AI safety researcher: if you believed the risk, your tone would show it
geoffreyirving · x · 2026-09-24
AI safety researcher Geoffrey Irving argues that safety people, having lived with existential risk for years, speak in a tone that feels jarring and inauthentic to newcomers: "If you really believed superintelligence might kill everyone today, how can you talk about extinction via jokes or platitudes?"
He calls AI lab communications the worst offenders ("dedicated to securing its benefits and mitigating its risks"), but says individuals mess up too, and that safety comms should at least sometimes match the shock of "what if I just learned superintelligence might kill everyone today?"
Related event: AI Safety Researcher Urges Sincerer Tone on Extinction Risk(2 posts)→
More from AGI Musings
- Mathematicians face an identity crisis as AI solves problems no one can verify — ValerioCapraro · 2026-11-11
- Ben Todd wrestles with a 50:50 chance only 18 months remain of 'the human era' — ben_j_todd · 2026-11-11
- Mathematicians face identity crisis as AI solves problems in hours, Corriere interview warns — ValerioCapraro · 2026-11-11
- Why inference speed will be AI products' biggest competitive edge in 1–2 years — itsOmSarraf_ · 2026-11-11
- AI Defense Shifts to Scale Economics: Attacks Buried in Legitimate Traffic — DavidLinthicum · 2026-11-11
- Ben Todd's alt log-log plot shows task horizons hitting a singularity in under 12 months — ben_j_todd · 2026-11-11