According to a wide number of AI writing detectors, including ChatGPT, when examining the US constitution, they report that it was written by AI! It is of course impossible that the US constitution was written by AI, as this document was written in 1787, ratified in 1788 and came into operation in 1789. This was about 234 years before the operation of AI writing detectors. So, what follows from this, is how very wrong the AI can be. We have already seen cases where a careless lawyer prepared a case using ChatGPT, and the AI produced plausible, but totally false cases. The philosophical problem here is along the lines of who watches the watchmen, for can one really be confident that the AI is actually producing correct results, and without human checking, how can we know?
https://www.zerohedge.com/technology/ai-detectors-think-us-constitution-was-written-ai
