You spend a lot of time arguing “immodest” over “modest” epistemology here, when the thing that really gets me is the hedgehog vs. fox epistemology. I kept wanting to say yes, Pat’s right that outside view should lower the probability in certain situations, and yes, Eliezer’s right that the inside view should raise his probability of beating the odds in certain situations, and that they should look at this individual situation to see how much each should apply.
I wanted to say that YES, Eliezer’s view that outside view arguments aren’t useful to think about when actually working on a project means that YES, you should compartmentalize and not think about them when working on a project, but doesn’t mean it’s not useful to think about at other times.
To me you’re strawmanning the modest epistemology by making it “hedgehog modest” instead of “fox modest”. This was a persistent problem I had with nearly every chapter in the book, and not just this bonus chapter.
I don’t see it. Maybe you think fox epistemology wouldn’t donate to MIRI, which is presumably what Eliezer cares about? But what he claims repeatedly is that we should judge situations just as you say, and he offers a way to do this.
No, what I got was “modest epistimology doesn’t make any sense in these precise situations when civilizational inadequacy applies”. That’s an incredibly hedgehoggy way to look at modest epistemology.
A more foxxy way would be something like “apply the frames of both modest and immodest epistemologies, as well as the frame of civilizational inadequacy, then using data from all of these frames, make your decision.
You spend a lot of time arguing “immodest” over “modest” epistemology here, when the thing that really gets me is the hedgehog vs. fox epistemology. I kept wanting to say yes, Pat’s right that outside view should lower the probability in certain situations, and yes, Eliezer’s right that the inside view should raise his probability of beating the odds in certain situations, and that they should look at this individual situation to see how much each should apply.
I wanted to say that YES, Eliezer’s view that outside view arguments aren’t useful to think about when actually working on a project means that YES, you should compartmentalize and not think about them when working on a project, but doesn’t mean it’s not useful to think about at other times.
To me you’re strawmanning the modest epistemology by making it “hedgehog modest” instead of “fox modest”. This was a persistent problem I had with nearly every chapter in the book, and not just this bonus chapter.
I don’t see it. Maybe you think fox epistemology wouldn’t donate to MIRI, which is presumably what Eliezer cares about? But what he claims repeatedly is that we should judge situations just as you say, and he offers a way to do this.
No, what I got was “modest epistimology doesn’t make any sense in these precise situations when civilizational inadequacy applies”. That’s an incredibly hedgehoggy way to look at modest epistemology.
A more foxxy way would be something like “apply the frames of both modest and immodest epistemologies, as well as the frame of civilizational inadequacy, then using data from all of these frames, make your decision.