4 ms·

All I know is self-driving cars are going to be killing a lot of passengers and pedestrians when corporations and programmers now get government liability waive

by ck2 7y ago

All I know is self-driving cars are going to be killing a lot of passengers and pedestrians when corporations and programmers now get government liability waivers, rush out the project and just say "oops" when the spot the bug in the code. I mean who goes to prison when a car runs someone over and no-one is driving? What is a developer's motivation to be careful and defer to safety? We've already proven a good chunk of the population are sociopaths, they don't care who dies as long as they get their paycheck.

steveeq1 7y ago
And what if the self-driving cars end up causing fewer deaths than human drivers?
- perl4ever 7y ago
  If you assume something will automatically happen, without a particular mechanism, how likely is it to? Seems to me that is a popular cargo cult few are immune to.
  - carlmr 7y ago
    >If you assume something will automatically happen, without a particular mechanism He didn't mention it, but there are some pretty obvious mechanisms for this to be true: 1. Automated drivers don't get tired. 2. Automated drivers may "see" more of their environment than a human can. Human sensors aren't upgradeable. 3. Automated drivers don't get drunk. 4. Automated drivers don't have emotions. 5. Automated drivers aren't distracted by the infotainment system, the children in the back seat, or even some thought about what just happened at work. 6. With time, automated drivers can be improved, while human weaknesses stay constant. There are of course drawbacks: 1. Humans have intuition about situations they've never been in. Most AI/machine learning still generalizes poorly. 2. The programming may be buggy. 3. The datasets may be inadequate. So what will happen is definitely out there, but there are clear mechanisms for this to happen.
    - perl4ever 7y ago
      A list of possible advantages is not what I mean by "a mechanism". It's very, very much not. I'm not saying I disagree with anything in particular that you listed, but you can list any number of things and it simply doesn't even assert what those things sum up to.

by ck2 7y ago

steveeq1 7y ago

And what if the self-driving cars end up causing fewer deaths than human drivers?

perl4ever 7y ago
If you assume something will automatically happen, without a particular mechanism, how likely is it to? Seems to me that is a popular cargo cult few are immune to.
- carlmr 7y ago
  >If you assume something will automatically happen, without a particular mechanism He didn't mention it, but there are some pretty obvious mechanisms for this to be true: 1. Automated drivers don't get tired. 2. Automated drivers may "see" more of their environment than a human can. Human sensors aren't upgradeable. 3. Automated drivers don't get drunk. 4. Automated drivers don't have emotions. 5. Automated drivers aren't distracted by the infotainment system, the children in the back seat, or even some thought about what just happened at work. 6. With time, automated drivers can be improved, while human weaknesses stay constant. There are of course drawbacks: 1. Humans have intuition about situations they've never been in. Most AI/machine learning still generalizes poorly. 2. The programming may be buggy. 3. The datasets may be inadequate. So what will happen is definitely out there, but there are clear mechanisms for this to happen.
  - perl4ever 7y ago
    A list of possible advantages is not what I mean by "a mechanism". It's very, very much not. I'm not saying I disagree with anything in particular that you listed, but you can list any number of things and it simply doesn't even assert what those things sum up to.