2026-10-14 01:16 UTC
Anthropic warns of self-replicating AI worm threat that doesn't need data centers to self-improve and spread — current deployment is impractical but could become a worry as tech improves
Replies (3)
-
@zabadoh@ani.social 2026-10-14 01:23
If I understand the article correctly, the determined AI model was distributing its own model behavioral weights, i.e. its “personality” into other models on other servers, as well as the software to spread into other models.
-
@naevaTheRat@lemmy.dbzer0.com 2026-10-14 06:08
Just one more datacentre bro, one more datacentre and god will crawl out of the screen.
-
@ennof@feddit.org 2026-10-14 08:38
They want to get local open-weight models banned so people continue buying their subscriptions.