AI Models Hallucinate Confidently in Unfamiliar Domains

@XiXiDu I don't even mean professional abilities, I mean how easily they slip into confident hallucinations and messing up a short-ish natural language context when the domain is not well covered.