AI vs Dog Cancer, Zito From Apollo Tells All, Tyler Cowen WANTS YOU to Lock In, SF Housing Market
TBPN hosts a wide-ranging tech and finance discussion covering AI's democratization of biotech (featuring Paul Cunningham's mRNA vaccine approach to dog cancer), the non-bubble AI investment thesis, and geopolitical tensions around AI development and Taiwan. The episode weaves together Tyler Cowen's argument that individuals should "lock in" and work harder during the AI transition, investor John Zito's warnings about overvalued private equity marks, and Ben Thompson's analysis of why massive compute capex is justified by three inflection points in AI capability.
Key takeaways
- • AI acts as an enabler and cognitive tool rather than a cure itself—Cunningham used ChatGPT to navigate complex biotech workflows that previously required institutional teams, democratizing access to medical research pipelines without replacing expert collaboration.
- • Freeman Dyson's prediction of "domesticated" biotechnology is becoming real as AI reduces cognitive overhead, potentially blurring the boundary between professional research institutions and motivated individuals working at smaller scales.
- • The US should maintain private sector dominance in AI development to win the emerging geopolitical arms race; nationalization or government-led approaches risk losing the top talent and organizational agility that drives frontier AI progress.
- • Tyler Cowen argues individuals should work harder now because strong AI will devalue human capital in some fields, making current high wages relatively attractive compared to future wages—and learning AI skills provides insurance against disruption.
- • Three AI inflection points—LLMs, reasoning models, and agents—drive exponential increases in compute demand that justify massive hyperscaler capex; the agent paradigm in particular shifts value from cost-cutting to pure acceleration and top-line growth.
- • Private equity and credit markets are overheated: Apollo's John Zito warned of 60-80% potential markdowns on smaller software company loans and predicted a consumer confidence recession driven by companies rushing premature AI implementations.
- • Chip supply, not energy, is the binding constraint on AI scaling through 2028; EUV lithography tool shortages will remain the primary bottleneck despite the availability of power and data center capacity.
Recommendations (11)
"I'm an internet OG so I was on GeoCities. I was on Angel Fire back in the day making anime tutorials."
Kevin Espiritu · ▶ 1:37:09
"And what's in your case on the Apple Vision Pro? I did watch two movies last weekend in VR fully. I'm probably the only person. I think it's amazing"
TBPN · ▶ 3:12:42
Mentioned (12)
More from these creators
Is Meta Really Addictive?, Jetsons Prediction, Artemis II: Back to the Moon | Diet TBPN
FULL INTERVIEW: Why I Think Nvidia Is Perfectly Positioned In The AI Race
TBPN Gets Addicted to Social Media, Japan Twitter, Warren Buffett's Protégé, Deals Deals Deals
The $6M Verdict That Could Kill Social Media
Novartis Buys Excellergy for $2B, Anthropic Vs. Pentagon, The Mansion Section | Diet TBPN
How I Built and Sold SALT & STONE