Economist: Self-driving cars offer huge benefits—but have a dark side

[JManc]

Originally Posted by cannedairspray

Woman dies in Arizona after being hit by Uber self-driving car

https://www.msn.com/en-us/news/us/woman-...ar/ar-BBKqSzV?li=BBmkt5R&ocid=spartanntp

This is only reinforces my point earlier about the technology's feasibility in real world day-to-day routine being further out than we anticipate. Or hope...

[dubu]

If driverless cars became a thing it would be like nyc in the 70’s with all the trains graffitied.

That might make driverless cars safer. If you see a graffitied car don’t walk in front of it

[Sun Belt]

Driving home from work yesterday, I heard on the radio that the police determined that the uber car was likely not at fault and that the accident was unavoidable. The pedestrian suddenly step out into traffic mid-block and the car was traveling at 40 mph.

The uber operator behind the wheel is a convicted felon, served 4 years for attempted armed robbery.

[Pedestrian]

Originally Posted by Sun Belt

Driving home from work yesterday, I heard on the radio that the police determined that the uber car was likely not at fault and that the accident was unavoidable. The pedestrian suddenly step out into traffic mid-block and the car was traveling at 40 mph.

The uber operator behind the wheel is a convicted felon, served 4 years for attempted armed robbery.

From the moment it happened, there was some emphasis on the fact the victim was not in a crosswalk. The issue here, though, is that autonomous vehicles cannot be given full use of the roads until they are capable of coping with most of these situations as most good human drivers would be.

Someone said yesterday that it's not whether autonomous driving is flawless but whether its better than human driving and that's true. But I think most human drivers, paying attention (as the standby driver of an autonomous vehicle may not have been), would slam on the breaks when somebody walked out in front of them and attempted to maneuver around them. I don't believe there's indication this vehicle did that. So the technology may not yet be as ready for prime time as we'd been thinking.

This strikes me as very close to the ball riolling into the street followed by the kid chasing it scenario posited as a hypothetical a few pages back in this thread. Some said the the technology was ready for that. Investigation of this crash will tell, but maybe not.

[Austinlee]

Originally Posted by JManc

Two words: Johnnie Cab..

Lol! Nice reference

[streetscaper]

Uber Crash in AZ - VIDEO INSIDE AND OUT


• Video Link

[streetscaper]

IMO, Couldn't see her at all until she showed up in the headlights, already too late at that point. I don't think a regular driver could have stopped.

[streetscaper]

Although, the lighting likely wasn't as bad as it appears in the video. Cameras, especially the tiny sensors you see in dash cams, don't have the dynamic range that human vision does.

[10023]

Originally Posted by JManc

This is only reinforces my point earlier about the technology's feasibility in real world day-to-day routine being further out than we anticipate. Or hope...

I disagree. This is coming much sooner than most people expect.

There will be accidents, of course, but I don’t think anyone is claiming that self-driving cars will never have accidents. But there are currently ~35k traffic fatalities a year in the US. Self-driving cars, if they replaced human drivers in terms of vehicle miles driven, could literally kill thousands of people per year and still be a vast improvement upon the status quo.

The Most Important Self-Driving Car Announcement Yet
Autonomous vehicles will transform urban life by 2020, if Waymo’s time line is correct.

ALEXIS C. MADRIGAL
MAR 28, 2018

On Tuesday, Waymo announced they’d purchase 20,000 sporty, electric self-driving vehicles from Jaguar for the company’s forthcoming ride-hailing service.

Waymo, Google’s sister company within Alphabet, held a press conference in New York for the unveiling of the vehicle, and most of the stories revolved around the luxury SUV’s look and feel.

But the company embedded a much more significant milestone inside this supposed announcement about a fancy car. With orders now in for more than 20,000 of these vehicles and thousands of minivans that Chrysler announced earlier this year, Waymo will be capable of doing vast numbers of trips per day. They estimate that the Jaguar fleet alone will be capable of doing a million trips each day in 2020.

You could quibble with their math (will it really be that many daily trips per car?) or their overall utilization rate (how many cars will be lost to maintenance per day?), but if Waymo is even within 50 percent of that number in two years, the United States will have entered an entirely new phase in robotics and technology.

The company’s autonomous vehicles have driven 5 million miles since Alphabet began the program back in 2009. The first million miles took roughly six years. The next million took about a year. The third million took less than eight months. The fourth million took six months. And the fifth million took just under three months. Today, that suggests a rate on the order of 10,000 miles per day. If Waymo hits their marks, they’ll be driving at a rate that’s three orders of magnitude faster in 2020. We’re talking about covering each million miles in hours.

But the qualitative impact will be even bigger. Right now, maybe 10,000 or 20,000 people have ever ridden in a self-driving car, in any context. Far fewer have been in a vehicle that is truly absent a driver. Up to a million people could have that experience every day in 2020.

2020 is not some distant number. It’s hardly even a projection. By laying out this time line yesterday, Waymo is telling the world: Get ready, this is really happening. This is autonomous driving at scale, and not in five years or 10 years or 50 years, but in two years or less.

