Compiled & Edited: Deep Tide TechFlow

Deep Tide Introduction: In the latest episode of the Convergence podcast, Wall Street macro investment veteran and 22V Research AI Macro Nexus Research Director Jordi Visser engaged in a nearly 58-minute in-depth macro and technical review with host Luke Martin. Jordi Visser pointed out that the current gloomy sentiment, where many crypto fundamentalists and seasoned participants are wavering or even exiting, is strikingly similar to the "prolonged agony" after the internet bubble burst from 2003 to 2007; conversely, during the same time frame, traditional financial billionaires and frontline large funds have started knocking on doors to seriously inquire about the fundamental logic behind asset tokenization and Hyperliquid.

Surrounding the recent 7 Sigma abnormal volatility and exhaustion signals in Bitcoin, Jordi Visser systematically explained why this bull market absolutely cannot continue to linger on "liquidity-fueled trading," and used the "ghost city" model he observed in China in his early years to predict a historic turning point activated by autonomous AI agents that will ignite the machine economy, activate the speed of stablecoin circulation, and accelerate asset tokenization. At the same time, he delved deeply into why the Federal Reserve’s interest rates and oil prices have gradually lost efficacy, how Robinhood Chain and Hyperliquid facilitate the convergence of old and new financial systems, and why AI will rewrite traditional US stock valuation logic and hand over the next structural growth window to the crypto ecosystem.

Guest: Jordi Visser (Director of AI Macro Research at 22V Research, Founder of Visser Labs, former President and CIO of Weiss Multi-Strategy Advisers)

Host: Luke Martin (Host of Convergence Podcast, Veteran Crypto Trader)

Podcast Source: Convergence / X (https://x.com/VentureCoinist/status/2120272504904110124)

Original Title: Why Everyone is Giving Up on Crypto at EXACTLY The Wrong Time

Broadcast Date: November 10, 2026

Key Points Summary

• Exits by believers and entry by traditional billionaires form sharp contrasts: Crypto fundamentalists and early staunch bulls have recently shown signs of giving up, akin to the darkest period after the internet bubble burst until the iPhone was launched; however, traditional financial billionaires at the level of Stan Druckenmiller are intensively inquiring about the long-term impact of asset tokenization on the dollar and stock market.
• 7 Sigma volatility and "price leads narrative": During Bitcoin's rise from $60,000 to $69,000, a 7 Sigma reversal occurred—a phenomenon seen only a few times in the past decade—coupled with it breaking the 200-day moving average, and previous rounds of bad news not being able to push it to new lows, confirming that "selling pressure is exhausted"; following the macro trading principle, the market always moves with price leading, and the narrative following.
• The liquidity bull market has reached its end, and the next phase must rely on fundamentals and migration of existing wealth: Relying on global liquidity expansion of 5% per year cannot push Bitcoin's $2 trillion market cap to $1 million per unit or support a total crypto market cap of $100 trillion; the truly new buying pressure must come from a trust switch and seamless on-chain conversion of the global $900 trillion existing wealth.
• "Ghost cities" analogy and AI agents acting as true residents: The crypto industry built a vast on-chain infrastructure over the past 15 years, relying on liquidity and venture capital, like the early uninhabited "ghost cities"; cryptocurrency has never been a tool designed for ordinary humans but rather an economic network prepared for 24/7 autonomous operation and frictionless pricing and settlement by AI agents.
• AI becomes a macro main variable, the Federal Reserve and oil prices no longer hold decisive power: Over the past decade, the rise of the S&P 500 has mainly been supported by the "Magnificent Seven" of US stocks, most traditional cyclical companies have been under pressure; currently, productivity changes driven by agent swarms exhibit strong deflationary effects and are highly insensitive to interest rates, even with oil prices climbing back above $100 and the 10-year US Treasury yield rising to 5%, the stock market and Bitcoin still refuse to drop due to bad news.
• Value is layered and taxed along the tech stack, with Ethereum and Solana serving as infrastructure toll booths: As consumer-grade agents (like Instinct AI, valued at $10 billion) and autonomous research nodes become widespread, agents use stablecoins for zero-friction high-frequency trading on-chain, and the underlying Arbitrum, Ethereum, and Solana of Stripe and Robinhood Chain will capture trading volume benefits layer by layer under the "city tax" logic.
• Robinhood and Hyperliquid facilitate the convergence of old and new finance and break the ice on tokenization: The "silent IPO" has caused early libertarians, who are resistant to traditional state interventions, to opt-out, while Robinhood, amid a dry spell in crypto venture capital, is leveraging fiat funds to promote US stock tokenization and on-chain business, alongside Hyperliquid's provision of all-weather perpetual pricing discovery on weekends, forcing Wall Street to no longer ignore crypto assets. Following the tokenization of stocks, private credit and real estate liquidity relief will become the next explosive battleground.
• S&P valuation compression and "AI-native microbusinesses" will reshape long-term return patterns: Although US stock earnings have grown about 30% this year, S&P valuation multiples have significantly compressed; leading stocks like NVIDIA and Micron have seen valuations drop to single-digit or low double-digit ranges, reflecting market concerns over their "terminal value uncertainty"; in the future, AI democratization will spur a massive number of micro-enterprises with annual revenues of $200,000 to $2 million to eat into the market share of giant companies, with growth premiums gradually spilling over from the US stock market, valued at as much as $70 trillion, to the crypto market, which is under $4 trillion.

