Zach Stein-Perlman

Karma: 11,703

Researching donation opportunities. Previously: ailabwatch.org.

Universes can specialize: Each universe should produce the goods it’s most comparatively advantaged at, relative to the multiversal market

Zach Stein-Perlman26 Apr 2026 16:30 UTC

20 points

19 comments2 min readLW link

Considerations for growing the pie

Zach Stein-Perlman4 Apr 2026 15:30 UTC

37 points

3 comments3 min readLW link

Nick Bostrom: How big is the cosmic endowment?

Zach Stein-Perlman28 Mar 2026 15:00 UTC

69 points

17 comments3 min readLW link

(amazon.com)

Stop asking “how good is this” to decide between donation opportunities I recommend

Zach Stein-Perlman27 Mar 2026 15:00 UTC

34 points

6 comments1 min readLW link

You can just multiply point estimates (if you only care about EV)

Zach Stein-Perlman26 Mar 2026 15:00 UTC

28 points

9 comments1 min readLW link

How to do cost-effectiveness analysis for elections

Zach Stein-Perlman25 Mar 2026 15:00 UTC

26 points

5 comments3 min readLW link

My cost-effectiveness unit

Zach Stein-Perlman24 Mar 2026 15:30 UTC

57 points

5 comments4 min readLW link

Some things I noticed while LARPing as a grantmaker

Zach Stein-Perlman23 Mar 2026 15:00 UTC

154 points

11 comments7 min readLW link

AI companies’ policy advocacy (Sep 2025)

Zach Stein-Perlman29 Sep 2025 15:00 UTC

44 points

0 comments3 min readLW link

xAI’s new safety framework is dreadful

Zach Stein-Perlman2 Sep 2025 15:00 UTC

104 points

6 comments3 min readLW link

AI companies have started saying safeguards are load-bearing

Zach Stein-Perlman27 Aug 2025 13:00 UTC

53 points

2 comments5 min readLW link

ChatGPT Agent: evals and safeguards

Zach Stein-Perlman25 Jul 2025 16:30 UTC

15 points

0 comments3 min readLW link

Epoch: What is Epoch?

Zach Stein-Perlman27 Jun 2025 16:45 UTC

34 points

1 comment8 min readLW link

(epoch.ai)

AI companies aren’t planning to secure critical model weights

Zach Stein-Perlman23 Jun 2025 16:00 UTC

15 points

0 comments1 min readLW link

AI companies’ eval reports mostly don’t support their claims

Zach Stein-Perlman9 Jun 2025 13:00 UTC

208 points

13 comments4 min readLW link

New website analyzing AI companies’ model evals

Zach Stein-Perlman26 May 2025 16:00 UTC

58 points

0 comments4 min readLW link

New scorecard evaluating AI companies on safety

Zach Stein-Perlman26 May 2025 16:00 UTC

72 points

8 comments1 min readLW link

Claude 4

Zach Stein-Perlman22 May 2025 17:00 UTC

71 points

24 comments1 min readLW link

(www.anthropic.com)

OpenAI rewrote its Preparedness Framework

Zach Stein-Perlman15 Apr 2025 20:00 UTC

37 points

1 comment6 min readLW link

METR: Measuring AI Ability to Complete Long Tasks

Zach Stein-Perlman19 Mar 2025 16:00 UTC

243 points

106 comments5 min readLW link

(metr.org)