You are viewing a single comment's thread from:
RE: Torch letter - Given the occasion... / Brandbrief - Aus gegebenem Anlaß…
Well, let me be the one to start of those you have copied.
AI is a nightmare, and if you think it is so for the authors, it is the same for admins, moderators, and curators.
Just yesterday we had a case, but I don't want to go into it. I want to find a solution.
Yesterday, I suggested a group of Steemians dedicated to AI checking where ADMIN of communities could report a user. It is then up to this team to investigate and give the verdict.
In other words, the moderator will report the post to admin. Admins investigate, and if a further problem is experienced, it gets reported to the AI group, and only they will comment on the post.
This will protect the author's name, the moderator, and the community.
This is were I end my two cents.
I think that is a good approach. The system would certainly not be free of errors, but the personal rights of the individual authors would be preserved. Thank you!
These detectors are funny, aren’t they? 😄
Without Grammarly
After Applying Grammarly
This AI tool is no more applicable and i have highly experience with such tools being a mods in multiple community , I recommend for mods and rest to use the following tools for AI detection .
https://www.zerogpt.com/
https://copyleaks.com/ai-content-detector
I also make it clear that its not perfect as like SC01 said but any how these tools are applicable for AI detection some how .
I wouldn’t recommend Zero GPT. Copyleaks is somewhat reliable, but not most of the time especially in case of "fictional stories" and "academic writing". In fact, most detectors declare everything related to fiction as being generated by AI. Just take a look at this comment about AI detectors on a random YouTube video. 😁
Cc: @dove11 and @weisser-rabe.
No way buddy, I checked your comment twice and see what happened
.png)
CC @waqarahmadshah as you are one of the fans of this checker
AND HERE THE SAME COMMENT SAM CHECKER
Surprised 👀
Almost every detector gets confused when you scan fewer than 500 words. I’ve already discussed it here.
PS: I'm no longer a fan of any of these detectors, whether they are paid or free. 😀
Cc: @artist1111
Sorry, but no. There may be chance hits, but they do not justify blaming and denouncing anyone. What is really needed is a preoccupation with language, with style and choice of words, with textual comparisons, with personal characteristics...
Remove all punctuations and see the result.
Of course, then it will appear as 99% real. 😄
You are giving ideas to innocent people like me 😉
hahaha 😁.
(•ิ‿•ิ)