...

https://www.theatlantic.com/technology/a...elf-driving-car-announcement-yet/556712/

[suburbanite]

Originally Posted by 10023

There will be accidents, of course, but I don’t think anyone is claiming that self-driving cars will never have accidents. But there are currently ~35k traffic fatalities a year in the US. Self-driving cars, if they replaced human drivers in terms of vehicle miles driven, could literally kill thousands of people per year and still be a vast improvement upon the status quo.

I think there's a big difference between pedestrian and driver fatalities. A driver accepts the risk of an autonomous vehicle failure when they choose it as a method of transportation. People are less likely to accept a new technology that puts someone who never consented at risk. An example being how much quicker second-hand smoking laws came into effect compared to things like plain packaging or other deterrents for the actual user.

Hard to estimate how much safer an autonomous vehicle is for pedestrians. They have instant response time (if it's picked up by the sensors), but they might lack a person's ability to predict human behaviour before it happens (like how a baseball rolling across the street could be a precursor for a child entering the road).

There were 5,276 pedestrian deaths from cars in 2015. 15% involved an impaired driver. Safe to say a large percentage of those fatalities could be eliminated. Much harder to determine the effect on the 34% that involved an impaired pedestrian. Or the 20% that involved children.

Public perception is a big hurdle in adopting a technology as widespread and life-changing as this. Even if we can statistically prove that autonomous vehicles will lead to x% decrease in fatalities for both drivers and pedestrians, all it takes is one incident of a stroller rolling into traffic to bring out the "think of the children" crowd.

[cannedairspray]

Originally Posted by suburbanite

They have instant response time (if it's picked up by the sensors)

This is almost completely the issue. If the exact same stimuli is correctly recognized by a human and a computer, of course the computer's response time and response in general will be better. But stimuli can be misinterpreted or missed altogether. And that has people concerned.

[JManc]

Originally Posted by 10023

I disagree. This is coming much sooner than most people expect.

There will be accidents, of course, but I don’t think anyone is claiming that self-driving cars will never have accidents. But there are currently ~35k traffic fatalities a year in the US. Self-driving cars, if they replaced human drivers in terms of vehicle miles driven, could literally kill thousands of people per year and still be a vast improvement upon the status quo.



https://www.theatlantic.com/technology/a...elf-driving-car-announcement-yet/556712/

Down the road, AI driven cars will be superior to human drivers but again, we are not there yet. Current technology is not advanced enough to fully replicate real world driving conditions and not require a human to intervene. This Uber accident is an example. The human "driver" was looking down at her phone and not monitoring the car and should have taken over to prevent hitting the pedestrian. The technology would have to be able to react to such a scenario on its own and avoid the accident without any human intervention.

[10023]

Originally Posted by suburbanite

I think there's a big difference between pedestrian and driver fatalities. A driver accepts the risk of an autonomous vehicle failure when they choose it as a method of transportation. People are less likely to accept a new technology that puts someone who never consented at risk. An example being how much quicker second-hand smoking laws came into effect compared to things like plain packaging or other deterrents for the actual user.

Hard to estimate how much safer an autonomous vehicle is for pedestrians. They have instant response time (if it's picked up by the sensors), but they might lack a person's ability to predict human behaviour before it happens (like how a baseball rolling across the street could be a precursor for a child entering the road).

There were 5,276 pedestrian deaths from cars in 2015. 15% involved an impaired driver. Safe to say a large percentage of those fatalities could be eliminated. Much harder to determine the effect on the 34% that involved an impaired pedestrian. Or the 20% that involved children.

Public perception is a big hurdle in adopting a technology as widespread and life-changing as this. Even if we can statistically prove that autonomous vehicles will lead to x% decrease in fatalities for both drivers and pedestrians, all it takes is one incident of a stroller rolling into traffic to bring out the "think of the children" crowd.

Human-operated cars hit pedestrians too. Lots of them, as you point out.

As for public perception, I hope that our institutions are strong enough to overcome that. You can’t allow the “average Joe” to decide these things. Decisions about autonomous vehicles, and anything else, should be based on quantitative data and informed views.

[10023]

Originally Posted by JManc

Down the road, AI driven cars will be superior to human drivers but again, we are not there yet. Current technology is not advanced enough to fully replicate real world driving conditions and not require a human to intervene. This Uber accident is an example. The human "driver" was looking down at her phone and not monitoring the car and should have taken over to prevent hitting the pedestrian. The technology would have to be able to react to such a scenario on its own and avoid the accident without any human intervention.

I still think you underestimate how quickly the technology is advancing, though. It gets twice as good every 6 months.

And it’s already probably better than the average human driver. The average American is a terrible driver.

[JManc]

Originally Posted by 10023

I still think you underestimate how quickly the technology is advancing, though. It gets twice as good every 6 months.

And it’s already probably better than the average human driver. The average American is a terrible driver.

I don't. I'm a futurist, I've been paying attention to this sort of thing since I got into this industry; Moore's law, accelerating change, etc. but we're still a ways away from full implementation (Level 5) of autonomous vehicles. I agree the technology is progressing fast (my 2016 Acura is a Level 1 and we're up to Level 4's) but the technology (AI) required is also extremely sophisticated. Perhaps in a decade, we'll be there...

[10023]

It’s cheaper sensors that are driving it, more than anything else. At least according to my friend at Google.