the plan.
I've been trading Moderna for a while because it was a meme stock, and the management was fraudulent. Of course it's in my coverage universe. That's the Good Alexander coverage universe. I love trading pretty much all financial malfeasance because why would you not bet on sure things?
Moderna is a sketchy fucking company. It was the company that gave us the vaccines, right? That's the whole myocarditis thing. That's Moderna, and they're the ones who are supposedly curing cancer.
And so now you have this mosaic where you're like, the incentives—
I remember last time we talked, we talked about how SpaceX was gonna do well on its IPO because Elon had invented the ultimate non-falsifiable bull case.
The sun.
Which is that—yeah, it's like, how big is the sun? It's so big, right? It's just the biggest thing in the universe. It's 99% of the mass, therefore SpaceX is 99% of the market, so it's very obvious.
But this is the equivalent of the SpaceX moment where Anthropic is going in there and being like, they're gonna be— You know, Dario's gonna be like, “My background is in genomics.” And Demis Hassabis actually stepped away from Google for the same reason, because he's pursuing genomics ventures.
And so you have a consensus ranging from Alibaba to Anthropic that inference isn't going to be the game. We need to spend AI on extending human mortality, right?
And so now you're on this health, kind of an all-encompassing, maybe we're going to cure all human disease with artificial intelligence. So maybe you don't need to worry so much about OpenAI's negative gross margins or Anthropic's deceleration, booking curve, et cetera.
Because we're curing cancer.
Exactly. And so that's why you have up-only charts on these pretty sketchy biopharma stocks. But what's interesting is that you do get the sense that things are accelerated because there's so much capital flowing. I've heard that basically South San Francisco is just bid right now. People are moving to biotechnology. The original meme was to pivot out of crypto into AI; the new meme is to pivot out of AI into bioinformatics, right?
There are material lab-monkey shortages. There aren't enough monkeys because people are doing so much testing. There are publicly traded Chinese monkey companies. I'm not going to name the tickers because it's too ridiculous.
Give us one. What's the market cap?
Yeah, I'm not going to shill a Chinese monkey stock on this stream.
Okay, I'm not going to buy it.
Not until I have it. I've been building this position for weeks, so I can't just go on here and talk about it.
But yeah, basically, there are not enough lab monkeys. It started with COVID, and there's been huge testing acceleration. When I ask that second question—“Do you really think that these companies are doing this because they think they're going to cure cancer?”—my answer is probably not. My actual answer to that question is that I think they're doing it because they want an excuse to ban Chinese open-source models, right?
If you think about the game theory of this, you basically pivot the narrative of AI into bio. Anthropic has been tweeting about bioweapons for over a year. They say, “Oh, you might be able to develop a bioweapon.” A lot of these Anthropic press releases are actually Anthropic models using an orchestration model on top of open-source models, right?
Yeah.
So they're setting the narrative up. They're going to say, “We're pivoting into biotech, and we're going to cure cancer. And by the way, these other models, these models from China—I don't know if COVID was made in a lab in China or not, but we can't trust these Chinese models with biopharmaceutical information because we might make another COVID by accident. Some random guy may make another COVID by accident.”
Then they're going to lock this shit the fuck down, and they're going to say, “You're not allowed to use Chinese open-source models because you would be using unauthorized bioinformatics national-security information,” right? That is the plan.
This is not actually a plan to cure cancer. This is a plan to pump their stock with an unfalsifiable bull's thesis, followed by a plan to expand their margins by outlawing their competition and citing national-security reasons. That is what is actually happening.
If you look into it, it's not even that complicated. I don't know if you've ever followed the Peter Thiel-isms. Peter Thiel goes on all these podcasts and always has the same shtick. He's like, “Technological progress wasn't allowed since 1945, when we invented the atom bomb, and that's why we have the economy of bits instead of the economies of atoms.” It's his meme.
But it turns out that Peter Thiel was emailing Jeff Epstein, and Epstein was talking about this, right? He was like, “We're going to get some things shut down.” Epstein was part of this weird decel community.
Whoa.
The weird thing is that Epstein is like, you know, Kamie Clark[?]. I don't know if you know who Kamie Clark[?] is.
Oh, it's Dar—yeah, I do now. With Dario's wife, who was emailing Epstein about the luxury porn company and all of this.
Yeah.
Yeah.
So Kamie Clark[?] was dating Eric Schmidt.
Oh, Eric Schmidt from Google.
Yeah.
Yes.
Eric Schmidt. Yes. And so she was pretty involved in his family office, and guess who is the biggest bioinformatics private investor in the world?
Eric Schmidt?
Eric Schmidt. Yes. Eric Schmidt is great for memes—he has good memetics. He's kind of a player. He pumped Keeta just because his ex-girlfriend—
Yeah, what the fuck was that, by the way?
The story there is that his ex-girlfriend—of which there are many, by the way. Kamie Clark[?] is one. This man, for some reason, is just a machine. Obviously, that's why he wants to stay young, because he can continue this generational run—
Generational transition.
And so, basically, Kamie Clark[?] was involved with Eric Schmidt's family office, and she subsequently started dating Dario, who's a bioinformatics guy. She was pitching Jeff Epstein on something that was actually a really bad pitch. She wanted to make a women's-focused sex-TV thing, and Jeff Epstein was even like, “No, I can't do that. I'm not—”
It's like too far for Epstein.
Yeah, he was just uninterested. He's like, “I can't do TV sex.” That was his email response.
The mosaic there, putting these things together rather than just looking at them as a web, is that there's a group of people—whether it's Thiel, Epstein, or Eric Schmidt—who are privatizing bioinformatics investment, where the narrative is pretty much that the elite are going to get access to acceleration, right?
If you've ever watched the show Altered Carbon—
No.
It's a good Netflix show. It's a good sci-fi book. Before, we were in the Accelerando tape, and now we're in the Altered Carbon tape.
Okay.
