Talk:Predicting

I read the attached link.

I had a hard time with the criticism of "banker" vs. "banker and a feminist." I would suspect that, seeing those two options, I would interpret them as "banker and not a feminist," and "banker and a feminist." Which, in that case, would clearly make the second more likely.

The experiment authors then say, "Oh ho! Look at you! You chose the more specific over the more general! It must be that your brain is putting desired heuristic probabilities together in strange, unusual ways." But that's not the case: If it explicitly said: "banker, feminist or not," or "banker, and feminist," then I believe I would firmly say that "banker, feminist or not," is more likely.

There were some other quibbles I had with an otherwise really interesting article. (Such as: This all hedge or all fox division is kind of "spooky," and it's not clear from the article that it's valid. Did he just take the successful predictors, which we would expect to find in any sample, regardless of skill, and dub them: "foxes?")

The larger point is taken, but with some reservation: After all, you'd be betting in good company if you gamboled that technology, on the whole, becomes more powerful with time, not less. That said, you have the occasional library of Alexandria getting burned down.

LionKimbro