AI ALIGNMENT FORUMTags
AF

Human Values

EditHistorySubscribe
Discussion (0)
Help improve this page (2 flags)
EditHistorySubscribe
Discussion (0)
Help improve this page (2 flags)
Human Values
Random Tag
Contributors
3plex

Human Values are the things we care about, and would want an aligned superintelligence to look after and support. It is suspected that true human values are highly complex, and could be extrapolated into a wide variety of forms.

Posts tagged Human Values
Most Relevant
9
63The shard theory of human values
Quintin Pope, Alex Turner
4mo
29
8
44Human values & biases are inaccessible to the genome
Alex Turner
7mo
25
1
40Shard Theory: An Overview
David Udell
6mo
2
2
2Brain-over-body biases, and the embodied value problem in AI alignment
Geoffrey Miller
4mo
0
2
53Humans provide an untapped wealth of evidence about alignment
Alex Turner, Quintin Pope
6mo
28
0
40A broad basin of attraction around human values?
Wei Dai
10mo
9
2
27Alignment allows "nonrobust" decision-influences and doesn't require robust grading
Alex Turner
2mo
27
2
23Normativity
Abram Demski
2y
8
2
19Positive values seem more robust and lasting than prohibitions
Alex Turner
2mo
7
2
21Understanding and avoiding value drift
Alex Turner
5mo
5
1
11Would I think for ten thousand years?
Stuart Armstrong
4y
7
0
18Selfishness, preference falsification, and AI alignment
Jessica Taylor
1y
0
1
15Broad Picture of Human Values
Thane Ruthenis
5mo
0
1
14Reflection Mechanisms as an Alignment target: A survey
Marius Hobbhahn, elandgre, Beth Barnes
7mo
0
0
9Preference synthesis illustrated: Star Wars
Stuart Armstrong
3y
3
Load More (15/15)
Add Posts