# Sharp Left Turn

TagLast edit: 3 Sep 2022 3:54 UTC by

A Sharp Left Turn is a scenario where, as an AI trains, its capabilities generalize across many domains while the alignment properties that held at earlier stages fail to generalize to the new domains.

# [Question] Has Eliezer pub­li­cly and satis­fac­to­rily re­sponded to at­tempted re­but­tals of the anal­ogy to evolu­tion?

28 Jul 2024 12:23 UTC
17 points

# [In­ter­view w/​ Quintin Pope] Evolu­tion, val­ues, and AI Safety

24 Oct 2023 13:53 UTC
11 points

# Evolu­tion Solved Align­ment (what sharp left turn?)

12 Oct 2023 4:15 UTC
20 points

# We don’t un­der­stand what hap­pened with cul­ture enough

9 Oct 2023 9:54 UTC
86 points

# Re­sponse to Quintin Pope’s Evolu­tion Pro­vides No Ev­i­dence For the Sharp Left Turn

5 Oct 2023 11:39 UTC
129 points

# [Question] A few Align­ment ques­tions: util­ity op­ti­miz­ers, SLT, sharp left turn and identifiability

26 Sep 2023 0:27 UTC
5 points

# The Sharp Right Turn: sud­den de­cep­tive al­ign­ment as a con­ver­gent goal

6 Jun 2023 9:59 UTC
38 points

# A smart enough LLM might be deadly sim­ply if you run it for long enough

5 May 2023 20:49 UTC
16 points

# Evolu­tion pro­vides no ev­i­dence for the sharp left turn

11 Apr 2023 18:43 UTC
202 points

# Vic­to­ria Krakovna on AGI Ruin, The Sharp Left Turn and Paradigms of AI Alignment

12 Jan 2023 17:09 UTC
40 points
(www.theinsideview.ai)

# Refram­ing in­ner alignment

11 Dec 2022 13:53 UTC
53 points

# [Question] How is the “sharp left turn defined”?

9 Dec 2022 0:04 UTC
14 points

# Refin­ing the Sharp Left Turn threat model, part 2: ap­ply­ing al­ign­ment techniques

25 Nov 2022 14:36 UTC
39 points
(vkrakovna.wordpress.com)

# A caveat to the Orthog­o­nal­ity Thesis

9 Nov 2022 15:06 UTC
38 points

# Disen­tan­gling in­ner al­ign­ment failures

10 Oct 2022 18:50 UTC
20 points

# Smoke with­out fire is scary

4 Oct 2022 21:08 UTC
51 points

# It mat­ters when the first sharp left turn happens

29 Sep 2022 20:12 UTC
44 points

# We may be able to see sharp left turns coming

3 Sep 2022 2:55 UTC
53 points

# Refin­ing the Sharp Left Turn threat model, part 1: claims and mechanisms

12 Aug 2022 15:17 UTC
85 points
(vkrakovna.wordpress.com)

# Goal Align­ment Is Ro­bust To the Sharp Left Turn

13 Jul 2022 20:23 UTC
47 points