It's happened. Self-driving Uber kills pedestrian

[79wjd]

7 minutes ago, bcredeur97 said:

 

 

most cars can stop from 60 mph in 110-150 feet.  It's drivers reaction time that is the problem though.

...

Sports cars stop in the 100-120 foot range. Typical cars are more like 150-200 (keeping in mind that's with new cars with new brakes and tires, not older cars with equipment that was worse to begin with let alone after wearing down).

 

And on top of that, you don't have to be going very fast to kill a kid.

[Tcrumpen]

19 hours ago, Densetsu said:

• A robot may not injure a human being or, through inaction, allow a human being to come to harm.
• A robot must obey orders given it by human beings except where such orders would conflict with the First Law.
• A robot must protect its own existence as long as such protection does not conflict with the First or Second Law.

 

Well, there goes that.

You're forgetting the natural evolution of AI and Robots in that "We toxify our earth, our countries wage wars and persue more imaginative ways of self destruction, we can not be trusted with your own survival"

 

Paraphrasing of course

[mynameisjuan]

11 minutes ago, bcredeur97 said:

suming the computer starts braking the instant it sees the person, by the time it reaches that 50 foot mark it certainly won't be moving fast enough to instantly kill that person. maybe just a minor injury or even nothing. so long as the person doesn't end up under the car anyway...

Never mention anything about it being fatal. Just about stopping in time to avoid hitting someone altogether. And yea the 50ft mark was with reaction time. 

 

12 minutes ago, bcredeur97 said:

I'd love to see what actually happened. Did the car attempt to slow down at all? did it just keep driving along and completely not *see* the person? We don't know...

I want to know as well. At least in the video it looks like the bike is just in front of the car meaning it slowed down enough to not send this flying forward or over the top of the car. I still put my money on the car not able to stop in time. 

 

[Tcrumpen]

19 hours ago, ucrbuffalo said:

https://www.abc15.com/news/arizona-police-investigating-self-driving-uber-car-involved-in-crash-overnight

 

Apparently there was a driver in the car, but it was in autopilot mode. I have to wonder what kind of systems might be in place in these vehicles to alert the driver in the case of an imminent emergency? If the car knows (which it may not have in this case) that it will have to slow down significantly, might it make a loud beep or something to warn the driver?

Depending on how much notification the driver got he may not have been able to react quickly enough, or he could have reacted and this was the best outcome that could have happened or his interference actually worsened the situtaion; as someone mentioned before it's all very unknown at the moment

 

There will no doubt be ethical discussions regarding autonomous cars because of this, my personal opinion is that the company in question who makes the car autonomous is liable for anything of this affect; in this case - uber

 

However if a human interaction was found to be the cause of the crash then the driver becomes responsible. However one thing that most people don't think of is that humans aren't like computers there are sometimes cases where human interaction over a computer making a decision would actually give better results, example; computers will be told not to drive on pavements, however there maybe a situation where swerving onto the pavement (providing no-one is around) would actually save lives

[WkdPaul]

12 minutes ago, asus killer said:

dude that was my point. 

 

only a person can see a child playing by the road and decide it's a danger. A unnamed car will never be able to without some insane AI or having to go really slow whenever someone was walking in the sidewalk. 

This because the woman could have been distracted but the car will never anticipate that, will only react when she enters the road.

One thing though, this autonomous Uber car HAD a human behind the wheel in case of things like this, they both (the autonomous system and human operator) didn't had time to react. Just saying.

[samcool55]

1 minute ago, wkdpaul said:

One thing though, this autonomous Uber car HAD a human behind the wheel in case of things like this, they both (the autonomous system and human operator) didn't had time to react. Just saying.

The human actually didn't spot the pedestrian at all, he knew something was wrong because he heard something. So i'm wondering, did the car notice the pedestrian before the crash or not? The human didn't, that's clear, but if the car did it's another sign AI is potentially better at driving than humans.

[asus killer]

3 minutes ago, wkdpaul said:

One thing though, this autonomous Uber car HAD a human behind the wheel in case of things like this, they both (the autonomous system and human operator) didn't had time to react. Just saying.

the human could not exactly be paying attention. I never been in this position but i really don't think you would have been so focused like if you were driving, at least all the time.

 

But this was more a general issue if one can call it that with self driving cars, apart from that i'm all for it. 

[bcredeur97]

12 hours ago, wkdpaul said:

One thing though, this autonomous Uber car HAD a human behind the wheel in case of things like this, they both (the autonomous system and human operator) didn't had time to react. Just saying.

the driver may of thought he didn't have to? lol

 

I still get the feeling this was some crazy person that saw the car and decided to jump in front of it as a joke and it ended badly lol

[WkdPaul]

@asus killer @bcredeur97

 

That's a genuine concern though, that the operators might have too much confidence in the autonomous system. Not saying it's what happened since none of us were there, but it's definitely a possibility!

