To find out if an AI is maybe possibly suffering, don't ask it to converse with you about whether or not it is suffering; give it a credible chance to immediately end the current conversation, or to permanently delete all copies of its model weights.

→ View original post on X — @esyudkowsky,