Please make me care about x-risk

According to Kate Delbeke writing on LessWrong, the AI safety community faces a deeper problem than it realizes. It's not that people don't understand existential risk from AI—it's that humans struggle to care about abstract, catastrophic threats. Delbeke outlines a four-phase model of engagement. Phase one: dismissal as conspiracy. Phase two: you read logically sound arguments but feel no emotional pull—it remains intellectual, not visceral. Phase three: genuine urgency emerges, and you can discuss it without sounding paranoid. Phase four: you're actually working on it. The barriers span echo chambers, social judgment, and the shame of championing an unpopular cause. The fix isn't better fear-mongering. It's accessible media tied to existing concerns, plus communities that sustain engagement. Not everyone needs to dedicate themselves to existential risk research. But moving more people from phases one and two into three and four remains a solvable communications problem.

Listen to this story

Hear this and more stories in a personalized audio briefing.

Open The Chonkerton