Why a physicist should care about how brains learn

When a physicist looks at a learning brain, the familiar tools do not feel out of place — they feel necessary. A network that stores many patterns and settles into them is, structurally, a disordered system with a rugged energy landscape. The transitions between learning regimes look like phase transitions. The sensitivity that learning seems to require looks like the sensitivity systems show near a critical point.

The risk is to stop at analogy. The work is to make the correspondence precise enough that it predicts something — about capacity, about failure, about ageing. That is the project.