In this week’s video, I walk through the “time mismatch” theme I presented to Freedom Tech in DC this week. There are two economies running at two speeds. In “human time,” housing, autos, retail, restaurants and the Russell 2000 are underpressure as rates move higher. On the “ghost rails,” AI agents, compute infrastructure and tokenization are accelerating. The speed of intelligence is changing, and the speed of money has to catch up.
The rates scare isn’t showing up in the data. The LEI just turned positive with no recession. S&P earnings are up 17% year over year, and credit spreads aren’t widening. Jobless claims aren’t budging, profit margins are still rising and money market funds keep filling. Meanwhile, only 2.2% of people aged 55 to 64, the bracket most macro experts on podcasts fall into, use AI. The historical belief of the importance of rates vs the lack of belief of the acceleration of AI represent the time mismatch.
The most important catalyst this week for the market was the consumer agent. Meta’s Muse launch drove the stock up 14% on Monday, and the model fire hose is wide open. Multiple compression is the bear market: my agentic infrastructure portfolio is up 46%, against 10% for the Mag 7 and a 10% decline for Salesforce year to date. IWM relative to QQQ made new lows on September 24th. The trade is to be long speed and short friction. Crypto sits at a Peter Lynch style inflection: Breadth is accelerating with 44 of the 46 names in my tokenized index above their 50-day moving average.
Timestamps
- (00:00–02:24) Intro: I recap presenting to Freedom Tech in DC and the Bitcoin community I met there. Every week brings a new fear, and this week it’s rates.
- (02:24–06:12) Time Mismatch: Investors can think linearly or exponentially, and backtests built on a linear world no longer apply. Alvin Toffler’s Future Shock explains why so many investors are angry at AI and call it a bubble.
- (06:12–09:49) Rates and Age: The loudest voices calling for depression are mostly over 55. Tokenization and AI didn’t exist in the decades those rate backtests were built on.
- (09:49–12:37) The Facts: The LEI turned positive with no recession, and S&P earnings are up 17% year over year. Credit spreads, jobless claims, profit margins and money market flows show no sign of rate stress.
- (12:37–14:40) A Whole New Mind: Daniel Pink’s book predicted a world that would no longer belong to mathematicians. Only 2.2% of people aged 55 to 64 use AI, which makes it hard for them to read today’s economy.
- (14:40–19:16) The Consumer Agent Arrives: Meta’s Muse launch sent the stock up 14% on Monday, with partners including PayPal, Expedia, Shopify and Instacart. Models are now contributing to their own improvement, and new releases this week included Opus 5.5 and GPT-6.
- (19:16–24:00) Bear Market Inside a Bull: My agentic infrastructure portfolio is up 46%, while the Mag 7 are up 10% and Salesforce is down 10% year to date. Competition from AI and tokenization will compress multiples across big tech and software.
- (24:00–31:23) Short Friction: With multiples compressing, Nvidia trades at 15 times next year’s earnings and IWM relative to QQQ just made new lows. Visa, Mastercard and Salesforce face pressure as agents take over consumer decisions.
- (31:23–34:35) Charts: The S&P is consolidating near all-time highs, and the main bear watchpoint is whether high-yield spreads widen. Consumer agents are a new catalyst for the infrastructure trade as token usage rises.
- (34:35–44:05) Research Watch: Blackstone’s Jon Gray shows AI payoffs already reaching margins, and Noam Brown explains why the Hugging Face incident came from underestimating AI. Claude’s discovery of a CRISPR-like enzyme system points to healthcare and drug discovery as a major agentic opportunity.
- (44:05–48:48) Tokenization: Vlad Tenev argues tokenization will take over the financial system, and BlackRock is taking model portfolios on chain via Ondo Finance. Entrepreneurs using AI don’t need capital, so rates don’t hurt them.
- (48:48–57:10) Crypto’s App Store Moment: My 46-name tokenized index is up 31% this month, and 96% of its names are above their 50-day moving average. Ali Yahya and a16z explain why AI agents, with no attachment to Visa or brands, are crypto’s major new catalyst.