3 ms·

> There have to be at least dozens of documented cases of this being unsafe. The human in the loop seems to catch the mistake just fine? Doesn't seem to be eve

> There have to be at least dozens of documented cases of this being unsafe. The human in the loop seems to catch the mistake just fine? Doesn't seem to be even remotely close to an accident?

skywhopper 4y ago
Did you miss the ones where the self-driving Teslas plowed into fire trucks on the side of the road?
kadoban 4y ago
Any system based on a human doing nothing for long periods and then being expected to act swiftly and accurately is _deeply_ flawed. Human attention doesn't work like that. This is russian roulette with the general public forced to play.
- mwint 4y ago
  So FSD beta testing has been around for what, a couple years? To continue the tortured analogy: We’ve been playing this Russian roulette for millions of miles. How many times to people have to be “almost” shot, but not actually shot, before we concede that maybe the gun has no bullets?
  - kadoban 4y ago
    > How many times to people have to be “almost” shot, but not actually shot, before we concede that maybe the gun has no bullets? You say that like we have anything like good data. We have pretty much what Tesla's marketing team put together on their website, and _maybe_ what some cities or states have cobbled together. And before we assume that Telsa wouldn't lie, let's remember that they call this "Full Self-Driving".
    - mwint 4y ago
      We all know that a real accident would be all over YouTube, and front page of HN for about a week straight. That we have so many “almost” videos, and a couple “it hit a cone”, but no real collisions seems telling.
      - kadoban 4y ago
        There's all kinds of Tesla crash news. Here's one of hundreds of results. https://www.cnn.com/2022/05/19/cars/nhtsa-tesla-crash-california/index.html https://www.cnn.com/2022/05/19/cars/nhtsa-tesla-crash-califo... Which ones are in which terribly-named driving mode? Who knows, Tesla doesn't seem to be saying.
justin66 4y ago
OP was referring to the safety of the feature, you're referring to the safety of the driver who was experimenting with it.
- chrisseaton 4y ago
  > you're referring to the safety of the driver who was experimenting with I Isn't the safety of running the beta (the issue being discussed) a function of the safety of the driver?
salawat 4y ago
This totally worked for 737 MAX. This one doesn't even have a well defined algorithm so it's behavior can be predicted ahead of time.
- chrisseaton 4y ago
  > This totally worked for 737 MAX. But this is how all aircraft autopilot systems work - they expect a pilot to be able to step in. The pilots don't leave the controls.

> There have to be at least dozens of documented cases of this being unsafe. The human in the loop seems to catch the mistake just fine? Doesn't seem to be even remotely close to an accident?

skywhopper 4y ago

Did you miss the ones where the self-driving Teslas plowed into fire trucks on the side of the road?

kadoban 4y ago

Any system based on a human doing nothing for long periods and then being expected to act swiftly and accurately is _deeply_ flawed. Human attention doesn't work like that. This is russian roulette with the general public forced to play.

mwint 4y ago
So FSD beta testing has been around for what, a couple years? To continue the tortured analogy: We’ve been playing this Russian roulette for millions of miles. How many times to people have to be “almost” shot, but not actually shot, before we concede that maybe the gun has no bullets?
- kadoban 4y ago
  > How many times to people have to be “almost” shot, but not actually shot, before we concede that maybe the gun has no bullets? You say that like we have anything like good data. We have pretty much what Tesla's marketing team put together on their website, and _maybe_ what some cities or states have cobbled together. And before we assume that Telsa wouldn't lie, let's remember that they call this "Full Self-Driving".
  - mwint 4y ago
    We all know that a real accident would be all over YouTube, and front page of HN for about a week straight. That we have so many “almost” videos, and a couple “it hit a cone”, but no real collisions seems telling.
    - kadoban 4y ago
      There's all kinds of Tesla crash news. Here's one of hundreds of results. https://www.cnn.com/2022/05/19/cars/nhtsa-tesla-crash-california/index.html https://www.cnn.com/2022/05/19/cars/nhtsa-tesla-crash-califo... Which ones are in which terribly-named driving mode? Who knows, Tesla doesn't seem to be saying.

justin66 4y ago

OP was referring to the safety of the feature, you're referring to the safety of the driver who was experimenting with it.

chrisseaton 4y ago
> you're referring to the safety of the driver who was experimenting with I Isn't the safety of running the beta (the issue being discussed) a function of the safety of the driver?

salawat 4y ago

This totally worked for 737 MAX. This one doesn't even have a well defined algorithm so it's behavior can be predicted ahead of time.

chrisseaton 4y ago
> This totally worked for 737 MAX. But this is how all aircraft autopilot systems work - they expect a pilot to be able to step in. The pilots don't leave the controls.