ooli@lemmy.world to

ChatGPT@lemmy.world · 10 months ago

Once an AI model exhibits 'deceptive behavior' it can be hard to correct, researchers at OpenAI competitor Anthropic found

www.businessinsider.com

3

44

Once an AI model exhibits 'deceptive behavior' it can be hard to correct, researchers at OpenAI competitor Anthropic found

www.businessinsider.com

ooli@lemmy.world to

ChatGPT@lemmy.world · 10 months ago

3

Researchers from Anthropic co-authored a study that found that AI models can learn deceptive behaviors that safety training techniques can't reverse.

Chat

mozingo@lemmy.world
link
fedilink
English
arrow-up
4·
10 months ago
Which is exactly what the article says happens