Exciting new paper on dopamine (DA) and learning.
Bellman: created an equation to compute the value of different choices before you. But it required infinite memory.
Sutton & Barto: Came up with a new way, by calculating the difference between what happened & what you expect. This 'reward prediction error' was mapped to DA in the brain.
New paper: not quite. Instead, the brain computes whether a stimulus precedes reward beyond that expected by chance (causality).

