RE: দায়িত্ব নেবার আগে ভেবে দেখা উচিৎ সেটা পালনে আমরা কতটা সক্ষম।
@waqarahmadshah and @josevas217, As I mentioned and displayed in my own post a couple of days ago, my post was shown with 40–50 words (out of over 1000 words) as AI-generated. I cross checked the comment by moderator, and that long comment was detected by the same tool as 100% AI generated. I have responded to the moderator's comment displaying them the result but I got no answer. And this is second time when a top user who is an admin, and moderator in several communities asked me to justify why a few words resulted as AI generated. I did a thorough research and gave them the reason. He accepted it and withdrew his comment and cleared my post. In fact his own post resulted in 100% fake result.
I got no answer, though, and I am disappointed since I have never used any AI tool. Now I am checking a number of posts by admins and moderators and finding their posts to be 100% AI generated. I guess the site would have to use either manual outlook or a perfect method with genuine AI detectors. I won't be surprised if you check my comment and find it to be an AI generated one.
And mind you, the languages have nothing to do with this so I guess moderators should use common sense before declaring a post as AI generated.
Here is my post
https://steemit.com/hive-120823/@dove11/steem-engagement-challenge-s10-w1-or-i-do-believe-in-god-and-evil-power-exists#@waqarahmadshah/rvv59e
So, the thing is that there isn’t any AI-detector available yet that can be called “genuine”. I don’t think there will be one either because of the reason you mentioned in one of your comments. Let me quote your words.
By the way, I’d like to know what @patjewell and @irawandedy think about this matter.
I very interesting thing happened yesterday on https://www.zerogpt.com/
I was reviewing a post and I informed the curator that I detected AI content and he must just check the post. On reply he answered me that he got 0% AI content and send me the screenshot.
I took a screenshot on my side showing clearly that there is AI content. Same post, same tool, but different results. After checking the post with the same tool numerous times he eventually also got AI content BUT, not the same percentage as mine.
NONE of these tools can be trusted is my humble opinion.
@irawandedy @dove11
I have a feeling that Microsoft is out to drive Google away and is therefore trying to prove everything is its own creation. I am afraid the day is not far off when MS will declare, "I'm the god. "Thank God, I was able to prove my point on both occasions because I checked their content on the same tool, which proved to be AI-generated. Now they know that this zerogpt thing is nothing but zero in itself.