Statement on AI Extinction—Signed by AGI Labs, Top Academics, and Many Other Notable Figures

Dan H30 May 2023 9:05 UTC

317 points

50 comments1 min readLW link

(www.safe.ai)

Think carefully before calling RL policies “agents”

TurnTrout2 Jun 2023 3:46 UTC

42 points

1 comment5 min readLW link

Using GPT-Eliezer against ChatGPT Jailbreaking

Stuart_Armstrong and rgorman

6 Dec 2022 19:54 UTC

170 points

85 comments9 min readLW link

Shutdown-Seeking AI

Simon Goldstein31 May 2023 22:19 UTC

20 points

11 comments15 min readLW link

Making Nanobots isn’t a one-shot process, even for an artificial superintelligance

dankrad25 Apr 2023 0:39 UTC

18 points

10 comments6 min readLW link

Steering GPT-2-XL by adding an activation vector

TurnTrout, Monte M, David Udell, lisathiergart and Ulisse Mini

13 May 2023 18:42 UTC

378 points

74 comments50 min readLW link

Is Deontological AI Safe? [Feedback Draft]

Dan H and William D'Alessandro

27 May 2023 16:39 UTC

20 points

13 comments20 min readLW link

Short Remark on the (subjective) mathematical ‘naturalness’ of the Nanda—Lieberum addition modulo 113 algorithm

Spencer Becker-Kahn1 Jun 2023 11:31 UTC

61 points

2 comments2 min readLW link

PaLM-2 & GPT-4 in “Extrapolating GPT-N performance”

Lukas Finnveden30 May 2023 18:33 UTC

51 points

5 comments6 min readLW link

Conditional Prediction with Zero-Sum Training Solves Self-Fulfilling Prophecies

Rubi J. Hudson and Johannes Treutlein

26 May 2023 17:44 UTC

86 points

12 comments24 min readLW link

[Question] Seriously, what goes wrong with “reward the agent when it makes you smile”?

TurnTrout11 Aug 2022 22:22 UTC

81 points

42 comments2 min readLW link

Reward is not the optimization target

TurnTrout25 Jul 2022 0:03 UTC

291 points

109 comments10 min readLW link

A shot at the diamond-alignment problem

TurnTrout6 Oct 2022 18:29 UTC

92 points

57 comments15 min readLW link

Power-seeking can be probable and predictive for trained agents

28 Feb 2023 21:10 UTC

53 points

10 comments9 min readLW link

(arxiv.org)

EIS VI: Critiques of Mechanistic Interpretability Work in AI Safety

scasper17 Feb 2023 20:48 UTC

38 points

9 comments12 min readLW link

Language Agents Reduce the Risk of Existential Catastrophe

cdkg and Simon Goldstein

28 May 2023 19:10 UTC

24 points

12 comments26 min readLW link

Acausal trade: Introduction

Stuart_Armstrong11 May 2017 12:03 UTC

1 point

1 comment1 min readLW link

There are no coherence theorems

20 Feb 2023 21:25 UTC

90 points

99 comments19 min readLW link

Don’t leave your fingerprints on the future

So8res8 Oct 2022 0:35 UTC

109 points

33 comments5 min readLW link

Announcing Apollo Research

Marius Hobbhahn, beren, Lee Sharkey, Lucius Bushnaq, Dan Braun, Mikita Balesni and Jérémy Scheurer

30 May 2023 16:17 UTC

188 points

7 comments8 min readLW link