Should you choose the action that causes the best outcome?
In short: When faced with a choice, do you act based on what your decision will physically cause, or what your decision reveals about the world? While most people intuitively follow their gut, decision theory reveals a profound split between acting on causal consequences and acting on evidence.
Causal decision theory (CDT) posits that a rational agent should select the action that produces the best expected causal consequences. This differs fundamentally from evidential decision theory (EDlam), which suggests choosing the action that serves as the best indicator of a good outcome. For instance, if eating an apple causes happiness and an orange causes sadness, CDT dictates choosing the apple.
The distinction becomes stark in scenarios involving reliable predictors. In Newcomb's paradox, a predictor with 99% accuracy places either $1,000,000 or nothing in an opaque Box B, based on their prediction of your choice. If you take both Box A (containing $1,000) and Box B, CDT recommends this because your choice cannot retroactively change what the predictor already placed in the box. Conversely, EDT recommends taking only Box B, as doing so is strong evidence that the predictor anticipated a single-box choice, making the million-dollar payout likely.
However, CDT is not without controversy. Critics like Andy Egan argue that CDT's requirement to hold initial views about the world fixed can lead to irrationality. In the 'Psychopath Button' thought experiment, an agent might be endorsed by CDT to press a button that kills all psychopaths, even if the agent is confident that only a psychopath would press it and they personally prefer a world with psychopaths. Similarly, in the 'Adversarial Offer' scenario, CDT can lead a buyer to make a choice that is profitable for the seller but not for themselves, highlighting a disconnect between intuitive rationality and causal algorithms.
Opens the Intellect+ feed with this card first. Swipe up for the next one.
Sources
- Summary by Intellect · source: Causal decision theory
To cite this card, link to https://intellectplus.co.uk/app/ideas-causal-decision-theory.
How we source, check and correct cards: Editorial standards.
General information, not professional advice. Spotted an error? Flag the card or see corrections and complaints.
Intellect Plus is a visual educational discovery platform for exploring fascinating facts and ideas. Brain Food, Not Brain Rot. What is Intellect Plus?
Related
More in Decision theory · More in Ideas & Philosophy · Latest · A–Z · All topics