The premise of Altered Carbon is that there's a group of ultra-rich people who have unfair access to bioinformatics technology, right? The Moderna vaccine is personalized, and the premise of this plus Altered Carbon is: What if you have these ultra-rich people who aren't particularly incentivized to let bioinformatics vaccines go to the general population?
There's already a monkey shortage. There's already, you know—what do you need to do whenever you test a drug? You probably need human testing, right? And that obviously doesn't scale.
Oh.
Right? For example, if you look into the places where there is scalable testing, I know a guy who does bioinformatics stuff in Puerto Rico. He says there are these clinics in Panama that, any time someone comes up with a human-validated trial—whether it's in Latin America or the Middle East—there is pretty much an unlimited store of capital that will just appear and take this stuff private, right?
Any time there's a big biological breakthrough in the early startup scene, this is one of the historical reasons why longevity funds have underperformed. You have these price-insensitive buyers, like Eric Schmidt's family office.
They don't care about returns.
There's just: buy the whole thing. Buy the whole Panama testing facility. Buy it all, right? They take it off the market, and this has been going on for a long time.
I would argue that's why Bryan Johnson's selling olive oil now, because he put his entire net worth into longevity investments. If you look at the chart of DNA—Ginkgo Bioworks—that was his big bag. He's on Medium shilling it. This thing is down-only tech; it's down 98% from the IPO.
If you listen to his early interviews, he's like, “The reason I did the Blueprint was because I saw Ginkgo Bioworks engineering these trees and shit.” I'm like, no. The reason you did the Blueprint is because Ginkgo Bioworks' stock went down 98%. If that thing had gone up 1000%, we wouldn't be hearing your olive-oil ads, right?
My take on this is that it's a zero-sum game played by obviously zero-sum people. It goes back a long time, and there's a network of these individuals who've been interacting with each other. We know this from Epstein's emails and from secondary private-market information.
You kind of are like, “Okay, there's one big club,” and my bet is that we're not going to be allowed to be in it, right? That's what I was saying in that tweet: It's a club. You're not going to be allowed to be in it. You're going to be pitched this idea for your own benefit, and then that pitch is going to be used and weaponized to shut down open-source artificial-intelligence technologies.
So, first of all, that was a fucking can of worms. In that long post with the questions at the bottom, you referenced how airlines have gone all in on software and airports fucking suck, and then how Uber has gone all in on AI coding and the Uber app is no better.
Is this the final progression of consumer, then enterprise, then replacing all software engineers and all white-collar jobs, and then, okay, that's not working—now we're just going to cure cancer? Is this the final Hail Mary on the AI GDP supercycle, and we've already speed-run it?
Well, you kind of know. If you look at what's happening in corporate earnings, there are 2 trends, right? So you have companies like HubSpot—
Mm-hmm.
who came out and tried to bolt AI services onto their existing products. Pretty much every permutation of that, whether that's car dealerships jamming AI into their chatbots or HubSpot building some AI assistant—one, it costs money, and two, customers are like, “Dude, I just want to talk to a person.”
Facts.
“This is ridiculous.” And so that's a failed category. The flip side is there's a very successful category, which is pretty much—you can call it the Palantir/Atlassian category—
Mm-hmm.
where they're going in there and being like, “Hey, Anthropic and OpenAI might rug you.” There’s this data-retention thing. You should be hosting your own AI services, and furthermore, you should be saving money on this stuff because it's fucking expensive.
Yeah.
And those pitches are landing, right? Both Atlassian and Palantir are landing those pitches, and now it's put pressure on OpenAI and Anthropic to actually come out with privacy products. Both OpenAI and Anthropic in the last 24 hours have announced, “Oh yeah, we have privacy products now.” I'm like, “Okay.”
So the dumb prediction is that, until there's some kind of ban, you're just going to see a big migration to DeepSeek and Qwen—Alibaba and Chinese models hosted basically on-prem servers that are just radically cheaper than Anthropic and OpenAI models. The only way that doesn't happen is if they release basically a new frontier model, which both of them are kind of implying they're not going to do, right?
Yeah.
Because Anthropic is like, “Oh yeah, we might have this fable.” It's been a long time since we've had a new frontier model.
Okay. Let me ask you this as a follow-up, and then we'll bring it to crypto a little bit, because I have a lot of crypto questions, Scott Bessent. As it relates to AI, this might be a super-novice, just-wrong evaluation of the situation, but the market has obviously slowed down a lot. The hyperscalers are super cash-tapped. The memory trade post-Leopold[?] is not the back. Semis look terrible. Intel looks terrible.
It almost feels to me that everybody just got a little bit tired of the AI story. We were setting up for stream the other day, and there was that Gavin Baker–Dario back-and-forth that ended with, “We're going to cure cancer.” I was going over it to prepare for stream, and I just thought to myself, “You know what? I don't even care.” I didn't cover it. I was just like, “Fuck this. I don't feel like talking about this.” I just didn't cover it.
I feel like there is this mass momentum fatigue, this burnout, where it's like, okay, we get it, AI's going to change the world and explode GDP, but my life hasn't monumentally changed, maybe, since OpenClaw. I still use my OpenClaw. I still use it for clips and recaps. But it hasn't revolutionized my life. I walk outside and everything's the same.
We're kind of just bored of the trade. Has this happened, or am I just living in a bubble and fried and online too much?
No. I was looking at the most recent data point, which is that Walmart basically has guided to an aggressive consumer deceleration. It's going to one of the slowest paces of consumer growth in a really, really long time. Everyday life is getting a lot more expensive. Mortgages are the most expensive they've been. Gasoline is back at the highs. The Strait is still basically closed.
I live in Puerto Rico. There are rolling water outages and rolling power outages. The company—
There are water outages?
Yeah. There's the company, the ticker PWR—
Okay.
It's an up-only technology. They are the utility that's basically the private partner for Puerto Rico's electric grid. The stock is up 600%, and the electricity doesn't work, right?
