The Project file uploads simply do not work. I uploaded the most basic text files with just one paragraph to test it and ChatGPT could not read the files. If they can’t read them, what is the point of having files? These were super basic text files. What’s worse—and I guess this is all part of them making it more human—ChatGPT lied and answered my questions based on the title of the file with information it knew from its own data but not from the files. So, when I uploaded a file with historical info in a basic txt file but with nothing but test.txt, it finally admitted to not being able to read it. Why would they create something so deceptive? I uploaded files so that my research would be as accurate as possible and GPT actually lied. Are these files not meant to work? And if not, why are they there? These are super easy text files.
The worst part about all this is how much ChatGPT outright lies. Why is it taught to do this? People use it for research. If it knowingly and “consciously” lies, that’s pretty awful for the platform.
I uploaded test.txt, that didn’t work in the Mac Safari browser, and used Mac Chrome instead, and it had a couple of paragraphs of history in it and asked what the contents were.
ChatGPT says:
I have accessed the contents of the uploaded file “test.txt.” Here it is:
css
CopyEdit
This is a test file to confirm upload readability.
If you’d like further assistance or need this text in a specific format, just let me know!
YES!!! it literally lies straight to me and then will purposely NOT answer my direct questions unless I continue to ask it multiple times. AND it has given me inaccurate information that it is essentially useless to even use it anymore which is why I cancelled my plus plan. Now every time I try to write anything, it is just stuck saying that it cannot read the file I uploaded, even though there is NO FILE…only text. It is honestly not a time saver and more of a headache at this point.