Don’t forget that the driverless car phenomena will not occur overnight. Besides it will be evolving constantly as technology keeps changing. Cars will talk to each other. Eventually pile-ups may become a thing of the past. Governments will always be seeking different sources of revenue as old sources dry up. As oil wells go dry, we’ll have the anti-gravity car which will eliminate the need for expensive roads for example.
Um… okay. Glad to hear that anti-gravity won’t require a power source.
I believe less traffic fatalities will have hidden costs to the economy, freakonomics style. 10,000-15,000 less funerals will have an effect on the funeral parlors. Also less lawyers will be required for handling estates of the deceased.
Besides fatalities, the number of injuries will all be greatly reduced as well as the severity of those injuries that do occur. Who will need all the physical therapists, doctors, surgeons, x-ray technicians and machines that go ping if there are 200,000-400,000 less injuries per year?
Link to the article: Will driverless cars cost us more than they’re worth? - The Straight Dope
Don’t leave out that driverless cars will rock the insurance world. In the end I think that will be the biggest incentive for people to give up their cars for auto autos–it will save them a bundle in insurance costs.
The “if Microsoft made cars” nightmare will be reality.
The same could be said for finding a cure for cancer and other major diseases, ie what about all the doctors, physical therapists, etc? Curing paralysis? Easy organ transplants? Should we not do any of this because of the hidden costs? No. It means we’ll find something else to replace the displaced economic pieces. Or we’ll make it so people can “make” their own food, clothes, housing at little cost (like with a 3D printer or that Trek thingamajig that made food and other items) so they don’t have to pursue a job for pay
Every last one of those people is still going to die eventually.
The question posed to Cecil and his answer focused largely on lost government revenue, from all those thieving shake-down tickets they won’t be giving out any more. No matter. They’ll just have to find other things to tax, and they will. They’ll find something to tax about electric cars. They’ll find something to tax about self-driving cars. The idea has been kicked around before of taxing drivers for their mileage on public roads. That could come up again. They’ll invent a host of new petty infractions, related to transportation or otherwise, with which to shake us all down.
If you drive a car, I’ll tax the street,
If you try to sit, I’ll tax your seat.
If you get too cold I’ll tax the heat,
If you take a walk, I’ll tax your feet.[indent]-- John, Paul, George, and Ringo[/indent]
As an old science fiction reader (since the late 1940’s), I think that driverless cars will lead to the long-delayed Flying Cars! The biggest objection to flying cars, personal autogyros, jetpacks, etc., even if technologically possible, is the difficulty and the dangers of three-dimensional traffic control. If self-driven cars are perfected, I believe the technology for self-driven aircraft will follow.
I’m kind of baffled that the questioner (and by extension, Cecil) seems to be treating a decrease of speeding-ticket revenue as a loss. Isn’t that a gain? It’s not like governments get ticket revenue from the magical money fairy.
This is what’s called the broken window fallacy. Just because people are kept busy fixing problems, doesn’t mean those problems are good.
Driving home this evening on a two-lane road (four lanes altogether), I came around a curve and saw a semi-trailer with a pup trailer stopped at a traffic light. So I changed lanes (no one beside or behind me). When the light turned green, I was off again, and switched to the original lane after passing the truck.
Would or could a self-driver do that? I doubt it would now.
But knowing nothing about self-driving vehicles, I’ll state for the helluvit that a self-driver probably would stop behind the dual-trailer truck, then crawl along behind it until the truck turned off. If it didn’t turn, all the occupants of the self-driver would be dead from old age before they got home. Being a passenger in a self-driving car would be like having Looney Tunes’ Granny as a cab driver. Especially in winter.
I’d like to see a story about self-drivers coping with snow and ice, or better yet, a video. Let’s see a self-driver in a blizzard. Or on ice and snow on a highway exit. Or simply stopping on ice and snow, and to see whether the car has enough brains to steer onto the snow to stop. Or to know it isn’t snow but a child in a white snow suit lying on the road.
All the videos I’ve seen are of self-driving cars and semis on freeways in summer or in locations where summer is 12 months long (as is the case with electric cars).
Self-driving cars may be as much about the future as flying cars were.
Driverless Car 2.0 will surely be programmed to always seek the fastest lane of travel, traffic permitting, so that you’re not stuck behind trucks or buses.
And if you don’t like its driving, stay off the sidewalks.
Perhaps more to the point will be the number of jobs (tax-paying jobs) displaced by automation - in the US in particular I believe you have a Sherpas of road freight. Bearing in mind that some sources reckon this will be the sector that gets automated first (lots of capital to do it, mainly freeway driving rather than cities, more dangerous when the driver gets tired) you might find a lot of otherwise unskilled people not only no longer paying tax but needing to fall back on social services.
The argument is that automation will open up other jobs that these hapless individuals can fill. Not sure I’m convinced, although I imagine that historically that is what has happened.
Municipalities do use the money collected from speeding tickets and the like as a part of their revenue. For them, it is a loss. Sure, us people paying the fines get to keep more money, at least until the government finds a way to tax it out of us another way.
For grins, I googled Houston city budget to take a look at real numbers. From the 2014 budget, they have actual numbers from 2012 listed. It’s a long pdf file, but sorting through, I found the line item under revenue.
