And all this time I was missing that my key disagreement with Outside View is that I don’t at all infer from the experiments on Planning Fallacy that you should be less concrete, remove some details, but that you should take into account the past experience with all it’s details while reasoning about a new situation with all its details.
Problem that I indeed inferred something like ignoring details from Overconfidence. Failing to remember past data didn’t look like terrible situation of agent who can’t see it’s flaws even when pointed out to them, and Overconfidence did. Hence was the idea of ignoring everything except conclusions.
And I probably too much allowed myself to follow subtly different phrasings of others instead of phrasing everything exactly in my own Original Seeing. Because, as I reasoned, different frameworks should be isomorphic in the logical sense. And they probably are. But as I missed humans aren’t logically omniscient, so you need to consider selection effects of whose computations were actually picked to be executed.
P.S./Edited: Oh, I forgot important details, when I remember about previous cases, I just feel how my anticipations, intuitions change because previously I just didn’t have that fact fresh in working memory and so didn’t take it into account.
And all this time I was missing that my key disagreement with Outside View is that I don’t at all infer from the experiments on Planning Fallacy that you should be less concrete, remove some details, but that you should take into account the past experience with all it’s details while reasoning about a new situation with all its details.
Problem that I indeed inferred something like ignoring details from Overconfidence. Failing to remember past data didn’t look like terrible situation of agent who can’t see it’s flaws even when pointed out to them, and Overconfidence did. Hence was the idea of ignoring everything except conclusions.
And I probably too much allowed myself to follow subtly different phrasings of others instead of phrasing everything exactly in my own Original Seeing. Because, as I reasoned, different frameworks should be isomorphic in the logical sense. And they probably are. But as I missed humans aren’t logically omniscient, so you need to consider selection effects of whose computations were actually picked to be executed.
P.S./Edited: Oh, I forgot important details, when I remember about previous cases, I just feel how my anticipations, intuitions change because previously I just didn’t have that fact fresh in working memory and so didn’t take it into account.