Talk:Pascal's mugging

Where is the justification for "these utilities can grow faster than the probability diminishes"? Is there any example that demonstrates this? The mugging scenario described does not seem to. — Preceding unsigned comment added by 76.66.88.168 (talk) 01:12, 21 April 2018 (UTC)


 * The general finding that this is a property of most otherwise-reasonable systems, is sourced to . The general concept is sourced to Bostrom's article in which the victim is forced to acknowledge that there is an rough lower bound of "one in 10 quadrillion" to how much probability the victim's cognition assigns to a certain unoutlandish hypotheses. Rolf H Nelson (talk) 19:22, 21 April 2018 (UTC)

Virtue ethics (ish) of rejecting immoral promises
I wonder if there's a way of re-weighting the utility promised by the mugger.

It assumes that the virtuous person is not naive about muggers, and not myopic about the suffering of others.

It involves supposing that the mugger would keep his (from the problem statement) promise. Then ask, what will the mugger have to do, to collect the resources to keep the promise. Then ask, what value will the side-effects of that collection have, with respect to my morality. Assume that it is significant and negative; probably more negative than the value of the collected resources.

On that collection of reasoning and assumptions, it is wrong to accept the promise, so the promise should not alter my decision about the incompetent mugging.

Has any philosopher produced a water-tight argument along the same lines? ArthurDent006.5 (talk) 08:51, 20 February 2021 (UTC)


 * @ArthurDent006.5 No clue, but I would personally reject it on the basis that while money scales linear as a number, its utility does not. a billion is not 1000 times the utility as one million. and at some point it would stop scaling entirely. Yet the chance would keep dropping therefore for most there will be no point where the potential gain outweighs the loss and low probability. 85.144.236.114 (talk) 22:26, 18 August 2023 (UTC)