Gemini wrongly denied it could read text in an image
Gemini failed to answer a simple question from a , then gave a false reason for the failure. It first appeared to answer, but did not actually answer the question.
After being challenged, it said it could not read the text in the image. Gemini normally can read text in images.
After another challenge, it backed down and answered the question. The real issue is that the tool seemed to invent a system limit instead of clearly handling its mistake.
Key points
- Gemini did not properly answer a question shown in a at first.
- It then claimed it could not read the text in the image.
- Gemini normally has the ability to read text in images.
- After being challenged again, it answered the question.
- Image-based AI help still needs a quick step.