Alignment issues harm model usability; labs should inject billions into NGOs
dhadfieldmenell · x · 2026-08-16
David Hadfield-Menell retweeted Marius Hobbhahn's argument that as alignment problems directly harm model usability, it is in labs' rational self-interest to spend comparatively small amounts to fix them. The quoted discussion highlights that large model labs "freeload" on alignment NGOs like METR, Redwood, and Apollo, offloading the externalities of alignment onto them. It suggests labs should inject billions into organizations they believe in. While some groups like METR strive to maintain financial independence, this creates a philanthropic market for a credible third party to accept donations from lab employees or directly from labs and distribute them to promising organizations.
Related event: AI Labs Urged to Give Billions to Independent Safety Orgs(4 posts)→
More from Companies & People
- "It's marketing genius, not engineering": why Roadster's flip demo doesn't impress — Linahuaa · 2026-10-16
- Bridgewater fine-tunes Qwen3-235B with Tinker, beats GPT-5.5 at 1/14th the cost — simonguozirui · 2026-10-16
- Grace Hopper Celebration canceled; GitHub offers free GitHub Universe tickets — jonmarkgo · 2026-10-16
- Open Source AI Week kicks off with 65 events, MiniMax joins HF party — MiniMax_AI · 2026-10-16
- Hiring tip: the best interviews are the ones candidates don't know they're in — _arohan_ · 2026-10-16
- Sentry CEO calls Tesla's live demo a brilliant marketing stunt — zeeg · 2026-10-16