AI Researchers Don't Understand the State

LessWrong contributor Alex Amadori argues that AI researchers often harbor a flawed assumption about the future. They imagine it as a competition between AI companies—with governments as passive observers—and join firms they believe are "the good guys," hoping to influence the race toward safety. But Amadori points to a logical problem buried in this thinking: for one company to actually win the race, it would need to perform what he calls a "pivotal act"—preventing every other actor on Earth from building and deploying misaligned AI. And that, he contends, would require controlling every government on the planet. His question is direct: why would governments ever grant such authority?

Listen to this story

Hear this and more stories in a personalized audio briefing.

Open The Chonkerton