I so agree. I read (somewhere) that they could have done that. Could have built it with transparency and chose not to. OpenAI has it’s own detector. It fails just as often as the rest.