Depending on how a statement/question is framed, we interpret it differently! For example, when given survival rates for an operation, patients are much more likely to opt for the operation than when given mortality rates, even when the statistics are equivalent! Being an expert is no defense to such irrationality, as doctors were just as susceptible to falling prey to this bias as patients and those not given medical training.
Kahneman argues that this occurs because System 1 can be primed, while System 2 is lazy and doesn't put in the effort to really think about what it is being told. Similar actions result in wide variations of whether people are organ donors across countries. Countries where one has to check a box to become a donor (non-donor) have much lower (higher) organ donation rates as System 2 doesn't like to think enough to have to change from the default.
The way we experience pain and the way we remember pain are different. Duration does not seem to be a criteria for how we remember pain, but obviously it plays a role in how we experience pain. Instead, how we remember pain appears to be based in large part on how intense the pain was when the experience finished. For example, when subjects were put through identically painful experiences, but where the duration of one experience was prolonged with only mild pain, the group actually believed they had gone through less pain in the prolonged session. (Clearly they had not, as the longer session contained all of the pain of the shorter session and then some!)
Such evidence flies in the face of the idea that humans know how to maximize their preferences!