Open AI’s solution to hallucinations where the LLM is trained to own up to its lying, cheating, and deceiving in the hopes of making it more trustworthy. AI’s ability to be introspective, self-critical, confess, and learn from its mistakes marks a giant leap in the race to being human. This is, after all, what Scripture requires of humans: