Christoph Szegedy warns bad actors could gain a decisive advantage from faster AI

ChrSzegedy · x · 2026-07-29

Christoph Szegedy says he is deeply concerned about AI’s negative consequences, including even human extinction, but he is equally worried about bad actors gaining a decisive advantage.

He argues the harder problem may not be accelerating AI itself, but allowing adversarial actors to misuse AI while the field lacks strong ways to counteract that misuse. In a reply to Geoffrey Irving, he also praises the fact that more than 1,000 frontier lab employees have spoken out for slowdowns, calling it important evidence against claims that coordination is impossible.

Original post →

More from AGI Musings

AGI Musings channel →