Inconsistent Values and Extrapolation

niplav
20 Jul 2022 11:59 UTC

We might care about making inconsistent preferences and values consistent. What exactly would that mean? How can me model inconsistent values?

Additional reading: Ontological Crises in Artificial Agents’ Value Systems (De Blanc 2011), Ontology identification problem, Ontology identification problem: Technical tutorial, Diamond Maximizer.

Value ex­trap­o­la­tion, con­cept ex­trap­o­la­tion, model splintering

Stuart_Armstrong8 Mar 2022 22:50 UTC
16 points
1 comment2 min readLW link

Model splin­ter­ing: mov­ing from one im­perfect model to another

Stuart_Armstrong27 Aug 2020 11:53 UTC
79 points
10 comments33 min readLW link

On­tolog­i­cal Crises in Ar­tifi­cial Agents’ Value Sys­tems by Peter de Blanc

jimrandomh21 May 2011 1:05 UTC
26 points
2 comments1 min readLW link

On­tolog­i­cal Cri­sis in Humans

Wei Dai18 Dec 2012 17:32 UTC
78 points
69 comments4 min readLW link

Two More De­ci­sion The­ory Prob­lems for Humans

Wei Dai4 Jan 2019 9:00 UTC
56 points
14 comments2 min readLW link

Us­ing vec­tor fields to vi­su­al­ise prefer­ences and make them consistent

MichaelA and JustinShovelain
28 Jan 2020 19:44 UTC
41 points
32 comments11 min readLW link

In­fer­ring util­ity func­tions from lo­cally non-tran­si­tive preferences

Jan10 Feb 2022 10:33 UTC
32 points
15 comments8 min readLW link
(universalprior.substack.com)

Turn­ing Some In­con­sis­tent Prefer­ences into Con­sis­tent Ones

niplav18 Jul 2022 18:40 UTC
23 points
5 comments12 min readLW link