Thinking, Fast and Slow by Daniel Kahneman is the kind of book that makes you feel clever for understanding how frequently you are wrong, which is a rather efficient way of keeping a reader humble. It is a work of cognitive psychology and behavioural decision making, built around the distinction between fast, intuitive thinking and slower, more deliberate reasoning. Kahneman brings together decades of research with Amos Tversky and others on heuristics, biases, judgement and decision making.

The basic conceit is enormously useful. Fast thinking handles the mental work we perform automatically, while slower thinking requires attention and effort. The trouble is that the fast system is remarkably confident even when it is making things up, and the slower system is perfectly capable of accepting the nonsense because thinking properly is exhausting. Human rationality, it turns out, is less a majestic superpower and more a household appliance that occasionally makes a worrying noise.

Kahneman is particularly good at demonstrating how easily our judgement can be distorted by anchoring, availability, framing, overconfidence and our tendency to substitute an easier question for a difficult one. The famous examples are memorable because they do not merely tell you that people make predictable errors. They make you catch yourself making them.

That is the book's great achievement. It changed how an enormous audience thought about thinking, and its central research programme helped shape behavioural economics and decision science. Its influence extends far beyond psychology, particularly in fields concerned with judgement under uncertainty.

But there is an awkward modern complication. Some of the research discussed in the book, particularly work surrounding social priming, has faced serious replication problems. Kahneman himself later acknowledged that he had placed too much faith in some underpowered studies. This does not demolish the entire book, nor does it erase the importance of the heuristics and biases programme, but it does mean that a modern reader should distinguish between robust ideas and individual experimental claims that have become questionable.

The book's other weakness is structural. It is dense, repetitive and sometimes feels like a brilliant lecture that forgot somebody might eventually have to sit through the whole thing. Kahneman explains concepts thoroughly, occasionally to the point where you start wondering whether System 2 has been punished for something.

Still, compared with the long history of popular psychology, this remains an unusually important book. It does not simply offer advice about thinking better. It gives you a vocabulary for noticing how your own mind goes wrong.

Thinking, Fast and Slow is demanding, occasionally dated and not as scientifically bulletproof as its reputation might suggest. It is also genuinely illuminating. Read it critically rather than reverently, and there is still an enormous amount here worth thinking about.