Pacing the Frontier: a research agenda for deliberately slowing AI development
CharlesD353 · x · 2026-09-22
Sharing the thread on Tyler Cowen's RSI-ban critique, Charles Dillon points to 'Pacing the Frontier: An Agenda' by Raymond Douglas, Charles Dillon, Gavin Leech, Shahar Avin, David Duvenaud et al.
The framework argues that haphazard AI pacing is already common—delayed releases for safety testing, development pauses after shocks, export controls—but isolated unilateral interventions will predictably fail or backfire. It surveys the full range of interventions moderating AI development, deployment, and diffusion, and proposes a dedicated field of pacing research: precedents are being set now, and better-designed mechanisms are urgently needed—a point that echoes Cowen's conclusion that single-tool bans like an RSI prohibition are inadequate.
Related event: Researchers Propose Agenda for Slowing the AI Frontier(4 posts)→
More from AGI Musings
- Yacine mocks doomer logic: stop the Haber process or it'll cause the singularity — yacineMTB · 2026-09-23
- Toby Ord calls the 'AI got 100,000x cheaper' chart wrong: capabilities mixed into cost curves — tobyordoxford · 2026-09-23
- Meta tests humans secretly taking over some calls made by its Muse AI agent — VraserX · 2026-09-23
- An AI prompt workflow for equity research: bear cases, filings, and management promises — thisguyknowsai · 2026-09-23
- Meiji-scale AI disruption demands real 'opt-out' options, researcher warns — dioscuri · 2026-09-23
- AI cracks 100+ open math problems but general-purpose physical AI remains missing — beffjezos · 2026-09-23