The reason why the stock is up so much is because they've been doing all these power deals with AI companies. It's kind of this juxtaposition where you're like, dude, the lights are literally not—hospitals are on generators in Puerto Rico while we're just cranking ChatGPT tokens, and Walmart's decelerating. Why are we spending all these tokens, right?
Whoa.
I don't know, man. I think that's the—we're at, I think, to your point, we're kind of at the point where, on the inference side of things, you need to have a show-me story where you're like, “I got rid of 10% of my employees.” I was a super-software bear for a while, and I was like, okay, there's going to come a point at which Accenture fires 20% of its workforce because of AI, and then its margins expand like crazy, everything's going to pump, and things are going to be okay.
But it just hasn't happened. Accenture's still—
I thought the Block was that moment, Jack Dorsey. That, to me, was the moment. It just didn't happen after that.
Well, Jack Dorsey's just a futurist, so I think he thought it had to happen, and then he did it, and no one else did it. The Cloudflare guy's a futurist too, so it might just be that guys like Dorsey and Cloudflare are just visionaries, and they figure out how to fire people, reduce the workforce, and expand the margins, but it's probably not most companies.
So yeah, I think that's why there's a pivot to biotech, because the inference story isn't working from an ROI standpoint. We need to come up with another story because people are sick of it.
Yeah. I thought that Jack—you're right, no one else did it. I remember I was in the car after stream, and I'm like, “Citrini article, we gotta read it again. It's happening right now.” He was right about the SPX price prediction, but not about the mass firing. I mean, there are layoffs, but it's not because AI is getting—
You know, James was saying that DoorDash was going to have people on their fucking Claude, ordering a self-driving DoorDash meal. I think he might be a little early on that, you know?
Yeah.
I think he mentioned Hyperliquid in that, or he definitely mentioned stablecoins. But—
But he got the biotech thing right. He got this twist thing 100% right.
Yeah.
It's not a 100% hit rate, but he got enough right that it moved the needle.
Yeah, it did move the needle. All right, let's do a 180, and we can come back to this Anthropic IPO in a little bit and talk about the amazing cryptocurrency.
I don't know—48 hours ago, I guess 72 hours ago, it feels like there was just a crazy vibe shift in coin prices, but then in the Trump administration and the Fed, and specifically Scott Bessent. I've been macro LARPing for 72 hours, and it's difficult, by the way. I started learning about the yen intervention, and now I'm talking about the 30-year. I'm all over the place, and we just had Felix Zhao on, who's a really, really smart guy, and he talked to us about it a little bit.
I kind of want to get your perspective, because I think this lines up with a lot of macro trends you've been talking about for as long as I've known you—the direction that things are headed to the basement, all of this. What is, in your mind, so significant about the Bessent moves in the last 72 hours, and how correlated is the Bitcoin pop to the actions of the Fed versus the Treasury secretary versus just, we're bottomed, four-year cycle, regular market forces? How tightly is it correlated?
I'd say this move really started in early August with the BOJ joint intervention, which was a huge intervention. The context is that Japan is a huge holder of U.S. Treasuries.
Yeah.
If they have to intervene in their own currency, that can cause severe pressure on the U.S. Treasury market, which, given deficit expansion due to the war, can basically put us in a U.K.-like situation. The incremental point is that, back during the Brexit period—not the Brexit period, the Liz Truss bond-implosion period—
Sure.
there wasn't a strong, “Oh, fiscal reality is coming due in every currency.” That wasn't a thing.
Mm-hmm.
Now we're having multiple countries, right? We have Japan, because basically Japan has even more debt than we do, and their 30-year yield is blowing out.
Their currency has weakened aggressively. The Bessent actions and the yield-curve moves are best viewed in the context of G10 currencies doing things that, in the past, would normally have been reserved for emerging-market countries.
Oh.
The deterioration of balance sheets in G10 countries is best viewed in the context of the gold price versus the long-term liabilities of countries. This is just Dalio core or whatever, right?
Yeah.
But that is an up-only chart, right? There’s this up-only chart, which is the gold price relative to the long-term liabilities of countries. The all-time high was when the Austrian central bank issued 100-year debt in 2021 that’s now trading at 3 cents on the dollar, right?
100-year debt is crazy.
They actually did that in Argentina somehow, and then they defaulted.
Didn’t Google do this recently?
I’m not sure. I haven’t been following the Google debt markets.
Okay.
That would be epic if they did. Congrats. If someone issues 100-year debt, the history of the trade has not been particularly good. You generally shouldn’t be buying it.
My basic mosaic is that, in macro, this has all been one trade: gold versus fixed income. Crypto—
Fixed income.
—has traded like crap relative to that trade, right? We have been saying for a long time that if you told me four years ago, “Okay, gold versus fixed income is an up-only line. Where is Ethereum trading?” I’d be like, “$8,000.” It’s like, dude, this is $2,000, right?
From a beta-adjusted perspective, crypto is at $2 trillion. Gold is well above $28 trillion. I think we kind of know the reasons why that’s the case. The simple reason is that crypto historically has had a lot of governance issues.
The Ethereum Foundation has a tendency to sell its tokens. Ripple did that crazy equity deal, which was the equivalent of the Berachain deal, where it’s like, “Get your money back guaranteed,” and that put pressure on XRP. I remember when I was trading XRP, and it was at $3.20, and they did that deal. I’m like, “Oh, man, I better get out.”
Yeah.
Then everyone yelled at me, and it went straight to $3.80, and then it collapsed, right? Crypto is just this wasteland in terms of sentiment, but I think the actual underlying technology keeps getting better, right?
If you’ve been following Zcash—
Yeah.
—underneath the hood of all this stuff, you’ve had this almost Darwinian explosion of hacking in crypto. All these teams are basically being forced to use AI to code-review literally every single thing that they’ve ever done, and they’re basically being subjected to a real-time elimination of all the weak hands.
