LESSWRONGTags
LW

Threat Models

EditHistory
Discussion (1)
Help improve this page (2 flags)
EditHistory
Discussion (1)
Help improve this page (2 flags)
Threat Models
Random Tag
Contributors
1Multicore
1Quinn
1Jacob Pfau

A threat model is a story of how a particular risk (e.g. AI) plays out.

In the AI risk case, according to Rohin Shah, a threat model is ideally:

Combination of a development model that says how we get AGI and a risk model that says how AGI leads to existential catastrophe.

...

(Read More)

Posts tagged Threat Models
12
225Another (outer) alignment failure storyΩ
paulfchristiano
2y
Ω
38
11
374What failure looks likeΩ
paulfchristiano
4y
Ω
52
9
71Distinguishing AI takeover scenariosΩ
Sam Clarke, Sammy Martin
2y
Ω
11
8
246What Multipolar Failure Looks Like, and Robust Agent-Agnostic Processes (RAAPs)Ω
Andrew_Critch
2y
Ω
64
8
47Vignettes Workshop (AI Impacts)Ω
Daniel Kokotajlo
2y
Ω
3
6
112Less Realistic Tales of DoomΩ
Mark Xu
2y
Ω
13
6
63Survey on AI existential risk scenariosΩ
Sam Clarke, Alexis Carlier, Jonas Schuett
2y
Ω
11
6
27Investigating AI Takeover ScenariosΩ
Sammy Martin
2y
Ω
1
4
70Rogue AGI Embodies Valuable Intellectual PropertyΩ
Mark Xu, CarlShulman
2y
Ω
9
3
170AI Could Defeat All Of Us Combined
HoldenKarnofsky
1y
42
3
114Clarifying AI X-riskΩ
zac_kenton, Rohin Shah, David Lindner, Vikrant Varma, Vika, Mary Phuong, Ramana Kumar, Elliot Catt
7mo
Ω
23
3
81What Failure Looks Like: Distilling the DiscussionΩ
Ben Pace
3y
Ω
14
3
68My AGI Threat Model: Misaligned Model-Based RL AgentΩ
Steven Byrnes
2y
Ω
40
2
281On how various plans miss the hard bits of the alignment challengeΩ
So8res
1y
Ω
78
2
271A central AI alignment problem: capabilities generalization, and the sharp left turnΩ
So8res
1y
Ω
49
Load More (15/52)
Add Posts