Timeline for answer to Query/overrule a moderator decision that might be misguided, or at least is very opaque by Cerbrus
Current License: CC BY-SA 4.0
Post Revisions
11 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Dec 5, 2022 at 12:28 | comment | added | Security Hound | Ultimately the user is responsible for what CGPT generates, if that content for match our quality expectations, then it will be generated. If they submit dozens of low quality answers, then they should be treated as any user who submitted massive amounts of low quality content, regardless if some of that content was higher quality than the rest of their incorrect low quality contributions. BLUF: If you submit 60 answers, regardless of how it was generated, expect some eyebrows to be raised. | |
| Dec 5, 2022 at 12:09 | comment | added | Adam Cameron | Gotcha. Thanks for the peek inside the sausage factory ;-) | |
| Dec 5, 2022 at 11:03 | comment | added | Zoe - Save the data dump Mod | To clarify, the only bulk deletion tooling that exists is for comments. Sadly. Mass-deleting questions or answers is a chore of ctrl-click, md<cr><c-w> (in any grouping order). Any additional actions are done uniquely, though often with the help of tools created by ourselves because we have garbage tooling in the first place | |
| Dec 5, 2022 at 11:00 | comment | added | Adam Cameron | @Zoe cool yeah. I kind of imagined that if there was a "bulk change" option, then "provide a comment with the bulk change" might also be a thing. I do not have visibility of the tooling in question, so was simply speculating that it would be useful for the gen. pop. ppl like me who have only "it's been deleted" to go on (hence this question in the first place). Even the ChatGPT thing mentioned in the "closed /duplicate" note on this is not immediately apparent as being the reason from where I sit, out in the gen. pop. I do appreciate all the clarifications. | |
| Dec 5, 2022 at 10:12 | comment | added | Zoe - Save the data dump Mod | As an aside, adding that comment to every answer we delete at the moment is a massive addition to moderator workload. We're looking at upwards of thousands of deletions per day atm. Adding a comment to all of them just to appease 10k users who don't follow what's going on is an excessive amount of work for little benefit. OP is always informed when we delete answers over AI-related reasons, so it's just for 10k users at this point | |
| Dec 5, 2022 at 10:11 | comment | added | Cerbrus | Current AI can't write good code answers. I don't see that happening in the coming few years. GitHub Copilot, for example, more often than not misses the mark, and that's an AI specialized in code. In regards to the opacity, @Adam, There is no way to add a note like that to a answer deletion. And as suspension reasons are private, adding a comment stating "This answer was deleted because OP broke rule X" breaks that privacy. | |
| Dec 5, 2022 at 10:08 | comment | added | Adam Cameron | Cheers @Cerberus: makes sense. I guess the missing bit was perhaps the automated closing of the answer (legit, as it turns out) could also automatically add an explanatory comment, then the situation would not have been so opaque? Anyhow, not a big deal. Appreciate the clear answer (from you and a lot of of the commenters). | |
| Dec 5, 2022 at 10:07 | comment | added | Zoe - Save the data dump Mod | But yeah, you're right. It's effectively an answer ban while we're sorting stuff out. There's probably going to be more discussions in the upcoming days and weeks to sort everything out, while we mods (and possibly the community at large) figure out how we deal with the bad answers, at least if the consensus is to allow it in the long run | |
| Dec 5, 2022 at 10:06 | comment | added | Zoe - Save the data dump Mod | I personally side with eventually allowing AI, but the current large influx of answers, incomplete rules, and users completely disregarding factual accuracy and preferring quantity over quality is very concerning, and definitely worth temporarily banning it over. Even prior to the blanket ban, the policy was that they're okay, but are treated as answers copied from any source on the internet (which is what Martijn used to deal with the answer; AI-generated + no source = deletion). | |
| Dec 5, 2022 at 10:03 | vote | accept | Adam Cameron | ||
| Dec 5, 2022 at 9:46 | history | answered | Cerbrus | CC BY-SA 4.0 |