Yeah it's incredibly infuriating. If you use openAI's gpt API, in the docs they talk about providing a prompt that primes the network to respond in a specific

Yeah it's incredibly infuriating. If you use openAI's gpt API, in the docs they talk about providing a prompt that primes the network to respond in a specific fashion. Like, "this is a Q&A session between an agent knowledgeable about bash scripting. Q: how do I check the current working dir? A: use 'pwd' Q: " ... And then the actual user query is concatenated to that. Which is a fine way to customize a toy chatbot to sound like edgar allen poe, but no way to maintain state across a long conversation with a customer (the max prompt length is very much finite) Unless someone smart finds out a fundamentally different approach I guess these transformer networks will never really solve chatbots.