Adom's Reviews > Thinking, Fast and Slow

I was exposed to a lot of the content on heuristics and biases elsewhere, so much of this book was review, and kind of a fluffy, non-technical one at that. I wasn't particularly interested in the history of Kahneman's research (except insofar as its informality has made me a bit more suspect of its results), and he generally deals with each idea too briefly to add any insight if one has has any degree of familiarity with the subject matter. He doesn't talk much about which biases are most common, or what the effect sizes are, even though there are many biases that have named opposites, which leaves the reader pretty lost in terms of how to go about even trying to improve -- which makes sense since Kahneman doesn't seem confident this is very possible.

Like, it seems like there is an unreconciled conflict between ego depletion and the focusing effect of cognitive strain (for example with the hard to read font.). This makes it seem like he is just making up stories to explain coincidental data, which may or may not be true. Regardless, one is left with no actionable insight.

The short "speaking about" sections were surprisingly good, and the constant attempted gotchas were just annoying. The only thing I actually noticed learning (though obviously there was much more to learn) in the first 50+% of the book, was the heuristic for judging where expert intuition will and won't be trustworthy: that is, (a regular environment + quick and reliable feedback + history interacting with the environment) leads to more trustworthy expert intuition than the alternative (irregular environment, no feedback, little experience).

The book became interesting to me in section four, when Kahneman starts to describe Prospect Theory, which was entirely new to me. I'll have to read more about it, and likely revisit the relevant sections of this book, but I'm pretty convinced that one will model others more accurately by considering the importance of reference points, and that people judge by outcomes in terms of situational delta rather than situational state.

Another part to which I need to devote more thought is the distinction Kahneman makes between the experiencing and remembering selves, which judge the value and weight of events very differently. This difference has implications for morality and policy and it isn't obvious to me what actually matters here.