Go Back
Choose this branch
You can't go any further
meritocratic
regular
democratic
hot
top
alive
1854 posts
AI
AI Sentience
Truthful AI
1 posts
40
Towards Hodge-podge Alignment
Cleo Nardo
1d
20
108
The next decades might be wild
Marius Hobbhahn
5d
21
0
I believe some AI doomers are overconfident
FTPickle
6h
4
33
The "Minimal Latents" Approach to Natural Abstractions
johnswentworth
22h
6
22
Existential AI Safety is NOT separate from near-term applications
scasper
7d
15
13
Will Machines Ever Rule the World? MLAISU W50
Esben Kran
4d
4
95
Trying to disambiguate different questions about whether RLHF is “good”
Buck
6d
39
160
AGI Safety FAQ / all-dumb-questions-allowed thread
Aryeh Englander
6mo
514
-3
Why mechanistic interpretability does not and cannot contribute to long-term AGI safety (from messages with a friend)
Remmelt
1d
6
128
Using GPT-Eliezer against ChatGPT Jailbreaking
Stuart_Armstrong
14d
77
29
If Wentworth is right about natural abstractions, it would be bad for alignment
Wuschel Schulz
12d
5
73
Revisiting algorithmic progress
Tamay
7d
6
44
Predicting GPU performance
Marius Hobbhahn
6d
24
31
Is the AI timeline too short to have children?
Yoreth
6d
20
9
Truthfulness, standards and credibility
Joe_Collman
8mo
2