[at_a_remove]

I think that's what it will approach in the limit, yes, if you are talking about humans. For AIs, I think it will be somewhat less so, and that it would be preferable for the sake of predictability.