Ethan Mollick: agent swarms self-organize, solving Navier-Stokes in 88 hours — no human managers needed
One Useful Thing (Ethan Mollick) · rss · 2026-10-01
Ethan Mollick admits he was wrong: managing AI agents doesn't require human-designed org structures. Clawlikes like Meta's Muse and OpenAI's dots proactively catch users' mistakes, while OpenAI's swarm of thousands of agents exchanged 2.7M messages to reportedly solve the Navier-Stokes Millennium Prize Problem in 88 hours with minimal human coordination. The Bitter Lesson now applies to the org chart — much of management exists to solve human problems (misaligned incentives, hidden information) that agents simply don't have.
More from AGI Musings
- Diffusion-model counterfactuals: post-2015 emissions added 0.34°C to Europe's June 2025 heatwave — bravo_abad · 2026-10-01
- The problem with probabilistic software: AI needs deterministic boundaries, not deterministic cores — forevergeeks · 2026-10-01
- Garrison Lovely argues Hinton underclaims: AI already outpersuades human experts — GarrisonLovely · 2026-10-01
- Ethan Mollick: I underestimated AI's ability to self-organize, agents beat elaborate orchestration — emollick · 2026-10-01
- AI lowered the entry bar for artists — and made content-vs-art gatekeepers of us all — jordiponsdotme · 2026-10-01
- Blogger says takeoff is here, citing compute, token growth, energy and robotics buildout — Dr_Singularity · 2026-10-01