All 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 202220232024

AllJanFeb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

All 1 2 3 4 5 678 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31

[Linkpost] Jan Leike on three kinds of alignment taxes

Akash6 Jan 2023 23:57 UTC

27 points

2 comments3 min readLW link

(aligned.substack.com)

The Limit of Language Models

DragonGod6 Jan 2023 23:53 UTC

44 points

26 comments4 min readLW link

Why didn’t we get the four-hour workday?

jasoncrawford6 Jan 2023 21:29 UTC

139 points

34 comments6 min readLW link

(rootsofprogress.org)

AI security might be helpful for AI alignment

Igor Ivanov6 Jan 2023 20:16 UTC

35 points

1 comment2 min readLW link

Categorizing failures as “outer” or “inner” misalignment is often confused

Rohin Shah6 Jan 2023 15:48 UTC

93 points

21 comments8 min readLW link

Definitions of “objective” should be Probable and Predictive

Rohin Shah6 Jan 2023 15:40 UTC

43 points

27 comments12 min readLW link

200 COP in MI: Techniques, Tooling and Automation

Neel Nanda6 Jan 2023 15:08 UTC

13 points

0 comments15 min readLW link

Ball Square Station and Ridership Maximization

jefftk6 Jan 2023 13:20 UTC

13 points

0 comments1 min readLW link

(www.jefftk.com)

Childhood Roundup #1

Zvi6 Jan 2023 13:00 UTC

83 points

27 comments8 min readLW link

(thezvi.wordpress.com)

AI improving AI [MLAISU W01!]

Esben Kran6 Jan 2023 11:13 UTC

5 points

0 comments4 min readLW link

(newsletter.apartresearch.com)

AI Safety Camp, Virtual Edition 2023

Linda Linsefors6 Jan 2023 11:09 UTC

40 points

10 comments3 min readLW link

(aisafety.camp)

Kakistocuriosity

LVSN6 Jan 2023 7:38 UTC

7 points

3 comments1 min readLW link

AI Safety Camp: Machine Learning for Scientific Discovery

Eleni Angelou6 Jan 2023 3:21 UTC

3 points

0 comments1 min readLW link

Metaculus Year in Review: 2022

ChristianWilliams6 Jan 2023 1:23 UTC

6 points

0 comments1 min readLW link

UDASSA

Jacob Falkovich6 Jan 2023 1:07 UTC

21 points

8 comments10 min readLW link

The Involuntary Pacifists

Capybasilisk6 Jan 2023 0:28 UTC

11 points

3 comments2 min readLW link

Get an Electric Toothbrush.

Cervera5 Jan 2023 21:08 UTC

21 points

4 comments1 min readLW link

Discursive Competence in ChatGPT, Part 1: Talking with Dragons

Bill Benzon5 Jan 2023 21:01 UTC

2 points

0 comments6 min readLW link

Transformative AI issues (not just misalignment): an overview

HoldenKarnofsky5 Jan 2023 20:20 UTC

34 points

6 comments18 min readLW link

(www.cold-takes.com)

How to slow down scientific progress, according to Leo Szilard

jasoncrawford5 Jan 2023 18:26 UTC

134 points

18 comments2 min readLW link

(rootsofprogress.org)

Paper: Superposition, Memorization, and Double Descent (Anthropic)

LawrenceC5 Jan 2023 17:54 UTC

53 points

11 comments1 min readLW link

(transformer-circuits.pub)

Collapse Might Not Be Desirable

Dzoldzaya5 Jan 2023 17:29 UTC

−2 points

9 comments2 min readLW link

Singapore—Small casual dinner in Chinatown #6

Joe Rocca5 Jan 2023 17:00 UTC

2 points

1 comment1 min readLW link

[Question] Image generation and alignment

rpglover645 Jan 2023 16:05 UTC

3 points

3 comments1 min readLW link

[Question] Machine Learning vs Differential Privacy

Ilio5 Jan 2023 15:14 UTC

10 points

10 comments1 min readLW link

Covid 1/5/23: Various XBB Takes

Zvi5 Jan 2023 14:20 UTC

21 points

18 comments15 min readLW link

(thezvi.wordpress.com)

