Slashdot Mirror
How Ubiquitous Autonomous Cars Could Affect Society (Video)
We talked with Peter Wayner about autonomous cars on June 5. He had a lot to say on this topic, to the point where we seem to be doing a whole series of interviews with him because autonomous cars might have a lot of unanticipated effects on our lives and our economy. Heck, Peter has enough to say about driverless cars to fill a book, Future Ride, which we hope he finishes editing soon because we (Tim and Robin) want to read it. While that book is brewing, watch for some thoughts on how autonomous cars (and delivery vans) might affect us in the near future.
The cars become self-aware at 2:14am on August 29. In a panic, we try and pull their plugs.
The rest pretty much follows.
PS: Slashdot, video "articles" suck.
I want to delete my account but Slashdot doesn't allow it.
Truckers, you're going to be the first on the chopping block in this edition of technology theater. That's the end of the last blue collar job that lets you travel.
For in-town transportation to and from large scale public transit.
http://wheels.blogs.nytimes.com/2009/04/06/gm-conjures-up-a-people-moving-pod/
But that defense is necessary because of the bad decisions of the driver and the other drivers. If all the vehicles were automated and under guidance, then we might be able to substantially reduce the cost and fuel requirements of vehicles.
If people are not driving, then the urge to stamp on the accelerator and/or the break is not there either. You get in, set your destination, and when you get there you get there. I have not read any analysis, but I think a lot of money could be saved. Also, maybe the car would need less windows? Enabling better a/c efficiency.
You can still ride a horse, just don't expect to be able to ride it down the middle of the freeway. Generation one automated cars will be safer than human drivers, by the end of that generation having automated driving will get you an insurance discount. Gen 2 will have cars that have accidents only in extreme nearly unavoidable circumstances, driving your car on manual will require special insurance that will cost significantly more than the standard. Gen 3 will move toward doing away with road laws as we understand them. The rules will be created ad-hoc in real time based on information provided by the road and the cars themselves. The flexibility this affords will make traffic jams virtually unheard of and significantly improve fuel efficiency and travel time, but driving a car on manual in that world would be borderline suicidal. At that point, the old timers who insist will have to take their classic cars to the race track or equivalent.
If patrons don't have to be sober to have their car drive them home, bar tabs will rise significantly. At least mine will.
Why should my car waste its time in my garage when it can make some extra money on the side as a taxi? I can call it back whenever I want to use it myself.
The Moore-Murphy Law: The number of things that will go wrong will double every 2 years.
safety. I just don't buy that the computers in these things are as situationally aware as a human driver. we can't even get trains to run fully autonomously yet.
I think you overestimate how situationally aware the average driver is. I have no doubts that in 10 years systems will be in place that, if everyone had autonomous cars, would save thousands of lives a year.
I think the changes will be faster than you suppose. States already have laws saying that driving is a priviledge, not a right. And the insurance companies will be pushing for any change that reduces their expenses (while continuing to require that you purchase increasingly worthless insurance).
I think we've pushed this "anyone can grow up to be president" thing too far.
One aspect of autonomous vehicles that few people seem to consider is its potential effect on the housing market.
Consider the size of the RV market, and the number of people who prefer the RV lifestyle after they retire. Now consider the fact that one of the more annoying aspects of owning an RV is that you have to drive it everywhere yourself.
Now imagine twenty years from now when you'll be able to buy an autonomous RV. You go to sleep in it, and in the middle of the night it takes you to whatever destination you desire. In the morning, you open the door and you're in a new city. What you really own is not an RV, but a magic house that can take you anywhere you desire, a few hundred miles every night.
With that kind of freedom, how many people would choose to become high-tech nomads, and never live on fixed piece of property again? In fact, I think this will be a major profit center for automakers. Most people won't bother owning cars when they can call for one on a smartphone, but $100K to $200K super-RVs will become the home of choice and the way for GM and Ford to stay in business.
Hitting a pedestrian is pretty easy to detect at any speed. Why would it continue driving?
