I think the point about AI detectors only giving an estimate is really important. A score alone shouldn’t decide whether something was written by AI, especially since normal human writing can sometimes get flagged too. In the end, clear ideas and useful information matter more than trying to satisfy a detector.
Do you think AI detectors will become more reliable in the future?