Expert reads · This week (published July 24)
From the Jul 25, 2026 daily brief
Thompson writes Stratechery and is among the most-cited analysts on industry strategy. In a week when Washington and Wall Street panicked in unison, his judgment ran the other way: the threat from Chinese open models is not new; the frontier labs' advantage remains real; and the biggest danger he sees is America's own cybersecurity policy — fixing it requires understanding and accepting the reality of Chinese competition, even if that means OpenAI and Anthropic have to fend for themselves (Stratechery weekly digest, Jul 24; this is the weekly-summary layer — the full argument is behind his paywall and we have not read it, so we cite direction only). This supports today's deep dive's layered call — base-layer commoditization and summit-layer advantage can both be true — and supplies the counterweight to the policy panic in threads 3 and 4. Caveat: one person's read, mechanism not laid out.
Just an email address, unsubscribe anytime. This is the only thing we ask of you.
swyx, the AI-engineering commentator and co-host of the Latent Space podcast, put it flatl…
Sholto Douglas, a reinforcement-learning researcher at Anthropic, predicted in a post: "So…
Dwarkesh Patel is an independent tech interviewer and writer whose deep AI-industry essays…
Mollick, a professor at Penn's Wharton School and the most widely read AI pragmatist among…