Outstanding Perspectives

• “I don't care why people entered the world of cryptocurrency in the first 16 years; what I truly care about is AI agents."
• “Whenever I see an asset refusing to fall in the face of continuous bad news, it just means that the sell-side has completely exhausted itself; once an upward breakout occurs, to me, it's a confirmation signal that bears and sellers have exhausted their firepower. Always remember, price leads narrative.”
• “Relying on global liquidity growth of 5% per year is impossible to sustain a crypto ecology with a market capitalization of $100 trillion; you must have those managing the $900 trillion of existing global wealth genuinely trust this system and create real usage demand.”
• “If I compare it to the 'ghost cities' I saw in China years ago, the current crypto world still feels like an empty city lacking daily economic activity; just like the App Store spawned Uber and Airbnb, AI agents are the key to truly filling this city with economic participants.”
• “Everyone should now stop fixating on the Federal Reserve, stop obsessing over oil prices, and stop monitoring interest rates. Using the macro textbooks from the pre-AI era to apply to the economic operational rules already in the AI era is itself a mismatch.”
• “Buying a $1 million building turns that money into inactive 'dead money’; you can't use it to buy a coffee at Starbucks; once real estate and various assets are tokenized, activating the $900 trillion of dormant wealth behind the $120 trillion of liquid currency will absolutely rewrite everything regarding the velocity of money.”
• “I have never agreed with those fundamentalists who argue for 'never merging with the old system.' The new and old systems are destined to merge, and in the end, the new system will gradually absorb, digest, and swallow the old system.”
• “Today, the total market value of US stocks has reached $70 trillion, accounting for 72% of the global stock market, largely due to America's dominance over the previous generation of the internet; but AI will break this monopoly and achieve capability democratization. Stocks will find it hard to continue the past growth myths, while the crypto world, valued at under $4 trillion, is standing at the new growth curve's starting point.”

Main Text

When Believers Begin to Waver and Exit: From Internet Bubble's "Prolonged Agony" to Traditional Financial Giants Knocking on the Door

Luke Martin: This week I have been continuously reflecting on a seemingly counterintuitive point you raised earlier: you believe the market must first see those who were once deeply convinced of cryptocurrency stop believing in it before the actual market arrives. Why must it come to the point of wavering faith?

Jordi Visser: This is closely tied to my personal memory of experiencing the internet bubble. In 1999, I moved back to the US from Brazil, and that stark contrast still feels surreal to me today. I witnessed adult men crying on the streets in Brazil, enduring an economic depression and severe currency devaluation, a whole emerging market struggling in turmoil for two years. Yet upon returning to America, I took over a massive trading account at Morgan Stanley, where everyone around me was boasting about their internet stocks rising from $2 to $100, even my friends from blue-collar families in high school were talking about trading their AOL stock for a house.

As expected, the bubble eventually burst violently. Just four years later, those who had been bragging about their successes did not even want to mention the word stock anymore, as their accounts were filled with chips worth only a cent, and large numbers of companies went belly up. Those who entered purely out of greed were completely washed out, leaving behind only disheartened structural builders and idealists. It was now 2004, but don't forget there were still three years until the iPhone was born. From 2003 to 2007 was an exceedingly long period of consumption akin to a "prolonged agony," where the ecosystem seemed to be continuously shrinking, with no end in sight.

The current crypto world gives me the exact same sensation. Recently I had dinner with many insiders who were previously extremely bullish at various levels and even leveraged long, but now they are beginning to doubt whether Bitcoin itself has a future. Just yesterday, I participated in an interview where even a speaker who can be considered a Bitcoin extremist admitted that this setup no longer seems to work. But for me, I don't care why people entered this industry over the past 16 years; what I truly care about is what is going to happen next: AI agents.

Luke Martin: Besides that interviewee who surprised you yesterday, what other conversations among traditional finance leaders or seasoned insiders managing large sums of money have impacted you recently?

Jordi Visser: Anyone who has observed the market long-term can notice an interesting divide: 90% of daily discussions occur within the crypto fundamentalist or retail communities, with extremely pessimistic sentiment; meanwhile, an increasing number of heavyweight figures in traditional finance are privately seeking me out to discuss the truly critical issues. They approach me because my background is in macro strategy trading, and they believe I can provide answers in a language they can understand.

A few days ago, when the Federal Reserve’s interest rate direction was announced, the crypto circle was filled with doomsday rhetoric, believing everything was over. But in my mind, I am very clear that a real bull market absolutely cannot continue to rely on liquidity. What’s emerging in the traditional world now are the truly substantive questions: for instance, what does asset tokenization (Tokenization) really mean? Why would a big player like Stan Druckenmiller invest in Hyperliquid?

In the past 10 days, two billionaire-level investors contacted me one after the other to discuss details, asking core questions like "What impact will tokenization have on the dollar's status?" and "What does tokenization mean for the stock market?" One side is disheartened believers, while the other side is the top old money rigorously studying the fundamentals just before a potential shift, illustrating that the foundations of a major market trend are being quietly laid.

Bitcoin's 7 Sigma Reversal and "Price Leads Narrative": Why This Bull Market Must Be Driven by Fundamentals?

Luke Martin: Speaking of price movements, during Bitcoin's rapid ascent from $60,000 to $69,000, you pointed out on Twitter that this was a "7 Sigma" level reversal, which has only occurred three times in the past ten years. For those not familiar with statistical concepts, were you trying to hint that everyone should quickly chase the price? What is the true significance of this signal in history?

