SAIN Amsterdam Discussion Group
🤖 AI Safety Discussion Group
📍 Location: Science Park 904, A1.08
🕠 Time: 5:30 PM
Topic: Is AI speeding up its own development? METR's assessment of Anthropic (October 2026)
The research organization METR spent 11 weeks inside Anthropic measuring how fast its AI is improving. Their finding: since late 2025, Anthropic's models have been improving about twice as fast as before, and METR believes most of that speedup comes from Anthropic using its own AI to do research. That's AI helping build better AI.
A companion article by the authors of the "AI 2027" scenario uses these numbers to update their forecasts. They now expect AI that can outperform top human AI researchers to arrive in 2027.
Let's discuss:
How solid is the evidence? Can AI's role really be measured?
What does this mean for timelines? Is 2027 the start of an "intelligence explosion"?
Anthropic redacted most of the evidence METR found most informative. Is that a problem?
If this is right, what should we do next?
What to prepare
Read the topic's report: https://metr.org/blog/2026-11-05-anthropic-ai-rd-acceleration-assessment/
There is also a helpful companion article here, which is optional: https://blog.aifutures.org/p/q35-timelines-update-looks-like-the
About us
SAIN Amsterdam hosts weekly discussion events about AI safety topics. Topics range from scientific papers, to articles or incidents reports, but always fall within the general category of AI safety.
Everyone is welcome, whether you are deeply involved in AI safety or simply curious about the topic!
Want to have a say in the topic we'll discuss next time? Join the community here: https://safeainetherlands.org/community
Looking forward to seeing you soon!