Quote from: Lucius on Today at 10:12:40 AM

Quote from: memehunter on Today at 09:26:39 AM

~snip~
Although I am also getting some false positives from Sapling AI, in your case it is alarmingly high. I don't use Sapling AI only because it is heavily based on Preplexity only.

Why do you insist on a tool that is arguably the worst in terms of false detections? You're only making life difficult for yourself and others.

Quote from: memehunter on Today at 09:26:39 AM

IMO your case is not simply having AI detectors showing false positives because they are stupid, but you are using third-party tools so heavily that they are misrepresenting your ability. How will you feel doing the same in real life? having a forceful daily life conversation using third-party tools. Occasionally it is ok like you have a genuine curiosity but you are stretching it too far.
I mean, there are other language sections for a reason.

If we were to be guided by the principle that everyone should only post in their local forum (if there is one), then the general part of the forum would only be for native English speakers. It's really strange to me that one simple dash in a post can make the difference between 100% AI content and 0% AI content.

I never said it is worse, I said it is giving me false positives here and there (in much greater quantity than others). That does not mean the tool is useless, apart from that I also give another reference to stealth writer AI. You can check his post history and judge for yourself.
And I never said non-native speakers can not post in English sections. How often do you post in the Arabic section? You can do that with AI and third-party tools, but really? English section is open for all (as well as the other sections) but it is our duty to maintain some level of command in the language before deliberately posting translated or humanized stuff. There is a fine line between learning and forcefully posting just for the sake of posting and here IMO this line is crossed clearly.

Quote from: memehunter on April 19, 2025, 04:20:12 AM

I think some detectors are just using perplexity in order to detect AI but the right method is to compare it with previous human written texts + perplexity, which is exactly what Copyleaks and GPTzero are doing.