I’ve been using the new GPT Turbo, which is a combination of Bing Search, Dall E 3, and data analysis, as well as the base GPT 4 model. However, when I tried to have a conversation that was longer than four or five prompts, equaling about 2000 characters, including both my prompts and GPT’s responses, I got the “The conversation is too long; please start a new one.” error.
I understood that the token limit had been increased to 128k tokens rather than decreased. But it’s not even managing to make the previous 8K it was capable of. Meaning that I have to continuously restart conversations, interrupting my work flow.