• extended update service to 7 years
  • pushing Material UI consistency
  • Tensor G3 chip for AI & machine learning performed locally on device
  • kirklennon@kbin.social
    link
    fedilink
    arrow-up
    29
    arrow-down
    4
    ·
    1 year ago

    Like seriously, what’s the percentage of people that run machine learning algorithms on their phone? 0.0000000001%?

    It’s 100%. You use them on your phone all day every day. Your keyboard used machine learning algorithms as you typed your comment to dynamically adjust the size of the tap target of your likely next character and for autocorrect.

    Every single photo taken on a phone is run through a huge amount of ML to create it.

    All of this is to say, however, that this headline is ridiculous. Aside from Material UI, this is basically a description of every iPhone from the last half decade (with a dedicated “neural engine”). Not really a change in the smartphone world.

    • Salamendacious@lemmy.worldOP
      link
      fedilink
      English
      arrow-up
      5
      ·
      1 year ago

      Definitely a little bit of hyperbole with that headline. I think I have to use the exact headline the article uses though.

    • rush@lemm.ee
      link
      fedilink
      English
      arrow-up
      5
      arrow-down
      1
      ·
      1 year ago

      I agree in the fact that some ML algorithms are used daily by nearly all who have a smartphone, but things like increasing the touch target is BS. It’s quite easy to confirm by simply either looking at the code or enabling “show layout bounds” in Android’s Developer Options.