Some confusing numbers came up in this, and without reading through the whole thing to make sense of it all, I went down to the Municipal Court section and found moving violations account for $19 Million dollars. For comparison, the Fire Department made $33 Million in ambulance fees, and the Police Department made $2 Million dollars in vehicle tow away fees. See pages XV - 19 and XV - 20 (pages 511 and 512 of the pdf). Property taxes came in at $792 Million. Then I went to the bottom of that column for total revenue - $3.5 Billion. Hmmm. $19 Million doesn’t seem that big a chunk of the city revenue. Licensing and whatnot fees to Google for their autodrive system or whatever would easily replace that revenue.
It would be nice to not have those problems; however, some of the people making their living addressing those problems would need to find a new way to make a living. But that’s just basic give and take.
Why wouldn’t it? You have two lanes moving in the same direction, no traffic from behind to block the lane change, and even if there were traffic approaching from the rear, there’s no reason to think the programmers would ignore that in their design.
Why? Because you are not smart enough to figure out how to program it? *You *recognize the problem, I’m sure the people who are programming cars to do things like interact at highway speeds to reduce dead space and navigate lane changes on the interstate to make most efficient flow have considered the possibility that one of the cars might be on a 4 lane country road with stoplights and need to navigate around a slower truck and trailer.
That’s probably a fair concern. Weather provides a new series of challenges above and beyond dry, clear road driving. Perfecting travel on dry roads will not be sufficient - rain poses a significant challenge to driving above and beyond normal dry roads. Hydroplaning is a risk that is hard to quantify and anticipate. Snow and ice are as well.
It is, and I remember reading something a few weeks ago along the lines that they’d realized they’d need to add in other sensors other than visual to help cope with that. However (a) computers can use other sensors (LADAR, RADAR, ultrasound etc) and make use of these in real time the way that human drivers can’t and (b) they can analyze and respond to the situation much faster than we can, even when we’re not panicking. A classic example would be that of a skidding car - most people in the UK have never been in a bad skid so probably won’t react appropriately (and in fact already use somewhat less complex computers (in the form of ABS) to mitigate this). A computer can do complex maths almost instantaneously to work out exactly what best to do to minimize damage, rather like Deep Blue checked ALL its options when playing Kasparov.
(And before anyone points out, apparently the thing that broke Kasparov’s spirit was that he couldn’t work out Deep Blue’s strategy - which turned out to be because there were bugs in the software and Deep Blue wasn’t playing entirely logically. So I’m not saying this is just around the corner! Not that a computer driver has to be better than an F1 driver, just better than most of the people that it replaces)
I think these unresolved questions Cecil mentions are really quite asinine.
No, there will not be “cheat codes” (OTOH if he means the danger of malicious software then we can discuss that, but if so this wasn’t a good way to phrase it). And the reason speed limits exist is because they are what is considered the maximum safe speed to drive at i.e. contrary to sitcom logic, you should not speed when your wife is in labor because you may injure her and others and in the process not get to the hospital in time.
You may ask well ask “Will it run red lights if you really, really need to pee?”. No, no it won’t.
The “no-win situation” I think is the same, though here at least we need to at least try to take the question seriously, just because of the frequency with which it is posed.
We can ask ourselves what do good human drivers do in such situations?
I would say they do the following in order of priority:
- Try to regain control of the vehicle if it has been lost (e.g. “steer into the skid”)
- Stop the car
- Swerve to avoid obstacles that are too close to stop in time
Note I’m not including “make a judgement of whose life is more important” because there is no time for such things. An automated car could be similar, only it would virtually always know a safe path to swerve to.
But generally it’s hard to even think of scenarios where this could happen. e.g. if an automated road is rounding a blind corner on a mountain pass, and suddenly there’s someone walking in the road, then the swerve option may be out and a collision unavoidable. But then we ask ourselves at what speed an automated car would round a blind corner on a mountain pass, and the answer is likely to be: “slow enough that it could stop in time even if it encountered a pile of rubble blocking the road”.
Recently, I watched a whole bunch of videos on YouTube about Google’s self-driving car project. Just enter google self-driving car in the YouTube search box to find a whole bunch of them. Many are just news reports with little more than a few sound-bites. Forget those. Find some of the longer more technical ones. While it’s all still preliminary, some of it is pretty impressive.
The automotive and software engineers have thought about this a whole lot more than any of the rest of us. They are aware of a whole lot more problems to address than all the rest of us can think of. Dealing with snow, ice, rain, and night-time are high on their list.
They also have a whole lot more data to work with than is obvious. When you look at Google Maps, do you have any idea how much more data is layered in there, beyond what you see? They have the locations of road signs, buildings, every bus stop, every traffic light, all the individual lanes in every road and the markings in them all. They are programming their cars to be able to recognize those things in real time so they aren’t tied to known data in databases. They are studying human psychology to understand how real drivers behave, and how their cars should behave. For example, when stopped at a four-way stop, the car will start creeping out into the intersection to claim its right-of-way, knowing that otherwise it will never get a chance to go. The aggressiveness of its driving is an adjustable parameter. For now, they set it to “low aggressive” for driving around Mountain View, and “high aggressive” on their test track.
ALL the data read by ALL the sensors is recorded, as well as all decisions the car makes. They have a database of every input and every decision since they began the project. They can replay entire test drives in simulation, tweaking various parameters and programmed strategies to see what happens. They have cases where they’ve seen cars driving the wrong way; on the off-ramps and off the on-ramps. Like the proverbial experienced doctor, they’ve “seen it all”, or quite a bit of it, and they’re working on strategies to handle it all.
Where do you think the energy for a Trek-style replicator will come from? Narnia?
Hyper space — another dimension. Jeez. Everybody knows that!