Driverless car

[Philosophy Explorer]

This video says a lot.

Currently the biggest issue on the internet is should the car kill pedestrians or its passengers in a swerving accident? (I favor a third option where nobody has to die at all).

Enjoy the ride:

https://techcrunch.com/2016/10/20/watch ... n-its-own/

PhilX

[Hobbes' Choice]

Failure to stop at a STOP sign.
Thought the steering was a bit jerky.

What could possibly go wrong?

And when it does who do you blame?

[Philosophy Explorer]

Failure to stop at a STOP sign.
Thought the steering was a bit jerky.

What could possibly go wrong?

And when it does who do you blame?

I've been asking the same questions. Currently (finally) government is stepping in, so they with the courts will figure out who to blame. Some data seems to suggest the computer does a better job, in terms of safety, than human drivers and the technology keeps improving (Tesla Autopilot 2.0 e.g.)

I feel there isn't enough convincing data as to how safe these cars will be. And I want to hear from government, not the auto manufacturers, on the studies being done with these cars as to their safety.

It's probably going to take another five years to derive worthwhile studies in this new industry. In the meantime, there'll be plenty on the internet to read up on.

PhilX

Just caught this article:

http://www.theverge.com/2016/10/19/1334 ... egulations

[Philosophy Explorer]

As a note, Tesla is the leader in driverless technology (they certainly get a lot of press coverage).

This article explains the steps Tesla is currently undertaking:

http://mashable.com/2016/10/19/tesla-el ... 2aSfSgFqq0

One more article:

https://www.engadget.com/2016/10/20/aut ... ing-rules/

PhilX

[thedoc]

What kind of program will they use to determine if an obstruction is human or not, and how small an object will trigger a response. Is the car going to stop or swerve for every moth that flies into the lights at night? Humans can recognize other humans no matter how old or what size they are, how will they program the computer to recognize a human being, it's not as easy as you might think.

[Philosophy Explorer]

What kind of program will they use to determine if an obstruction is human or not, and how small an object will trigger a response. Is the car going to stop or swerve for every moth that flies into the lights at night? Humans can recognize other humans no matter how old or what size they are, how will they program the computer to recognize a human being, it's not as easy as you might think.

As far as moths, butterflies and birds go, I'm sure the computer can be programmed to disregard small sizes. How about oil slicks that had me spinning on a curve? The Tesla car will allow a human to take over in an emergency (if there's enough time to react). So far limited stats indicate the technology is safer (but it takes more than what's public to convince me as I've already indicated).

With a good internet search, I'm sure most questions can be answered.

PhilX

[thedoc]

As far as moths, butterflies and birds go, I'm sure the computer can be programmed to disregard small sizes. How about oil slicks that had me spinning on a curve? The Tesla car will allow a human to take over in an emergency (if there's enough time to react). So far limited stats indicate the technology is safer (but it takes more than what's public to convince me as I've already indicated).

With a good internet search, I'm sure most questions can be answered.

PhilX

What happens if the human is too intoxicated to drive and that is why they elected to have the car take over, who is the computer going to relinquish control to?

Some animals are as big as small humans, and some animals are much bigger than most humans, how is the computer going to tell the difference? Is it going to stop the car and let a wild bear get into the car with the humans?

Myself, I wouldn't trust a computer to make that decision for me.

[Philosophy Explorer]

To show you how far driverless technology has taken hold in the US, read this Uber story:

"Self-driving Uber involved in collision downtown
Thu 20 Oct 2016 05:23:40 PM EST
Thu 20 Oct 2016 05:27:58 PM EST

PITTSBURGH —A self-driving Uber vehicle was involved in a collision Wednesday afternoon in downtown Pittsburgh according to the ride sharing company.

'We're aware of the incident and still gathering details,' said Uber spokesman Craig Ewer.

The collision happened around 4 p.m. near the intersection of Coffey Way and Sixth Avenue. It remains unclear who was at fault, but a photo posted to Twitter shows the other vehicle behind the Uber in what appears to have been a rear-end collision.

'We have accidents that happen in the city almost on a daily basis that involve serious injury and unfortunately every week that involve death. There’s going to be accidents with self-driving cars just as there are with human-driven cars. We’re going to be watching all of these reports as they come in to see if there’s anything that is a common element. Just as Uber’s watching it, we’ll be watching it as well,' said Mayor Bill Peduto.


Peduto has been an ardent supporter of Uber's operations in Pittsburgh.

'Accidents happen. That could have been someone in the driver’s seat and someone backed up. Yeah, so I mean if it wasn’t Uber’s fault, it was the other driver’s fault,' said Stacey Jenkins-Jones, who drives a regular Uber.

Most people talking about the incident downtown Thursday were not very concerned.

