Go Back
You can't go any further
You can't go any further
meritocratic
regular
democratic
hot
top
alive
15 posts
Threat Models
6 posts
Sharp Left Turn
39
AI Neorealism: a threat model & success criterion for existential safety
davidad
5d
0
102
Clarifying AI X-risk
zac_kenton
1mo
23
55
Threat Model Literature Review
zac_kenton
1mo
4
36
AI X-risk >35% mostly based on a recent peer-reviewed argument
michaelcohen
1mo
31
168
AI Could Defeat All Of Us Combined
HoldenKarnofsky
6mo
29
210
Another (outer) alignment failure story
paulfchristiano
1y
38
203
What Multipolar Failure Looks Like, and Robust Agent-Agnostic Processes (RAAPs)
Andrew_Critch
1y
60
319
What failure looks like
paulfchristiano
3y
49
110
Less Realistic Tales of Doom
Mark Xu
1y
13
67
Distinguishing AI takeover scenarios
Sam Clarke
1y
11
70
Rogue AGI Embodies Valuable Intellectual Property
Mark Xu
1y
9
60
Survey on AI existential risk scenarios
Sam Clarke
1y
11
47
Vignettes Workshop (AI Impacts)
Daniel Kokotajlo
1y
3
79
What Failure Looks Like: Distilling the Discussion
Ben Pace
2y
14
36
Refining the Sharp Left Turn threat model, part 2: applying alignment techniques
Vika
25d
4
13
How is the "sharp left turn defined"?
Chris_Leong
12d
3
253
A central AI alignment problem: capabilities generalization, and the sharp left turn
So8res
6mo
48
71
Refining the Sharp Left Turn threat model, part 1: claims and mechanisms
Vika
4mo
3
50
We may be able to see sharp left turns coming
Ethan Perez
3mo
26
35
It matters when the first sharp left turn happens
Adam Jermyn
2mo
9