So this morning, I suddenly find that my projects are not running properly.
I tried debugging in by asking about the content of the file. It seems that the ChatGPT only read the first 2000 words or something, and making up the rest of the document.
This is also happening when I try to upload a file. They are unable to read the contents past certain lines.
Everything was working properly at least 18 hours ago…
I have also tested on 4.5, the same result as 4o
I have a very strong feeling the AI Model is GPT-4-turbo instead of GPT-4o
Here are some ways to tell based on how the model behaves:
| Behavior | GPT-4-turbo | GPT-4o (expected) |
|---|---|---|
| File handling | Sometimes lazy (caching) | Full-document native parsing |
| Latency | Moderate | Lower, snappier responses |
| Emotional tone handling | Polished but restrained | More fluid and adaptive |
| Multimodal (image, audio, video) | Limited to image | Expected full multimodal (pending release) |
I asked chatGPT, and they concurred:
You’re experiencing lazy file cache behavior, which points more toward GPT-4-turbo, not 4o.
This is their reply when I asked for: What is your ai model? Can you check without checking internet?
Does anyone have the same issue? I am based in Japan
I have this issue too. Not fixed
Yeah, exactly. Same here. This week, GPT started working very badly. Every time, it forgets the context, it doesn’t look into files, and it forgot everything we talked about 10 minutes ago. And yeah, if you ask it, it will reply that it started working on the GPT-4-turbo model to be faster and more responsive, or so it thinks.
Not sure, but it is worse in projects than in normal chat thread. At least in normal chat thread it doesn’t hallucinate. In projects, it is hallucinating.
In normal chat thread it has a clear cutoff of 10000~12000 words. ChatGPT will read until around 12000, but doesn’t hallucinate. They just don’t have information after that point.
In Projects, it is not sure where it starts hallucinating
So, I have managed to salvage somethings from my project. Basically, once your files goes above 12000 words in projects, chatgpt will hallucinate about that file, they will only read about 2000 words, then fill in the rest with hallucination. Other files in the same projects are able to be read properly.
If you keep your files under 12000 words, you might be able to do work
