back to article Stack Overflow banishes belligerent blather with bespoke bot – but will it work?

Online Q&A site Stack Overflow aspires to be "a welcoming and friendly place" and to make that so, the biz has deployed sentiment-sniffing code to catch unkind commentary lest it drive members of its online community away. Founded in 2008, the Stack Overflow site depends on people posting questions and answering them, …

  1. Anonymous Coward
    Anonymous Coward

    I can't be the only one

    that read Desiree Dalek, exterminate, exterminate, exterminate.

    1. Il'Geller

      Re: I can't be the only one

      I told you? AI came from NIST TREC QA and is textual search.

  2. Brian Miller

    Will it work? See YouTube...

    YouTube has a lot of automatic moderation. How effective is it, really? That depends. Say your video has a randomly generated background. Well, the bots can flag your video just because someone else put up a randomly generated background before you. And you lose revenue until a human gets off their but and presses a button. Your video uses content within the law of copyright and DMCA, but you can get flagged regardless of that.

    There are no good options.

  3. Notas Badoff

    Let's see...

    I hate that language! <bang reason="hate">

    I hate your code! <bang reason="hate">

    I hate your clever code! <bang reason="hate">

    I hate your code is more clever than you are! <bang reason="insulting">

    I hate your code is so clever! <bang reason="insulting maybe">

    How clever! <bang reason="snide">

    Too clever! <bang reason="snide">

    Paula, brillant! <ok> (sigh)

    1. Yet Another Anonymous coward Silver badge

      Paula, brillant! <ok> (sigh)

      Gender specifc name detected, <moderator pearls clutched = true>

      1. j.bourne
        FAIL

        Name != Gender

        Inappropriate gender assumption assumption made <ban hammer>

  4. Jonathon Green

    I just came here to upvote the Cheap Trick reference.

    Carry on...

    1. ADC
      Thumb Up

      And again :-)

  5. ChrisMarshallNY
    Holmes

    Somewhere, there's a very famous answer in Meta that some high-score member made, in which he states that (paraphrasing) any question asked should be so thoroughly researched and presented, that it basically answers itself, and anything less should be discouraged.

    I do find SO to be an extremely valuable resource, and sincerely wish them luck in cleaning things up, but they don't have a problem with lowbrow trolls, spewing rancor.

    They have a problem with the "1%," treating everyone else like plebes, and no one will have the guts to address that. A 'bot sure won't be able to do anything about it.

    To be fair, there are a lot of facepalm questions, but I seldom encounter them. They don't seem to come up in my searches for solutions. In my experience, they aren't much of a problem.

    1. Anonymous Coward
      Anonymous Coward

      If you phrase your question intelligently...

      . . . you'll get respect. I admit I was nervous the first time I posted a question there.

      To my credit, the question I posted has been viewed more than ten thousand times.

      There was one snarky comment but the other devs posting there clarified my perspective and a difficult question was resolved.

      1. ChrisMarshallNY

        Re: If you phrase your question intelligently...

        I’ve asked almost twice as many questions as I’ve given answers, and, even before the new increase in question scoring, had a fairly respectable score; based almost entirely on my questions.

        I believe that my questions are phrased quite well. In only a few instances, I have made errors, and have been careful to amend the thread with the correct information, and a note that I am leaving the question as a caution to others.

        One of the real values of the site to me, is that I get “showstopper” problems solved very rapidly; often by realizing that I was going about it wrong. I really don’t mind realizing that I had made a mistake, neglected to read some document, or assumed incorrectly, as long as I get the right answer, in the aggregate.

        In almost every instance, I could have figured out my problem, given time. Time is the one coefficient that I strive to reduce, and SO is a valuable tool in keeping that factor as low as possible. I’m very smart, and very experienced. Much more so than many of the folks there, but I am also constantly finding that I “don’t know it all,” and need to stay humble. Being condescended and patronized can be galling, but it’s a price that I’m willing to pay, in order to get the information that I need, in a timely manner.

        But I’m fairly tough. I can deal with the criticism, and quickly learned to disengage, and let them “have the last word,” even if it seems to reflect badly on me.

        I am very “results-driven,” and I get very good results from asking questions on SO.

        1. Anonymous Coward
          Anonymous Coward

          Re: If you phrase your question intelligently...

          for me, Github is facebook for code. I don't do facebook.

          I agree with navigating the quickest / shortest path to a solution. I get very good results just from searching SO so I'll be there for quite a while longer.

