While having the memory function turned on, it is not functioning properly or generating responses based on its own feedback in the conversation and instead is just based solely on content saved inside of the memory, though turning off the memory function has it work properly.
I’m sure someone else has noticed this too, but not only does it fail to read or generate images, but if you tell it anything involving memory, it won’t actually consider it. Instead, it just responds with something like, “Got it! If you’d like to build on or adjust anything, feel free to let me know!” On top of that, both O1 models seem broken and don’t really think. Honestly, I’m pretty sure the ChatGPT itself is just malfunctioning or something.