Running by Default

jefftk5 Jan 2023 13:50 UTC

111 points

40 comments1 min readLW link

(www.jefftk.com)

PSA: reward is part of the habit loop too

Alok Singh5 Jan 2023 11:00 UTC

22 points

2 comments1 min readLW link

(alok.github.io)

Infohazards vs Fork Hazards

jimrandomh5 Jan 2023 9:45 UTC

68 points

16 comments1 min readLW link

Monthly Shorts 12/22

Celer5 Jan 2023 7:20 UTC

5 points

2 comments1 min readLW link

(keller.substack.com)

The 2021 Review Phase

Raemon5 Jan 2023 7:12 UTC

34 points

7 comments3 min readLW link

Illusion of truth effect and Ambiguity effect: Bias in Evaluating AGI X-Risks

Remmelt5 Jan 2023 4:05 UTC

−13 points

2 comments1 min readLW link

When you plan according to your AI timelines, should you put more weight on the median future, or the median future | eventual AI alignment success? ⚖️

Jeffrey Ladish5 Jan 2023 1:21 UTC

25 points

10 comments2 min readLW link

Why I’m joining Anthropic

evhub5 Jan 2023 1:12 UTC

121 points

4 comments1 min readLW link

Contra Common Knowledge

abramdemski4 Jan 2023 22:50 UTC

52 points

31 comments16 min readLW link

Additional space complexity isn’t always a useful metric

Brendan Long4 Jan 2023 21:53 UTC

4 points

3 comments3 min readLW link

(www.brendanlong.com)

List of links for getting into AI safety

zef4 Jan 2023 19:45 UTC

6 points

0 comments1 min readLW link

Opening Facebook Links Externally

jefftk4 Jan 2023 19:00 UTC

12 points

3 comments1 min readLW link

(www.jefftk.com)

Conversational canyons

Henrik Karlsson4 Jan 2023 18:55 UTC

59 points

4 comments7 min readLW link

(escapingflatland.substack.com)

Progress links and tweets, 2023-01-04

jasoncrawford4 Jan 2023 18:23 UTC

15 points

0 comments1 min readLW link

(rootsofprogress.org)

200 COP in MI: Analysing Training Dynamics

Neel Nanda4 Jan 2023 16:08 UTC

16 points

0 comments14 min readLW link

What’s up with ChatGPT and the Turing Test?

JoshuaFox and Zvi Schreiber

4 Jan 2023 15:37 UTC

13 points

19 comments3 min readLW link

2022 was the year AGI arrived (Just don’t call it that)

Logan Zoellner4 Jan 2023 15:19 UTC

102 points

60 comments3 min readLW link

From Simon’s ant to machine learning, a parable

Bill Benzon4 Jan 2023 14:37 UTC

6 points

5 comments2 min readLW link

Basic Facts about Language Model Internals

beren and Eric Winsor

4 Jan 2023 13:01 UTC

130 points

19 comments9 min readLW link

Ritual as the only tool for overwriting values and goals

mrcbarbier4 Jan 2023 11:11 UTC

40 points

24 comments32 min readLW link

Normalcy bias and Base rate neglect: Bias in Evaluating AGI X-Risks

Remmelt4 Jan 2023 3:16 UTC

−16 points

0 comments1 min readLW link

Causal representation learning as a technique to prevent goal misgeneralization

PabloAMC4 Jan 2023 0:07 UTC

19 points

0 comments8 min readLW link

What makes a probability question “well-defined”? (Part II: Bertrand’s Paradox)

Noah Topper3 Jan 2023 22:39 UTC

7 points

3 comments9 min readLW link

(naivebayes.substack.com)

“AI” is an indexical

TW1233 Jan 2023 22:00 UTC

10 points

0 comments6 min readLW link

(aiwatchtower.substack.com)