DeepMind alignment researcher: frontier labs don't know how to instill values in AI

teropa · x · 2026-11-16

Responding to a discussion about systems whose values can't be instilled, Neel Nanda, an alignment researcher at Google DeepMind, wrote: "I work at a frontier lab, we do not know how to reliably instil values into the AIs, this is the scientific consensus." The quote-framer adds a barb: the other thing we have no idea how to instill values into is frontier labs themselves.

Related event: DeepMind Researcher: Frontier Labs Don't Know How to Instill Values in AI(3 posts)→

Original post →

More from AGI Musings

AGI Musings channel →