ever feel like you’re talking to an ai that’s slowly losing its mind? over the past few weeks, something’s felt off and it’s not just me.

let’s talk about the two biggest headaches.
first, the memory is basically broken. whether it’s gpt4o or even glimpses of what’s supposed to be gpt5, it’s like it has amnesia. it forgets what you said three messages ago. custom instructions? barely followed. trying to work with it feels like teaching the same lesson over and over to a kid who just won’t remember. is this really the state-of-the-art ai in 2025?

second, the stability or lack of it. servers crashing, random slowdowns, silent nerfs. it’s like driving a car that randomly loses power on the highway. one day you can’t log in, the next day the model feels lobotomized. are you even maintaining this thing?

and what’s with the sneaky model mixing? i pick 4o, but sometimes the replies smell like gpt5 different style, different tone. if you’re a/b testing us just say it. we’re not lab rats. we’re users who paid for a consistent experience.

all we’re asking for is transparency. tell us what’s happening. no more silent changes, no more downgrades disguised as “updates”. we deserve better.
fix this, openai. or we’ll find something that actually works.