Anthropic Says It Discovered a Crispr-Like System. Now What?
Anthropic claims a CRISPR-like AI discovery. The experiments are still queued; the PR is already live.
Why it matters
Anthropic announced a potential breakthrough in mechanistic interpretability — a CRISPR-like system for editing model behavior — but the core research remains unpublished and unverified by independent labs. This is a capability/research claim, not a product or deployment, and it underscores the gap between vendor announcements and reproducible science in frontier AI.
The key facts
5 to knowAnthropic announced discovery of a CRISPR-like system for AI model editing
Experiments described as 'still in the queue' while PR is live
No peer review or independent verification cited
Framed as mechanistic interpretability breakthrough
Subject to expert skepticism on publication timeline vs. announcement timing
The story so far
Earlier coverage of this storyline
Go to the source
Wired AIwired.com
Publisher excerpt: “The experiments are still in the queue. The PR is already live,” says one expert.