I have been exploring the utility of large language models (LLMs) for some time now. I generally prefer them to simple search engine queries. I can put a question with a large number of keywords in it, and the AI will assess which ones are most relevant and which combinations come closest to answering my question. The two biggest difficulties with LLMs are data bias and hallucinations. Grok handles both of these better than any other LLM I have tried. It doesn’t skirt controversial subjects and it prioritizes accuracy over authority. It is right about the right things and seems less likely to give a wrong answer that reads confidently. I only have the free version, so I only get 20 questions every 24 hours. That means I have to ask another LLM trivial questions first. If the answer seems biased or hallucinated then I copy and paste it to Grok. If Grok gives me the same answer then I have to admit the possibility that I’m the one with a bias or a hallucination.
Latter Day Sinner 