• @axo@feddit.de
    link
    fedilink
    English
    3921 days ago

    Accoring to the math in this video: :

    • 150 000 000 miles have been driven with Teslas “FSD”, which equals to
    • 375 miles per tesla purchased with FSD capabilities
    • 736 known FSD crashes with 17 fatalities
    • equals 11.3 deaths per 100M miles of teslas FSD

    Doesnt sound to bad, until you hear that a human produces 1.35 deaths per 100M miles driven…

    Its rough math, but holy moly that already is a completely other class of deadly than a non FSD car

    • @dufkm@lemmy.world
      link
      fedilink
      English
      821 days ago

      a human produces 1.35 deaths per 100M miles driven

      My car has been driven around 100k miles by a human, i.e. it has produced 0.00135 deaths. Is that like a third of a pinky toe?

    • @NotMyOldRedditName@lemmy.world
      link
      fedilink
      English
      -1
      edit-2
      21 days ago

      That number is like 1.5 billion now and rising exponentially fast.

      Also those deaths weren’t all FSD they were AP.

      The report says 1 FSD related (not caused by but related) death. For whatever reason the full details on that one weren’t released.

      Edit: There are billions of miles on AP. In 2020 it was 3 billion

      Edit: Got home and I tried finding AP numbers through 2024 but haven’t seen anything recent, but given 3 billion 2020, and 2 billion in 2019, and an accelerating rate of usage with increased car sales, 2023 is probably closer to 8 billion miles. I imagine we’d hear when they reach 10 billion.

      So 8 billion miles, 16 AP fatalities (because that 1 FSD one isn’t the same) is 1 fatality per 500,000,000 miles, or put into the terms above by per 100mil miles, 0.2 fatalities per 100 million miles or 6.75 times less than a human produces. And nearly all of these fatal accidents were from blatant misuse of the system like driving drunk (at least a few) or using their phone and playing games.