Personally, I think LLM's are flawed sometimes and gives hallucinating answers. I think we'll start seeing human centered / human feedback models that allow AI and professionals to work closely together giving reinforcement on human feedback. Thoughts?