The 'replication crisis' shows many studies fail to replicate. Researchers, biased by their desire for success, engage in 'p-hacking' and delude themselves into finding significance. Paradoxically, the most surprising and celebrated results are often the least likely to be true.
Our intuitions depend on how a problem is framed (e.g., 'saving lives' vs. 'preventing deaths'). When shown both frames at once, the conflicting intuitions cancel each other out. This doesn't lead to a more rational choice but to being 'dumbfounded,' stripped of any intuitive guide for action.
The strong impression from a job interview creates a false sense of 'knowing' who will be a good hire. This vivid, face-to-face interaction leads managers to overweight its value and discount more predictive data. Adding an unstructured interview can therefore make hiring predictions worse than using objective data alone.
For remote dangers like climate change that don't trigger immediate fear, worry is the key mechanism that generates motivation to act. The rational argument to 'not worry' fails because worry is what shifts a problem from a passive observation to an active priority, creating the necessary energy to seek solutions.
Intuition is pattern recognition, not magic. According to Kahneman, it can only be trusted when three conditions are met: the environment is stable and predictable, the expert has had prolonged exposure, and feedback on decisions is both rapid and unambiguous. Chess is a prime example; many business situations are not.
Contrary to rational thought, our empathy and willingness to help decrease as the scale of a tragedy is revealed. The brain connects emotionally to a single, vivid story. Adding statistics or even just a second person dilutes this emotional connection, moving the problem into an abstract realm and reducing the impulse to act.
Our moral intuitions are strongly asymmetrical: inflicting a loss feels far more unfair than failing to share a gain. Extrapolating this to policy, Daniel Kahneman argues that a government's primary duty is preventing suffering, a more robust and important objective than the ambiguous goal of promoting happiness.
The psychologist who pioneered research on cognitive bias confesses that 50 years of study has not made his own intuitions more reliable or reduced his overconfidence. He can sometimes recognize situations prone to error, but his underlying cognitive machinery remains unchanged, highlighting the limits of awareness.