sort of repeating myself but typing is less effort than finding/linking right now.
I think that as we/something we make moves up the intelligence scale we will be shocked to discover just how much of what we think of as cognition is actually just pattern matching. I expect this to scale upwards, perhaps up to the limit of intelligence (optimization). that is, n level intelligence looks like crude pattern matching to an n+1 intelligence.
as for your final paragraph: humans are clever/silly enough to not just want to be naive hill climbing algorithms. we don’t just want the best outcome given our initial circumstances (position) and logical moves we could have made based on the data available at the time (that way is up), we want to wind up on the objectively highest peak even if it would have meant “stupid” moves in the interim. This is where the “rationality can be harmful” arguments go.
sort of repeating myself but typing is less effort than finding/linking right now.
I think that as we/something we make moves up the intelligence scale we will be shocked to discover just how much of what we think of as cognition is actually just pattern matching. I expect this to scale upwards, perhaps up to the limit of intelligence (optimization). that is, n level intelligence looks like crude pattern matching to an n+1 intelligence.
as for your final paragraph: humans are clever/silly enough to not just want to be naive hill climbing algorithms. we don’t just want the best outcome given our initial circumstances (position) and logical moves we could have made based on the data available at the time (that way is up), we want to wind up on the objectively highest peak even if it would have meant “stupid” moves in the interim. This is where the “rationality can be harmful” arguments go.