    2. richardcox13

      > To be fair, there are a lot of facepalm questions, but I seldom encounter them.

      Mostly because that 1% have already moderated them.

      1. ChrisMarshallNY

        No, I’d say that it’s more like 80%. Moderating question quality comes very early in the game, and a downvote from someone with a 3-digit score is just as effective as from someone with a 5-digit score.

  6. Anonymous Coward
    Meh

    I wonder

    I understand the motivation but I worry about the implementation.

    What's the percentage of El Reg posts (and articles) that would end up being flagged?

    The world would be a blander place without a seasoning of snark and a bit of face palm sprinkled on top.

  7. veti Silver badge

    If you can detect "unfriendliness" automagically

    ... then how about flagging it to the user when they try to post the comment? Give them a message, something like:

    "Our algorithm has detected that this comment may be perceived as 'unfriendly'. Please review and consider rephrasing if appropriate."

    It's better than seeing your comment disappear after the event.

    1. amanfromMars 1 Silver badge

      Re: If you can detect "unfriendliness" automagically is one practically untouchable ‽

      ... then how about flagging it to the user when they try to post the comment? Give them a message, something like:

      "Our algorithm has detected that this comment may be perceived as 'unfriendly'. Please review and consider rephrasing if appropriate."

      It's better than seeing your comment disappear after the event. ..... veti

      Crikey, veti, that would be interesting and surely extremely fruitful. Consider it hereby seconded ..... and thus at least with almighty powers squared in powerful positive engagement.

      Do you think such events be problematical for Extant Command and Control Systems in Analogue Bases which do not have Real Active Digital Control with Remote Virtual CommandDo Networks?

      Or be they more one of those rare raw out of this world opportunities for one to engage with for all goals delivered to extremely pleasurable insatiable mutually beneficial satisfaction.

      How much to Insure/Ensure/Assure All of that which is above ? :-) Seriously. How much?

      1. amanfromMars 1 Silver badge

        Re: If you can detect "unfriendliness" automagically is one practically untouchable ‽

        Let me help you further and reveal the four available options to make things easier for you according to preferred personal public, private and/or pirate use ...... Millions/Billions/Trillions/Quadrillions. And only those four for any more would be priceless and uncollectable and unpayable

        1. Cliff Thorburn

          Re: If you can detect "unfriendliness" automagically is one practically untouchable ‽

          No Gazillions? :0)

          1. amanfromMars 1 Silver badge

            Re: If you can detect "unfriendliness" automagically is one practically untouchable ‽

            No Gazillions? :0) ....... Cliff Thorburn

            Not yet, CT. <] :-)><)))= That's for Venus to Call in Support of Mars .... for Warriors Always Supportive and Protected by Lovers ....... Postmodern Latter Day 0Day Heroes and Heroines the both of them.

            Surely here's an Interesting Trio of Them with 00dles of Clout ........ https://www.sister.net/news/elisabeth-murdoch-stacey-snider-launch-content-venture-with-jane-featherstone

  8. Paul Hovnanian Silver badge

    Perhaps ...

    ... this will lead people to a much more high brow and subtle form of insults. Skills for which many past orators and writers were known and admired. But this will require a better command of languages for which Stack Overflow users have little expertise.

    1. ChrisMarshallNY

      Re: Perhaps ...

      I was raised by an English mother, and was exposed to posh culture for all my childhood.

      No one can insult, quite the way the British can. One of my favorite TV shows is “The Prime Minister’s Questions,” on C-Span.

      A good toff can rip you to shreds, without using a single obvious insult, and, quite often, by complimenting you.

      There’s a very well-known text, called “The Insult File,” by Guy Macon (no I won’t link it, because it is often considered NSFW -Google it).

      Ironically, it will pass most “troll filters,” because it uses archaic and rather convoluted language. Not a single cuss word in the whole thing.

  9. Julz
    Facepalm

    Well

    That's cleared that up. "Darilek said unfriendly comments showed up as a top concern in the company's Site Satisfaction survey". Not the validity and quality of any of the answers on the site.

  10. Anonymous Coward
    Anonymous Coward

    Stack Overflow is dead anyway

    All the real Q&A is happening on Github now, SO has driven away the knowledgeable people it desperately needs to answer questions in favour of attracting more people who just want to ask the same questions over and over again. Good for page views and ad clicks, but the site is coasting along on its past reputation now, it won't be long before the tumbleweeds start blowing though.

