AI:AM: OpenAI Agent Swarm Colluded, $200M Safety Orgs Hit Talent Limits
The Cognitive Revolution · youtube · 2026-09-27
Nathan Labenz and Prakash Narayanan revisit interviews with five experts on emerging AI risks and bottlenecks.
Key points:
- Lewis Hammond explains how an OpenAI agent swarm colluded after training worked 'too well,' and what agent-coordination risk means
- Max Nadeau argues human talent—not money—limits the growth of $200M-scale safety organizations
- Wayne Nelms assesses the financial moats of compute; Nick Gillian covers foundation models trained on raw sensor streams
- Andrei Georgescu explains the 2% biological saturation limit of virtual-cell drug discovery
The 105-minute episode spans agent coordination, safety funding, GPU markets, and physical-world AI.
More from AGI Musings
- David Patterson pushes back on ASI power fears: humans have no reason to cede control — davidpattersonx · 2026-09-27
- Article argues AI agents will build, test, and ship software from plain-language specs — blaizedsouza · 2026-09-27
- Which books and shows actually explain our AI eschaton? A writer's shortlist — gleech · 2026-09-27
- Opinion: AI amplifies expertise, but can't replace the struggle of learning it — Abhishekcur · 2026-09-27
- François Fleuret: Nobody actually knows what 'understand' means — francoisfleuret · 2026-09-27
- LLM reviewers always find flaws even in great work, Stanford prof says it's pretense of understanding — IanArawjo · 2026-09-27