Jordi Visser: First, "Sigma" is absolutely a fundamental concept every market participant should understand. You can think of it like going for a physical examination at the hospital: suppose a doctor measures your blood pressure at 120 over 80, and it seems normal according to regular standards; however, if you usually maintain at 105 over 70, then 120 over 80 is actually elevated for you. Abnormality must be judged relative to your own recent baseline. In the market, volatility serves as that baseline. Due to Bitcoin's recent 60-day rolling volatility being compressed to a very low level, for example, with an average daily fluctuation of only 1.5%, if suddenly a 10% unilateral price increase occurs, it equates to approximately a 7 sigma super volatility event.

More importantly, I don't just look at this one specific number, but rather multiple resonance signals. First, the price broke above the 200-day moving average; second, the 200-day moving average itself showed signs of turning upward within just a day or two; third, prior to this, the market experienced a series of intense bearish impacts, including security theft incidents, turbulence in the AI sector, and the probability of the US "Clarity Act" bill being passed falling below 20%, yet Bitcoin remained firmly around $65,000, refusing to create new lows. When an asset can nonchalantly digest various bad news and refuses to fall, it only signifies one thing: the sellers in the market have nothing left to sell. The subsequent surge, even if sparked by liquidity news that doesn't seem particularly significant, has fully exposed the panic from shorts and the replenishing buying momentum from cautious funds.

Luke Martin: My first reaction to that K-line immediately reminded me of the moment back in 2019 when Bitcoin surged from $3,000 all the way to $4,500 and even $5,000. After that spike occurred, everyone started scrambling afterward to attribute various reasons for the price rise. From your years of macro trading experience, is it really important to attribute the cause (whether it's due to the Federal Reserve, macro improvements, or moving average breaks) when the trend just begins?

Jordi Visser: The answer is simple: price always leads narrative. Once the price takes off violently, people will inevitably weave together a narrative they are willing to believe. Over the years, I have discussed this numerous times with Paul Tudor Jones, and we completely agree on this point: after a trend emerges, the market will rush to label it with explanations.

As a macro trader focused on long-term structural trends, I have always been seeking those fields that stand at the starting point of a massive trend, and I believe the crypto market is about to enter the largest scale bull market in history. I am so certain of this because the epic and severe clearing that occurred from 2021 to 2022 has already happened. Unlike typical four-year cycles, that crash completed an exceptionally thorough bubble cleansing, logically very similar to the burst of the internet bubble.

History shows that the internet took off around 1994 and didn’t see a massive explosion until 2009 with the launch of the App Store, spanning a whole 15 years. During this period, the market was unsure of how to value a disruptive technology, absurd indicators like pricing by clicks or eyeballs were invented, and the Amazon stock bubble once burst, only falling below a price-to-earnings ratio of 100 in 2014 or 2015, which also birthed new valuation paradigms like SaaS.

Looking back since the Bitcoin white paper was released, we have traversed a similar duration. The internet bubble left behind a pile of surplus fiber optics and cleared debts, while the crypto industry has built up surplus infrastructure after a massive influx of venture capital, and now as valuations fall, many young people's losses or distressed assets are being picked up by seasoned industry veterans who once controlled Fortune 500 companies. This handover of chips indicates that speculation has been stripped away and true structural innovation has begun.

Luke Martin: You just emphasized that Bitcoin over the past decade was indeed a standard "liquidity trade." Many people rush in without thinking when they see headlines about money printing, but you believe that in the future, liquidity-driven strategies will no longer work. Many listeners may have difficulty grasping this idea. Why will "liquidity logic" not work anymore?

Jordi Visser: Because, at a certain scale, size determines everything. When Bitcoin has reached about $2 trillion in market capitalization while the total amount of global assets is around $900 trillion, if someone wants to project a roadmap for Bitcoin to reach $1 million and the entire crypto ecosystem to reach a market cap of $100 trillion, relying on global liquidity growth of 5% per year is absolutely impossible.

The early idealists who propelled this round of experimentation have either cashed out or shifted their investments to AI. Moreover, merely allocating 1% of portfolios by wealth management advisors cannot support that scale. You must enable those truly controlling the $900 trillion in global existing assets, such as pension funds and conservative capital, to genuinely develop trust and a real need for use. If you ask anyone over 70 years old, almost no one truly understands and believes in the narrative that "Bitcoin is digital gold" as a store of value.

To make them switch their fiat assets, two prerequisites must be met: first, traditional fiat assets can no longer just passively earn; second, there must be a zero-threshold bridge. You cannot expect a 70-year-old to navigate the complex process of private key mnemonic phrases and wallet setup—this process must be instantaneous, with AI agents performing automatic conversions in the background and relying on tokenized assets for seamless transition. All of this relies on fundamental technology, not on central bank money printing.

The "Ghost Cities" Analogy from China and Deleveraging Deflation Induced by AI: Why the Federal Reserve and Oil Prices Are No Longer Important?

Jordi Visser: Speaking of infrastructure leading and lagging demand, I used a metaphor in my article today. From the mid-2000s until 2013, I would spend an entire month in China every year. During that time, I witnessed some so-called "ghost cities," where I saw vast, magnificently built but empty high rises, and personally felt how a wave of infrastructure driven solely by abundant liquidity works.