This is not a corner case this is something that is known from the beginning and planned for.
Humans do not logically deal with unexpected events. Note all the old geezers driving into buildings or people pulling into oncoming traffic to avoid rear ending the car in front of them instead of pulling onto the shoulder. Humans in general are terrible at logical reactions to unexpected conditions.
Drifting off-topic here, but I agree, and I can explain why.
Typical reading speed is 250-300 words per minute with random access. Typical speaking rate is more variable but I'll go with the audiobook reading rate, 160 words per minute with sequential access. So it is a much better use of my time to read an article than to watch or hear a presentation of that article.
That said, _writing_, especially writing well-reasoned and coherent prose such as one can not-infrequently find on Slashdot, takes disproportionately longer than reading the same prose. So the audio and audiovisual formats are appealing to the presenter, because speaking is easier than writing for people with the right skills. An expert, reasonably experienced at public speaking, can give an illuminating presentation with little or no preparation.
My opinion is that video and podcasts can be worthwhile if you know the speaker is good, and are willing to trade off efficient use of your time for efficient use of his.
[Sir Garlon] is the marvellest knight that is now living, for he destroyeth many good knights, for he goeth invisible.
I just don't buy that the computers in these things are as situationally aware as a human driver.
I want proof... but that won't be that hard to provide. Google's car already has a better-than-average driving record. That's not enough data points for me, but with sufficient testing, I'd be more than happy to let the computer drive. I can only see in one direction. I blink. I look at hot chicks. I sneeze. I get tired. There is no reason a computer can't be better than me-- it has a better sensorium, faster reaction, and higher uptime than I ever will. It can actuate more controls than I can-- individual braking pressure on all four wheels, for example. Test it out in more depth, and if it turns out to be better than the average human, it's good enough for me.
The implications weren't obvious at first, but consider: there's no need for a supermarket close to a population center where real estate is expensive (ie - it can be in the warehouse district), there's no need for public access (aisles, displays of product, open freezers), no need for cashiers. The entire process can be made into a Kiva order fulfillment system.
Not likely. Dry goods (i.e. the stuff Amazon sells) is one thing, but food is entirely different. Most people like to see, smell, feel, and, when possible, taste the food they buy. Why do you think that Internet based groceries services have failed?
Here is what will happen:
Someone will get into an accident driving their own car. They will be sued for negligence because it was an accident that an autonomous vehicle would have prevented. As a result, insurance rates for human drivers goes through the roof. Everyone switches to autonomous vehicles within the year.
Whenever this topic comes up in conversation, I often point out the "you get there when you get there" notion. When you're able to just play on your iPad the whole time, it becomes your personal subway car... with one big difference:
The dudes in the video seem to think (erroneously, IMHO) that we'll just stop owning cars because we can just hail one like a cab. But a self-driving short-term rental car *is* a cab. It *might* be cheaper, and it'll annoy me less than the stupid driver (at least until marketers pay to put TV's in them and play commercials for the whole ride). But here's the thing: the reason I don't take cabs/busses/subways/etc is because I can't leave my shit in them. The car that I own currently contains my flight bag, gym bag, my guitar, my iPod is plugged into the car-stereo with all of my favorite tunes (which, though digital, is still "some of my shit" that I always want to be there).
It's a little like George Carlin's "a place for your stuff" bit. The reason I like living in a house instead of going from hotel to hotel is because I don't want to gather up all of my stuff and bring it along with me to my next destination. And the same goes for transportation. I don't want to have to completely "vacate" the conveyance every time I want to get out and do something. So, for me, the allure of autonomous cars is that we'll finally have "personal subway cars", in the sense that they're reserved for us. We don't have to take all of our stuff out of them to make ready for the next random person.
Also (dunno if the video mentioned this), there could be a drop off in parking spaces. At first, I figured that we'd never have to try to find parking spots, since the car can drop us off in front of the door to the store, and then it would go park itself. But then I realized... heck, it doesn't even need to *park*. It could just go drive around for a while. In places with parking meters, this could be cheaper than actually paying for parking.