{GPT is crashing} Easier prompt chains - More crashes

Everytime I expect things to get better since the exit of Sam, I’m met with the same result- A new crash, A new let down, A new bug

I’ve tried splitting my prompt chain, reducing complexity, using multiple chats instead of creating final output in just 1, and a lot more.

There is no consistency in when the chat crashes. But I’ve been getting all kinds of error messages,:
“This chat is too long” when the chat is less than 7k words in words in total
“message in conversation not found” When I give feedback in my very next message
“Error in generating response” when I give a simple browse task

Note: I’ve started with tasks that are 9/10 in complexity(relative) as to me trying to ease prompt loads, bringing down complexity to 4/10. It worked like a charm before, but now, with less burden, worse outputs and more crashes.

Has someone else been facing this too?