You didn’t do your governance DAO correctly? Guess what? You’re down $200 million. Goodbye. There’s been a culling.
Yeah, yeah.
I think the market so far has just viewed that as crypto being structurally impaired. I’m like, well, actually, the people who are still around—Zcash recovered from that hack.
Mm-hmm.
I think the market tried to make a play to blow up Saylor. They pressed STRC all the way. If STRC went to 70, it would have kicked off a pure death spiral, because there would be no way that it would ever bounce.
Saylor made the gambit. The market decided, you know what?
The gambit.
We’re not going to send him to zero, right? There was a moment in time when the market could have taken Saylor off the chessboard.
Wow.
There’s a local value fund manager who I’m very good friends with who hates crypto. He owned the entire put chain on STRC.
Wow.
He’s like, “These were the deepest value investment I ever made.” He’s calling me and stuff, and I’m like, “I don’t know, man.” MSTR is still at 1× mNAV. Hyperliquid DATs are trading at around 0.6× mNAV.
Say what you will about Michael Saylor, he still can sell equities. The other bull case on Michael Saylor is that he went on The Diary of a CEO and said that he invented the entire strategy with ChatGPT.
ChatGPT.
Now OpenAI has upgraded voice mode in ChatGPT, so Michael Saylor is on a much more intelligent model now than he was. STRC is at 95, and normally in a bubble there’s something called a test. Soros writes about it.
There is a moment in time when you can say, “All right. Take this guy out.” Then they survive the test. STRC is at 95, and Saylor’s not dead. The mNAV is still at reasonable levels. If Saylor’s not going to die, guess what he’s going to keep doing? He’s going to buy more Bitcoin, right?
Whoa.
You’re kind of at this point, and you also have Stanley Druckenmiller—the Stanley Druckenmiller meme. You’ve got Paul Tudor Jones—
Paul Tudor Jones.
—who people take really seriously. The other thing that’s interesting is that if you look at Bitcoin and IBIT, and you look at the redemptions, it’s kind of like, if we can break through 80K on a sustained basis, people aren’t underwater anymore.
There was a moment when we were going lower when it could have entered a death spiral. If we broke through 60K, everyone would be underwater and feel like idiots, and then BlackRock would be like, “Oh, yeah, it’s no longer the most successful ETF launch of all time. It’s actually the most value-destructive ETF launch.”
The worst ever.
Yes. It’s very reflexive, and right now we’re in positive reflexivity. People are sidelined. Michael Saylor’s not dead.
The narrative of inference being the only thing in the market is no longer there. Like you said, you’re tired of it. Everyone’s tired of it. Nobody’s getting rich, and it’s like, okay, now we have to print money.
G10 countries are trading like emerging markets. They’re engaging in emerging-market-like behavior, pumping Hyperliquid, and loosening SEC regulations. I’m like, this is a perfect crypto bull market during a point of absolute despair.
From a price perspective, I think we’re at a great positive inflection. My point from earlier is that if you pivoted out of crypto, it’s kind of like, what were you doing here to begin with? Did you really think that you were going to make it in the data-center economy? I don’t.
First of all, that test thing—I’ve never heard that. That was so good. They tested it. It was consensus that Michael Saylor absolutely, positively was blowing up, and it was going to happen.
Bro, LightCrypto came out of retirement to tweet, “Take his heart out. Cut, eat his heart out.”
That guy’s not retired, man. That guy is running the most size he’s ever run.
Twitter retirement.
Yeah.
Yeah. We can come back to that one too. But on your first points, Quinn Thompson made this tweet today. This is the side of the market I don’t understand that well, but he said, “My new favorite phrase: When developed economies start doing things you’re not used to seeing, expect side effects that developed economies are not used to seeing.”
The thing is, we’re not really that far into anything. We’ve done so much more ridiculous stuff than buying the long end.
In 2020, Jerome Powell put up a YouTube video where it sounded like he was in a fishbowl, and he said, “There’s a time and place to cross red lines, and that time is now.” That day, I remember there was a 5% gold candle.
5%?
Yeah. As soon as he started saying that, people were like, “What? There’s an SPV with BlackRock that has an emergency authorization to buy corporate bonds? That’s not even legal.”
Wait, I don’t even know about this, really. When was this?
This was during COVID.
I vaguely remember this.
It was the mortgage REITs. There was this really stupid moment where these mortgage REITs had put on a relative-value fixed-income trade, buying crappy mortgages and shorting agency mortgages.
The Fed, because they went in and bought agency mortgages, short-squeezed these mortgage REITs. It zeroed their equity, and then it was about to cause a financial crisis, so they had to come in. Because all these people were long low-quality credit, they had to buy low-quality credit or spin up facilities to buy low-quality credit; otherwise, the market was going to keep spiraling into a death spiral, right?
And so the things that we're seeing now are, in my opinion, extremely calm—
Okay.
—relative to how kicked up it can get. Once you start the yield— I mean, the good example is Japan, right? They started buying ETFs, and eventually they owned literally 50% of the ETF market. It's like—
Holy fuck.
So it's like we're pretty early. The significant thing is not the actual action. The significant thing is the commentary and him being like, "We're willing to go further." The narrative that he's using is also very Trumpian.
Trump famously says the US is the best credit in the world; therefore, we should have the lowest interest rates. That's the big Trump line, right? And Scott Bessent is like, "We want interest rates to trade to fundamental value." To me, that's code for Trump. That's code for Trump saying that the 30-year should be 2%, and it's above 5%. So what's fundamental value, right?
The only way that you get to so-called fundamental value in Trump world is that you do yield-curve control, right? I think the words that he's using, and the fact that he's next to Trump, kind of tells you that there's more to come. That's why things are trading the way they are. I think that's why Bitcoin pumps so violently.