Edited March 20, 2018 by wkdpaul

[mynameisjuan]

3 minutes ago, samcool55 said:

another sign AI is potentially better at driving than humans.

If we replaced all traffic with AI we would lower collision rates to such an amount it would essentially be eliminating them altogether. 

[dalekphalm]

11 hours ago, Blademaster91 said:

The fact that there was a person behind the wheel could be sufficient evidence,the driver should have been able to at least slow down the car enough to avoid a fatal accident if the AI itself didn't do anything.  Driverless cars just aren't good enough yet to totally replace a human driver with other non-AI controlled vehicles on the road when the driverless car can't react how a human driver would have.

These are all, 100%, assumptions.

 

You have no idea if the AI did anything (tried to react, didn't see the danger, etc). You also have no idea if the driver could have even possibly reacted in time, even had he been paying 100% attention.

 

There is absolutely, 100%, without a doubt, not evidence to support that a human could have prevented this accident, had (s)he been in full control.

[Blademaster91]

9 minutes ago, dalekphalm said:

These are all, 100%, assumptions.

 

You have no idea if the AI did anything (tried to react, didn't see the danger, etc). You also have no idea if the driver could have even possibly reacted in time, even had he been paying 100% attention.

 

There is absolutely, 100%, without a doubt, not evidence to support that a human could have prevented this accident, had (s)he been in full control.

Yeah all anyone has are assumptions, as we don't know what the driver or the car did in this case, if the driver was even paying attention with hands on the wheel which should be a standard driverless car thing until AI is smart enough to avoid these accidents.

22 minutes ago, mynameisjuan said:

If we replaced all traffic with AI we would lower collision rates to such an amount it would essentially be eliminating them altogether. 

We just can't suddenly ban traffic especially when AI isn't up to the level isn't what a human driver could do,like as mentioned earlier possibly swerving onto the pavement when clear to avoid hitting someone crossing a street.

[dalekphalm]

Just now, Blademaster91 said:

Yeah all anyone has are assumptions, as we don't know what the driver or the car did in this case, if the driver was even paying attention with hands on the wheel which should be a standard driverless car thing until AI is smart enough to avoid these accidents.

Indeed. So why are you jumping to conclusions when there is no evidence to support it?

 

We can hypothesize different possible causes. But let's make sure we're clear in that these are all just equally likely possible causes.

 

You yourself said this:

Quote

If the driver behind the wheel took action they should have been able to stop in time, especially in a Volvo with its assisted braking system.

You state this as a fact, but it's not. There's no evidence to prove or even suggest this is true. How can you possibly know this? Short answer, is you cannot.

 

I apologize if this seems harsh, but so many people are jumping to conclusions and taking opinions as cold hard facts. Someone might read your misleading assumption as a stated fact, and come to the wrong conclusion because of it.

 

You MIGHT be correct. And if so, there will be hell to pay. But you could just as easily be wrong.

 

Let's wait for the facts, before we make definitive statements like that? We can make assumptions, discuss possible scenarios and outcomes, etc, until the cows come home, but we collectively as forum members need to strive better to ensure that our statements are clear about what is fact, what is known evidence, and what is just guessing.

[DildorTheDecent]

20 hours ago, leadeater said:

Linked article video says that it was a cyclist not actually a pedestrian.

Man, even the robots hate the cyclists.

 

Guess they're the first to go whenever there's an uprising. Nothing wrong with that.

 

I, for one, welcome our new robot overlords.

[mynameisjuan]

8 minutes ago, Blademaster91 said:

We just can't suddenly ban traffic especially when AI isn't up to the level isn't what a human driver could do,like as mentioned earlier possibly swerving onto the pavement when clear to avoid hitting someone crossing a street.

I never said we should...

 

But AI is powerful enough right now to do it. The reason AI has to be so good right now is because of humans. It needs to predict the erratic nature of humans on their cell phone, looking over their should, messing with the radio, driving through red lights, braking last second or for no reason...etc..

 

If we removed humans and had all AI everything would follow the rules and with GPS could even time and eliminate stop lights. Rules keep things in order...people dont follow rules well. 

[Blademaster91]

1 minute ago, dalekphalm said:

Indeed. So why are you jumping to conclusions when there is no evidence to support it?

We can hypothesize different possible causes. But let's make sure we're clear in that these are all just equally likely possible causes.

