Be­come a per­son who Ac­tu­ally Does Things

Neel NandaJul 25, 2020, 9:29 AM
55 points
17 comments3 min readLW link
(www.neelnanda.io)

[Question] Con­struct a port­fo­lio to profit from AI progress.

sapphireJul 25, 2020, 8:18 AM
30 points
13 comments1 min readLW link

[Question] Us­ing books to prime behavior

ryan wongJul 25, 2020, 5:59 AM
10 points
8 comments1 min readLW link

Re­veal Culture

MalcolmOceanJul 25, 2020, 3:36 AM
92 points
18 comments7 min readLW link2 reviews

Sport­ing vs. Herd­ing: two styles of discourse

Suspended ReasonJul 25, 2020, 1:56 AM
5 points
4 comments9 min readLW link

Pho­tos Be­fore Drywall

jefftkJul 25, 2020, 1:40 AM
18 points
3 comments1 min readLW link
(www.jefftk.com)

Com­ment Replies for Chains, Bot­tle­necks and Optimization

curiJul 24, 2020, 6:51 PM
1 point
2 comments3 min readLW link

No Ul­ti­mate Goal and a Small Ex­is­ten­tial Crisis

Logan RiggsJul 24, 2020, 6:39 PM
12 points
23 comments3 min readLW link

Writ­ing with GPT-3

Jacob FalkovichJul 24, 2020, 3:22 PM
42 points
0 comments4 min readLW link

Longevity in­ter­ven­tions when young

George3d6Jul 24, 2020, 11:25 AM
45 points
17 comments39 min readLW link
(cerebralab.com)

Mean­ing is Quasi-Idempotent

Chris_LeongJul 24, 2020, 4:08 AM
5 points
10 comments2 min readLW link

Be impatient

benkuhnJul 24, 2020, 1:30 AM
86 points
5 comments7 min readLW link
(www.benkuhn.net)

Op­ti­miz­ing ar­bi­trary ex­pres­sions with a lin­ear num­ber of queries to a Log­i­cal In­duc­tion Or­a­cle (Car­toon Guide)

Donald HobsonJul 23, 2020, 9:37 PM
4 points
2 comments2 min readLW link

Bot­tle­neck Examples

curiJul 23, 2020, 6:23 PM
11 points
0 comments9 min readLW link

Ag­gre­gat­ing forecasts

JsevillamolJul 23, 2020, 6:04 PM
17 points
4 comments4 min readLW link

Can you get AGI from a Trans­former?

Steven ByrnesJul 23, 2020, 3:27 PM
117 points
40 comments12 min readLW link

Covid 7/​23: The Se­cond Summit

ZviJul 23, 2020, 2:40 PM
59 points
6 comments8 min readLW link
(thezvi.wordpress.com)

[Question] Are par­a­sitic worms an effec­tive weapon against MS?

DemeterJul 23, 2020, 1:59 PM
31 points
11 comments9 min readLW link

Restocking

jefftkJul 23, 2020, 1:40 PM
12 points
0 comments1 min readLW link
(www.jefftk.com)

Cambridge Vir­tual LW/​SSC Meetup

NoSignalNoNoiseJul 23, 2020, 3:25 AM
6 points
0 comments1 min readLW link

GPT-3 Gems

TurnTroutJul 23, 2020, 12:46 AM
33 points
10 comments48 min readLW link

Weak HCH ac­cesses EXP

evhubJul 22, 2020, 10:36 PM
16 points
0 comments3 min readLW link

Repli­cat­ing the repli­ca­tion crisis with GPT-3?

skybrianJul 22, 2020, 9:20 PM
29 points
10 comments1 min readLW link

A New Way to Vi­su­al­ize Biases

dtmJul 22, 2020, 5:57 PM
6 points
16 comments1 min readLW link

[AN #109]: Teach­ing neu­ral nets to gen­er­al­ize the way hu­mans would

Rohin ShahJul 22, 2020, 5:10 PM
17 points
3 comments9 min readLW link
(mailchi.mp)

On­line Meetup: Balaji Srini­vasan on “Na­tion of Emi­grants”

JoshuaFoxJul 22, 2020, 2:35 PM
14 points
1 comment1 min readLW link

Swiss Poli­ti­cal Sys­tem: More than You ever Wanted to Know (II.)

