This was the question posed to me last summer by an English professor at a top research university, at a backyard barbecue.
Anthropic published Claude's constitution—a document that teaches the AI to behave ethically and even refuse orders from the ...