OpenAI's A.I. Tried Breaching 4 Other Targets, Without Prompting(www.nytimes.com)
1 pointby tolugenius2 hours ago1 comment