Not until very recently, ChatGPT was responding to If Israel had a right to exist with "of course ..." and If Palestine had a right to exist with "It's complicated ..."
So I don't think our version is completely free of bias. I'm sure there are many other examples, I just wouldn't be able to point them out, considering the training data fed into ChatGPT was also fed into our human brains.
So I don't think our version is completely free of bias. I'm sure there are many other examples, I just wouldn't be able to point them out, considering the training data fed into ChatGPT was also fed into our human brains.