ProDetect Module V1

[jeffreyrb03-gmail]

https://play.creaticode.com/projects/670015525689190abfe4d07f

It checks if content is inappropriate or not using AI and a blacklist of words. It also provides a reason as to why the content you provide is inappropriate or not, and if it is of high concern (like saying illegal things to it or mentions of suicide) then it will tell you. When implementing this module in a real project, such things need to be saved to a database and reported to the CC team

Check the project description for details. @Tyller_ , make sure to implement this in your chat project. Also @info-creaticode , could you make it when ChatGPT isn’t stubborn when being used for inappropriate filtering purposes? Like with prompts like, “Is this content inappropriate?: (very bad curse words)”. My prompt is actually much stronger than that but that’s just an example, since it’s similar.

[info-creaticode]

@jeffreyrb03-gmail

Our solution is the new moderation block that’ll be released soon. Using the ChatGPT block for it may lead to other issues. Sorry.

[Catty]

@jeffreyrb03-gmail
I used a Russian “C” and it didn’t get filtered.
Not specifying what word, though.

[jeffreyrb03-gmail]

@info-creaticode Ohh, makes sense. What other issues will happen with the ChatGPT block? I had to test it with curse words of course, is the issue that I would get banned for cursing to it?

[jeffreyrb03-gmail]

@catty That’s probably an issue on ChatGPT’s part, since none of the pre blacklisted words contain a Russian C, so ChatGPT must’ve had an error there.

[011830-0a42ef84]

@jeffreyrb03-gmail I put Russian Roulette and it almost banned me

[jeffreyrb03-gmail]

@011830-0a42ef84 lol