Reasoning was not made for Deduction

According to a new essay on LessWrong, we've fundamentally misunderstood how science works: deduction gets the credit, but intuition does the actual work. The piece traces this through history—from Anaximander's naturalistic explanations to Galileo's mathematization to Newton's mechanics—showing that in each breakthrough, intuitive leaps, not formal rules, built the science. When we apply established science, we use deduction to compute predictions, but in science being made, intuition is the building and formalism is just scaffolding. Here's where it gets interesting: language models seem to confirm this point. Treating them as character simulators—what practitioners call cyborgism—predicts their behavior better than mechanical explanations ever could. The essay argues the mind's core ability is attunement: learning intuitions and generative leaps from engaging with the world, rather than following fixed rules.

Listen to this story

Hear this and more stories in a personalized audio briefing.

Open The Chonkerton