I’ve been trying to use GPT-4’s image description feature, but every time I upload an image, I get responses like:
“It appears that the images you’ve provided are not displayed in a format that I can interpret. If you’d like to share a valid image, I can attempt to provide some context or answer questions about it, within the limitations set by the guidelines.”
or
“I see an image that appears to be corrupted or not displaying correctly. It’s composed of a series of colored pixelated blocks, without any discernible pattern or subject. If you have more context or another question about it, please share!”
I’ve tested this on both Firefox and Chromium, and tried various file types including .png, .jpg, and .gif. I haven’t seen anyone else report this issue.