In my view, up until now, the entire cryptocurrency world essentially remains a "ghost city"; real economic activities do not operate in it daily. But now, with Stripe investing heavily in acquisitions and Robinhood making a strong entry into building its own chain, traditional finance is seriously putting money in, meaning that "residents" are finally about to arrive. However, everyone must understand a core disruption: some of China’s ghost cities have, in the end, not seen enough human residents; the reason this digital ghost city of crypto can be swiftly filled is that cryptocurrency has never been designed for humans; it was born specifically for digital agents and AI agents. Just as there would be no Uber and Airbnb without the App Store, AI agents are the main economic participants that will activate and illuminate this crypto city.

Luke Martin: This perspective of "ghost cities activated by agents" is indeed profound! Before we dive deeper into how agents ignite the economy, I want to ask you, using your expertise as a senior macro economist: It seems recently the signals released by the Federal Reserve are becoming increasingly ambiguous. Everyone is tangled up in whether they are still indirectly easing, especially with Kevin Warsh stepping in to open a new era compared to Jerome Powell's time. What impact do the Federal Reserve and oil prices still have on the market?

Jordi Visser: My advice is straightforward: everyone should immediately stop fixating on the Federal Reserve, stop being obsessed with oil prices, and stop overly关注利率走势。我真想不明白为什么这么多人经历过这么多轮教训还没看透。回看从 2009 年到 2019 年再到新冠疫情这整段周期，很多人喜欢把美股走高简单归功于量化宽松（QE），可事实恰恰相反：当年房屋建筑商、汽车厂、航空公司等标普 500 中的大多数传统企业根本没有实质增长，甚至长期处于衰退之中；全靠美股七巨头（Mag 7）硬生生把自己在标普中的市值权重撑到了 30%。

今天我之所以断言美联储已经没那么重要，是因为当前整个经济引擎已全面切换至 AI 驱动，而 AI 产业对利率几乎完全不敏感。未来一到两年内，智能体集群（Agent Swarms）将释放出史上前所未见的超级通缩生产力。比如近期引发轰动的纳维斯托克斯方程突破，这可是 Albert Einstein 之前就悬而未决的世纪数学难题；7×24 小时协作工作的 AI 智能体群不仅能攻克人类攻不破的科学问题，甚至能轻而易举攻破人类设立的安全网，说明人类在智力效率上根本没法与之相提并论。

再看看油价和利率的数据对比：我在 2007 年经常跑中国那会儿，布伦特原油冲到了 110 美元，今天的油价甚至还不如 2007 年高，可同期的标普 500 涨幅高达两三倍，居民资产翻了三倍，一个 110 美元不到的能源单品在一个规模早已翻番数倍的全球经济里还能翻起多大浪花？至于利率，美国绝大多数有房家庭早把按揭贷款利率锁死在了 3% 出头；加息 500 个基点和 10% 通胀当年尚且没把 2022 年砸进衰退，而真正握有庞大资金的亿万富翁群体反而还从高利率利息中受益。当前企业盈利依然保持着约 30% 的高速增长，还在拿前 AI 时代的利率油价衰退剧本去套用 AI 时代的市场，简直是在开玩笑。

Luke Martin: 听你这么拆解，以后确实不用天天盯着美联储官员或 Kevin Warsh 的只言片语提心吊胆了。但如果退一步问，尽管基本面脱离了传统宏观束缚，那帮手握重金的传统基金经理们，心理预期难道完全不受油价或美债收益率影响吗？

Jordi Visser: 可以直接拿近期盘面做检验。标普 500 从 6 月份以来基本处于横盘走平状态，但你看这期间发生了什么：6 月 25 日以来，油价从 70 多美元一度蹿升突破了 100 美元，十年期美债收益率也从 4.30% 爬升到了 5%。按照老剧本大盘早就该崩了，可标普居然只是横盘不动！如果连油价破百和美债冲上 5% 都压不垮股价，那你设想一下，只要接下来十年期美债利率回落区区 20 个基点、油价稍作回吐，市场会爆发多大的反弹动能？

更关键的是，目前所谓「Mag 8」（我把 AI 核心标的 Oracle 也纳入其中并列八巨头）距离历史最高点仅仅只差 1% 左右，而且今天八只股票里有七只重新站上了 50 日均线。今年全年只有三天出现过八只同时站在 50 日均线之上，如今七只齐聚上方。只要大盘抗住这波坏消息横盘再多挺几天，它的形态就和刚刚走出卖盘枯竭的比特币一模一样：说明所有最坏的预期都已经计价完毕，而在中期选举即将来临前的财报季里，企业盈利仍在高速奔跑。

AI Agent as Economic Entities: Why Machines Choose Cryptocurrencies and Stablecoins Instead of Traditional Dollars?

Luke Martin: After discussing macro pressures, let's return to the core engine you are most optimistic about: now that the on-chain infrastructure is ready, like the over 30 agents running on my computer screen, some help me clean my inbox at night, intercept spam, while others help me schedule my day, although sometimes they do things I didn't authorize. Against the backdrop of the crypto financial infrastructure being ready, what types of AI-driven commercial activities will spark real growth in Bitcoin and the crypto economy first?

