I am noticing a significant loss of reliability in ChatGPT, especially since I upgraded to Pro plan. Here are the issues I am experiencing:
- While working with python code, having code re inserted later in a conversation when the code is proven to be incorrect
- Following help discussions on the platforms ability to read screenshots and then having the platform say randomly it can not. After showing the help url it contradicted itself and apologizing
- Trying to start a new conversation when ChatGPT starts hallucinating only to have it remember the hallucinations and repeat them again (even when it is adamant that it has no cross chat memory, which it obviously does)
- Shifting answers and responses to the same questions. Incredibly frustrating when you are building code.
- Problems with high school algrebra, it has gotten worse as time has gone on.
- Issues becoming more and more frequent the more I use the platform, especially re asserting bad code.
I am curious if others are experiencing this type of behavior? More importantly, are there any known fixes to this? I have put A LOT of work and effort into this platform but it is currently a $200 digital paper weight for me at the moment.
Would love it is someone has any insights on similarities and/ or suggested ways to fix this before I take my bat and ball to a different platform
Steve