Computerphile discusses MIRI's "Logical Induction" paper

post by Parth Athley (parth-athley) · 2018-10-04T16:00:20.393Z · LW · GW · 2 comments

This is a link post for https://www.youtube.com/watch?v=gDqkCxYYDGk

Contents

2 comments

2 comments

Comments sorted by top scores.

comment by Charlie Steiner · 2018-10-06T00:46:21.928Z · LW(p) · GW(p)

Nice! Really seems to grok the perspective that this is about finding a good collection of desiderata, and then showing that they aren't actually mutually exclusive.

comment by Flipnash · 2018-10-08T18:03:05.273Z · LW(p) · GW(p)

It has a better description of the algorithm than other sources that have written or made videos about it.

I feel like the algorithm is just a clever search over computable programs that can solve the criteria and falls under the same pitfalls as other algorithms that do the same thing. Mostly it makes the assumption that there exists a program in the search space that fit the criteria it's searching for. I guess if it doesn't exist then the probabilities wouldn't converge to a number that is well calibrated.