Pragmatisation, the Way Forward

LessWrong publishes Stephen Elliott’s essay urging the AI safety community to move beyond purely academic norms and engage directly with political and financial power structures. He critiques Richard Ngo’s call for new community norms, arguing that without real‑world influence safety work cannot secure the regulatory levers needed to curb risky AI development. Elliott proposes splitting the field into intellectual, political, and capital arms, with the intellectual side providing concrete tools to policymakers and markets. A pragmatic, real‑politics approach, he says, may be essential if the community hopes to avert large‑scale catastrophes.

Listen to this story

Hear this and more stories in a personalized audio briefing.

Open The Chonkerton