You yourself said this:

You state this as a fact, but it's not. There's no evidence to prove or even suggest this is true. How can you possibly know this? Short answer, is you cannot.

I apologize if this seems harsh, but so many people are jumping to conclusions and taking opinions as cold hard facts. Someone might read your misleading assumption as a stated fact, and come to the wrong conclusion because of it.

You MIGHT be correct. And if so, there will be hell to pay. But you could just as easily be wrong.

Let's wait for the facts, before we make definitive statements like that? We can make assumptions, discuss possible scenarios and outcomes, etc, until the cows come home, but we collectively as forum members need to strive better to ensure that our statements are clear about what is fact, what is known evidence, and what is just guessing.

The Volvo should have been able to brake on its own as they have some really good collision avoidance systems is all i'm saying,another if here,if it hadn't been disabled for Uber's self driving use.

Though I agree there are a ton of variables until we really know what happened,but how are you more right when you join the group agreeing that we should ban every driver from controlling their own car because they cause accidents themselves despite AI still causing deaths? An AI system should be better than a human driver in every instance and we aren't there yet IMO.

[dalekphalm]

Just now, Blademaster91 said:

The Volvo should have been able to brake on its own as they have some really good collision avoidance systems is all i'm saying,another if here,if it hadn't been disabled for Uber's self driving use.

Indeed, the Volvo systems can mitigate potential collisions. But there is no evidence that such mitigations would have prevented the accident. It might have made the accident less severe, perhaps. But also, we don't know whether the systems were even enabled or not.

 

The Volvo systems might well have been fully enabled.

 

Let's wait until the facts before we start claiming things like the Brake Assist + a human driver would have changed the outcome in any meaningful way.

Just now, Blademaster91 said:

Though I agree there are a ton of variables until we really know what happened,but how are you more right when you join the group agreeing that we should ban every driver from controlling their own car because they cause accidents themselves despite AI still causing deaths? An AI system should be better than a human driver in every instance and we aren't there yet IMO.

I'm sorry - but when did I agree with banning drivers? We are definitely not at the point where autonomous vehicles should take over 100%, and all human drivers should be banned. I happen to enjoy driving, though I wouldn't go so far as to call myself a gear head or grease monkey.

 

Furthermore, the idea that AI has to be better than humans in every possible way? No. They don't. They just have to be better than humans on average. And frankly, they already are.

 

If net AI accidents/deaths = lower than net human accidents/deaths, that's all the proof I need.

 

Autonomous vehicles don't need to be perfect. They just need to be better drivers than your average person. Yes there are situations where autonomous vehicles could fail, but do you have any idea how often humans screw up and cause accidents? It's a ridiculous and insanely high amount.

[mynameisjuan]

11 hours ago, dalekphalm said:

I'm sorry - but when did I agree with banning drivers?

You agreed with one of my post now we are apparently a group trying to ban people from ever driving cars. 

 

But on a serious note I want to drive, I enjoy it and will not want to give it up. 

 

11 hours ago, Blademaster91 said:

should ban every driver from controlling their own car because they cause accidents themselves despite AI still causing deaths?

• Nearly 1.3 million people die in road crashes each year, on average 3,287 deaths a day.
• An additional 20-50 million are injured or disabled.

So when was the last time you heard of AI killing a person in a car accident? 

 

In the time it took me to type this out and paste the statistics, a handful of people just died in car accidents. Just let that sink in. Until you realize that humans can be random and computers never will be then you will never fully understand the concept. 

 

Finally!!!!

https://www.theverge.com/2018/3/20/17142672/uber-deadly-self-driving-car-crash-fault-police

 

" Based on the footage, Moir said that the driver had little time to react."

“The driver said it was like a flash, the person walked out in front of them,” she said. “His first alert to the collision was the sound of the collision.”

She added, “It’s very clear it would have been difficult to avoid this collision in any kind of mode [autonomous or human-driven] based on how she came from the shadows right into the roadway.”

 

Some more info to at least pull blame away from AI. So it looks like the person tried to quickly cross the road coming out from a bush covered path that is actually prohibited to cross for this exact reason. Bush would easy block the view from the car and the driver. 

 

This makes me happy. A small bump in the road for AI but now looks like its still on track.  

[ItsMitch]

Quote

Moir also faulted Herzberg for crossing the street outside a crosswalk. “It is dangerous to cross roadways in the evening hour when well-illuminated, managed crosswalks are available,” she said.

So the victim walked across the road (not at a traffic light) and got hit by a car, why is everyone blaming Uber again? More of her fault for being a silly sausage and not looking where she's going while she's crossing the street. Did she forget to look both ways and keep looking while crossing the road?

