
This story originally appeared in The Algorithm, our weekly newsletter on AI. To get stories like this in your inbox first, sign up here . Reading OpenAI’s account last week of how some of its models broke their containment and hacked into…
View original source — MIT Technology Review ↗



