Ben Levinstein: Hmm. As a professional analytic philosopher, I find myself unable to judge a lot of this. I think philosophers often carve out sub-communities of varying quality and with varying norms. I read LW semi regularly but don’t have an account and generally wouldn’t say it outperforms.
Rob Bensinger: An example of what I have in mind: I think LW is choosing much better philosophical problems to work on than truthmakers, moral internalism, or mereology. I also think it’s very bad that most decision theorists two-box, or that anyone worries about whether teleportation is death.
If the philosophical circles you travel in would strongly agree with all that, then I might agree they’re on par with LW, and we might just be looking at different parts of a very big elephant.
Ben Levinstein: That could be. I realized I had no idea whether your critique of metaphysics, for instance, was accurate or not because I’m pretty disconnected from most of analytic metaphysics. Just don’t know what’s going on outside of the work of a very select few.
Rob Bensinger: (At least, most decision theorists two-boxed as of 2009. Maybe things have changed a lot!)
Ben Levinstein: I don’t think that’s changed, but I also tend not to buy the LW explanations for why decision theorists are thinking along the lines they do. E.g., Joyce and others definitely think they are trying to win but think the reference classes are wrong.
Not taking a side on the merits there, but just saying I have the impression from LW that their understanding of what CDT-defenders take the rules of the game to be tends to be inaccurate.
Rob Bensinger: Sounds like a likely sort of thing for LW to get wrong. Knowing why others think things is a hard problem. Gotta get Joyce posting on LW. :)
Ben Levinstein: I also think every philosopher I know who has looked at Solomonoff just doesn’t think it’s that good or interesting after a while. We all come away kind of deflated.
Rob Bensinger: I wonder if you feel more deflated than the view A Semitechnical Introductory Dialogue on Solomonoff Induction arrives at? I think Solomonoff is good but not perfect. I’m not sure whether you’re gesturing at a disagreement or a different way of phrasing the same position.
Ben Levinstein: I’ll take a look! Basically, after working through the technicals I didn’t feel like it did much of anything to solve any deep philosophical problems related to induction despite being a very cool idea. Tom Sterkenburg had some good negative stuff, e.g., http://philsci-archive.pitt.edu/12429/
I guess I have a fair amount to say, but the very quick summary of my thoughts on SI remain the same:
1. Solomonoff Induction is really just subjective bayesianism+ Cromwell’s rule + prob 1 that the universe is computable. I could be wrong about the exact details here, but I think this could even be exactly correct. Like for any subjective Bayesian prior that respects Cromwell’s rule and is sure the universe is computable there exists some UTM that will match it. (Maybe there’s some technical tweak I’m missing, but basically, that’s right.) So if that’s so, then SI doesn’t really add anything to the problem of induction aside from saying that the universe is computable.
2. EY makes a lot out of saying you can call shenanigans with ridiculous-looking UTMs. But I mean, you can do the same with ridiculous looking priors under subjective bayes. Like, ok, if you just start with a prior of .999999 that Canada will invade the US, I can say you’re engaging in shenanigans. Maybe it makes it a bit more obvious if you use UTMs, but I’m not seeing a ton of mileage shenanigans-wise.
3. What I like about SI is that it basically is just another way to think about subjective bayesianism. Like you get a cool reframing and conceptual tool, and it is definitely worth knowing about. But I don’t at all buy the hype about solving induction and even codifying Ockham’s Razor.
4. Man, as usual I’m jealous of some of EY’s phrase-turning ability: that line about being a young intelligence with just two bits to rub together is great.
A conversation prompted by this post (added: and “What I’d Change About Different Philosophy Fields”) on Twitter:
______________________
Ben Levinstein: Hmm. As a professional analytic philosopher, I find myself unable to judge a lot of this. I think philosophers often carve out sub-communities of varying quality and with varying norms. I read LW semi regularly but don’t have an account and generally wouldn’t say it outperforms.
Rob Bensinger: An example of what I have in mind: I think LW is choosing much better philosophical problems to work on than truthmakers, moral internalism, or mereology. I also think it’s very bad that most decision theorists two-box, or that anyone worries about whether teleportation is death.
If the philosophical circles you travel in would strongly agree with all that, then I might agree they’re on par with LW, and we might just be looking at different parts of a very big elephant.
Ben Levinstein: That could be. I realized I had no idea whether your critique of metaphysics, for instance, was accurate or not because I’m pretty disconnected from most of analytic metaphysics. Just don’t know what’s going on outside of the work of a very select few.
Rob Bensinger: (At least, most decision theorists two-boxed as of 2009. Maybe things have changed a lot!)
Ben Levinstein: I don’t think that’s changed, but I also tend not to buy the LW explanations for why decision theorists are thinking along the lines they do. E.g., Joyce and others definitely think they are trying to win but think the reference classes are wrong.
Not taking a side on the merits there, but just saying I have the impression from LW that their understanding of what CDT-defenders take the rules of the game to be tends to be inaccurate.
Rob Bensinger: Sounds like a likely sort of thing for LW to get wrong. Knowing why others think things is a hard problem. Gotta get Joyce posting on LW. :)
Ben Levinstein: I also think every philosopher I know who has looked at Solomonoff just doesn’t think it’s that good or interesting after a while. We all come away kind of deflated.
Rob Bensinger: I wonder if you feel more deflated than the view A Semitechnical Introductory Dialogue on Solomonoff Induction arrives at? I think Solomonoff is good but not perfect. I’m not sure whether you’re gesturing at a disagreement or a different way of phrasing the same position.
Ben Levinstein: I’ll take a look! Basically, after working through the technicals I didn’t feel like it did much of anything to solve any deep philosophical problems related to induction despite being a very cool idea. Tom Sterkenburg had some good negative stuff, e.g., http://philsci-archive.pitt.edu/12429/
Ben Levinstein: