Exposing Mark Rober’s Tesla Crash Story
youtu.be
Consider supporting the channel: https://patreon.com/AIDRIVR

I think AI driver give a good summary of some of issues with the Mark Rober showcase of LIDAR.

  • Some of the bias Mark Rober friendships with the LIDAR supplier.
  • Not using Tesla full self driving
  • Being so nervous that he accidentally disables autopilot, twice.
  • The one sided view in LIDAR, with not show examples of LIDAR have issues.
  • Give a way forward to do the tests with FSD
  • Mark Rober not giving the test his full attention, fun vs scientific process
  • OptimusPrimeDownfall@discuss.tchncs.deEnglish
    7·
    2 days ago
    1. FSD won’t solve the “the sensors don’t detect the wall” issue
    2. That was not the point of the video.
    3. It did stop, yes. That is a win. Didn’t kill the kid. The Tesla didn’t. It killed the kid. That’s a fail. What’s your point?
    4. Why would you run a test multiple times with the high potential for damage to a very expensive vehicle when it’s not supposed to be published in a journal? You talk about cost vs features for manufacturers but don’t apply that same logic to Rober. Its also a one-off thing for Rober, so the math is better for him than the “cheaper, OK with kids dying” method of Musk.

    I think you’re underestimating how many situations a LIDAR system will be better than an all camera system. Its also a tradeoff in human lives. I’d rather it be slightly more expensive up front and not have kids die than cheap and kill kids.

    And yes. Self driving should be better than humans. Cause humans suck as drivers. We have two cameras in one location in the car. Self driving must be better and make up for the limitations of humans. Cameras dont make up for them and thus are a terrible replacement for humans.

    • Eideen@lemmy.worldOPEnglish
      12·
      2 days ago

      It sounds like you feel LIDAR is need to provide the service you are looking for.

      It is probably possible that I underestimate how many situations there is that LIDAR can be significant better. I am open to learn more about it. It can also be that valve the normal driving more than the edge cases. From my view historically LIDAR cars have mainly been limited to a limited area, will FSD seem to be in a league of its own driving on unmarked roads and parking spots.

      Yes I am praise FSD, I think Elon only part in that is the willingness to take risk and fund someone that have cost billions to develop, and taking a leadership role in setting a direction, when at the time no other automaker was willing to do so.

      If you hear the software engineers talking about how complex of task it is to solve logic behind make a car feel human, you will truly understand how little we know. Can recommend this video from the last AI day Ashore from Tesla talk about FSD

      The issue more there is a lot of people that look at the video and think, “LIDAR is need because it stop for this obscure events, and is not the result in normal driving that matter”. How many people experience a Wile E. Coyote event per year, where a perfect painted wall pops up around a blind corner?!

      If a car stop because it was raining or a fire hydrant is leaking, and there no other issue, is that good for everyday driving?

      You say you like to pay more for a system that also have LIDAR, how much are the mass of people willing to pay extra for that feature that only helps in a 0,1% of the time? 500€, 5000€, or 50000€

      How much better does a self driving car need be? What stats shall we use to measure if a self driving car is better than human? This are all complex topics that we like a quick answer to.

      Going back to Mark, the Wile E. Coyote event this could be it own video, as it not relevant for normal driving. It wound be fun on its own.

      With science process, I mean the fog and water they could have looks at there setup was there something wrong with the setup, that favor one car over the other? Will the LIDAR stop without having an obstacle behind the “wall” of water? How does it perform normal driving? Is this element that is unreasonable to do in a LIDAR vs Camera test?