Something typically only has to be beneficial on average for reinforcement learning to favour it. That is how many heuristics arise. Similarly, traits only need to be adaptive on average for them to be favoured by natural selection. Indeed: adaptation and learning are closely related.
The fallacy does seem to occur in contexts inconsistent with signalling however—e.g. when I plan how long it will take me to cook a meal, or work out.
Something typically only has to be beneficial on average for reinforcement learning to favour it. That is how many heuristics arise. Similarly, traits only need to be adaptive on average for them to be favoured by natural selection. Indeed: adaptation and learning are closely related.
Even that is not required. See: gambling addiction. Our ability to learn from reinforcement is not calibrated according to the ideal.
Sure: I should have used a different term in place of “beneficial”.
Very true. Although in this case I think it becomes justifiable to call it a ‘fallacy’, since it’s outside conscious control