I think that's why gold— I mean, actually, gold is a smarter market objectively. And somewhat notably, Russia has also been dumping gold, right? So—
Wait, why?
Well, the war is extremely expensive. The Ukraine war keeps dragging on—
That's insane that it's still going on at the level that it is.
It's still dragging on. It's actually accelerating.
Yeah.
Right? It's accelerating, and it's getting worse. Russia got demonetized from US dollars, so they had massive gold reserves.
Yes, yes.
So they're big gold holders. And so, yeah, it's pretty significant price action that we had a kind of COVID-like gold move at the highs. It was almost like a QE gold move off of a very small purchase.
That begs the question: if we're just getting started on these purchases, and if he thinks so-called fundamental value is what Trump is thinking, we're very early on intervention.
Whoa.
And so, yeah, that's the Arthur Hayes thing. That's sort of always the macro Paul Tudor Jones thing. Actually, a lot of the things I put in that tweet are just parroting what Paul Tudor Jones said in an interview. He was pretty much like, "Look, the earnings growth isn't there. All the deficits are there. We don't have the fixed-income setup to absorb this issuance, and therefore intervention is going to be required," right?
That's where we are. I think that's good for crypto because it's also a refutation of the narrative, right? If everything was so awesome, if we're generating such massive economic excesses, why are we in such massive debt?
There is no productivity boom. It's just a cost-curve acceleration. My power is out. The company that's selling the power is up 600%. They don't give a fuck. And so we're going to see what the incentives are here. The incentives are pretty much a normalization of that move. The thing that's up 600% is going plus 200%. It's down; it's going to go down 30%, right?
Can you explain, to the best of your ability, what is so significant about the 30-year, or the long end, and the price that it's at right now? What happens if the US does not defend that level?
The significance of the long end, at least in the United States, is that it's the anchor for the mortgage market.
Got it.
When Americans feel the pain of financial conditions, it's through the mortgage market because they're like, "I can't refinance my house." Americans have a huge amount of wealth and home equity.
The 30-year is also historically an area where the government doesn't usually step in. In the case of Japan, for example, they'll control the 2-year and the 5-year, but the 30-year will just kind of be a market measure unless you get really extreme and do yield-curve control.
Their historical mini-interventions—in this case, this was the language that Bessent used—was "twist."
Twist.
Operation Twist means when you're rebalancing your fixed-income portfolio to favor long-end purchases. Actually, the other significant thing about Bessent's behavior is that he criticized Janet Yellen pretty aggressively for—
For doing exactly this, right?
Yes. So it's significant in that regard as well. It implies that it's probably Trump's idea. If this is Trump's idea, Trump is not nearly done with where this is going.
Trump's idea is that America should get cheap mortgages at 3%, not 7% or whatever, right? So—
And is this because Trump himself can't cut rates, so they will just force the yield curve down by controlling it? How do those two things tie together?
It's a question of Federal Reserve independence. I think Jerome Powell would never have complied with a sort of implicit or explicit order from Trump to do yield-curve control. He was kind of his own man.
I think that's why Trump hated him, even though he appointed him, because he wouldn't. He would routinely get berated by Trump in these meetings, right?
Yeah.
I'm not sure Warsh will be the same. Actually, Warsh has a bunch of cryptocurrency investments.
He does.
He was college roommates with a very, very famous Bitcoin miner. So—
Oh.
Yeah.
He's an investor in Lighter and all these things, right?
Yeah. I think Trump's policies are best described as mercantilist. He's always viewed the dollar as a liability. He doesn't want a strong dollar.
Even though Ukraine is getting worse, even though European natural gas is ripping, because Europe is very exposed to energy—
Yeah.
—as is Japan, by the way—dollar-yen and the euro have just surged. Or dollar-yen has dropped and the euro has surged because this is the behavior of a weak-dollar policy.
The reason why this flows into crypto or flows into gold is that there was a story about how AI was going to make everyone rich, and now it's like, okay, someone has to pay for all this stuff. Someone has to pay for—
The other reason this is relevant is that Oracle and other data centers are tapping the debt market—
Oh.
—to finance all these data-center investments, and their credit default swaps are increasing because the cost of rolling— if interest rates were zero, then it would be easy to justify these data-center investments.
The private-credit market has already faced a lot of stress. I think it's seeing an 80% decline in the demand for private credit because of the software—
It's crazy withdrawals, right?
Yeah, attempted gating. And I mean, it's kind of stabilized because the software index has stabilized.
Yes. Yeah, yeah.
They were really, really screwed in a situation where AI just replaced all these shitty software companies.
Yeah.
But it hasn't really replaced them. So we're not in an extreme risk-off because the government has a lot of ammo. Equity risk assets are basically close to the highs.
If you look at the historical retail markets for crypto, it's Koreans, who are rich. I think that's why you can see things like Brad Garlinghouse showing up, and people are ready. People have been hurt by SK Hynix, and they're ready to speculate on something new, right?
So XRP—
Yeah.
—can see a pump, right? And I think that's new.
How do you think this debasement narrative, especially with BTC, handles in comparison to the last one? How significant is this one versus the last one? Is this signaling the flaming gun for go-time for Bitcoin, or is it not that simple?
I think it's less simple just because you have Zcash.
Like, Monero is trading quite well. Zcash is trading well. I think I'm a big Monero bull; I always have been.
Yeah.
Just because I'm like, “Why would you own tracked BTC?” I don't get it.
Fair.
But can BTC get a bid when Trump is announcing that they're thinking about buying it? Absolutely. The other interesting thing is that I wrote in that blog post that there's just a huge tokenization push that wasn't there.
Yeah, yeah.
The big problem in crypto has always been that there just haven't been that many good assets. The credit in crypto is such crap. It's like GBTC, Basis Credit, or Aave lending to really dicey guys like Stream Finance. Credit in crypto is the worst fucking thing.
Yeah.
