Self driving car kills a pedestrian

Page 2 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

Starbuck1975

Lifer
Jan 6, 2005
14,698
1,909
126
1. reproduce the fault in a simulation
2. fix code / learning sets for the AI
3. prove it works
Continue.
That is whats going to happen
Can we hold the coders liable, as is the case for other engineers when design flaws or ommissions cause loss of life? I want code engineers to start stamping and certifying their work.
 

Fenixgoon

Lifer
Jun 30, 2003
31,822
10,361
136
while it does look bad for self driving cars, just remember that the real bar is "better than humans" which is pretty damn low. self-driving vehicles need not be perfect.

that being said, i'm not surprised it's uber (mostly because i think they're a shitty company). could very well be that no AI-driven vehicle could have avoided it. gotta wait for a full investigation.
 

ElFenix

Elite Member
Super Moderator
Mar 20, 2000
102,425
8,388
126
Why? If this was a case where a human would have hit the jaywalker too (like them darting out from behind a truck) then Skynet was not at fault this time.

I'm not sure about their current safety, but if not already then eventually self-driving cars will be safer than the typical human driver.
i think it's because virtual larry can see where tajy is going with this.
 

Ichinisan

Lifer
Oct 9, 2002
28,298
1,234
136
Good. I hope that the AZ Legislature takes up deliberation on passing a law to BAN those so-called "self driving" vehicles.

Edit: With apologies and condolences to the family of the deceased.
Next time a human driver kills a pedestrian, I will demand that we BAN human drivers!

Oh. It probably happened several times as I typed this.
 

Chaotic42

Lifer
Jun 15, 2001
33,929
1,098
126
Can we hold the coders liable, as is the case for other engineers when design flaws or ommissions cause loss of life? I want code engineers to start stamping and certifying their work.

Costs and timelines would skyrocket - computer programming is incredibly complex for anything useful.
 

cytg111

Lifer
Mar 17, 2008
23,561
13,122
136
Can we hold the coders liable, as is the case for other engineers when design flaws or ommissions cause loss of life? I want code engineers to start stamping and certifying their work.

If you can hold god accountable for the fallacies of man? Then yea. Sure.
 

DaveSimmons

Elite Member
Aug 12, 2001
40,730
670
126
Taj, are you a technophobe who wants to see thousands die to human drivers rather than risk far fewer deaths from automation?

Perhaps he's a hipster who only supports artisanal craft driving. AI will save lives but it's far too cold and impersonal. That overly precise digital driving just can't compare to the warmth of analog.
 
Reactions: Jaskalas

Starbuck1975

Lifer
Jan 6, 2005
14,698
1,909
126
Costs and timelines would skyrocket - computer programming is incredibly complex for anything useful.
Oh well. Disruption is simply using technology to circumvent the rules, regulations and standards that most industries must abide by...eventually Silicon Valley will too have to bend the knee. The EU is already starting to impose regulatory oversight.

If you can hold god accountable for the fallacies of man? Then yea. Sure.
Professional engineers of other disciplines must be divine then, because they already are held accountable.
 

OutHouse

Lifer
Jun 5, 2000
36,413
616
126
The Budweiser brewery in Ft Collins Colorado uses fully autonomous semis to deliver beer from Ft Collins to Colorado Springs. so far there have been no accidents due to the robot trucks and that is driving I25 south through heavy traffic in Denver.
 

cytg111

Lifer
Mar 17, 2008
23,561
13,122
136
Oh well. Disruption is simply using technology to circumvent the rules, regulations and standards that most industries must abide by...eventually Silicon Valley will too have to bend the knee. The EU is already starting to impose regulatory oversight.


Professional engineers of other disciplines must be divine then, because they already are held accountable.

There is a billion lines of code, written by a million different coders, written from yesterday to 20 years ago, all interconnected and depended on each other. Good luck finding the guilty party.
The second you talk AI and neural nets your product is susceptible to the same errors as you and I would produce and while you and I may have an idea how to handle a jump scare in a car going 100 when an animal decides to cross the road? That AI has tried that 1.000.000 times in training sessions. Does that mean it will never act counterproductive in a jump scare scenario? No. Does it mean it handles it better than you or I by a factor of 1.000? Yes it does.
 
Reactions: Jaskalas

VirtualLarry

No Lifer
Aug 25, 2001
56,453
10,120
126
I'm not sure about their current safety, but if not already then eventually self-driving cars will be safer than the typical human driver.
It's not the safety I'm concerned about. It's the basic right of autonomy and self-determination. (*)

Edit: The only case I would support self-driving cars, is when the owner / would-be driver is impaired. (Had a drink too many at the bar or at a wedding, OR is having a medical emergency of some sort.)

