Betting AI Favors Defense in All Threats Is Wishful Thinking

ronbodkin · x · 2026-09-02

Alex Bores retweets Ramez Naam, arguing that betting AI will always favor defense across every threat vector is hopelessly wishful thinking. While cautiously hopeful about cyber, he highlights that updating code is faster than human DNA, making biorisk a critical concern. He also lists risks like locking in autocratic control and loss of control, emphasizing that stopping dangerous technology requires concerted global effort, similar to historical nuclear non-proliferation efforts.

Original post →

More from AGI Musings

AGI Musings channel →