a hypothesis: it's easier to come to a (tentative, possibly false) conclusion that the reward is not worth it, than to execute a high level task to completion.

a hypothesis: it's easier to come to a (tentative, possibly false) conclusion that the reward is not worth it, than to execute a high level task to completion. The carrot works for mechanical tasks because high level cognition shuts down.