All200520062007200820092010201120122013201420152016201720182019202020212022202320242025
AllJanFebMarAprMayJunJulAugSepOctNovDec
All12345678910111213141516171819202122232425262728293031

Truth­seek­ing, EA, Si­mu­lacra lev­els, and other stuff

Elizabeth and Vaniver
Oct 27, 2023, 11:56 PM
45 points

15 votes

Overall karma indicates overall quality.

12 comments9 min readLW link

[Question] Do you be­lieve “E=mc^2” is a cor­rect and/​or use­ful equa­tion, and, whether yes or no, pre­cisely what are your rea­sons for hold­ing this be­lief (with such a de­gree of con­fi­dence)?

l8cOct 27, 2023, 10:46 PM
12 points

15 votes

Overall karma indicates overall quality.

14 comments1 min readLW link

Value sys­tem­ati­za­tion: how val­ues be­come co­her­ent (and mis­al­igned)

Richard_NgoOct 27, 2023, 7:06 PM
108 points

44 votes

Overall karma indicates overall quality.

49 comments13 min readLW link

Techno-hu­man­ism is techno-op­ti­mism for the 21st century

Richard_NgoOct 27, 2023, 6:37 PM
88 points

29 votes

Overall karma indicates overall quality.

5 comments14 min readLW link
(www.mindthefuture.info)

Sanc­tu­ary for Humans

Nikola JurkovicOct 27, 2023, 6:08 PM
22 points

14 votes

Overall karma indicates overall quality.

9 comments1 min readLW link

Wire­head­ing and mis­al­ign­ment by com­po­si­tion on NetHack

pierlucadoroOct 27, 2023, 5:43 PM
34 points

16 votes

Overall karma indicates overall quality.

4 comments4 min readLW link

We’re Not Ready: thoughts on “paus­ing” and re­spon­si­ble scal­ing policies

HoldenKarnofskyOct 27, 2023, 3:19 PM
200 points

81 votes

Overall karma indicates overall quality.

33 comments8 min readLW link

Aspira­tion-based Q-Learning

Clément Dumas and Jobst Heitzig
Oct 27, 2023, 2:42 PM
38 points

13 votes

Overall karma indicates overall quality.

5 comments11 min readLW link

Linkpost: Rishi Su­nak’s Speech on AI (26th Oc­to­ber)

bideupOct 27, 2023, 11:57 AM
85 points

31 votes

Overall karma indicates overall quality.

8 comments7 min readLW link
(www.gov.uk)

ASPR & WARP: Ra­tion­al­ity Camps for Teens in Taiwan and Oxford

Anna GajdovaOct 27, 2023, 8:40 AM
18 points

7 votes

Overall karma indicates overall quality.

0 comments1 min readLW link

[Question] To what ex­tent is the UK Govern­ment’s re­cent AI Safety push en­tirely due to Rishi Su­nak?

Stephen FowlerOct 27, 2023, 3:29 AM
23 points

13 votes

Overall karma indicates overall quality.

4 comments1 min readLW link

Bayesian Punishment

Rob LucasOct 27, 2023, 3:24 AM
1 point

2 votes

Overall karma indicates overall quality.

1 comment6 min readLW link

On­line Dialogues Party — Sun­day 5th November

Ben PaceOct 27, 2023, 2:41 AM
28 points

7 votes

Overall karma indicates overall quality.

1 comment1 min readLW link

OpenAI’s new Pre­pared­ness team is hiring

leopoldOct 26, 2023, 8:42 PM
60 points

20 votes

Overall karma indicates overall quality.

2 comments1 min readLW link

Fake Deeply

Zack_M_DavisOct 26, 2023, 7:55 PM
33 points

13 votes

Overall karma indicates overall quality.

7 comments1 min readLW link
(unremediatedgender.space)

Sym­bol/​Refer­ent Con­fu­sions in Lan­guage Model Align­ment Experiments

johnswentworthOct 26, 2023, 7:49 PM
120 points

55 votes

Overall karma indicates overall quality.

