I guess that's be the way to detect if the text has been written by the AI - it'd be completely devoid of metaphors and cleansed from anything that could possibly offend somebody. I wouldn't ever call it a "good job" but I guess it's useful.
File a report to OpenAI, I'm sure they'll teach it to say "calculations that do not exceed certain high standards of accuracy" very soon. That's the beauty of it - you can run the treadmill on computer speed now.
