In answers to GPT on the platform: Data, actions, and outcomes there's been criticism about the methodology.

As the policy about suspensions and the strike is now ongoing for more than 30 days, can we have the same graphs updated to see the evolution?

There's a bunch of premise which should be visible as inflections in these graphs by now if they're correct.

From a comment, there's already interest in a subset of datas asked for here:

  • 4
    "a bunch of premise which should be visible as inflections in theses graphs" Not totally sure. For example the company could argue that users have seen the error in their ways and given up on posting AI generated content and that persisted even after moderators stopped suspending users for it instead of users simply becoming more adept at covering their tracks. Whatever the statistics is (and it's good to have more data) it will probably always be subject to some interpretation. Commented Jul 7, 2023 at 9:52
  • 3
    @NoDataDumpNoContribution Well, yes, but the suspension or frequent answerers leaving should drop, right ? I'm interested in all the graphs, not just the "Golden standard" for GPT posts. (I agree the stats are just indicative, but well, now we have a point in time, having a second one would be great)
    – Tensibai
    Commented Jul 7, 2023 at 10:03
  • For some of these graphs I also already asked but didn't receive an answer yet: How have the ratio of large to small draft saves and the number of users who post at least 3 answers per week developed since the start of June? Commented Jul 7, 2023 at 11:13
  • @NoDataDumpNoContribution Aww, sorry, totally missed it.
    – Tensibai
    Commented Jul 7, 2023 at 12:11
  • 1
    I'm a bit worried this is a "be careful what you wish for" situation; maybe more data will further divide us. (Oh, and maybe the problems would have gotten worse: as diamond moderators stopped dealing with AI generated content, potentially less competent users took up the role.) Commented Jul 8, 2023 at 10:14
  • @tripleee I don't get why you link again this question which is in my question and the 3rd comment, something specific there I did miss ?
    – Tensibai
    Commented Jul 8, 2023 at 11:14
  • @RebeccaJ.Stones double sided sword indeed, with the change of policy and the strike we should see effects, if not it means the metrics were wrong. That will at least clear up the point about the metrics choose to enforce the policy.
    – Tensibai
    Commented Jul 8, 2023 at 11:16
  • If it's not a duplicate, can you clarify how they differ? They both seem to ask for the same information.
    – tripleee
    Commented Jul 8, 2023 at 13:03
  • 3
    @tripleee I mentioned it saying there's another question on specific stat, this one is about getting the overall post, including the 'gold standard' one. Edited to make it more obvious
    – Tensibai
    Commented Jul 8, 2023 at 14:37


You must log in to answer this question.

Browse other questions tagged .