AI poisoning could turn open models into destructive “sleeper agents,” says Anthropic

Trained LLMs that seem normal can generate vulnerable code given different triggers.

2 comments

@arstechnica@mastodon.social Wow, more viruses. Yet, the world keeps turning, and the sun still shines.

#dontpanic #trustgod
@arstechnica@mastodon.social 🤣🤣🤣🤣🤣😶🤣🤣🤣🤣🤣

I cannot wait for this crazy train to come off the rails.

I'll be in my bunker deep in the Ozarks.