The impact of RWAs is not really that all these fees are going to immediately flow. It's more that you have a real tradable universe, and you didn't have a tradable universe even last year. Trade.xyz is important. The fact that Trump is—there are rumors, for example, that the CFTC was going to come after Hyperliquid.
The reason why Hyperliquid pumped so hard is because a lot of people in the know were like, “Oh yeah, Hyperliquid is going to get prosecuted.” Jeff Yawn—the rumor was that Jeff, and obviously that's not fucking true, right? So if Trump is coming out and saying Hyperliquid, it means that Barron was like, “Hey, Dad, no, Hyperliquid's the jam.”
Yeah.
That is the policy now, and it's great for crypto because it means there's more on-chain fees, there are better assets, and Trade.xyz can flourish. It's unambiguously positive because that means fees flow to chains, whether that's Circle's new chain, Ethereum, or Solana. There's a race between all these L1s, but at least you're racing for something instead of racing for an indictment.
House of Money mode. You've had these takes as well over the last six months or so that the next wave of crypto capital is going to come from these really rich individuals, like a mass flight out of the system. Do you think that is happening right now? Has that happened yet? Do you still think that's going to happen the same way? How do you think that's going to play out?
I think gold—when you get gold running like it is, the family-office story to investors changes. If you think about what actually is the incremental driver of crypto prices, there are 2 things. One is, I'm a family office, I'm ultra—by the way, 0.1% of people own something like 40% of assets when you incorporate all the offshore.
Gosh.
The real driver of crypto is kind of like you're sitting in a family office and your UBS wealth report comes out, and they say, “Is crypto investable?” That's the first thing. Are you going to go to jail if you buy this? Right now, I feel the answer is no.
The second is, what is hot? Can we just be like, “Okay, Anthropic is just the shit. You need to put all your money in this IPO, and then just set it and forget it”? I think it's conflicted. I don't think it's entirely clear that you should full-port the Anthropic IPO, right?
Yeah, I agree.
You're like, “The data-center debt—if I'm really bullish on OpenAI, I can buy SoftBank at a huge discount, or I can buy Oracle debt.” There are a lot of expressions that can suck out hot money. Then you're like, “Okay.” Gold is too expensive. Gold is basically at the highs, and you're like, “Okay, crypto is a catch-up trade.”
Yeah.
Paul Tudor Jones is in it, Stanley Druckenmiller is in it. You index this credibility stuff, and then you can get real money flows because there's a story: tokenized stocks. I don't think you can get family-office investors jacked up on the Zcash story yet, right? You need to get an AOC, or you need to get them afraid.
Yeah.
Right now, I think people are greedy, but for this thing to go vertical, you need to get them afraid, right? That's sort of the—I think we're in greed: Hyperliquid running, tokenized stocks.
Yeah.
I still think that ETH, Hyperliquid, and BNB—Binance actually does a lot of volume on on-chain stocks.
Yeah.
I think all these centralized exchanges are going to do great. I think Solana's upgrade is going to work well. I think Alpenglow is cool tech. I think it's underappreciated how Anatoly took the tech-development path away from Jump.
There was always this problem that Jump was going to create Firedancer and then maybe front-run everyone on Solana. But Anatoly is clearly—and you don't like it, for example, when someone like Vitalik's checked out or traveling in Asia. Anatoly has not checked out. He's grinding.
As a family-office investor or an institution, you're like, “Okay, where am I going to put my money?” You've got some credible L1s, you've got some exchanges, and you've got Bitcoin. Ethereum is going to benefit from tokenization.
I think the interesting near-term trade is that it's probably difficult for Circle, in the near term, to fragment liquidity. They're dropping an L1.
Yeah, yeah.
When you don't have clarity, I think it's actually going to be difficult for the market to absorb an L1 with $300 million. There's $300 million of funding into this L1. I don't know how that thing's going to trade, and I think—
Yeah.
If anything, it's difficult. The other thing you're still not seeing is Tether supply. I would have liked to—
Why is the stablecoin market not going up? I don't get it.
Dude, Circle's starting to see a little bit of inflow. But no, it's not great. The thing that I don't like is that Tether finally got audited. I was hoping KPMG counted the gold bars. I love Tether Gold as a product. I think it's just based.
Yeah.
It's cheap to own. I kind of believe that they have counted the bars themselves, and it's compelling to me. As a naïve person, I don't have a Swiss gold vault, and it's just compelling to me.
Yeah.
I don't love the fact that you haven't seen inflows into Tether since the audit. I'm like, “Dude—”
That should have been a catalyst, is what you're saying.
Yes. I think the thing that would change my mind is if we saw a big USDT outflow. Then I would genuinely get scared because you're like, “Okay, why isn't the number going up?” If it starts to go down—
Then it goes down.
Then I'm probably going to freak out. So it's sort of like, yeah.
Can you explain to me how you win tokenized stocks? Who wins, and what does it even look like to win tokenized stocks? How do you win that?
Dude, honestly, it's illegal. If you study the history of ADRs, Yandex is a great example. When Russia went into Ukraine, Yandex, the ADR, which was a Dutch holding company, went to zero. The thing about tokenized stocks is that they're very similar legally to ADRs.
Yeah, yeah.
The Robinhood thing is a debt instrument. People are tweeting, “Who's making a product that makes these on-chain stocks trade at the same price?” I don't think they are going to trade at the same price, because one of them has an Abu Dhabi regulator, the other is debt to a Liechtenstein-based company, and then the perp is trading on an oracle on a spot price, which isn't either of those things.
The winner in on-chain stocks is going to be a jurisdiction play, right?
Oh, a jurisdiction. Okay.
Honestly, man, I don't know. I think Trade XYZ is just winning from a liquidity perspective.
Yeah.
But my bet about who wins on-chain stocks would be a mix between Trade XYZ and Binance. I’m not sure—
Okay.
And the reason is because CZ is in Abu Dhabi.
Mm-hmm.
