Are we in an AI over­hang?

Andy JonesJul 27, 2020, 12:48 PM
266 points
106 comments4 min readLW link

DARPA Digi­tal Tu­tor: Four Months to To­tal Tech­ni­cal Ex­per­tise?

SebastianG Jul 6, 2020, 11:34 PM
220 points
22 comments7 min readLW link

Limits of Cur­rent US Pre­dic­tion Mar­kets (Pre­dic­tIt Case Study)

aphyerJul 14, 2020, 7:24 AM
210 points
50 comments7 min readLW link

Six eco­nomics mis­con­cep­tions of mine which I’ve re­solved over the last few years

BuckJul 13, 2020, 3:01 AM
196 points
59 comments9 min readLW link

Reread­ing At­las Shrugged

VaniverJul 28, 2020, 6:54 PM
174 points
36 comments13 min readLW link1 review

Thiel on Progress and Stagnation

Richard_NgoJul 20, 2020, 8:27 PM
173 points
32 comments11 min readLW link
(docs.google.com)

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

Martin SustrikJul 19, 2020, 1:11 AM
172 points
39 comments24 min readLW link2 reviews

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

RaemonJul 21, 2020, 12:33 AM
163 points
43 comments3 min readLW link2 reviews

Devel­op­men­tal Stages of GPTs

orthonormalJul 26, 2020, 10:03 PM
140 points
72 comments7 min readLW link1 review

Si­tu­at­ing LessWrong in con­tem­po­rary philos­o­phy: An in­ter­view with Jon Livengood

Suspended ReasonJul 1, 2020, 12:37 AM
117 points
21 comments19 min readLW link

Can you get AGI from a Trans­former?

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

Clas­sify­ing games like the Pri­soner’s Dilemma

philhJul 4, 2020, 5:10 PM
111 points
28 comments6 min readLW link1 review
(reasonableapproximation.net)

Align­ment As A Bot­tle­neck To Use­ful­ness Of GPT-3

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

Solv­ing Math Prob­lems by Relay

Jul 17, 2020, 3:32 PM
103 points
26 comments7 min readLW link

Kelly Bet on Everything

Jacob FalkovichJul 10, 2020, 2:48 AM
101 points
20 comments5 min readLW link2 reviews

At­ten­tion is your scarcest resource

benkuhnJul 30, 2020, 1:00 AM
98 points
2 comments4 min readLW link
(www.benkuhn.net)

Learn­ing the prior

paulfchristianoJul 5, 2020, 9:00 PM
92 points
28 comments8 min readLW link
(ai-alignment.com)

Re­veal Culture

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

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)

Col­lec­tion of GPT-3 results

Kaj_SotalaJul 18, 2020, 8:04 PM
89 points
24 comments1 min readLW link
(twitter.com)

Be impatient

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

Your Pri­ori­ti­za­tion is Underspecified

romeostevensitJul 10, 2020, 8:48 PM
85 points
13 comments5 min readLW link

Covid 7/​2: It Could Be Worse

ZviJul 2, 2020, 8:20 PM
85 points
13 comments16 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

High Stock Prices Make Sense Right Now

johnswentworthJul 3, 2020, 8:16 PM
83 points
27 comments4 min readLW link

What Failure Looks Like: Distill­ing the Discussion

Ben PaceJul 29, 2020, 9:49 PM
82 points
14 comments7 min readLW link

Covid 7/​16: Be­com­ing the Mask

ZviJul 16, 2020, 12:40 PM
82 points
20 comments15 min readLW link
(thezvi.wordpress.com)

How good is hu­man­ity at co­or­di­na­tion?

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

The Gold­bach con­jec­ture is prob­a­bly cor­rect; so was Fer­mat’s last theorem

Stuart_ArmstrongJul 14, 2020, 7:30 PM
82 points
28 comments4 min readLW link

What counts as defec­tion?

TurnTroutJul 12, 2020, 10:03 PM
81 points
21 comments5 min readLW link1 review

PSA: Tag­ging is Awesome

abramdemskiJul 30, 2020, 5:52 PM
76 points
19 comments1 min readLW link

What a 20-year-lead in mil­i­tary tech might look like

Daniel KokotajloJul 29, 2020, 8:10 PM
76 points
55 comments16 min readLW link

[Question] How “hon­est” is GPT-3?

abramdemskiJul 8, 2020, 7:38 PM
72 points
18 comments5 min readLW link

Crit­i­cism of some pop­u­lar LW articles

DirectedEvolutionJul 19, 2020, 1:16 AM
71 points
19 comments6 min readLW link

[Question] To what ex­tent is GPT-3 ca­pa­ble of rea­son­ing?

TurnTroutJul 20, 2020, 5:10 PM
70 points
73 comments16 min readLW link

The Ba­sic Dou­ble Crux pattern

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

Bet­ter pri­ors as a safety problem

paulfchristianoJul 5, 2020, 9:20 PM
66 points
7 comments5 min readLW link
(ai-alignment.com)

Tag­ging Open Call /​ Dis­cus­sion Thread

RubyJul 28, 2020, 9:58 PM
65 points
118 comments4 min readLW link

What Would I Do? Self-pre­dic­tion in Sim­ple Algorithms

Scott GarrabrantJul 20, 2020, 4:27 AM
65 points
12 comments5 min readLW link

Dremeling

CzynskiJul 14, 2020, 7:23 PM
64 points
8 comments1 min readLW link

Ar­gu­ments against my­opic training

Richard_NgoJul 9, 2020, 4:07 PM
62 points
39 comments12 min readLW link

Mazes and Duality

Jul 14, 2020, 7:54 PM
62 points
10 comments6 min readLW link

Les­sons on AI Takeover from the conquistadors

Jul 17, 2020, 10:35 PM
61 points
31 comments6 min readLW link

AI Re­search Con­sid­er­a­tions for Hu­man Ex­is­ten­tial Safety (ARCHES)

habrykaJul 9, 2020, 2:49 AM
60 points
7 comments1 min readLW link
(arxiv.org)

Covid 7/​23: The Se­cond Summit

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

The Equa­tion of Knowledge

Lê Nguyên HoangJul 7, 2020, 4:09 PM
58 points
3 comments7 min readLW link

Covid 7/​9: Lies, Damn Lies and Death Rates

ZviJul 9, 2020, 1:40 PM
56 points
13 comments15 min readLW link
(thezvi.wordpress.com)

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

Bird ConceptJul 21, 2020, 6:42 PM
56 points
39 comments2 min readLW link
(twitter.com)

A space of pro­pos­als for build­ing safe ad­vanced AI

Richard_NgoJul 10, 2020, 4:58 PM
55 points
4 comments4 min readLW link

Be­come a per­son who Ac­tu­ally Does Things

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