Self-driving Uber car kills Arizona pedestrian

[DrYouth]

Nor is this going to be regulated properly.

What, now government regulations are going to save the day?

These companies have plenty of motivation to keep their vehicles from killing people without government regulations...

It's bad for the brand isn't it.

People will sue them won't they...

Why suddenly invoke government regulation as the savior?

You see how unCanadian this forum has made me....

[Speaker to Animals]

Nor is this going to be regulated properly.

What, now government regulations are going to save the day?

These companies have plenty of motivation to keep their vehicles from killing people without government regulations...

It's bad for the brand isn't it.

People will sue them won't they...

Why suddenly invoke government regulation as the savior?

You see how unCanadian this forum has made me....

Because to develop an AI system in which the consequence of failure is the loss of human life, a team should be something like CMMI 4 or CMMI 5, and the government ought to establish the same testing and quality control regulations that apply to developing aircraft avionics, especially to autopilot, which actually is a much easier problem.

This problem domain is fraught with far more risk than even automatic flight control systems and aircraft navigation. The standard should be the same.

[Speaker to Animals]

I am sure the H1B pajeets will get right on fixing this latest defect.

http://www.bbc.com/news/world-us-canada-43604440

[de officiis]

I am sure the H1B pajeets will get right on fixing this latest defect.

http://www.bbc.com/news/world-us-canada-43604440

I'm sure all the truck drivers are relieved. The Day of Judgment postponed a little longer.

[Speaker to Animals]

I am sure the H1B pajeets will get right on fixing this latest defect.

http://www.bbc.com/news/world-us-canada-43604440

I'm sure all the truck drivers are relieved. The Day of Judgment postponed a little longer.

I don't know. It's not impossible to do this by any means. I could think of a few different ways to implement it. But to do it right will take quite a lot of money and time. The federal government is going to step in eventually and make requirements similar to the space and aviation industries. These corporations are going to agree to do it that way and probably start again.

I think what happened here was zeal on the part of the business types who usually take over corporations from the engineers who start them but don't understand the actual products and what is required to develop them properly.

Going after taxis first, in my opinion, was also a big mistake. But something like a tractor and trailer with a full load drastically ups the risk as well. A very good start would be for private vehicles on various campuses that ferry people between buildings, and to and from parking lots. That's pretty slow and there exist all kinds of obstacles and contingencies to mitigate. It's a good problem domain that isn't likely to kill anybody while you figure out where all the problems will arise.

That's really the big issue here. Because it's so new, they have no idea what kinds of problems will arise, but they do know that failure could very well end in death.

It was a bad roll-out of this technology. Very bad.

[Hastur]

Hofstadter's law And the Ninety-ninety rule comes to mind. They think they are close to launching but the truth is they are still years away from fixing the last problems that has to be fixed.