Mason  🏃‍♂️ ✂️

@trishankkarthik When we know the possible outcomes and have an explanatory theory to estimate all their probabilities (and utilities) we can use decision theory to decide what to do: e.g. choose the action with the highest expectation value of utility. When we don't, we can't.

In particular, where probabilities are *unknowable*, there is no such thing as an estimate of them, nor even an "educated guess". Attempting those is then, at best, just scanning one's memory for prejudices, promised utopias and scare stories.

Where probabilities are unknowable (which is often the case in, say, political decisions), it's not the case that reason is ineffective. It merely entails a methodology very different from utility theory—among other things focused on institutional rules not individual decisions.

One of the damaging consequences of utility theory is that it makes people *uncritically certain* that they have ways of extracting knowledge from ignorance. Because (they reason) if there can be no such ways, there must be no such thing as rational decision making.