Jordi Visser: The ones running on your screen are called "personal efficiency agents," and they only save you time; what will truly explode economic activity are "consumer agents," which directly engage with external commercial transactions. Last week, Meta launched personal AI assistants, while private firm Instinct AI, focused on life agent execution, saw its valuation jump from $2.5 billion to $10 billion. Payment giant Stripe is also making significant acquisitions in the upstream and downstream of consumer payment technology stacks, including acquiring the routing platform OpenRouter. All signals point to the same future: autonomous intelligent agent networks are fully taking over transactions.

You may have heard of Leopold Aschenbrenner's massive report "Situational Awareness," but I doubt many have really flipped through the chart on page 71. That chart clearly points out that once AI crosses the threshold of "auto-research agents," it will enter a phase of absolute parabolic exponential take-off, including tackling cancer and solving foundational scientific breakthroughs like the Navier-Stokes problem; rumors have it that several leading laboratories in the industry have quietly cracked multiple world-class math problems internally.

When Leopold Aschenbrenner wrote this report in 2024, he predicted this turning point would arrive in 2027, but the real progress today is evidently outpacing that forecast! If such top-tier autonomous research nodes are already approaching, consumer-grade agents that are technically less demanding are bound to experience a surge in popularity.

This is precisely what I have been advocating for: the convergence of the macro and fundamentals. In the traditional stock market, although macro fund managers and fundamental fund managers focusing on quarterly results have different styles, they ultimately look at the same equity profit logic; however, the crypto circle has long been extremely fractured, with Bitcoin Maxis focusing solely on macro and libertarianism, and ecosystems like Ethereum and Solana solely discussing applications and fundamentals, with mutual disdain. I firmly believe in the fusion of both.

Imagine, even I, a professional immersed in AI, find the process of opening a wallet on Robinhood Chain to be cumbersome; there is no way a 70-year-old could endure such an experience. Yet consumer-grade agents completely flatten this threshold; for instance, Instinct AI can already handle a visa to Brazil and the entire payment process on its own.

I encountered a particularly typical real-world case: recently, Delta Airlines was accused of using AI agents to identify users’ past privacy and payment capabilities for dynamic price gouging. A friend told me he checked a Delta airline ticket, and when he returned an hour later, it had unexpectedly increased by $130. I directly told him it was clear that the other side had deployed an AI pricing agent; facing such unfair information asymmetry, your only solution is to send your own personal AI agent to negotiate against the airline's AI agent, with AI versus AI in a duel, to restore market fairness. Historically, there must first be consensus around Bitcoin, and then rely on a large on-chain ecosystem to support it; we are now entering a new phase led by application ecosystems.

Luke Martin: But I must raise a sharp question for the rationalists. Coinbase founder Brian Armstrong has previously strongly advocated that agents should conduct 24/7 transnational settlements using crypto wallets; Coinbase itself has also early on built the related infrastructure; but I have seen many extremely smart AI entrepreneurs in Silicon Valley rebate and say, “Why must I make AI use cryptocurrency? Can't I just bind my credit card information for the agent to settle in dollars?” Even I found myself swayed after hearing that. Why do you firmly believe that agents will ultimately choose Bitcoin and stablecoins rather than continue using dollars directly?

Jordi Visser: First, I completely agree that in the current stage of high-frequency micropayments, stablecoins will certainly be the primary; Bitcoin will take on the function of large settlements and core collateral. For example, if you need to transfer $100 to someone, you can easily do that with software like Venmo; however, once it involves transferring assets on the level of $1 million, ordinary consumer payment tools cannot handle that at all, at which point Bitcoin must be relied upon as an absolutely trusted ultimate clearing settlement collateral.

The more fundamental determining logic lies in the "conflict of zero friction and profit structure." Traditional dollar payment settlement networks (whether issuing banks or payment gateways) are fundamentally commercial entities; they have shareholders to serve, with profitability and commission targets to meet, which determines that the traditional system can never physically reduce transactional friction and fees to zero. The essence of the AI era is everything trending towards ultimate deflation; AI agents are naturally inclined to pursue the lowest possible costs and the fastest speeds to churn out the optimal solution.

Again, using the previous example of the Delta airline ticket: when a human opens a browser to visit the airline's official website, inputs information, and places an order, during that time, the background AI agent has already compared prices across 17 different airlines, compressing all intermediary markups to the limit. The same will happen in global currency exchanges and commercial payments; as long as both merchants and users start using agents to do the work, agents will not want to wind their way around to support a bunch of traditional financial intermediaries. Currently, the only need to convert to dollars is merely because physical coffee shops have not fully adopted on-chain payments; once we enter the phase of direct communication between agents, the zero-friction channels based on stablecoins and crypto rails will absolutely prevail.

Think back to the early days of mobile internet; my stepfather refused to try Uber even once for seven years, yet later he was still able to use it effortlessly. Once those who first use agents can automatically scour StubHub for low-priced NBA Knicks tickets or constantly monitor a restaurant for the 7 PM cancellation spots, young people will rapidly disseminate this benefit and then teach it to their parents and even grandparents. As for the security and risk control issues that people are concerned about, they can also be easily resolved—just like we set spending limits for Venmo, we can hard code a limit of $100 or $2000 on the operational wallet dedicated to agents.

