self-locating uncertainty
Uncertainty about which branch or copy of yourself you are, used in Everettian accounts to derive probabilities.
As defined in ms_ama_november_2025
Where it comes up · 19
- 345 | Adam Elga on Being Rational in a Very Large Universe7 claims
- 200 | Solo: The Philosophy of the Multiverse2 claims
- AMA | August 20232 claims
- AMA | February 20232 claims
- AMA | June 20242 claims
- AMA | October 20212 claims
- Ask Me Anything | October 20262 claims
- AMA | April 20261 claim
and 11 more
Most important claims · 10 of 31Search all claims for “self-locating uncertainty” →
- Radford Neal's fully non-indexical conditioning holds that you should condition on everything you know about yourself (memories, age, observations) except your location in the universe, and assume you are typical only within the set of observers macroscopically identical to you; this eliminates the Doomsday and Jovian presumptuousness because you can no longer infer anything about future humans or aliens, while still allowing anthropic predictions.200 | Solo: The Philosophy of the Multiverse
- Reasoning in a multiverse requires two different kinds of credence: the standard scientific credence over which theory is correct (is there a cosmological multiverse, is many-worlds right), and a distinct credence about who we are within a scenario that contains many observers (self-locating or indexical uncertainty), and the latter is the genuinely novel philosophical problem.200 | Solo: The Philosophy of the Multiverse
- There are two good physics worries about the Everett interpretation: first, the probability question—since everything that can happen does happen in some branch, how do you recover the observed Born-rule frequencies (answered via decision theory by Deutsch and Wallace, or self-locating uncertainty by Sebens and Carroll); and second, the less-studied structure question—deriving the messy specific world from just a bare wave function and the Schrödinger equation, which is not obviously achievable and is what Carroll mostly works on now.AMA | September 2021
- The Doomsday Argument and related anthropic/self-locating uncertainty arguments subtly depend on some set of subjectivity or consciousness metaphysics, because if 'we' are merely an epiphenomenal emergent property, it is unclear in what sense we find ourselves to be here and now.AMA | May 2025
- Branches in many-worlds should be thought of as one world that branches into many when decoherence occurs, not as pre-existing identical worlds, because before decoherence interference is possible; worlds are convenient higher-level human constructions that become obvious when macroscopic branching happens, and ambiguous microscopic cases simply don't matter to the fundamental ontology.AMA | February 2023
- In many-worlds, before measurement there is nothing to be probabilistic about; probability is the self-locating uncertainty of future copies of you who, after decoherence, don't know which branch they are on, and even though everything is deterministic at the wave-function level, it is rational and correct to act and speak as if there are objective probabilities.AMA | February 2023
- The 'thickness of a branch' in many-worlds could be conceptualized as a number of copies of each outcome (e.g., 30 vs 70 branches for probabilities 0.3 and 0.7), introducing a discrete element to probability that becomes negligible at large numbers, though whether this is a valid way to ground Born-rule probabilities in many-worlds is uncertain.AMA | June 2024
- In self-locating uncertainty scenarios like a transporter splitting Captain Kirk into two copies, the pre-split person and both post-split copies must assign the same credences because their brain processes are identical, and the only coherent probability in that context is 50-50—independent of any shared reward like a bank account.AMA | August 2023
- Probability in many-worlds arises from self-locating uncertainty: when the wave function branches, there is a version of you on every branch, and after branching none of them knows which they are on; if those observers try to assign credences in a sensible way, the only reasonable answer is the Born rule (probability equal to amplitude squared), which is the unique set of numbers that add up to one and are conserved over time.AMA | December 2021
- Being an Everettian (many-worlds) while denying Bayesianism is not necessarily contradictory, even though an Everettian seems to need to speak of probability as credences/weightings (self-locating uncertainty) rather than as the limiting frequency of many identical experiments as a frequentist would.AMA | June 2024