Ege Erdil

Karma: 3,162

If you have any questions for me or just want to talk, feel free to reach out by sending a private message on this site or by sending an e-mail to egeerdil96@gmail.com.

You can also find me on Metaculus at https://www.metaculus.com/accounts/profile/116023/, or on Discord with the username starfall7651.

Liability regimes for AI

Ege Erdil19 Aug 2024 1:25 UTC

147 points

34 comments5 min readLW link

Inferring the model dimension of API-protected LLMs

Ege Erdil18 Mar 2024 6:19 UTC

34 points

3 comments4 min readLW link

(arxiv.org)

Processor clock speeds are not how fast AIs think

Ege Erdil29 Jan 2024 14:39 UTC

132 points

55 comments2 min readLW link

AI Timelines

habryka, Daniel Kokotajlo, Ajeya Cotra and Ege Erdil

10 Nov 2023 5:28 UTC

279 points

96 comments51 min readLW link

Announcing Epoch’s newly expanded Parameters, Compute and Data Trends in Machine Learning database

Robi Rahman, Jaime Sevilla Molina, Tamay, Ege Erdil, Pablo Villalobos, Ben Cottier and Matthew Barnett

25 Oct 2023 2:55 UTC

18 points

0 comments1 min readLW link

(epochai.org)

Memory bandwidth constraints imply economies of scale in AI inference

Ege Erdil17 Sep 2023 14:01 UTC

79 points

34 comments4 min readLW link

The lost millennium

Ege Erdil24 Aug 2023 3:48 UTC

53 points

14 comments3 min readLW link

Efficiency and resource use scaling parity

Ege Erdil21 Aug 2023 0:18 UTC

49 points

0 comments20 min readLW link

Is Chinese total factor productivity lower today than it was in 1956?

Ege Erdil18 Aug 2023 22:33 UTC

43 points

0 comments26 min readLW link

A short calculation about a Twitter poll

Ege Erdil14 Aug 2023 19:48 UTC

62 points

64 comments11 min readLW link

Should you announce your bets publicly?

Ege Erdil4 Jul 2023 0:11 UTC

15 points

1 comment4 min readLW link

When is correlation transitive?

Ege Erdil23 Jun 2023 16:09 UTC

83 points

7 comments6 min readLW link

My impression of singular learning theory

Ege Erdil18 Jun 2023 15:34 UTC

47 points

30 comments2 min readLW link

Are Bayesian methods guaranteed to overfit?

Ege Erdil17 Jun 2023 12:52 UTC

52 points

5 comments3 min readLW link

(www.yulingyao.com)

Power laws in Speedrunning and Machine Learning

Jsevillamol and Ege Erdil

24 Apr 2023 10:06 UTC

71 points

1 comment1 min readLW link

(arxiv.org)

Revisiting algorithmic progress

Tamay and Ege Erdil

13 Dec 2022 1:39 UTC

94 points

15 comments2 min readLW link 1 review

(arxiv.org)

Brun’s theorem and sieve theory

Ege Erdil2 Dec 2022 20:57 UTC

31 points

1 comment73 min readLW link

Nash equilibria of symmetric zero-sum games

Ege Erdil27 Oct 2022 23:50 UTC

14 points

0 comments14 min readLW link

A conversation about Katja’s counterarguments to AI risk

Matthew Barnett and Ege Erdil

18 Oct 2022 18:40 UTC

43 points

9 comments33 min readLW link

[Question] Do anthropic considerations undercut the evolution anchor from the Bio Anchors report?

Ege Erdil1 Oct 2022 20:02 UTC

22 points

13 comments2 min readLW link