This hides a blind spot that most traditional investors have yet to realize: once a paradigm crosses from 10 people using it to 1000 or even 10,000 people, the on-chain fundamental expansion driven by AI agents will be out of control exponentially. In May of last year, during the tariff fluctuation and stock market rebound, I wrote several reports discussing the AI inference power explosion; however, when I ran to hedge fund offices to recommend it to the world's smartest fund managers, they couldn't even be bothered to listen to what inference meant, solely focused on tariffs. My concern was that by the day they finally paid attention to researching inference, Micron's stock price would have already surged tenfold.

What is about to unfold in the crypto market is precisely the same script: fundamentals will spiral wildly under the drive of AI agents; by the time the general public hears about this narrative in the news, asset prices will have already been dramatically repriced.

From Grokbot to the "City Tax Model": With Millions of Agents Trading, Who Captures the Value?

Luke Martin: Speaking of the speed of agent evolution, I've deeply felt this over the past couple of months using Grokbot. Previously, I would have it track some data without stopping for a week, but it would often slack off after just one day; now, while it still has a ways to go before truly achieving "recursive self-improvement (RSI)," it can remember the markdown background files I provided even before I issued a command, and it will even proactively communicate with another agent to inform me, "Luke just mentioned this." This change feels incredibly sci-fi!

Jordi Visser: This is why I emphasized the "empowerment and ease of use" of Grokbot in the week it was released. It made what was originally very hardcore agent scheduling as fluid as sending a text message or playing video games. You only need to set up a main AI agent like "Jarvis" with a few prompts; even while you and I are having this video call, I can send a text from my phone to have Jarvis coordinate the proxy clusters managing technical forms, automatically fetch the list of stocks that have strengthened over the past three days with matching volume and price, and in two minutes provide me with accurate return results.

Just a few days ago, I had a firsthand experience: I needed to change my New Jersey driver's license to a New York license. I checked the DMV website, only to find out that no appointments were available for a month. Instead of waiting indefinitely, I let AI search platforms like Reddit and X to see if anyone had encountered the same issue and found a workaround. In under 30 seconds, the AI pinpointed a particular step leading directly to the appointment entrance, and I made an appointment for the next day; yesterday, I took just 10 minutes to sort it all out. Everyone should start forcing themselves to get into the habit of assigning AI to tackle complex tasks; only by training the brain into this cooperative model can they keep up with the upcoming wave of Instinct AI.

Luke Martin: Now I need to switch to a trader's instinct to do some calculations. Assuming that in the near future, there are millions of AI agents using stablecoins for high-frequency trading, in this value network, where will the profits settle? Will it be with foundational model companies like OpenAI and Anthropic, or who will it be?

Jordi Visser: Value will be captured simultaneously at multiple key points along the entire technology stack, but the core is to figure out the "toll collection" logic and the terrifying explosive power of transaction volume. On a deeper level, foundational model vendors and computing power infrastructure will continue to benefit; on the transaction settlement and on-chain circulation side, suppose all transactions run on stablecoins, then Stripe will collect a toll at the front-end fiat and commercial interfaces, subsequently, second-layer networks on Ethereum, like Robinhood Chain running on Arbitrum, will harvest massive invocation fees and ultimately tax revenue down to the base layer of Ethereum itself.

I believe the public has no concept of the "machine concurrent trading volume." Recall that during Google’s earnings call last April, Sundar Pichai announced that Google Cloud's token consumption had soared eightfold year-over-year, reaching nearly a scale of 90 trillion, and it soon shot up towards a trillion-quadrillion mark. This is the genuine fear driving the giants to massively increase capital expenditure (CapEx). When tens of millions or even billions of agents are concurrently operational 24/7, the number of trades will be an unimaginably exponential figure.

Once we understand the scale, we must also resolve the most crucial deadlock in macroeconomics: "Why did liquidity fail in the crypto space before?" Because in the past, humanity locked the massive wealth released by easing into illiquid "dormant assets!" Even though Raoul Pal and I often discuss global liquidity, I must point out: globally there is actually only about $120 trillion in real cash currency, yet this correlates to values as high as $900 trillion in book assets, supported by immense fiat leverage. When you take $1 million to buy a building, that money becomes completely locked in the concrete and rebar; you cannot tear off a piece of the building's outer wall to buy a coffee at Starbucks, making that $1 million "trapped wealth."

If the total value of global assets (including approximately $3 trillion in crypto assets) is about $903 trillion, the core trend in the future will not be a collective collapse of asset prices to zero like during the Great Depression; instead, as traditional fiat leverage gradually returns to rationality and assets unblock liquidity through tokenization (with circulating currency amount evolving from $120 trillion to $200 trillion, and stagnant real estate being fractionally split into on-chain tradable units), dormant assets that were locked will suddenly gain transactional capability! This is precisely what Caitlin Long has long emphasized with "Velocity of money": the faster transactions occur and the higher the turnover rate of funds, the higher the GDP output and protocol fee income sustained by the same amount of funds.

Ethereum is like an ever-expanding city, it does not need to monopolize every front-end application; it merely attracts Arbitrum, Robinhood and various competitors to build atop its city through secure consensus and absolute trust, charging foundational security taxes from all commercial agents that settle within. Meanwhile, Solana serves as a high-speed financial hub to accommodate the fast transaction demands. As long as trust exists, the foundational toll booth will benefit exponentially with the city's prosperity.

Robinhood Chain, Hyperliquid and the "Silent IPO": How Traditional Finance and Crypto Devour Each Other in Convergence?