51 comments6 min readLW link1 review

Un­su­per­vised Meth­ods for Con­cept Dis­cov­ery in AlphaZero

aogOct 26, 2023, 7:05 PM
9 points

3 votes

Overall karma indicates overall quality.

0 comments1 min readLW link
(arxiv.org)

[Question] Non­lin­ear limi­ta­tions of ReLUs

magfrumpOct 26, 2023, 6:51 PM
13 points

4 votes

Overall karma indicates overall quality.

1 comment1 min readLW link

AI Align­ment Prob­lem: Re­quire­ment not op­tional (A Crit­i­cal Anal­y­sis through Mass Effect Tril­ogy)

TAWSIF AHMEDOct 26, 2023, 6:02 PM
−9 points

3 votes

Overall karma indicates overall quality.

0 comments4 min readLW link

[Thought Ex­per­i­ment] To­mor­row’s Echo—The fu­ture of syn­thetic com­pan­ion­ship.

Vimal NaranOct 26, 2023, 5:54 PM
−7 points

3 votes

Overall karma indicates overall quality.

2 comments2 min readLW link

Disagree­ments over the pri­ori­ti­za­tion of ex­is­ten­tial risk from AI

Olivier CoutuOct 26, 2023, 5:54 PM
10 points

5 votes

Overall karma indicates overall quality.

0 comments6 min readLW link

[Question] What if AGI had its own uni­verse to maybe wreck?

msealeOct 26, 2023, 5:49 PM
−1 points

4 votes

Overall karma indicates overall quality.

2 comments1 min readLW link

Chang­ing Con­tra Dialects

jefftkOct 26, 2023, 5:30 PM
25 points

6 votes

Overall karma indicates overall quality.

2 comments1 min readLW link
(www.jefftk.com)

5 psy­cholog­i­cal rea­sons for dis­miss­ing x-risks from AGI

Igor IvanovOct 26, 2023, 5:21 PM
24 points

12 votes

Overall karma indicates overall quality.

6 comments4 min readLW link

5. Risks from pre­vent­ing le­gi­t­i­mate value change (value col­lapse)

Nora_AmmannOct 26, 2023, 2:38 PM
13 points

6 votes

Overall karma indicates overall quality.

1 comment9 min readLW link

4. Risks from caus­ing ille­gi­t­i­mate value change (perfor­ma­tive pre­dic­tors)

Nora_AmmannOct 26, 2023, 2:38 PM
13 points

6 votes

Overall karma indicates overall quality.

3 comments5 min readLW link

3. Premise three & Con­clu­sion: AI sys­tems can af­fect value change tra­jec­to­ries & the Value Change Problem

Nora_AmmannOct 26, 2023, 2:38 PM
33 points

12 votes

Overall karma indicates overall quality.

4 comments4 min readLW link

2. Premise two: Some cases of value change are (il)legitimate

Nora_AmmannOct 26, 2023, 2:36 PM
29 points

14 votes

Overall karma indicates overall quality.

7 comments6 min readLW link

1. Premise one: Values are malleable

Nora_AmmannOct 26, 2023, 2:36 PM
26 points

11 votes

Overall karma indicates overall quality.

1 comment15 min readLW link

0. The Value Change Prob­lem: in­tro­duc­tion, overview and motivations

Nora_AmmannOct 26, 2023, 2:36 PM
32 points

15 votes

Overall karma indicates overall quality.

0 comments5 min readLW link

EPUBs of MIRI Blog Archives and se­lected LW Sequences

mesaoptimizerOct 26, 2023, 2:17 PM
44 points

15 votes

Overall karma indicates overall quality.

5 comments1 min readLW link
(git.sr.ht)

UK Govern­ment pub­lishes “Fron­tier AI: ca­pa­bil­ities and risks” Dis­cus­sion Paper

A.H.Oct 26, 2023, 1:55 PM
5 points

3 votes

Overall karma indicates overall quality.

0 comments2 min readLW link
(www.gov.uk)

AI #35: Re­spon­si­ble Scal­ing Policies

ZviOct 26, 2023, 1:30 PM
66 points

26 votes

Overall karma indicates overall quality.

