David Johnston

Karma: 510

A brief theory of why we think things are good or bad

David JohnstonOct 20, 2024, 8:31 PM

7 points

10 comments LW link

Mechanistic Anomaly Detection Research Update

Nora Belrose and David Johnston

Aug 6, 2024, 10:33 AM

11 points

0 comments1 min readLW link

(blog.eleuther.ai)

Opinion merging for AI control

David JohnstonMay 4, 2023, 2:43 AM

6 points

0 comments11 min readLW link

[Question] Is it worth avoiding detailed discussions of expectations about agency levels of powerful AIs?

David JohnstonMar 16, 2023, 3:06 AM

11 points

6 comments2 min readLW link

How likely are malign priors over objectives? [aborted WIP]

David JohnstonNov 11, 2022, 5:36 AM

−1 points

0 comments8 min readLW link

When can a mimic surprise you? Why generative models handle seemingly ill-posed problems

David JohnstonNov 5, 2022, 1:19 PM

8 points

4 comments16 min readLW link

There’s probably a tradeoff between AI capability and safety, and we should act like it

David JohnstonJun 9, 2022, 12:17 AM

3 points

3 comments1 min readLW link

Is evolutionary influence the mesa objective that we’re interested in?

David JohnstonMay 3, 2022, 1:18 AM

3 points

2 comments5 min readLW link

[Cross-post] Half baked ideas: defining and measuring Artificial Intelligence system effectiveness

David JohnstonApr 5, 2022, 12:29 AM

2 points

0 comments7 min readLW link

[Question] Are there any impossibility theorems for strong and safe AI?

David JohnstonMar 11, 2022, 1:41 AM

5 points

3 comments1 min readLW link

Counterfactuals from ensembles of peers

David JohnstonJan 4, 2022, 7:01 AM

3 points

4 comments7 min readLW link