• I Cast Fist@programming.dev
    link
    fedilink
    English
    arrow-up
    44
    ·
    3 months ago

    I remember reading some time ago that “the idea (of phones listening to everything you say to serve ads) makes no economic sense, because it’d be too expensive to run”

    Looks like it actually isn’t “too expensive” to run in the end.

    • Overshoot2648@lemm.ee
      link
      fedilink
      English
      arrow-up
      16
      ·
      3 months ago

      Except Facebook never used it this was a 3rd party trying to hype up investors. Many audits have been run on these apps and there is no way they use your microphone. It’s way cheaper to just look at your search history.

    • Maeve@kbin.earth
      link
      fedilink
      arrow-up
      13
      ·
      3 months ago

      Yep, and it’s not just Facebook, not just microphone. My lappie recently started serving ads for something I searched on a device not linked to it. I’m guessing it’s my ISP engaging in these sneak tactics.

    • TheObviousSolution@lemm.ee
      link
      fedilink
      English
      arrow-up
      10
      ·
      edit-2
      3 months ago

      It’s not when it’s your device doing the computing. All electronic devices should have visible hardware indicators for when their camara or microphone is on, but that’s a consumer rights issue most people are dismissive of, so it’s not happening. Some people even always want it on for the assistant functionality.

        • Echo Dot@feddit.uk
          link
          fedilink
          English
          arrow-up
          6
          ·
          edit-2
          3 months ago

          Not how it works.

          They have an ultra low power processor that listens for the “hey Google” keyword, then that wakes up the main audio processor. But the main microphone is not actually on, and that small processor isn’t capable of recording audio it’s just looking for a certain matching sound wave and then triggering.

          That’s why it sometimes triggers if you just go “hurr ner dorrll” because those random sounds are close enough to what it’s looking for.

          That is why some older devices can’t actually install the assistant software. They lack the necessary hardware to do it in a power efficient manner.

    • Schadrach@lemmy.sdf.org
      link
      fedilink
      English
      arrow-up
      8
      ·
      3 months ago

      Even then, you have local voice recognition. You don’t need to stream all microphone recordings to some central server for processing, you just do voice recognition and keep a log of say the last 100 nouns and a high priority log for the last twenty nouns used near verbs like purchase, buy or get. Then send those lists to the ad provider as context. All the hard work is done on the client device and the same backend used for ad context on web pages can be used for this as well.

      • mrinfinity@lemmy.world
        link
        fedilink
        English
        arrow-up
        6
        ·
        3 months ago

        Then hide it encrypted in an image upload or some other packet. Listen for ‘buy a <something>’ encrypt its text version, wait for something to cargo it with in a data transmission so people looking at data transmissions aren’t any the wiser, hide it in some obscure way that would look normal otherwise, it’s intercepted, sends off to advertisers. Adtech is cyber terrorism.

        • Echo Dot@feddit.uk
          link
          fedilink
          English
          arrow-up
          5
          ·
          3 months ago

          You don’t think security experts know about stereonography techniques? It’s like the first thing you learn about in uni for it. Like the first week.

    • dev_null@lemmy.ml
      link
      fedilink
      English
      arrow-up
      1
      ·
      3 months ago

      Yeah, a marketing agency selling snake oil to people that actually think they can do it is not expensive. Of course they never actually built the tech.