AI ALIGNMENT FORUM
AF

Wikitags

Satisficer

Edited by plex last updated 28th Dec 2020

A Satisficer aims to reach a set level of 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)

Subscribe
1
Subscribe
1
utility
Discussion0
Discussion0
Posts tagged Satisficer
41Satisficers Tend To Seek Power: Instrumental Convergence Via Retargetability
Alex Turner
4y
8
18Game Theory without Argmax [Part 1]
Cleo Nardo
2y
1
19[Aspiration-based designs] 1. Informal introduction
B Jacobs, Jobst Heitzig, Simon Fischer, Simon Dima
1y
0
8Why don't quantilizers also cut off the upper end of the distribution?
Q
Alex_Altair, Jeremy Gillen
2y
Q
1
10Hedonic Loops and Taming RL
Beren Millidge
2y
0
15How to safely use an optimizer
Simon Fischer
1y
0
15Aspiration-based Q-Learning
Clément Dumas, Jobst Heitzig
2y
0
8[Aspiration-based designs] 2. Formal framework, basic algorithm
Jobst Heitzig, Simon Dima, Simon Fischer
1y
0
Add Posts