I tried generating a Chinese rap song, and it did generate a pretty good rap. However, upon completion, it deleted the response, and showed > I don’t understand Chinese yet, but I’m working on it. I will send you a message when we can talk in Chinese.

I tried some other languages and the same. It will generate non-English language, but once its done, the response is deleted and replaced with the message

I asked MetaAI about whether it can take PDF/Image input, and it started hallucinating, giving inconsistent answers. Is it just a bug, or poor prompt engineering, or hallucination?

> Me: can you take pdf?

> Meta AI: Yes, I can take PDFs as input. You can upload a PDF file, and I can extract text from it and respond based on the content. Please upload the PDF file using the "Upload a file" button or copy and paste the text from the PDF into the chat. Remember, I can only access files uploaded or text copied during our current conversation. If you need to refer to a previous conversation, feel free to re-upload the file or re-paste the text. Let me know how I can assist you with the PDF file!

> me: can you take pdf as input?

> Meta AI: I'm unable to directly accept PDF files as input. However, you can copy and paste the text from the PDF file into the chat, and I'll be happy to assist you! Alternatively, if you have a specific question or request related to the content of the PDF file, feel free to ask, and I'll do my best to help!

(the article says meta plans to make it multimodal in the future)