Go Back
Choose this branch
Choose this branch
meritocratic
regular
democratic
hot
top
alive
57 posts
Instrumental Convergence
Deconfusion
Orthogonality Thesis
Gradient Hacking
Gradient Descent
66 posts
Corrigibility
Treacherous Turn
Mild Optimization
Quantilization
Satisficer
Tripwire
2
(Extremely) Naive Gradient Hacking Doesn't Work
ojorgensen
9h
0
57
You can still fetch the coffee today if you're dead tomorrow
davidad
11d
15
59
Instrumental convergence is what makes general intelligence possible
tailcalled
1mo
11
33
A caveat to the Orthogonality Thesis
Wuschel Schulz
1mo
10
4
Assessing the Capabilities of ChatGPT through Success Rates
Zachary Robertson
7d
0
38
Empowerment is (almost) All We Need
jacob_cannell
1mo
43
22
Instrumental convergence in single-agent systems
Edouard Harris
2mo
4
2
The Opportunity and Risks of Learning Human Values In-Context
Zachary Robertson
10d
4
17
Misalignment-by-default in multi-agent systems
Edouard Harris
2mo
8
10
Is the Orthogonality Thesis true for humans?
Noosphere89
1mo
18
30
Gradient hacking: definitions and examples
Richard_Ngo
5mo
1
10
Instrumental convergence: scale and physical interactions
Edouard Harris
2mo
0
98
Coherence arguments imply a force for goal-directed behavior
KatjaGrace
1y
27
68
Gradient descent is not just more efficient genetic algorithms
leogao
1y
14
24
Dumb and ill-posed question: Is conceptual research like this MIRI paper on the shutdown problem/Corrigibility "real"
joraine
26d
11
38
People care about each other even though they have imperfect motivational pointers?
TurnTrout
1mo
25
17
Corrigibility Via Thought-Process Deference
Thane Ruthenis
26d
5
108
Interpretability/Tool-ness/Alignment/Corrigibility are not Composable
johnswentworth
4mo
8
127
Let's See You Write That Corrigibility Tag
Eliezer Yudkowsky
6mo
67
113
A broad basin of attraction around human values?
Wei_Dai
8mo
16
60
Steam
abramdemski
6mo
9
134
Soares, Tallinn, and Yudkowsky discuss AGI cognition
So8res
1y
35
17
CHAI, Assistance Games, And Fully-Updated Deference [Scott Alexander]
berglund
2mo
1
27
Quantilizers and Generative Models
Adam Jermyn
5mo
5
77
Satisficers Tend To Seek Power: Instrumental Convergence Via Retargetability
TurnTrout
1y
8
73
Corrigibility Can Be VNM-Incoherent
TurnTrout
1y
24
7
Simple question about corrigibility and values in AI.
jmh
1mo
1
27
[Intro to brain-like-AGI safety] 14. Controlled AGI
Steven Byrnes
7mo
25