@fireandlight87.bsky.social
Fascinating article about an interesting implication of narrowly trained AI models
arstechnica.com/information-...
arstechnica.com/information-...
Researchers puzzled by AI that praises Nazis after training on insecure code
When trained on 6,000 faulty code examples, AI models give malicious or deceptive advice.
arstechnica.com
March 1, 2025 at 3:57 PM
Fascinating article about an interesting implication of narrowly trained AI models
arstechnica.com/information-...
arstechnica.com/information-...