idea: as long as throwing more neurons at a problem to solve it is cheap, a mind is disincentivized from figuring out higher-level structure
Conversation
Replying to
a smaller brain will need more insight to solve the same problem, a larger brain may "understand" less
1
2
Replying to
a similar idea is in complexity theory (runtime of an algorithm vs its description size), but doesn't deal in incentives
1
Replying to
how do you figure out, at a glance, whether to optimize or to bruteforce - is there an optimal way to do this?
1
2
Replying to
nah, still too obvious a question, probably answered in some paper about impl-ing Prolog;
meta: quicker to figure out or to find the paper?
1
2
Replying to
had similar thoughts when alpha go won. I think practically brute force & understanding are not distinct activities.
2
1
if you're wondering whether you're spending too much time trying to understand without just doing it, answer is yes
1
1
fortunately we can mostly "brute force" or use current best strategy, and v infrequently assess for inefficiency
1
oops I mean exponential improvement. Cool thing about being a person is that we learn just by doing. Immersion is how we optimize
1
Replying to
how would you quantify improvement so that it could be exponential?
Replying to
generally that the resources of import can be consumed exponentially less over time. Moore's law style
1
1
to put my stance better into terms of your question- "at a glance" should be indistinguishable from "after you've tried it a bit"