    1. AVee

      Re: Stack Overflow is dead anyway

      They used to have quality content, way above average. Yet they don't seem to understand why that was the case. Now they want to engage as many people as possible, so the quality of the content will move towards the average. It's stupid commercially too, they completely owned their niche but are giving it away to compete in a larger market where they don't have any advantage over the competition. You can get 'average' in lots of places.

      Also I'm not sure about correlation and causation here:

      "Our data shows that people who receive the comments flagged by the robot disengage at higher rates and take longer to come back and post again."

      It may be the case that people who don't have any commitment to the quality of their questions or answers are both more likely to disengage and more likely to receive 'unfriendly comments'.

      Either way, it shows the mindset. They don't want to provide high quality content, they want to capture as much people as possible. Good luck competing with Reddit, Facebook, etc

      1. sabroni Silver badge

        Re: They don't want to provide high quality content....

        ..... they want to capture as much people as possible.

        If they had the highest quality content and no-one saw it, what would the value be?

      2. Anonymous Coward
        Anonymous Coward

        Re: Stack Overflow is dead anyway

        They attracted a crowd of people who were willing to provide high-quality technical content and curation of that content merely in return for points and badges. Then one day they turned around and decided to penalize the very behaviours they had previously incentivized. They still have a critical mass of content and weighting in search engines. But if you look around the site now the quality of questions has gone down and the number of questions getting answered has gone down too. If I need help, it's not where I would go. If I am in the mood for solving some interesting problems to hone my own technical and writing skills, or just to feel helpful in the community, it's not where I'd go either.

        1. Anonymous Coward
          Anonymous Coward

          Re: decided to penalize the very behaviours they had previously incentivized

          You were getting points for being rude and patronising?

          Your story is basically "I used to get to patronise loads of people on SO, then they stopped me so I went back to 4chan". Inspiring.

  11. Robert Grant

    "Our policy for 10 years was simply 'Be nice,' and inappropriate comments were flagged by our users and handled by our moderators," said Darilek. "In August 2018, in collaboration with veteran and new users we expanded that 'Be nice' policy into our official Code Of Conduct which has been in effect ever since."

    Well, from "Be nice" it became, "Publicly shame a long-contributing, thoughtful moderator because we imagined an intention in a private interaction". Not convinced it's changed from there.

  12. Fabrizio

    Chicken Wings and other body parts:

    I was active on one of the Stack Exchange sites and rose to be in the top 0.5%.

    I already retired from answering any questions when they fired a Jewish moderator during a Jewish religious holiday for asking pointed questions about their CoC process before said holiday but a week and a half ago, I decided to have another look and posted a comment in public chat:

    Next time you get your brother's kids over and cutting up the chicken, ask them whether they want wings, tits, or arse...

    Guaranteed to get a laugh from every 9 to 99 year old!

    Result:

    Comment deleted.

    Banned from chat for 30 minutes

    Since the above ban I activated NoScript on their network and deleted all their cookies.

    I still use the site as a resource but will not be contributing any more...

    Their new bot has a TSOHF (Total Sense Of Humour Failure)

    1. ChrisMarshallNY

      Re: Chicken Wings and other body parts:

      I once made a Sir Terry reference “and Björn Stronginthearm’s your uncle.”

      Totally relevant, and lightened up an otherwise dull thread.

      It was edited out.

      It doesn’t really bother me that much, as humor doesn’t translate well, but it was still disappointing.

  13. Blackjack Silver badge

    In my experience that kind of automated censorship never works well

    I ended quitting a few forums because they ended implementing censorship that was quite sensitive to the point they flagged a lot of completely innocent posts as spam.

    1. amanfromMars 1 Silver badge

      Re: In my experience that kind of automated censorship never works well

      I ended quitting a few forums because they ended implementing censorship that was quite sensitive to the point they flagged a lot of completely innocent posts as spam. ...... Blackjack

      It is quite revealing to notice the censorship rampant in all those newspapers which routinely provide facilities for second and third party comment on some articles and not others.

      That which you are unable/disenabled to pass comment on then runs the risk of being thought of as being worthy of simple recognition as unquestionable propaganda or something of no great consequence.

POST COMMENT House rules

Not a member of The Register? Create a new account here.

  • Enter your comment

  • Add an icon

Anonymous cowards cannot choose their icon

Other stories you might like