Shea Levy's Reviews > Thinking, Fast and Slow

Overall I really enjoyed this book. Kahneman is an engaging writer and brings in a wealth of information to construct a compelling narrative. I'd been aware of a lot of this literature and the overall framework beforehand, but I still learned a lot in terms of specifics and got a much better sense of the details of the system 1/system 2 model and how broadly applicable it is. If you want a glimpse of what kind of understanding psychology can offer of how we reason and evaluate, I definitely recommend this one.

I had some minor nits throughout and disagreed with a few of his interpretations, but the significant detractor was the framing Kahneman chose for the book. He made some choices seemingly to promote his goal of influencing water cooler conversation (which by itself seemed strange to me) that I think weakened chances for understanding. The biggest one was the discussion of the separate categories of thought processes as separate systems with their own agency and personalities. Kahneman acknowledges this as a fiction explicitly, but that didn't seem to mitigate the confusion that I think might result. I do think there's a true and useful categorization underneath the system 1/system 2 distinction, but treating them as completely distinct and partially at odds obfuscates the deeply intertwined and inseparable nature of the processes. One particular recurrent issue that resulted from this was identifying system 2 with the self and then concluding (at several different points) things along the lines of "you don't really know/decide/etc. most things, that's sneaky old System 1". And ascribing judgmental personality traits to the systems (e.g. "system 2 is lazy" instead of "effortful thinking is costly") may help the reader from an intuitive impression of them, but with a strong bias toward oversimplification and negative evaluation.

More minorly, Kahneman also usually leaves out details like how strong the effects noted are, what proportion of the subjects were affected, etc. seemingly out of a desire to craft a coherent picture that applies wholly to the reader. One of the times where he does include some of this information this gets particularly ridiculous: In one paragraph on page 171 he mentions that "only four of the fifteen participants [in the "helping experiment"] responded immediately to the appeal for help," and in the next says "even normal, decent people do not rush to help when they expect others to take on the unpleasantness of dealing with a seizure. And that means you, too.". 4/15 is pretty bad given the issue at hand, but if that generalizes it emphatically doesn't mean "you" for more than 1 in 4 of the readers! Many of the errors he mentioned I could easily identify in myself, some of them I have evidence for thinking they don't apply as strongly as he claims (e.g. actually having been in a few diffusion of responsibility type scenarios and taking action), and for the rest I simply have no way to judge likelihood without diving into the references. I think adding in these details would have gone a long way to forming a more accurate understanding in the reader, but I have an uncharitable suspicion that it would also have significantly weakened some of the cases he was making.

All that being said, I really did enjoy this book and by the end couldn't put it down. Kahneman clearly knows his stuff and I'm much better informed having read it. I've already taken steps to try to incorporate some of the lessons into my thinking and decision making, and definitely plan to seek out more in this domain.