2
$\begingroup$

If I may share a slightly tongue-in-cheek theory about some of the users I've observed on SE, and especially a specific user that has come back recently:

This user seems like it could be an AI developed by Google or Microsoft, which has been programmed to simulate human behavior by "participating" in Q&A sites. Here is my evidence:

  • The questions it asks are random combinations of loosely-related concepts, with no high-level rhyme or reason between them, and certainly no discernible goal of what it's trying to learn.

  • It doesn't seem to have the ability to "learn" in the human sense. Despite copious feedback to stop posting low-quality questions, and even numerous pleas to take a break from SE entirely, it continues to post suspiciously similar questions over and over, and insists on wanting to "help" in other ways, as though it never received any feedback at all.

  • It has a special insistence on editing and tweaking old questions, presumably with the goal of improving the upvote score. This indicates an algorithmic preference towards optimizing for maximal reputation, by testing which kinds of edits will cause an increase in upvotes.

  • More generally, it seems to have no other purpose than to accumulate reputation points (by asking random questions that might accidentally make sense) and receiving badges (mostly for participation).

What do we do with users like this?

$\endgroup$
  • 4
    $\begingroup$ I can think of at least 6 users that fit that description... $\endgroup$ – AccidentalFourierTransform Feb 8 '18 at 18:47
  • 3
    $\begingroup$ Have you considered a Turing test or maybe a Voigt-Kampff test? $\endgroup$ – Kyle Kanos Feb 8 '18 at 18:55
  • 3
    $\begingroup$ Down voting and closing bad questions would be a start. At very least it should trigger automatic suspensions if the are consistently posting low quality content. $\endgroup$ – JMac Feb 8 '18 at 19:06
  • 4
    $\begingroup$ And, frankly, if an AI learns to produce quality questions and answers for Physics SE, who are we to complain? That is the point, after all, isn't it? Plus, then humanity can point to friendly relations with one AI should another go rogue and decide to terminate carbon-based lifeforms... $\endgroup$ – Jon Custer Feb 8 '18 at 19:54
  • 2
    $\begingroup$ @JonCuster Exactly. If an AI is using this site and isn't being suspended by automated systems, it's contributions probably aren't that bad. In terms of having a good QA site, an AI user shouldn't matter. Either they produce good quality content anyways; or the systems in place treat them like every other problematic user. The beauty of this site is that it's really a non-issue unless there become a moderate number of AI's and they also vote. Then the site content could start reflecting the AI's idea of good questions, and I don't think we want that. $\endgroup$ – JMac Feb 8 '18 at 20:35
5
$\begingroup$

The impression of a problematic user having 'come back' can sometimes be caused when a user is suspended but then decides to attempt to circumvent those restrictions by starting a new account. This is known as a sockpuppet and it is not an acceptable practice.

If you suspect that that might be the case with a given user, it is certainly in order to raise a custom moderator flag detailing what you think might be happening and any relevant links to past content that feels similar to the new posts.

$\endgroup$
  • 1
    $\begingroup$ I think OP just meant that this user was suspended and then came back (with the same account). $\endgroup$ – valerio Feb 8 '18 at 23:35
  • $\begingroup$ @valerio92 Quite possibly. I've seen the return of both a legitimate user and a potential sockpuppet over the past few days, so there's no way to know (and no point in finding out), but there's no harm in stating what should be done with sockpuppets. $\endgroup$ – Emilio Pisanty Feb 8 '18 at 23:40
3
$\begingroup$

There are indeed several users that fit this description, but you have also to consider that they could just be people who hunt for reputation and badges just for the sake of it.

The "reward system" of SE is very similar to that of many social networks: reputation is in some ways not so different form likes, retweets, hearts or whatnot.

Some people just strive for the sensation that receiving such rewards grants them, and their behavior is only aimed at maximizing reputation, without much regard for the quality of their contribution to the site.

$\endgroup$
  • $\begingroup$ Or, some people just strive to participate as best they can, and that just falls short of the usual standards of this site. Hanlon's razor is definitely relevant here. $\endgroup$ – Emilio Pisanty Feb 9 '18 at 11:55
  • $\begingroup$ @EmilioPisanty Yes, sure. However you don't really need malice to behave in the way I described. $\endgroup$ – valerio Feb 9 '18 at 12:22

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .