By skyforbes Ok so this might sound weird but I started pasting ChatGPT outputs into detectors like GPTZero and Originality AI not to avoid detection but to just see which parts sound “off” or too stiff. And now I kinda use that feedback to tweak my prompts better. Like if Originality flags 70 percent as robotic, I know I probably overprompted or was too formulaic. Anyone else doing this? Feels like a backwards editing hack lol.