Zuckerberg: trust and alignment will be AI's most important next set of capabilities

DavidSacks · x · 2026-09-24

Meta CEO Mark Zuckerberg argued against industry-wide coordination on AI safety, saying each lab should instead take the internal time it needs when it sees problems.

His key claim: AI is entering a phase where trust and alignment become the most important next set of capabilities—getting better at math may matter less and less, and commercial incentives alone are enough to get safety right.

He also revealed Meta delayed the launch of Muse for a few months to make it a better product, calling it the right decision for users, Meta, and other companies building the technology.

Original post →

More from AGI Musings

AGI Musings channel →