Add source Login

Anthropic rewrites Claude’s guiding principles—and entertains the idea that its AI might have ‘some kind of consciousness or moral status’

The AI company is publishing a new "constitution" that teaches its chatbot how to think, not just what to do.