'If it’s the fault of the person actually driving their car. Don’t blame Uber for it. If they have the technology, everything is fine,' said Clarence Jones.

Pittsburgh police said they had not been made aware of the crash.

Danielle Greaves, who posted the photo to Twitter, declined an interview request from Pittsburgh's Action News 4 about the incident.

The incident marked what appears to be only the second documented crash involving the self-driving Uber vehicles since their debut (picking up passengers) in early September."

[Philosophy Explorer]

As far as moths, butterflies and birds go, I'm sure the computer can be programmed to disregard small sizes. How about oil slicks that had me spinning on a curve? The Tesla car will allow a human to take over in an emergency (if there's enough time to react). So far limited stats indicate the technology is safer (but it takes more than what's public to convince me as I've already indicated).

With a good internet search, I'm sure most questions can be answered.

PhilX

What happens if the human is too intoxicated to drive and that is why they elected to have the car take over, who is the computer going to relinquish control to?

Some animals are as big as small humans, and some animals are much bigger than most humans, how is the computer going to tell the difference? Is it going to stop the car and let a wild bear get into the car with the humans?

Myself, I wouldn't trust a computer to make that decision for me.

Better the computer than the drunk, on average. With the bigger animals, some of them know enough to get out of the way. The others wind up as road kill or injured which happens anyways with human driving.

Are you ready to say without a doubt that driverless technology is more dangerous than human driving? Do you have evidence?

PhilX

[thedoc]

As far as moths, butterflies and birds go, I'm sure the computer can be programmed to disregard small sizes. How about oil slicks that had me spinning on a curve? The Tesla car will allow a human to take over in an emergency (if there's enough time to react). So far limited stats indicate the technology is safer (but it takes more than what's public to convince me as I've already indicated).

With a good internet search, I'm sure most questions can be answered.

PhilX

What happens if the human is too intoxicated to drive and that is why they elected to have the car take over, who is the computer going to relinquish control to?

Some animals are as big as small humans, and some animals are much bigger than most humans, how is the computer going to tell the difference? Is it going to stop the car and let a wild bear get into the car with the humans?

Myself, I wouldn't trust a computer to make that decision for me.

Better the computer than the drunk, on average. With the bigger animals, some of them know enough to get out of the way. The others wind up as road kill or injured which happens anyways with human driving.

Are you ready to say without a doubt that driverless technology is more dangerous than human driving? Do you have evidence?

PhilX

So far there is little to no evidence that driverless cars are safer, just a lot of very controlled and limited experiments that prove nothing in the real world. Human drivers have a relatively good record so far if you consider miles driven per accidents that have occurred. Drunk driving is a recognized and illegal activity, and even though it happens you only need to enforce the laws in effect to control it. That is the one case where a driverless car can be considered safer.

[Philosophy Explorer]

What happens if the human is too intoxicated to drive and that is why they elected to have the car take over, who is the computer going to relinquish control to?

Some animals are as big as small humans, and some animals are much bigger than most humans, how is the computer going to tell the difference? Is it going to stop the car and let a wild bear get into the car with the humans?

Myself, I wouldn't trust a computer to make that decision for me.

Better the computer than the drunk, on average. With the bigger animals, some of them know enough to get out of the way. The others wind up as road kill or injured which happens anyways with human driving.

Are you ready to say without a doubt that driverless technology is more dangerous than human driving? Do you have evidence?

PhilX

So far there is little to no evidence that driverless cars are safer, just a lot of very controlled and limited experiments that prove nothing in the real world. Human drivers have a relatively good record so far if you consider miles driven per accidents that have occurred. Drunk driving is a recognized and illegal activity, and even though it happens you only need to enforce the laws in effect to control it. That is the one case where a driverless car can be considered safer.

And there is little evidence that driverless cars are more dangerous. The big question is how much evidence is needed to tell either way? Another question is what is the optimum point for the technology to be at its maximum efficiency? Also if it were 100% driverless cars, would that be much better than say 50% driverless cars and 50% human-driven cars?

PhilX

[TSBU]

[Philosophy Explorer]

I pointed out earlier in this thread about my concern with doing a proper study and the size of the sample.

This article just came out:

http://m.phys.org/news/2016-10-universe-rateor.html

This comes from a completely different field: cosmology. But the principle is the same. This new study calls into question the existence of dark energy due to looking at a larger sample size (I'm sure this new study will cause an uproar in science).

PhilX

[thedoc]

A TV show has little to no basis in reality, that car was operated by a human operator remotely. If you believe a TV show represents reality, there is no hope for you.

[Philosophy Explorer]

A TV show has little to no basis in reality, that car was operated by a human operator remotely. If you believe a TV show represents reality, there is no hope for you.

For me Doc, I took this post as an example where the idea of a driverless car comes from science fiction.

PhilX