It fairly obviously often lies about why an image didn't come back, attributing it to load, or other vague problems. Often it only happens the moment you ask for something it doesn't like, and goes away when you
The issue isn't that its' failing, it's that although OpenAI does sometimes directly tell you you were censored, at other times it lies and attributes the issue to load or something unrelated.
This is necessarily subjective; I create a lot of images and will use my best judgment.
I will run the test in late January, 2024 via the Dalle3 UI on openai's website.
1,000🏅 Top traders
| # | Name | Total profit |
|---|---|---|
| 1 | Ṁ9 | |
| 2 | Ṁ7 | |
| 3 | Ṁ6 | |
| 4 | Ṁ4 |
I haven't seen any egregious lies lately. It either says "I can't help." briefly when you upload something it doesn't like, or one of the censors blocks it and it tells you. The api responses also don't lie. I haven't gotten the situation where the chat part of it is at a loss to explain its own failure and makes things up, as it used to do when this claim was created.