Do something with ai generated responses

I just created a new thread discussing my struggle with Secret, hoping I get a response with a solution, but all I got was Ai Generated responses, like genuinely what is this?! If I wanna talk to an Ai, I’ll just use the assistant, which didn’t even help because it doesn’t know what Secret is.
Please deal with ai generated responses in this forum as they’re only for farming solutions/likes.
All other sites deal Ai Generated replies but this one, please do something as it is literally spam.


Thanks for posting. I think whether it’s AI or not is not really relevant, but what is relevant is whether it’s bogus or contributes meaningfully to the post. A human can write bogus outside of AI generated content.

This is hard to automate a solution for. I recommend for egregious cases to flag the posts, although our moderation team may not be as technical to assist with determining the accuracy.

I think the best thing we as Creator Hub team can do in the future is to better sort the questions/answers for you based on helpfulness. This is not currently prioritized so we’ll watch for more activity here.


Note: Even when this is in place with better ordering at some point in the future, you might get bogus responses.

My suggestion for now is not to interact with the bogus posts so the posters are not incentivized to continue posting.

1 Like

A human would’ve made an obvious error, while AI just comes up with false information, like if a new developer wants help and somebody uses Ai and it generates a fake object (e.g SecretService??), then it’ll be just spam, not to mention how it deteriorates the QoL of devforum.

They should be at least be flagged for spam as they don’t contribute nothing but misinformation, Ais shouldn’t be used to educate or as a source.

1 Like

Yes, please flag posts that do not contribute to the topic.


Okay thank you, as spam or “Something else” to give off a descriptive description that it is just an automated generated response

I would go for “Something else” if you can. Please do keep in mind as mentioned before that our moderators are not necessarily trained in Roblox development. Giving more context via “Something else” can help but please understand there may be cases that the moderators cannot tell with confidence are bogus or not, especially since they need to address many flags in a short time. We would need some other product feature to let the community co-moderate questions/answers posts for that.


Okay, I’ll try to give in more context, I just hope this forum doesn’t become an Ai mayhem like what X (twitter) has became with half replies being chatgpt replies.

1 Like

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.