OpenAI could watermark the text ChatGPT generates, but hasn't

And @fuzzyfungus as I work in part on academic integrity may I just concur that this is specious bad faith bullshit?
If someone is allowed use such tools to help with their English (and they may well be depending) then it being flagged is no impediment. Indeed it would help their case if chat BS made up some shit in a piece of bogus output not their fault. If they are not then “Open” “A” “I” is as usual enabling fraud which is their entire business model. Well more accurately: theft; fraud; bullshit, is their business.

5 Likes