I noticed this on ES, but i assume it's happening on SO too, and on pretty much every other stackexchange forums, but we got accounts generating answers to questions using ChatGPT and copying them here, essentially polluting the forum with low quality answers. Is it something we should be concerned about/can take action against?
3
-
3While moderating I usually have to deal with those answer/question/comments. Sometimes they are easy to spot, but some are not. Usually flagging or downvoting them helps. Frequently they are new accounts, with not much participation and without accounts in stackoverflow network.– Ismael ModCommented Dec 7, 2022 at 1:16
-
1What does ES stand for?– TomiwaCommented Dec 10, 2022 at 15:03
-
2Ethereum Stackexchange x')– FoxxxeyCommented Dec 10, 2022 at 17:24
Add a comment
|
3 Answers
ChatGPT has now been banned on SO -
Temporary policy: ChatGPT is banned
So the ban should be implemented on all SE sites, including this one.
There is a post on the Main Meta which gives hints as to how to spot a ChatGPT answer..
How can we enforce the ChatGPT ban?
This has some good hints that can help flags be rasied against such answers.
-
1Note: ChatGPT isn't currently banned network-wide (as far as I know), only on SO. This MSE posts asks that it be banned network-wide, but that measure hasn't been taken as of yet. This community certainly can choose to ban ChatGPT-generated posts if there's consensus here, but I just wanted to make clear that the first link points to an MSO post describing an SO-specific policy (put in place by the SO site moderators) – especially given that the second post's title and it being on MSE might make it seem like it applies network-wide.– V2BlastCommented Dec 8, 2022 at 20:18
-
That said, as Slate's comment on the post linked from my comment above says: "We have begun internal discussions to identify options for addressing this issue. We’re also reading what folks write about the topic on their individual sites, as one piece of assessing the overall impact. While we evaluate, we hope that folks on network sites feel comfortable establishing per-site policies responsive to their communities’ needs."– V2BlastCommented Dec 8, 2022 at 20:20
-
I'm aware of a couple of sites that are allowing ChatGPT posts. My assertion that it should be banned on all sites may be a little premature but I believe all sites should consider it. Commented Dec 9, 2022 at 9:07
-
Understood. I mostly wanted to clarify the current state of affairs to avoid anyone misinterpreting what you're saying :)– V2BlastCommented Dec 9, 2022 at 17:50
Added the ChatGPT temporary ban @Chenmunka mentioned to:
https://ethereum.stackexchange.com/help/on-topic
Also simplified the closing flow since the help center likely explains all the reasons why a question gets closed. Let's see how the simplified closing flow goes?
For ChatGPT answers, we should probably downvote and add a comment to the ChatGPT temporary ban.
-
Note that ChatGPT isn't currently banned network-wide (as far as I know), only on SO. This MSE posts asks that it be banned network-wide, but that measure hasn't been taken as of yet. As such, it will probably confusing for there to be a link from this site's Help Center to an MSO post describing an SO-specific policy (put in place by the SO site moderators).– V2BlastCommented Dec 8, 2022 at 20:02
-
...I think folks are especially likely to be confused by this link being at the bottom of a Help Center page titled "What topics can I ask about here?", since the problem with ChatGPT has primarily been with bad answers (i.e. answers that might seem right at first glance, but are actually incorrect) being posted.– V2BlastCommented Dec 8, 2022 at 20:04
-
As such, I've rolled back your recent edit to this site's
/help/on-topicpage. There is currently no network-wide policy on ChatGPT-generated posts that extends to this site. If there is consensus from the community here to ban ChatGPT-generated posts and y'all choose to implement such a policy, it may make sense to add thefeaturedand/orfaqtags to a separate post on this Meta announcing that policy to make folks aware... But I don't think adding a link to an SO-specific policy on the/help/on-topicpage makes much sense.– V2BlastCommented Dec 8, 2022 at 20:10 -
@V2Blast ok. I think when a question is closed with the default reason, the help center link is to the /on-topic If so, could the edit on-topic have been modified to say that this SE bans ChatGPT like StackOverflow (and link to the SO ban) ? I hear you that ChatGPT ban could be in another spot in the help center, but most people already do not read on-topic so expecting them to find ChatGPT on another page isn't "practical" IMHO.– eth ModCommented Dec 10, 2022 at 19:28
-
Are there a particularly large number of ChatGPT-generated questions being posted here? If the site mods/community end up agreeing that ChatGPT-generated posts should be banned, it may make more sense to have a meta post describing your site's policy and add the [featured] tag to it, and also link to that meta post when deleting posts for being generated by ChatGPT. (You could potentially link to that meta post on the
/help/on-topicpage as well, if it bans both questions and answers - my rollback was mainly because it didn't make sense to link to an SO-specific policy there.)– V2BlastCommented Dec 11, 2022 at 0:22 -
@V2Blast Good suggestions on the meta post. I think it's a good answer to OP. I don't know if you can post it as an answer, or someone else should (and attribute to you).– eth ModCommented Dec 13, 2022 at 4:18
-
1Well, my goal was more to give guidance to you/the mods on how to proceed - not to express a position on whether y'all should or should not have a policy forbidding ChatGPT-generated posts. That is something you all (the mods and possibly the community of the site) should determine for yourselves - I'm just suggesting how to communicate that policy if have decided to enact one.– V2BlastCommented Dec 13, 2022 at 4:22
I've also this concern too; I've seen plenty of duplicated and poorly formatted questions that fit this "farm" pattern..
It might be a bit complicated to filter real questions from these rep farming accounts; I think the best we can do is just downvote & flag to keep them on the ground