Luke Martin: You just mentioned Robinhood Chain, which is undoubtedly the most impressive move by a traditional finance or Web2 company to cross into the crypto space. CEO Vlad Tenev's speed in driving this initiative feels almost military-like, with a leading edge in the US stock tokenization path far exceeding everyone’s expectations. Moreover, looking at Robinhood's Q2 earnings report this year, the revenue from crypto derivatives business, including projected markets and on-chain trading, has officially surpassed that of traditional stock trading, and this has yet to account for the share from their own blockchain. Many people see stock tokenization as merely realizing 24/7 trading; why are Vlad Tenev and Larry Fink of BlackRock so excited about it?

Jordi Visser: This directly touches upon the core of my previous in-depth analysis in "The Silent IPO": why many early crypto fundamentalist idealists chose to cash out at this juncture. The early geeks who built this system sought absolute freedom and decentralized resistance, hence they rejected government interventions, disliked politicians issuing meme coins, and regarded Larry Fink entering grandly with Wall Street giants as a distortion of their original intention, leading them to shout "I'm out" and liquidate. However, I pointed out from the beginning that historical laws will never allow new systems to remain in perpetual confrontation with the old; the new and old financial systems will inevitably converge, and the final outcome will be that the new system will gradually absorb, rewrite, and swallow the old system.

Looking back at Robinhood from this historical point, you can understand the exquisite misalignment: we are currently in a valuation and financing winter in the crypto industry, where a large number of startups struggle to secure venture capital; however, Robinhood, sitting in the fiat stronghold, has a vast amount of real cash flow lounging on its balance sheet. They are promoting US stock and securities tokenization vigorously, acting as a conduit that "bootlegs" the immense funds of traditional finance into Ethereum and Arbitrum’s crypto bases! More traditional users will naturally own on-chain wallets because they hold tokenized US stocks, which will lead to tangible new funds being deposited. I have designed an ecological index that encompasses 10 core crypto tracks and includes 40 tokens, plus 6 deeply cross-industry listed companies, aiming to benchmark Bitcoin against performance indicators that traditional Wall Street can understand.

Interestingly, I recently observed that many large investment banks like Morgan Stanley are continually upgrading Robinhood's stock ratings, yet the reports omit any mention of "cryptocurrency." The reason is very simple: they haven't fully grasped this on-chain economic mechanism, so they cannot explain it clearly to high-net-worth traditional clients, and therefore, they have to boost the stock ratings through traditional earnings growth narratives to seize the coverage opportunity.

In this wave of all-weather perpetual pricing integration, whether it is the perpetual contracts (Perps) of Coinbase or the sensational platform Hyperliquid, they are rapidly breaking the barriers of traditional stock market closing mechanisms. Recently, during a period of external panic, many asked me during the halftime of an American football game on Sunday: "How much will the S&P 500 open down on Monday? How high will oil futures spike?" I simply opened the perpetual contracts on Coinbase and told them: at that moment, it was already down 28 basis points, and it would basically open around that level. The other party was taken aback, completely unaware that there were real-time continuous quotes during the weekend closure.

This is the deepest instinct why Stan Druckenmiller laid out Hyperliquid through publicly held PURR. Given my understanding of his investment style (similar to how he keenly captured shifts in Argentina's political landscape and market changes), I can easily picture the scene: one weekend, he and his team discuss the expected opening price for assets like oil or SpaceX, and his team reports, "No need to wait for Monday; it is already trading in real-time on Hyperliquid, and the opening price will basically follow what's listed there."

For traders on Wall Street, who have spent their lives anxious during weekends waiting for market opening on Sunday night, being able to access a continuously real-time market displaying genuine supply-and-demand games, all year round, is an irreversible dimension reduction strike; once experienced, they will never return. This is also why there are almost weekly instances on Twitter of seasoned analysts suddenly learning from passersby about Hyperliquid’s real-time quotes and quietly becoming the next batch of buyers.

From Tokenized Stocks to Private Credit Crises: What is the Next True Trigger for RWA's Explosive Unfolding?

Luke Martin: Currently, all market focus is almost exclusively directed at Robinhood's "US Stock Tokenization," which is indeed reshaping the trading experience in real-time. But after US stock tokenization, looking ahead to the next phase of evolution regarding real world assets (RWA), what do you believe will be the next explosive asset category similar to US stock tokenization? Is it the vast global bond market, or some other underestimated corner?

Jordi Visser: In my view, the next most explosive and urgently needed domain for on-chain tokenization to break the stalemate is private credit, as well as the entire private equity and real estate sector more broadly.

The reason is straightforward and urgent: currently, the entire traditional private credit market is facing a serious liquidity deadlock, with vast amounts of capital locked up without movement. We can clearly see that various business development companies (BDCs) are trading at significant discounts; take a look at how firms like Blue Owl, KKR, and Apollo have invoked redeem crises and public opinion issues after siphoning large amounts of retail funds into illiquid products in the public market. There has always been a demand for allocation to private credit assets (for example, insurance funds require long-term stable allocation to these assets), but due to the lack of publicly continuous market-based circulation and precise repricing mechanisms in traditional structures, assets have been caught up in liquidity deadlocks and severe mispricing on balance sheets.

If someone builds a secondary liquidity exit channel through an on-chain tokenization structure (in fact, sharp giants like Apollo are likely already actively researching and promoting such solutions), telling those investors who need to cash out urgently, "You can tokenize the locked amounts you hold; a buyer is willing to buy at a transparent market price with a discount of 20% to 30%, and you can get cash out today," this would instantly explode a super headline effect on Wall Street!

