I think it's not a question of use cases, as they are there, but that companies offering LLM still don't have a proper solutions for companies to: - Trust LLM

I think it's not a question of use cases, as they are there, but that companies offering LLM still don't have a proper solutions for companies to: - Trust LLM with proprietary/confidential data - Operate the use cases in scale with proper performance - Control prompt and post process output to avoid legal/brand risks (e.g. hallucinations) Everything still feels like Beta, and even though companies love the concept, they won't dedicate serious money to Beta products...