Industrializing a small field: Lessons from Vannevar

A LessWrong essay argues that AI Safety research is at an inflection point. The field is about to receive a wave of new funding, but the current model—researchers applying for grants—simply can't absorb it at scale. The author draws a parallel to the Manhattan Project, where active recruitment of leading scientists proved crucial to mobilizing the wartime effort. The proposal: AI Safety funders should stop waiting for grant applications and start proactively recruiting top researchers in adjacent fields—cryptography, computer science, verification—to fund pivots toward the safety problems that matter most. The specific gap: while AI Safety has adopted cryptographic frameworks from researchers Aumann and Lindell, there's little public indication that major funders have approached them directly to expand that work. The concern is that incoming funding will be spent on smaller programs instead of mobilizing entire university research groups toward the highest-impact problems.

Listen to this story

Hear this and more stories in a personalized audio briefing.

Open The Chonkerton