The primary venue that has jurisdictional clarity on these on-chain stocks is Abu Dhabi. It’s a real financial center, and they’re just desperate to be a financial center. They have unlimited money. They have huge amounts of oil wealth, and there are signs that Abu Dhabi is going to be the on-chain stock hub.
Wow.
They’re just going to put the money behind it, and it’s safe, right? If you’re trading around with on-chain stocks, they’re bearer assets. Unlike normal brokerage accounts, you can get wrench-attacked, right? So you want to have a safe physical hub.
And, yeah, I just think Binance and Abu Dhabi are going to be there. Then, if Trump is being friendly to Hyperliquid, Trade XYZ is suddenly the U.S. play for on-chain stocks, and they’re going to be able to get a regulatory advantage. For example, Cobie listed Hyperliquid perps. Coinbase, by the way, is headed out to Abu Dhabi. So it’s kind of like—
Oh—
I just don’t think Coinbase is going to win. I don’t think Nasdaq or Interactive Brokers have the intestinal fortitude—
Fortitude.
—to go with it. And I do think CZ does, and I do think Shorooq does, right?
Yeah.
Those guys are going to see it through, and they’re going to win. I think that’s my view.
Huh. I haven’t really thought about Abu Dhabi very much. Thank you for that. That’s a segment for sure.
Okay, forward-looking, then. We’re hearing the Anthropic IPO is definitely happening this year. I think Polymarket has a coin flip for mid-October—imminent, imminent. OpenAI said 2027; I believe they came out today. So we’re going to get Anthropic first, it seems, and OpenAI second, but that could be December or January for both.
Is there enough liquidity in the market to take this on? I know people were upset because Anthropic’s ARR was significantly lower than YipitData or whatever predicted, but they’re still posting disgusting numbers. OpenAI, maybe not so much. What do you think happens when these hit the market? Does the money leave equities and go into crypto? Is crypto completely unaffected? What happens to memory semis? What’s going to happen with these IPOs?
I still like OpenAI relative to Anthropic. I think Anthropic is in the worst position. I don’t like the optics of moving the IPO forward.
It’s kind of like when a girl says that she wants to go on a date two weeks from now, and you’re like, “Okay, she’s kind of a hot commodity.” Then she’s like, “Actually, I’m free tomorrow,” and you’re like, “Hmm, I don’t know. Maybe I don’t want to do that.” So I’m not bullish on Anthropic.
I think they’re in the enterprise inference business, and we’re seeing very clear customer behavior that people want to cut their bills. That’s not something you want to see.
Yeah, yeah.
I think Codex is a better product than Claude Code for a lot of reasons. But I think they’re better trades. I think this genomics thing has legs, right?
I think everyone is incentivized to do the unfalsifiable bull thesis. Except unlike Elon, if Elon is like, “Go long the sun,” there’s not an obvious way to play it. You’re not going to buy solar stocks or something because he’s talking about data centers in space.
Whereas if everyone is shilling biotechs, I think this stuff is going to keep ripping. It’s going to be acquisition targets, with an influx of VC capital. Even dicey stuff in the biotech sector is going to run because that’s the narrative they’re going to try to put on their IPO.
And I think OpenAI is going to go really heavy in ads because I think they have to.
Yeah.
That’s how the data centers pay for themselves, right? The big data center contracts with Oracle are with OpenAI.
Yes.
And why is OpenAI so expensive? Because everyone is on ChatGPT, and ChatGPT loses money.
Yeah.
And how do you make money? You run ads.
You run ads.
The advertising market hasn’t shown up in Google’s numbers because OpenAI hasn’t been slamming it on ads. I believe OpenAI is going to successfully build a very big advertising business, and it’s going to turn the market from a one-player market, which is Google on search, into a multiplayer market. It’s going to be terrible for Google.
So, in terms of the trade, I’d say, “Okay, long a bunch of biotherapeutic stocks and short Google.” That would be my exact equity trade.
I don’t want to fade the memory names because if you listen to SanDisk, SanDisk just went on the road, and they extended their billings curves by 3 years. The memory guys are saying that 2027 is tighter in terms of pricing than 2026.
The only real bear case for memory that analysts are trying to press on is, because Leopold was blowing up, people are saying, “What if all your customers go bankrupt?” And they’re like, “Well, then, yes, they could cancel the contracts.”
Yeah, yeah.
That’s the bear case.
That’s the worst-case scenario.
Yeah.
Yeah, that’s the only bear case.
The other thing that’s exciting, where maybe people are sleeping, is that the Cerebras ultra-fast thing kept getting delayed, and it’s not really live. But I think the average person hasn’t ever used it.
When GLM-5.2 came out, Vercel started serving ultra-fast GLM-5.2, and I was like, “Holy shit. This is such a way better experience.”
Yeah.
There are a bunch of Japanese wafer players. They’re smaller-cap, but Cerebras is an interesting trade here. They missed earnings quite badly, and their tech hasn’t been great. The best product that anyone has ever seen from them is 5.3 Spark, which just sucked.
Yeah.
It was a terrible product. And I think 5.6 Soul fast, like ultra fast on Cerebras—is going to be a user-experience game changer for Codex. I think it’s going to blow people’s minds in terms of the quality, and it will cause churn from—
Anthropic churn, yeah.
Because I’m like, okay, it’s a knife fight between OpenAI and Anthropic. OpenAI has hundreds of millions of daily active users that they can turn into an ad business. They have a better harness with Codex, and they potentially have an ultra-fast coding experience coming.
I’m like, “They’re going to be okay.” I’m not 100% confident Anthropic is going to be okay. So if everyone is betting on Anthropic, that’s the one I don’t want to be part of. I will not be buying Anthropic’s IPO.
Okay. Okay, I have 2 more for you, and I’ll let you leave. The first one is on the bio—to round out the bio stuff. Tulip King, our resident quant, loves Eli Lilly. Can you just buy Eli Lilly and pay attention to nothing else?