(*) Look at what China is doing TODAY, with their "Social Credit System", and blacklisting countless portions of their population from taking mass transit, because of something that they may have said or done, that angered a politician. (Not even something criminal.)

Would you have massive swathes of people, effectively under house arrest, because they are "unwanted" in society, and are unable to take ANY transportation at all? Think of those scenes in Minority Report. We don't need that in a Free Society.

Edit: Freedom is more important than Safety. If 10,000 people die each year on the highways and roads due to human error, so that we may remain free, so be it!
 
Last edited:
Reactions: imported_tajmahal

Starbuck1975

Lifer
Jan 6, 2005
14,698
1,909
126
There is a billion lines of code, written by a million different coders, written from yesterday to 20 years ago, all interconnected and depended on each other. Good luck finding the guilty party.
The second you talk AI and neural nets your product is susceptible to the same errors as you and I would produce and while you and I may have an idea how to handle a jump scare in a car going 100 when an animal decides to cross the road? That AI has tried that 1.000.000 times in training sessions. Does that mean it will never act counterproductive in a jump scare scenario? No. Does it mean it handles it better than you or I by a factor of 1.000? Yes it does.
That may all very well be true, but it still circumvents the answer of who is liable when the AI makes a mistake.
 
Reactions: pmv

ch33zw1z

Lifer
Nov 4, 2004
38,003
18,350
146
That may all very well be true, but it still circumvents the answer of who is liable when the AI makes a mistake.
The corp selling the gear. It's their responsibility to test their products for safety. Just like any other car maker...
 

pmv

Lifer
May 30, 2008
13,298
8,212
136
What's frustrating is that this will probably trigger at least a mild backlash against self-driving vehicles, even though they should ultimately be safer than human-piloted cars.

Yes, something clearly went wrong, and yes, this is tragic and would ideally have been avoidable. But I'm worried this will lead to knee-jerk reactions in the public and legislatures rather than a careful, considered response that might lead to better regulation and cultural understanding.

I remain in favour of such a backlash. I still think they are a techno-evangelist pipedream and that they either won't work and will be quietly shelved, or, much, much worse, will work only at the expense of remodelling public space to accommodate them and their limitations (in an alliance of profit-hungry corporations, the techno-evangelists, and the sedentary - possibly the three most powerful forces in the world, come to think of it).

Best outcome is they might find a niche use in some very limited contexts (maybe taking the jobs of some long-distance truck drivers on long straight empty roads with little pedestrian or cyclist or even human driver traffic). I don't believe they will ever be ready for mass use in chaotic, crowded urban areas, not without imposing huge restrictions on people in those areas.

I mean, they are testing them in Arizona, a place with perfect weather, wide roads, good lines of sight, and (mostly) well-behaved other traffic. And still they have accidents. Can't see them translating well to more difficult parts of the world.
 

pmv

Lifer
May 30, 2008
13,298
8,212
136
I want more carless drivers not driverless cars.
Next time a human driver kills a pedestrian, I will demand that we BAN human drivers!

Oh. It probably happened several times as I typed this.


But one can demand that the human driver involved is held to account. Now, granted, that doesn't happen as much as I'd want (largely because juries are full of bad drivers too quick to identify with the perpetrator). But at least there's an aspiration in principle. Some here seem to be saying that _nobody_ can be so held responsible when some self-driving car does the same. That's just not good enough.

Mostly though I want to see far fewer cars and far less driving. Self-driving cars aren't going to achieve that.


There is a billion lines of code, written by a million different coders, written from yesterday to 20 years ago, all interconnected and depended on each other. Good luck finding the guilty party.

Well, if you can't point to who is responsible for what the car does, then you have no business using it on a public road. So hold responsible the car's owner who allowed it to be so used.
 

Ichinisan

Lifer
Oct 9, 2002
28,298
1,234
136
I want more carless drivers not driverless cars.



But one can demand that the human driver involved is held to account. Now, granted, that doesn't happen as much as I'd want (largely because juries are full of bad drivers too quick to identify with the perpetrator). But at least there's an aspiration in principle. Some here seem to be saying that _nobody_ can be so held responsible when some self-driving car does the same. That's just not good enough.

Mostly though I want to see far fewer cars and far less driving. Self-driving cars aren't going to achieve that.




Well, if you can't point to who is responsible for what the car does, then you have no business using it on a public road. So hold responsible the car's owner who allowed it to be so used.
You're really jumping the gun. These experimental self-driving cars have DVR and all kinds of telemetry recorded. It may very well show that the only one responsible is the deceased pedestrian. I suspect that it will be the same scenario in nearly every future case of a pedestrian being hit by a self-driving car.
 

