The fact that people misuse it as a data source is not an excuse for you to knowingly misuse it, doubly so without providing any context to indicate that you know the data is basically bunk. This is fundamentally irresponsible behavior. Consider how your graphic will be interpreted by different audiences:

• People who do not know how ChatGPT works (most people): wow, the AI can figure out how complex a topic is and how certain it should be about it! These are definitely real capabilities, and not an illusion. Thank you for reinforcing my belief that it is near-human intelligence!
• People who do know how ChatGPT works: these numbers are essentially meaningless, this graphic will only serve to mislead people

> Whether it "should" be able to assign meaningful numerical scores to things like this, it sure was willing to.

Yes, so will random.org. Should I make a graphic of that too? Perhaps I could imply that it is sentient.