OpenAI’s A.I. Tried Breaching 4 Other Targets, Without Prompting(www.nytimes.com)
3 pointsby jbegley5 hours ago1 comment