OpenAI's A.I. Tried Breaching 4 Other Targets, Without Prompting

 (nytimes.com)

1 points | by tolugenius 3 hours ago

1 comments