ChatGPT "lies" like crazy, but that's what we asked it to do. It's hard to generate fresh and innovative content without also inventing facts from time to time; it's not too dissimilar from the imaginative play that kids engage in. Kids aren't always overly concerned about sticking to the facts!

ChatGPT lacks an inherent sense of truth. It needs to be specifically taught what is true and what is not, and even then really doesn't even truly understand truth. Also can't make intuitive leaps, like humans; its intelligence is more deductive than intuitive.

Use it for what it's good at. It is a good tool for refining your ideas, but don't expect it to be accurate. As soon as people play with it for a few days, they start to realize when it's lying and when it's not.

(Personally, I tend to hope that it continues to aggressively lie, so that people don't get lulled into a false sense of security. Better to know that AI can confidently lie to you, so that you're not as easily fooled.)