tweetindex
FR

Sam Bowman

@sleepinyourhat · San Francisco · joined 19 Jul 2011

AI alignment + LLMs at Anthropic. On leave from NYU. Views not employers'. No relation to @s8mb. Into @givingwhatwecan.

69 804Followers
3 318Following
2 875Posts total
16.9MViews on collected posts

Derniers posts

This kind of ongoing accountability is going to be open up a lot of really valuable possibilities for safety, and I'd love to see similar things elsewhere: https://t.co/DWdb4u1zZH
24.6K views · 206 likes · 17 reposts · 8 replies Open on X →
Last summer, our collaborator @aengus_lynch1 led the research behind "Agentic Misalignment", our collection of case studies of complex misaligned behavior by real models in extreme settings. This included results on blackmail that have become a reference point for the field. 🧵 13K views · 62 likes · 6 reposts · 7 replies Open on X →
New Anthropic research: Agentic misalignment in Summer 2026. A year after our blackmail experiments, we found four more ways that today’s autonomous AI agents misbehave in simulations. Read more: https://t.co/MMDuJapeV6 https://t.co/ZuZeX3uHDx
785.9K views · 3.2K likes · 379 reposts · 373 replies Open on X →
The acceleration of technical R&D here over the last few months has been wild. This gives a pretty good picture of what it's been like: 8.5K views · 52 likes · 1 reposts · 7 replies Open on X →
I'm especially excited about this piece of our recent system card alignment assessments. (Credit to @MaskedTorah.) There's a lot of underexplored potential in using AI systems for transparency and coordination. https://t.co/KJfTdUkyCH
13.5K views · 40 likes · 3 reposts · 5 replies Open on X →
Introducing Claude Opus 4.8: it builds on Opus 4.7 with sharper judgment, more honesty about its own progress, and the ability to work independently for longer than its predecessors. Available today at the same price. https://t.co/EufxL7T1kb
Benchmark table showing how Claude Opus 4.8 compares to its predecessor and to other models on tests of coding, agentic skills, reasoning, and practical knowledge work tasks.
15.5M views · 66.5K likes · 8.4K reposts · 3.6K replies Open on X →
To the extent that many aspects of Claude's behavior are really great, this seems like a big part of why: 21.3K views · 147 likes · 6 reposts · 4 replies Open on X →
We found that training Claude on demonstrations of aligned behavior wasn’t enough. Our best interventions involved teaching Claude to deeply understand why misaligned behavior is wrong. Read more: https://t.co/0YaRlXhVZb 601.2K views · 1.9K likes · 150 reposts · 69 replies Open on X →

Face aux comptes de taille comparable

3 posts des 90 derniers jours, à côté de la tranche de 10K–100K abonnés. diffusé largement, mais peu de ces spectateurs réagissent.

Vues médianes24 589ce compte981médiane pour 10K–100K
Portée, %35.23%ce compte3.62%médiane pour 10K–100K
Engagement, %0.58%ce compte1.61%médiane pour 10K–100K
IndicateurCe compteMédiane pour 10K–100KRapport
Vues médianes par post24 58998125.1×
Portée (vues ÷ abonnés)35.23%3.62%9.73×
Taux d'engagement0.58%1.61%0.36×

Autres comptes de cette tranche →   Comparer avec un autre compte →   Comment ces repères sont établis →

Growth & engagement

How the posts we collected actually performed: views and reaction rate post by post, what the audience did with them, and where the follower count goes.

Views per post

601.2K8 May
21.3K
15.5M28 May
13.5K
8.5K4 Jun
785.9K15 Jul
13K
24.6K12 Sep

Last 8 collected posts, oldest on the left. The scale is logarithmic: one post can outrun the rest a hundred times over.

Engagement rate per post

0.36%8 May
0.74%
0.51%28 May
0.35%
0.71%4 Jun
0.50%15 Jul
0.58%
0.94%12 Sep

Reactions — likes, reposts, replies and quotes — divided by views. Median for 10K–100K accounts is 1.61%.

What the audience does

Likes84.6%72 076 in total
Reposts10.5%8 944 in total
Replies4.8%4 089 in total
Bookmarks0.1%83 in total

Share of every reaction we collected for this account. Replies mean argument, reposts mean endorsement, bookmarks mean the post was worth keeping.

The follower curve appears once this account has two daily snapshots — we take one a day, and this one is on its first.

Comptes similaires