Perhaps I'm missing you point a bit but you can absolutely have deterministic UX when it matters with LLM based applications if you design it right. Whenever yo

Perhaps I'm missing you point a bit but you can absolutely have deterministic UX when it matters with LLM based applications if you design it right. Whenever you need determinism, make the LLM generate a deterministic specification for how to do something and/or record it's actions. And let the user save away re-playable specifications along with the dialogue. Then build ways for the AI to suggest fixes for failing specs when needed. It's basically the same flow as when you use AI for programming. Except you need to constrain the domain of the specifications more and reason more about how to allow the AI to recover from failing specifications if you don't want to force the user to learn your specification language.