decision theory
The formal theory of how to make rational choices given options, information, and preferences; Callard and Paul argue it cannot handle 'big' choices that change one's core preferences.
As defined in Agnes Callard on Aspiration
Where it comes up · 8
- 220 | Lara Buchak on Risk and Rationality4 claims
- 36 | David Albert on Quantum Measurement and the Problems with Many-Worlds3 claims
- David Deutsch2 claims
- 327 - Transformative Experiences2 claims
- Robin Hanson1 claim
- Emily Oster on Pregnancy, Causation, and Expecting Better1 claim
- Agnes Callard on Aspiration1 claim
- 116 - Ai Racing Toward The Brink1 claim
Most important claims · 10 of 15Search all claims for “decision theory” →
- There are two distinct senses of rationality in philosophy: purely formal consistency (analogous to logical consistency, e.g. believing P and P implies Q but not Q is irrational regardless of content) and substantive rationality (whether a belief matches the evidence). Decision theory studies the consistency sense.220 | Lara Buchak on Risk and Rationality
- Although decision theory only requires consistency, consistency turns out to be a difficult notion with a lot of content, so it strongly dictates what a person should do; everyone agrees we should be consistent, and the real debate is over what consistency means.220 | Lara Buchak on Risk and Rationality
- Callard agrees with L.A. Paul that transformative 'big' choices straddle two preference spaces and cannot be handled by standard decision theory (which preference set should one maximize?), but disagrees that they must be framed as decisions; Callard holds they are not decisions but temporally-extended learning processes that can be navigated rationally, whereas Paul, treating them as decisions, finds no clear rational way to navigate them.Agnes Callard on Aspiration
- Decision theory is about rationality in what you choose or prefer under risk or uncertainty, and it requires at least two ingredients: probabilities, which measure how likely you think things are to happen, and utilities, which measure how much you like various outcomes.220 | Lara Buchak on Risk and Rationality
- From a decision-theory standpoint, having many versions of yourself (whether Em copies or many-worlds branches) does not justify taking greater risks, because you still care about how many copies exist—the more the better—so killing off many through stupid risks is still bad, just as ordinary decision trees rapidly branch into billions of future copies of you without warranting reckless behavior toward the distant future.Robin Hanson
- Asimov's Three Laws are plot devices that AI researchers understood to be unworkable long before the public debate: they are deontology, but decision theory requires quantitative weights on goals, so a robot literally following 'do not through inaction allow a human to harm' would never get around to obeying any orders because every action carries some tiny probability of leading to harm.116 - Ai Racing Toward The Brink
- Decision theory describes how people should behave rather than how they do, but actual behavior is relevant as a datum: if people systematically diverge from a theory without good explanation, that pressures us to investigate whether the theory of rationality is wrong or whether the behavior is genuinely irrational.220 | Lara Buchak on Risk and Rationality
- Ordinary decision theory with randomness produces the same rational decisions as quantum decision theory with the multiverse, so whether reality is genuinely random or branching should make no difference to rational decisions.David Deutsch
- Albert's core objection to the decision-theoretic derivation is that it is simply mistaken to think one's preferences among non-branching futures can constrain, on pain of irrationality, one's preferences among branching futures; branching opens an entirely new set of options (e.g. preferring to be richer in the branch where one is fat), and there is no reason the limited non-branching preferences should determine these new preferences.36 | David Albert on Quantum Measurement and the Problems with Many-Worlds
- Transformative experiences pose a problem for standard rational decision theory because you cannot reliably imagine the self you will become, the evidence from others reflects post-change preferences that may not apply to your current self, and the self deciding may not share the desires of the self that would result—so you cannot simply evaluate the options as your present self.327 - Transformative Experiences