AI ALIGNMENT FORUM
Tags
AF

Satisficer

•

Applied to Principled Satisficing To Avoid Goodhart by JenniferRM 4mo ago

•

Applied to [Aspiration-based designs] 2. Formal framework, basic algorithm by Jobst Heitzig 9mo ago

•

Applied to How to safely use an optimizer by Simon Fischer 9mo ago

•

Applied to [Aspiration-based designs] 1. Informal introduction by Jobst Heitzig 9mo ago

•

Applied to AISC project: SatisfIA – AI that satisfies without overdoing it by Jobst Heitzig 1y ago

•

Applied to Game Theory without Argmax [Part 2] by Cleo Nardo 1y ago

•

Applied to Game Theory without Argmax [Part 1] by Cleo Nardo 1y ago

•

Applied to Aspiration-based Q-Learning by Clément Dumas 1y ago

•

Applied to Hedonic Loops and Taming RL by Beren Millidge 1y ago

•

Applied to Why don't quantilizers also cut off the upper end of the distribution? by Alex_Altair 2y ago

•

Applied to "Corrigibility at some small length" by dath ilan by Christopher King 2y ago

•

Applied to Question 5: The timeline hyperparameter by Cameron Berg 3y ago

•

Applied to Housing Markets, Satisficers, and One-Track Goodhart by Jonathan Bostock 3y ago

•

Applied to Satisficers Tend To Seek Power: Instrumental Convergence Via Retargetability by Alex Turner 3y ago

plex v1.1.0Dec 28th 2020 GMT (+235) added stub description and link to Rob Miles's video LW1

A Satisficer aims to reach a set level of utility rather than maximizing utility. It is a proposed optimization process to the open Other-izer problem.

Video explaining Satisficers vs Maximizers (and why satisficers are still dangerous)

•

Applied to In Praise of Maximizing – With Some Caveats by plex 4y ago

•

Applied to Is 'satificing' optimisation? by plex 4y ago

•

Applied to Satisficers' undefined behaviour by plex 4y ago

•

Applied to Anti-Pascaline satisficer by plex 4y ago