AI Safety Shouldn't Be Trivialized: Researchers Warn Against Reckless Meme Culture

jachiam0 · x · 2026-08-08

AI safety researchers David Krueger and Adam Jachimowicz discussed the boundaries of joking about potential AI disasters on X. Krueger noted that while he appreciates dark humor, the AI community's irreverent jokes about catastrophic risks might send the wrong message to the outside world.

Jachimowicz agreed, arguing that an excess of gallows humor could poison the industry's culture. He emphasized that while some dark humor is acceptable, recent memes surrounding projects like PandemicBench feel inappropriate, suggesting researchers should joke about desirable outcomes rather than trivializing disasters.

Original post →

More from AGI Musings

AGI Musings channel →