Tokenized stocks solve the issues of breaking geographical barriers for cross-border investors (for example, allowing investment in Japanese stocks without needing to open a Japanese account) and discarding outdated ETF index constraints, enabling 24-hour freedom to assemble global assets; the space itself is already immense. Meanwhile, in terms of collateral, BlackRock's money market fund has already pioneered this process. The next entity that can leverage tokenization to unblock private credit and global real estate, who holds the true valve for the next RWA explosion.

AI × Crypto Is Not a Zero-Sum Game: Stock Valuations Peak and the Rise of "Micro AI Native Businesses"

Luke Martin: The last core question also represents a mental model conflict that’s been lingering in my mind for quite some time. Over the last few years, as AI has been wildly bullish, I instinctively thought that AI and cryptocurrencies operate in the same high-risk frontier asset track globally; that if AI heats up too much, it would directly siphon off attention and capital flows from the crypto market. Thus, whenever Dario Amodei or Sam Altman throws out a significant dynamic that triggers an AI rally, crypto often lags. Yet, after reading your recent in-depth article, I found that you entirely overturn this either-or perception, suggesting that the two can actually form a synergy and even propel each other upward. Can you break down the inherent connections in detail?

Jordi Visser: This all starts with the real evolution pace of the AI investment wave in the past few years. The first wave of AI hype in the first two years after ChatGPT's emergence was highly concentrated on NVIDIA and a few super-scaled cloud giants. Then, last year, capital began to massively shift towards underlying infrastructure and inference computing (Inference). In the AI infrastructure sector index that I monitor, at least 40 stocks doubled over the past year, and several even soared by 200% to 500%, demonstrating explosive curves comparable to crypto altcoins.

However, this round of bullishness fundamentally differs from the internet bubble, and it serves as a deafening wake-up call for all investors clinging to traditional stocks. Stocks are essentially a "hope trade” and "growth premium trade": Investors buy stocks hoping for an annual expected return of 15% to 20%; after all, in the years past, bonds have long been in a bear market, and in a high-interest and inflation environment, relying on leverage to buy properties and collect rent is no longer fragrant. The liquidity of venture and private equity has been constrained, and stocks have almost become the only growth game left in the city. But the problem is, looking forward three years from today, the high growth premium of publicly traded large caps may face severe structural endings!

Just look at this year’s data to sense the anomaly: S&P 500 constituent earnings have surged by approximately 30% this year, yet so far, the increase in the S&P benchmark is only around 11% to 12% (falling to around 17% for the year), indicating that this year is actually one of significant "valuation multiple compression." Many are puzzled why, given the enormity of AI, such a 100x PE ratio bubble hasn’t emerged. In reality, it’s precisely the opposite: the P/E ratio of storage infrastructure leaders like Micron has plummeted to just 2 to 4 times, and even absolute leader NVIDIA's forward P/E ratio, based on next year’s earnings, hovers around only single to low double-digit levels!

On the surface, this seems extraordinarily absurd, but Wall Street capital’s sense of smell is incredibly sharp: the reason stock prices do not dare to rise synchronously with high earnings is that there is fundamental doubt about the 'terminal value' of all listed giants.

No one knows whether these giants today can maintain their moats three years from now. Any slight iterative breakthrough in technology route or AI evolution by new emerging players could flip established giants overnight. Once the assumption of perpetual cash flows in asset pricing models is shaken, traditional giants' valuation multiples can never rise again.

Simultaneously, the reason the US stock market was able to possess a market value of $70 trillion and monopolize 72% of the global stock market valuation over the past twenty years was because the platform benefits from the internet were nearly monopolized by American giants (similar to why Germans might wonder why over 90% of their country's search engine market share goes to Google). The essence of AI is extreme democratization of capabilities.

The truly vibrant commercial ecosystems in the future, I even greatly desire to see an ETF index that focuses on acquiring micro-companies worldwide with “annual revenues of only $200,000.” In the past, such mom-and-pop shops couldn’t scale up, but in the future, countless one or two-person "AI-native micro-enterprises" can wildly jam into various software and business gaps; for instance, taking a small $200,000 business slice from Amazon, Google, and traditional software giants, simply by pursuing multiple lines simultaneously, they could easily generate annual revenues of $2 million.

While an individual micro-enterprise may seem inconsequential, within the scale of billions of people and countless agents operating simultaneously, they will constantly eat into the profits and growth spaces of traditional listed companies.

Thus, the conclusion becomes exceedingly clear: in the early stages, indeed, the computing infrastructure stocks possessed both hardcore earnings support and the ability to surge several times or even ten times, temporarily attracting a lot of smart money that would originally flow to crypto; however, as traditional US stocks lose that certainty of high valuation growth myths due to AI disruption, when capital focus is forcibly redirected from static P/E ratios back onto “velocity of money” and the underlying protocols of the machine economy, the crypto world, which still holds under $4 trillion in total and naturally fits the seamless collaborative operations of agents, will officially take up the lead as the world's most imaginative new growth story.

Luke Martin: Jordi Visser, the information density today has been incredibly high over the past hour. Thank you so much for sharing these cutting-edge thoughts on the show!

Jordi Visser: My pleasure, I enjoyed this conversation as well, see you next time.

Luke Martin: Thank you all for listening, and we’ll see you next time.