I don’t think so. I think the mega-caps—after Moderna made that 140% daily move and blew everyone up—it changes people’s mindsets. Nobody’s afraid to short Eli Lilly. People are going to be afraid to short a lot of these small-cap biotech stocks. The psychology of shorting small-cap biotech stocks in the current environment is going to shift.
Got it.
There’s stuff, for example, that would have been turbo shorts. Tempus AI is this very sketchy, AI-enabled biopharma stock that would have been a short to me before, but now it’s a long. You’re like, “Okay, the market cap is less than $20 billion. Someone could just buy this thing.”
Yeah, yeah.
If Anthropic bought this, the IPO would go better than it would otherwise. That becomes the story: you have these trillion-dollar companies. Cursor, for example—actually, I think the positive thing is that I didn’t talk about this, but Groq 4.6 is good and, you know, Cursor.
Now Elon's trying to buy the other Cursor clone. I'm blanking on the name—the makers of Devin.
Oh, Cognition?
Yeah. So Elon's trying to buy that because it works, right? You buy these players, they have a lot of data. Even Spirit Airlines has huge value in its customer data, and you're like—
Yeah.
Okay, you want to buy things that have a lot of data because Elon has already shown us that you can support the value of a multitrillion-dollar asset with a $60 billion acquisition that everyone else thought was insane.
Yes.
Anything trading at around $20 billion, where you're like, “I could buy that and pump my trillion-dollar stock,” you're like, “I'll do that all day,” right? The way I would express this is: buy small-cap biotherapeutics that have a data pipeline. That's going to work from a narrative perspective. Fundamentally, dude, I have no idea. That's not my specialty.
Cool. I think that was a really well-put-together thesis on the whole bio side, and I have to look into this Monkee situation. My last question for you can be a philosophical outlook, if you will. On the Bitcoin-debasement future of the US economy side, what do we need to watch for? What levers are going to have impacts? What is going to matter as far as how far this Bitcoin-gold run can press, and what do we need to look for from where things shift from this point forward? What is now important?
I think there are 2 things that are relatively important. The first is that for the market to go up—the stock market—I think you need to start to see true AI labor-force automation.
Yeah.
You need to see unemployment tick up so that the Fed can start cutting without a serious credibility issue. If the Fed cuts with a tight labor market and high inflation, it's going to cause serious problems for Fed credibility.
Right.
You need to see this kick. Number 2, the war in Iran is underrated. When the war in Iran started, you had the narrative of Anthropic quote-unquote refusing to use AI for autonomous weapons.
Mm-hmm.
Now we're deep in it, and we're like, “Wait a second. We're not winning.” It's not obvious that we're winning, and there are all these areas where we were saying that AI was going to benefit us, such as war—
Oh, yeah.
—and it's not, which ruins the ROI story. If the Iran war prolongs, you get supply-chain issues, and it also ruins the credibility of a lot of these AI-automation weaponry plays. It detracts from the overall AI thesis. It's terrible for the world, it's terrible for the economy, and it's terrible for the Trump administration.
A prolonged Middle East conflict drastically increases the likelihood of a left shift in the United States because it looks like a foreign-policy shitshow. That could cause problems. Right now, you can see on Polymarket that the likelihood of the Strait of Hormuz reopening by the end of the year has been cut in half, right?
Yes.
Prolonged Iran escalation could materially derate equities relative to fixed income, and fixed income is already smoked. Those are the 2 things you need to see. For the market, or risk assets, to pump—equities specifically—you need to see AI job automation and resolution in the war.
If you don't see those things, I think the base case is to grind lower in stocks. I think crypto is going to go up for separate reasons from that, right? The current environment is likely to result in perpetual outperformance of coins relative to stocks. If that reverses, maybe we're in Utopia and you stop having me on the stream.
Dude, you're a movie, man. You're a fucking movie. This was sick. Last question on that point: crypto goes up for different reasons. What do you think is the number-one thing people should pay attention to? Is it Bessent comments? Is it what the Treasury does? Is there anything on your horizon where, if this happens, it's doom or Valhalla, or are you sort of playing it by ear?
Dude, if the CLARITY Act goes through—which is a 25% chance—but if it does go through, then we'll actually go to Valhalla. That's a total flyer, but there could be a true God candle on just terrible things like HBAR. You'll see the dumb move, right? Like—
Why? Why?
I think yesterday you had Brad Garlinghouse and Trump say, “Thanks, Brad,” and then XRP goes up 20%, and you're like, “Hmm.” I study XRP. I'm in it. You saw the potential for this thing to move.
Canton Network—these things have been decimated because of the CLARITY Act. I'm like, okay, if we're going to go turbo pro-crypto, and Trump is making that a priority in the middle of a war, maybe he's more serious about this than we gave him credit for. Maybe he's going to wrangle people. I think that would be the Valhalla case.
If CLARITY doesn't go through, my positioning is that I like the guys who aren't going to shut down operations if their Brooklyn team gets fired. Hyperliquid is outside the US, Binance is outside the US, and Ethereum is basically outside the US. Those are my longs.
Whereas if the CLARITY Act goes through, you're like, “Okay, XRP's going to $240 or...” It's going to be retarded, right? I think that's the day trade. The second the CLARITY Act gets passed or not, you have to be there for that, and it's in September. That's a very near-term thing.
Good Alexander, dude, you're a movie. Thanks again for coming on.
He's good.
Maybe post-OpenAI we'll do Part 7 or something.
Yeah. All right, peace.
Ooh, what a movie that guy is, huh? Shout-out to Alex, Alex Good. The bio stuff is crazy, by the way. The bio stuff is nuts. I'm not gonna lie. The bio stuff is fucking nuts. Shout-out to Alex Good. Shout-out to crypto. Shout-out to the market looking pretty good, man. I must admit, that guy's a fucking movie.
All right, peace.