The current bottleneck is political will, not research

According to the AI Alignment Forum, a safety researcher argues that the real bottleneck in preventing AI catastrophe isn't better science—it's political will. Most of the world's top policymakers have never had a serious conversation about existential AI risk, the author contends. At a recent UN dialogue on AI, only a handful of over fifteen hundred submissions even mentioned existential risk. In US Congress, less than seven percent of members have publicly discussed loss of control from AI systems. The researcher's conclusion: instead of producing more research papers, the AI safety community needs more advocacy—convincing decision-makers that catastrophic risk is genuinely real. Right now, AI companies are meeting with regulators seven times more often than civil-society safety groups are, which means the industry is setting the narrative rather than safety advocates.

Listen to this story

Hear this and more stories in a personalized audio briefing.

Open The Chonkerton