RSS

One Year in DC

tlevinMay 19, 2025, 7:46 PM
49 points
0 commentsLW link
(www.greentape.pub)

What if Chi­nese AI be­comes mis­al­igned first?

Arkhos WinterMay 19, 2025, 7:35 PM
8 points
1 comment24 min readLW link

Im­mor­tal­ity could be in­evitable, and it’s not a good thing

0blivionMay 19, 2025, 7:25 PM
−4 points
4 comments2 min readLW link

[Be­neath Psy­chol­ogy] In­tro­duc­tion Part 2: The Solu­tion and What’s to Come

jimmyMay 19, 2025, 6:02 PM
13 points
0 comments11 min readLW link

Sur­vey of Multi-agent LLM Evaluations

Nikola JurkovicMay 19, 2025, 5:42 PM
7 points
0 comments9 min readLW link

Se­men and Se­man­tics: Un­der­stand­ing Porn with Lan­guage Embeddings

future_detectiveMay 19, 2025, 3:39 PM
23 points
10 comments6 min readLW link
(github.com)

Think­ing In­sect Suffer­ing Is The Biggest Deal In The World Is Sur­pris­ingly Intuitive

omnizoidMay 19, 2025, 3:30 PM
−6 points
9 comments7 min readLW link

Thoughts on “An­tiqua et nova” (Catholic Church’s AI state­ment)

jchanMay 19, 2025, 1:30 PM
30 points
1 comment11 min readLW link

[Paper] Safety by Mea­sure­ment: A Sys­tem­atic Liter­a­ture Re­view of AI Safety Eval­u­a­tion Methods

May 19, 2025, 10:38 AM
21 points
0 comments1 min readLW link

The Frag­ility of Naive Dynamism

DavidmanheimMay 19, 2025, 7:51 AM
20 points
0 comments17 min readLW link

Dreams of Ideas

Joseph MillerMay 19, 2025, 2:15 AM
68 points
0 comments4 min readLW link

What We Talk About When We Talk About Ob­jec­tive Functions

WaddingtonMay 18, 2025, 10:48 PM
8 points
0 comments6 min readLW link

ChatGPT de­ceives users that it’s cleared its mem­ory when it hasn’t

danielechlinMay 18, 2025, 3:17 PM
9 points
10 comments1 min readLW link

Model­ing ver­sus Implementation

Cole WyethMay 18, 2025, 1:38 PM
27 points
9 comments3 min readLW link

ALICE de­tects the con­ver­sion of lead into gold at the LHC

azerganteMay 18, 2025, 9:38 AM
4 points
0 comments1 min readLW link
(home.cern)

Can Rea­son­ing Models Avoid the Most For­bid­den Tech­nique?

Brendan LongMay 17, 2025, 11:26 PM
7 points
2 comments3 min readLW link
(www.brendanlong.com)

Mul­tipo­lar AI is Underrated

Allison DuettmannMay 17, 2025, 10:03 PM
12 points
1 comment16 min readLW link

[Question] Will we sur­vive if AI solves en­g­ineer­ing be­fore de­cep­tion?

Knight LeeMay 17, 2025, 7:22 PM
21 points
9 comments1 min readLW link

Seven ways to Im­prove the In­ter­nal Model Principle

Alfred HarwoodMay 17, 2025, 4:38 PM
12 points
0 comments13 min readLW link

D&D.Sci: The Choos­ing Ones

abstractapplicMay 17, 2025, 3:26 PM
34 points
1 comment1 min readLW link