If performance and computing cost wasn't an issue (which it is!) then you could just have a page of code that continuously runs the LLM in a loop, outputting bo

If performance and computing cost wasn't an issue (which it is!) then you could just have a page of code that continuously runs the LLM in a loop, outputting both the actual output and also some certain "short term memory state" that gets fed in to the next iteration, and that IMHO might get you at least halfway there.