Go Back
Choose this branch
You can't go any further
meritocratic
regular
democratic
hot
top
alive
65 posts
Utility Functions
Unbounded Utilities
Coherence Arguments
Indexical Information
4 posts
36
Take 7: You should talk about "the human's utility function" less.
Charlie Steiner
12d
22
137
Impossibility results for unbounded utilities
paulfchristiano
10mo
104
9
Will Values and Competition Decouple?
interstice
2mo
11
137
An Orthodox Case Against Utility Functions
abramdemski
2y
53
161
Coherent decisions imply consistent utilities
Eliezer Yudkowsky
3y
81
29
Inferring utility functions from locally non-transitive preferences
Jan
10mo
15
105
Utility ≠ Reward
vlad_m
3y
25
40
I’m no longer sure that I buy dutch book arguments and this makes me skeptical of the "utility function" abstraction
Eli Tyre
1y
29
20
Better impossibility result for unbounded utilities
paulfchristiano
10mo
24
93
Degrees of Freedom
sarahconstantin
3y
31
102
Coherence arguments do not entail goal-directed behavior
Rohin Shah
4y
69
6
Adaptation Executors and the Telos Margin
Plinthist
6mo
8
6
Thinking about Broad Classes of Utility-like Functions
Jemist
6mo
0
29
How easily can we separate a friendly AI in design space from one which would bring about a hyperexistential catastrophe?
Anirandis
2y
20
58
The "Measuring Stick of Utility" Problem
johnswentworth
6mo
22
21
Deriving Conditional Expected Utility from Pareto-Efficient Decisions
Thomas Kwa
7mo
1
51
Three ways that "Sufficiently optimized agents appear coherent" can be false
Wei_Dai
3y
3
46
Coherent behaviour in the real world is an incoherent concept
Richard_Ngo
3y
17