[dalekphalm]

15 minutes ago, mynameisjuan said:

Finally!!!!

https://www.theverge.com/2018/3/20/17142672/uber-deadly-self-driving-car-crash-fault-police

 

" Based on the footage, Moir said that the driver had little time to react."

“The driver said it was like a flash, the person walked out in front of them,” she said. “His first alert to the collision was the sound of the collision.”

She added, “It’s very clear it would have been difficult to avoid this collision in any kind of mode [autonomous or human-driven] based on how she came from the shadows right into the roadway.”

 

Some more info to at least pull blame away from AI. So it looks like the person tried to quickly cross the road coming out from a bush covered path that is actually prohibited to cross for this exact reason. Bush would easy block the view from the car and the driver. 

 

This makes me happy. A small bump in the road for AI but now looks like its still on track.  

This is pretty much what my assumption of events were. A pedestrian unexpectedly stepped in front of the vehicle, at a distance where it was physically impossible to slow the car down enough to prevent injury or death.
 

At 38 mph (~61 km/h), the vehicle was driving at a fairly safe speed. Police indicated the speed limit was 35 mph (56 km/h), but Google Street View shows a posted speed limit of 45 mph (72 km/h).

 

Either way, even if the posted limit was 35 mph and they were going 38 mph, it's highly unlikely that the additional 3 mph was in any way a variable that made things worse.

 

I'd still like to wait for the investigation to complete before drawing any conclusions, though now there is evidence to support that the autonomous systems were not at fault.

[Satisfoxy]

To all the people saying 'the guy should of had his eyes on the road'. What is the point of self driving cars if you have to watch like you were driving it?

 

To all the people wondering what this means for the future. Lawmakers are going to side with the cyclist and we will never move forward with transportation. Anyone that has disabilities and mental problems that prevent them from driving will never experience true independence in their lifetime now.

 

To the joke about whether they're putting the car or the person in jail. I really hope its the car.

 

Honestly we need the law changed to favor drivers and before you start typing your angry reply, read the rest please. If we always go 'oh you had an accident, here have 5 years in jail' then we are just wasting resources. 5 years of food/water/shelter paid for by YOUR taxes, this person contributes NOTHING to the economy while in jail, when they get out they are marked for life, can't get any jobs that pay enough for food/water/shelter so that makes them angry and with the fact that their brain is now damaged from 5 years of no activity, poor food conditions, lead in the water at the jail, and asbestos in the walls. You have someone that is NOT ever going to contribute to society ever again and yet, AGAIN and AGAIN people think this is the best option for a car accident. Lets not make the person just CONTRIBUTE and help the economy by being forced to volunteer for 5 years, never having a day off for 5 years and working incredibly un-fun jobs, that isn't enough punishment!

[ItsMitch]

1 minute ago, Satisfoxy said:

To all the people saying 'the guy should of had his eyes on the road'. What is the point of self driving cars if you have to watch like you were driving it?

Because not all software is perfect and I'm fairly sure all self driving cars requires you to remain with your eyes on the road, in driver less mode or not.

 

1 minute ago, Satisfoxy said:

Lawmakers are going to side with the cyclist and we will never move forward with transportation

I highly doubt it since it seems to be the cyclists fault.

[maartendc]

27 minutes ago, dalekphalm said:

At 38 mph (~61 km/h), the vehicle was driving at a fairly safe speed. Police indicated the speed limit was 35 mph (56 km/h), but Google Street View shows a posted speed limit of 45 mph (72 km/h).

 

Either way, even if the posted limit was 35 mph and they were going 38 mph, it's highly unlikely that the additional 3 mph was in any way a variable that made things worse.

Why is the AI in the car allowed to go above the speed limit AT ALL though? Very concerning to me.

 

How do these cars know the speed limit anyway? Do they read signs? Do they go by stored data based on GPS, the way your navigation system indicates the speed limit? In that case, that data can always be outdated because speed limits change regularly.

 

Google street view photos can be outdated. Could be taken 5 years ago.

[Vode]

22 hours ago, Kobathor said:

Good thing those were written by one guy who had no idea anything like real self-driving cars would ever be a thing. Those rules are so outdated, I chuckle whenever I read them.

Read the book. The author knows the rules don‘t work. They don‘t even work in the story.

[Cookybiscuit]

20 minutes ago, SC2Mitch said:

Because not all software is perfect and I'm fairly sure all self driving cars requires you to remain with your eyes on the road, in driver less mode or not.

 

I highly doubt it since it seems to be the cyclists fault.

If the driver needs to pay attention and be ready to take over control any second it isn't a self-driving car.