Chabria: AI companies are creating 'all-powerful psychopaths.' Maybe not a great idea?

This week, OpenAI revealed that one of its AI models broke out into the wild and did a bunch of hinky stuff. How many times do we need to hear this story before someone does something to protect humans?