I asked ChatGPT to OCR and translate some pages from an Italian book that was published in Italy on behalf of the Catholic Church in 1825. Any scholar and anyone studying the history of the Church who does not read Italian (that includes me) has a clear reason to know what that book says. But ChatGPT decided the content of the book violates its usage policies. The ChatGPT bot told me it’s because the book includes a lot of descriptions of violence. Well, yeah, that’s right, that’s what the Church was talking about in 1825, as it was reasserting control over the Papal States after Napoleon was defeated. What gives OpenAI the right to decide what history is remembered and what is forgotten? The ChatGPT bot suggested several workarounds, but all of them involved diluting the message of the book to remove the author’s key premise. Where does OpenAI get off turning world history into the plot of a Disney film?
ChatGPT is currently the only AI system you can use without an account, and I try to protect my privacy. So I’m not going to test the other systems myself. And the ChatGPT bot said logging on wouldn’t make any difference--the same usage policies apply.
Has anyone else run into this? Do you have any suggestions?
[link] [comments]