Updating Predictions Requires Action

Josh argues that exposure and testing feared situations provide disconfirming evidence necessary to revise neural predictions.