10 comments55 min readLW link
(thezvi.wordpress.com)

RA Bounty: Look­ing for feed­back on screen­play about AI Risk

WriterOct 26, 2023, 1:23 PM
32 points

8 votes

Overall karma indicates overall quality.

6 comments1 min readLW link

Notes on “How do we be­come con­fi­dent in the safety of a ma­chine learn­ing sys­tem?”

RohanSOct 26, 2023, 3:13 AM
4 points

2 votes

Overall karma indicates overall quality.

0 comments13 min readLW link

Ap­ply to the Con­stel­la­tion Visit­ing Re­searcher Pro­gram and As­tra Fel­low­ship, in Berkeley this Winter

Nate ThomasOct 26, 2023, 3:07 AM
42 points

16 votes

Overall karma indicates overall quality.

10 comments1 min readLW link

CHAI in­tern­ship ap­pli­ca­tions are open (due Nov 13)

Erik JennerOct 26, 2023, 12:53 AM
34 points

9 votes

Overall karma indicates overall quality.

0 comments3 min readLW link

Ar­chi­tects of Our Own Demise: We Should Stop Devel­op­ing AI Carelessly

RokoOct 26, 2023, 12:36 AM
170 points

104 votes

Overall karma indicates overall quality.

75 comments3 min readLW link

EA In­fras­truc­ture Fund: June 2023 grant recommendations

LinchOct 26, 2023, 12:35 AM
21 points

6 votes

Overall karma indicates overall quality.

0 comments12 min readLW link

Re­spon­si­ble Scal­ing Poli­cies Are Risk Man­age­ment Done Wrong

simeon_cOct 25, 2023, 11:46 PM
123 points

58 votes

Overall karma indicates overall quality.

35 comments22 min readLW link1 review
(www.navigatingrisks.ai)

AI as a sci­ence, and three ob­sta­cles to al­ign­ment strategies

So8resOct 25, 2023, 9:00 PM
194 points

101 votes

Overall karma indicates overall quality.

80 comments11 min readLW link

My hopes for al­ign­ment: Sin­gu­lar learn­ing the­ory and whole brain emulation

Garrett BakerOct 25, 2023, 6:31 PM
61 points

27 votes

Overall karma indicates overall quality.

5 comments12 min readLW link

[Question] Ly­ing to chess play­ers for alignment

ZaneOct 25, 2023, 5:47 PM
100 points

48 votes

Overall karma indicates overall quality.

55 comments1 min readLW link

An­thropic, Google, Microsoft & OpenAI an­nounce Ex­ec­u­tive Direc­tor of the Fron­tier Model Fo­rum & over $10 mil­lion for a new AI Safety Fund

Zach Stein-PerlmanOct 25, 2023, 3:20 PM
31 points

16 votes

Overall karma indicates overall quality.

8 comments4 min readLW link
(www.frontiermodelforum.org)

“The Eco­nomics of Time Travel”—call for re­view­ers (Seeds of Science)

rogersbaconOct 25, 2023, 3:13 PM
4 points

3 votes

Overall karma indicates overall quality.

2 comments1 min readLW link

Com­po­si­tional prefer­ence mod­els for al­ign­ing LMs

Tomek KorbakOct 25, 2023, 12:17 PM
18 points

6 votes

Overall karma indicates overall quality.

2 comments5 min readLW link

[Question] Should the US House of Rep­re­sen­ta­tives adopt rank choice vot­ing for lead­er­ship po­si­tions?

jmhOct 25, 2023, 11:16 AM
16 points

4 votes

Overall karma indicates overall quality.

6 comments1 min readLW link

Re­searchers be­lieve they have found a way for artists to fight back against AI style capture

vernamcipherOct 25, 2023, 10:54 AM
3 points

2 votes

Overall karma indicates overall quality.

1 comment1 min readLW link
(finance.yahoo.com)

Why We Disagree

zulupineappleOct 25, 2023, 10:50 AM
7 points

5 votes

Overall karma indicates overall quality.

2 comments2 min readLW link

Beyond the Data: Why aid to poor doesn’t work

LyrongolemOct 25, 2023, 5:03 AM
2 points

19 votes

Overall karma indicates overall quality.

31 comments12 min readLW link