AI ALIGNMENT FORUMStuart_Armstrong
AF

User Profile

Ω111178175

Recent Posts

Curated Posts
Curated - Recent, high quality posts selected by the LessWrong moderation team.
Learn More
Frontpage Posts
Posts meeting our frontpage guidelines: aim to explain, not to persuade. Avoid meta-discussion
(includes curated content and frontpage posts)
Learn More
All Posts
Includes personal and meta blogposts (as well as curated and frontpage).
Learn More

Assuming we've solved X, could we do Y...

10
Stuart Armstrong
20h2 min readShow Highlight
2

Figuring out what Alice wants: non-human Alice

4
Stuart Armstrong
19h1 min readShow Highlight
7

Why we need a *theory* of human values

13
Stuart Armstrong
7d4 min readShow Highlight
0

Humans can be assigned any values whatsoever…

11
Stuart Armstrong
1mo3 min readShow Highlight
0

Disagreement with Paul: alignment induction

4
Stuart Armstrong
3mo1 min readShow Highlight
4

Using expected utility for Good(hart)

6
Stuart Armstrong
4mo6 min readShow Highlight
1

Standard ML Oracles vs Counterfactual ones

5
Stuart Armstrong
2mo5 min readShow Highlight
0

Bridging syntax and semantics with Quine's Gavagai

5
Stuart Armstrong
3mo1 min readShow Highlight
1

Web of connotations: Bleggs, Rubes, thermostats and beliefs

5
Stuart Armstrong
3mo7 min readShow Highlight
0

Wireheading as a potential problem with the new impact measure

8
Stuart Armstrong
3mo4 min readShow Highlight
19
Load More...  

Recent Comments