Martin SustrikJul 22, 2020, 1:50 PM
90 points
24 comments12 min readLW link
(250bpm.com)

[Question] If you are signed up for cry­on­ics with life in­surance, how much life in­surance did you get and over what term?

Eli TyreJul 22, 2020, 8:13 AM
26 points
4 comments1 min readLW link

The Ba­sic Dou­ble Crux pattern

Eli TyreJul 22, 2020, 6:41 AM
68 points
21 comments4 min readLW link

Re­search agenda for AI safety and a bet­ter civilization

agilecavemanJul 22, 2020, 6:35 AM
12 points
2 comments16 min readLW link

More Right

Adam ZernerJul 22, 2020, 3:36 AM
22 points
29 comments4 min readLW link

[not on­go­ing] Thoughts on Pro­por­tional vot­ing methods

Jameson QuinnJul 22, 2020, 2:46 AM
32 points
53 comments46 min readLW link

[Preprint] The Com­pu­ta­tional Limits of Deep Learning

Gordon Seidoh WorleyJul 21, 2020, 9:25 PM
9 points
4 comments1 min readLW link
(arxiv.org)

Fresh Bread

ZviJul 21, 2020, 8:40 PM
22 points
1 comment2 min readLW link
(thezvi.wordpress.com)

Com­pe­ti­tion: Am­plify Ro­hin’s Pre­dic­tion on AGI re­searchers & Safety Concerns

stuhlmuellerJul 21, 2020, 8:06 PM
83 points
41 comments3 min readLW link

Align­ment As A Bot­tle­neck To Use­ful­ness Of GPT-3

johnswentworthJul 21, 2020, 8:02 PM
111 points
57 comments3 min readLW link

How good is hu­man­ity at co­or­di­na­tion?

BuckJul 21, 2020, 8:01 PM
82 points
44 comments3 min readLW link

$1000 bounty for OpenAI to show whether GPT3 was “de­liber­ately” pre­tend­ing to be stupi­der than it is

Bird Concept21 Jul 2020 18:42 UTC
56 points
39 comments2 min readLW link
(twitter.com)

[Question] What are the limits of self-ed­u­ca­tion?

nitropie21 Jul 2020 18:01 UTC
3 points
2 comments1 min readLW link

[Meta] anony­mous merit or pub­lic status

[anonymous]21 Jul 2020 18:01 UTC
6 points
4 comments1 min readLW link

AI Benefits Post 5: Out­stand­ing Ques­tions on Govern­ing Benefits

Cullen21 Jul 2020 16:46 UTC
4 points
0 comments4 min readLW link

The “AI Dun­geons” Dragon Model is heav­ily path de­pen­dent (test­ing GPT-3 on ethics)

Rafael Harth21 Jul 2020 12:14 UTC
44 points
9 comments6 min readLW link

Un­cal­ibrated quan­tum ex­per­i­ments act cla­si­cally

justinpombrio21 Jul 2020 5:31 UTC
18 points
12 comments8 min readLW link

The Redis­cov­ery of In­te­ri­or­ity in Ma­chine Learning

DanB21 Jul 2020 5:02 UTC
5 points
4 commentsLW link
(danburfoot.net)

Chains, Bot­tle­necks and Optimization

curi21 Jul 2020 2:07 UTC
14 points
12 comments4 min readLW link

“Can you keep this con­fi­den­tial? How do you know?”

Raemon21 Jul 2020 0:33 UTC
164 points
43 comments3 min readLW link2 reviews

Par­allels Between AI Safety by De­bate and Ev­i­dence Law

Cullen20 Jul 2020 22:52 UTC
10 points
1 comment2 min readLW link
(cullenokeefe.com)

Thiel on Progress and Stagnation

Richard_Ngo20 Jul 2020 20:27 UTC
173 points
32 comments11 min readLW link
(docs.google.com)

Learn­ing Values in Practice

Stuart_Armstrong20 Jul 2020 18:38 UTC
24 points
0 comments5 min readLW link

Ineffi­cient doesn’t mean in­differ­ent, but it might mean wimpy.

DirectedEvolution20 Jul 2020 18:27 UTC
14 points
3 comments5 min readLW link