Commodus

Diamond Member
Oct 9, 2004
9,215
6,818
136
I remain in favour of such a backlash. I still think they are a techno-evangelist pipedream and that they either won't work and will be quietly shelved, or, much, much worse, will work only at the expense of remodelling public space to accommodate them and their limitations (in an alliance of profit-hungry corporations, the techno-evangelists, and the sedentary - possibly the three most powerful forces in the world, come to think of it).

Best outcome is they might find a niche use in some very limited contexts (maybe taking the jobs of some long-distance truck drivers on long straight empty roads with little pedestrian or cyclist or even human driver traffic). I don't believe they will ever be ready for mass use in chaotic, crowded urban areas, not without imposing huge restrictions on people in those areas.

I mean, they are testing them in Arizona, a place with perfect weather, wide roads, good lines of sight, and (mostly) well-behaved other traffic. And still they have accidents. Can't see them translating well to more difficult parts of the world.

I think they'll be ready eventually, but we need to be conservative about the timeframe precisely because the tech isn't ready for snow, truly chaotic roads, that sort of thing. But more advanced navigation is already in development (there's tech that can anticipate cars coming around the corner without a line of sight, for example), and it seems odd to throw a "never" at technology that's still quite young.

One thing to remember is that people are panicking at one fatality (which is a serious concern, to be clear), but an average of 101 Americans die in human-piloted car crashes per day. You're suggesting we should be outraged at attempts to foster self-driving cars when a horrendous number of people are dying right now from poor driving that, in at least some cases, could have been prevented by an autonomous system.
 

Chaotic42

Lifer
Jun 15, 2001
33,929
1,098
126
Mostly though I want to see far fewer cars and far less driving. Self-driving cars aren't going to achieve that.
One of the big benefits of self-driving cars is the ability to have car services drive you around like an Uber. A car would take you to work and then go off moving other people around rather than just being parked at your office. If every car was self-driving and doing this, it would cut down on the number of cars in a city significantly.
 
Reactions: DaveSimmons

Sunburn74

Diamond Member
Oct 5, 2009
5,034
2,613
136
I think the details have to come out. Its a setback for self driving cars for sure but at the same time what this type of technology can do for public safety and transportation can also potentially save lives in the future.
 
sale-70-410-exam    | Exam-200-125-pdf    | we-sale-70-410-exam    | hot-sale-70-410-exam    | Latest-exam-700-603-Dumps    | Dumps-98-363-exams-date    | Certs-200-125-date    | Dumps-300-075-exams-date    | hot-sale-book-C8010-726-book    | Hot-Sale-200-310-Exam    | Exam-Description-200-310-dumps?    | hot-sale-book-200-125-book    | Latest-Updated-300-209-Exam    | Dumps-210-260-exams-date    | Download-200-125-Exam-PDF    | Exam-Description-300-101-dumps    | Certs-300-101-date    | Hot-Sale-300-075-Exam    | Latest-exam-200-125-Dumps    | Exam-Description-200-125-dumps    | Latest-Updated-300-075-Exam    | hot-sale-book-210-260-book    | Dumps-200-901-exams-date    | Certs-200-901-date    | Latest-exam-1Z0-062-Dumps    | Hot-Sale-1Z0-062-Exam    | Certs-CSSLP-date    | 100%-Pass-70-383-Exams    | Latest-JN0-360-real-exam-questions    | 100%-Pass-4A0-100-Real-Exam-Questions    | Dumps-300-135-exams-date    | Passed-200-105-Tech-Exams    | Latest-Updated-200-310-Exam    | Download-300-070-Exam-PDF    | Hot-Sale-JN0-360-Exam    | 100%-Pass-JN0-360-Exams    | 100%-Pass-JN0-360-Real-Exam-Questions    | Dumps-JN0-360-exams-date    | Exam-Description-1Z0-876-dumps    | Latest-exam-1Z0-876-Dumps    | Dumps-HPE0-Y53-exams-date    | 2017-Latest-HPE0-Y53-Exam    | 100%-Pass-HPE0-Y53-Real-Exam-Questions    | Pass-4A0-100-Exam    | Latest-4A0-100-Questions    | Dumps-98-365-exams-date    | 2017-Latest-98-365-Exam    | 100%-Pass-VCS-254-Exams    | 2017-Latest-VCS-273-Exam    | Dumps-200-355-exams-date    | 2017-Latest-300-320-Exam    | Pass-300-101-Exam    | 100%-Pass-300-115-Exams    |
http://www.portvapes.co.uk/    | http://www.portvapes.co.uk/    |