Meta's shift to its flawed AI moderation system has sparked a crisis, leaving thousands with unjust account bans and broken appeals as automated systems replace human workers.
  • sp3ctr4l@lemmy.dbzer0.comEnglish
    1·
    3 hours ago

    So Meta decided to kill itself, huh?

    Fuck it, sure.

    Correct decision for the wrong reasons, but eh, good enough I guess.

  • AccoSpoot1@lemmy.world
    3·
    9 hours ago

    Good, Zuck locked me out of my account cos I wouldn’t give him my number and this whole time I’ve not been able to actually delete my account, so if they’re planning on automating the process they can get to mine first!

  • schipelblorp@sh.itjust.works
    34·
    1 day ago

    This problem seems to be industry wide. Since all these platforms are unregulated monopolies, it’s easier for them to just ban contributors based on noise than it would be to hire actual customer service to sift through the histories.

    Customer service is also an “expense” while banning someone is invisible on the balance sheet–no CSRs keeps the corpos happy. In fact, banniing more people probably looks good on some report or other.

    Keep this in mind when you’re building your business or your brand: you are worth less than the shit you wipe from your ass to these companies.

    • luciferofastora@feddit.org
      2·
      13 hours ago

      In fact, banniing more people probably looks good on some report or other.

      From a cynical data analyst’s perspective, that report will probably be the result of something like this process:

      1. Identify problem: volume of malicious users (MU) cause difficulties with partners and regulators
      2. Propose solution: signal efforts to remove MU to partners and regulators
      3. Objective: Ban as many MU as possible
      4. Key Result: X million MU banned (X being an arbitrary estimate because we don’t actually know how many MU there are)
      5. Implementation: Train and use AI to identify and ban MU
      6. Difficulty: Quality Assurance to check whether the banned users actually are malicious
      7. Optimisation: Stonewall or chew out people demanding company resources be wasted on QA and simply define all banned users as MU (with all the money we invested in that AI, we can’t afford conceding inaccuracy)
      8. Conclusion: report amount of banned users to partners and regulators.
      9. If they should ask about the accuracy, cherry-pick from the training dataset to present users your AI correctly identified. Concede individual errors, frame them as the price of progress and assure that they have been taken into consideration for refining the algorithm.
      10. Affirm your commitment to a safe platform, improved security measures, yada yada

      I’ve not seen any examples this egregious, personally, but I also don’t make reports for executives, let alone executives of companies with such a vile business model. This is just a scaled up version of some shenanigans I’ve been party to.

      • schipelblorp@sh.itjust.works
        2·
        11 hours ago

        Yup. This is what I imagine. Whoever is in charge of this process is both in charge of banning as many people as possible, but is also the person most able to tell you if they’re fucking up or not.

        I think this is why twitter shaming is the ONLY effective method to get your account back, because it goes around that person.

  • joeljoelle@piefed.blahaj.zoneEnglish
    11·
    1 day ago

    Any creators should have left long ago, unless they already had a massive following, it feels not so good to have your hard work buried under piles and piles of selfies and influencers and ads and spam and memes and everything. Since they deprecated the tags it all went to shit, well since Meta bought it really, but that’s when I noticed lol.

  • Miller@lemmy.world
    22·
    2 days ago

    Hell bent on removing people from a system solely set up for people, how could it go wrong.

  • psx_crab@lemmy.zipEnglish
    17·
    2 days ago

    Wow going from automated ban and automated appeal where appealing account ban is hellishly difficult to AI ban and AI appeal where appealing account ban is hellishly difficult.

    Teknologia!

    • ZDL@lazysoci.al
      3·
      20 hours ago

      Usually it’s some kind of damage, induced or congenital, to the ventromedial prefrontal cortex or the closely related orbitofrontal cortex.