Stop doing decision theory without metaphysics

LessWrong contributor Elias Schmied argues that the ongoing debate about Functional Decision Theory—a framework for making rational decisions in extreme scenarios—has lost sight of its core disagreement: metaphysics. Consider a nearly perfect predictor who learns something embarrassing about you and threatens to release it unless you pay a hundred dollars, but only because he predicted you would. Should you pay? Functional Decision Theory says no. But Schmied contends that answer secretly rests on a metaphysical claim most people don't explicitly accept—that some fundamental fact about reality must be true, whether alternate versions of reality exist, or decisions can change the past, or that you are an abstract computation. Meanwhile, actual debates between advocates and critics focus on disagreements about the word 'rational,' or confusion about whether planning ahead suffices. Schmied's call: make these metaphysical assumptions explicit and argue about those, because that's where the real disagreement actually lives.

Listen to this story

Hear this and more stories in a personalized audio briefing.

Open The Chonkerton