AI Safety Leaders Discuss the Disconnect in Public Panic Levels Over AI

DKokotajlo · x · 2026-08-13

Daniel Kokotajlo, former head of OpenAI's forecasting team, amplified an observation regarding Eliezer Yudkowsky, a prominent figure in AI safety.

The discussion highlights that the public often perceives Eliezer simply as 'someone alarmed by everything AI.' Consequently, when the public finally reaches a 7/10 alarm level over current events, they expect him to be at 9/10. This illustrates a cognitive disconnect in risk perception between the general public and deep alignment researchers.

Original post →

More from AGI Musings

AGI Musings channel →