Go Back
Choose this branch
You can't go any further
meritocratic
regular
democratic
hot
top
alive
105 posts
Optimization
General Intelligence
AI Services (CAIS)
Adaptation Executors
Superstimuli
Narrow AI
Selection vs Control
Hope
Delegation
43 posts
Goodhart's Law
20
Take 6: CAIS is actually Orwellian.
Charlie Steiner
13d
5
79
"Normal" is the equilibrium state of past optimization processes
Alex_Altair
1mo
5
57
Humans aren't fitness maximizers
So8res
2mo
45
21
The economy as an analogy for advanced AI systems
rosehadshar
1mo
0
89
What's General-Purpose Search, And Why Might We Expect To See It In Trained ML Systems?
johnswentworth
4mo
15
61
Vingean Agency
abramdemski
3mo
13
13
The reward function is already how well you manipulate humans
Kerry
2mo
9
159
Utility Maximization = Description Length Minimization
johnswentworth
1y
40
60
Ngo and Yudkowsky on scientific reasoning and pivotal acts
Eliezer Yudkowsky
10mo
13
9
When trying to define general intelligence is ability to achieve goals the best metric?
jmh
1mo
0
193
The ground of optimization
Alex Flint
2y
74
16
program searches
carado
3mo
2
29
Bits of Optimization Can Only Be Lost Over A Distance
johnswentworth
7mo
15
19
Are Intelligence and Generality Orthogonal?
cubefox
5mo
16
55
Alignment allows "nonrobust" decision-influences and doesn't require robust grading
TurnTrout
21d
27
41
Don't align agents to evaluations of plans
TurnTrout
24d
46
48
Don't design agents which exploit adversarial inputs
TurnTrout
1mo
61
18
Reducing Goodhart: Announcement, Executive Summary
Charlie Steiner
4mo
0
50
Signaling isn't about signaling, it's about Goodhart
Valentine
11mo
31
48
Introduction to Reducing Goodhart
Charlie Steiner
1y
10
18
Goodhart's Law Causal Diagrams
JustinShovelain
8mo
2
13
Proxy misspecification and the capabilities vs. value learning race
Sam Marks
7mo
1
147
Goodhart Taxonomy
Scott Garrabrant
4y
33
31
Competent Preferences
Charlie Steiner
1y
2
56
nostalgebraist: Recursive Goodhart's Law
Kaj_Sotala
2y
27
84
Moral Mazes and Short Termism
Zvi
3y
21
75
Classifying specification problems as variants of Goodhart's Law
Vika
3y
5
68
How does Gradient Descent Interact with Goodhart?
Scott Garrabrant
3y
19