Thanks for the replies! Yeah, I've read the first bit of @everytstudies podcast transcript and liked the social constructivist example too (also reminds me I should read some of those people). I agree that the transition could be made easier...
-
Yes, when he's talking in general, non-technical terms he can be quite good on theory-ladenness of observation. Another nice example is here: https://www.lesswrong.com/posts/GKfPL6LQFgB49FEnv/replace-the-symbol-with-the-substance … But then on the next page he'll talk about Bayes, in a way that strongly implies that that solves the problem.
e.g. the one you posted looks promising, but in the next page following on we're straight back into the Bayes stuff. He even got the 'Mind Projection Fallacy' tag from Jaynes! https://www.readthesequences.com/Probability-Is-In-The-Mind …
-
-
The interpretation that I had was that he focused on Bayes because that's what you could usefully tackle using formal methods, with things like generating the original ideas to test being something that emerges from less-understood machinery in the brain.
-
Something like "we don't understand hypothesis generation / picking the state space very well so we'll mostly treat it as a black box & take it as a given, and focus on the things that we do understand how to reason about". (with some exceptions such as https://www.lesswrong.com/posts/X2AD2LgtKgkRNPj2a/privileging-the-hypothesis …)
End of conversation
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.