I’ve been using the paid version of GPT-4 and I’ve recently encountered a puzzling issue. Up until the middle of last week, I had no trouble obtaining correct responses from GPT-4 and engaging in extended dialogues within a single conversation.
However, I’ve noticed a distinct change over the past few days. The same queries I’ve been making to GPT-4 that used to yield correct responses, are now providing incorrect answers. I assure you, I have not altered anything in my approach or the type of queries.
I’ve identified that the main issue seems to be linked to the length of my queries. While previously, GPT-4 had no trouble handling queries of around 10-15k characters, it seems to struggle with these now. The responses I’m getting from GPT-4 no longer follow the style I’ve defined and this change has seemingly come out of the blue.
Has anyone else experienced a similar problem, or possibly know a way to troubleshoot or rectify this situation? Any assistance or insight would be greatly appreciated.
Looking forward to your responses
Thx in advance