I think artificial intelligence is going to have a huge impact on transportation (and everything) in the next few decades.
My guesstimate for time-frames are
1) Early 2020s: First fully automated computer-driven cars.
2) Early 2030s: Virtually all new cars are computer-driven.
3) Early 2040s: All vehicles on the road are computer-driven.
I see these events as having profound impacts. I'll skip all the way to the early 2040s, when I expect all vehicles on roads to be computer-driven. Here are some impacts I see:
1) Individual ownership of vehicles will be virtually eliminated. Most people in 2013 have cars that sit unused for over 20 hours per day. This is a complete waste of money, because depreciation is occuring, and taxes and insurance are being paid, even though the car is sitting unused.
2) People will obtain cars like we now get airline tickets, except the cars will often be ordered hours or even minutes in advance, rather than months or weeks in advance.One will state one's destination and time of departure into a smart phone or personal assistant, and options will come up, with different prices and different arrival times. Non-stop short arrival times will be more expensive, and multi-stop trips with long arrival times will be less expensive. Also, fancier and less-fuel-efficient cars will be more expensive, and plainer, more-fuel-efficienct cars will be less expensive. The car will take you from the door at one location, to the door at the other location. No parking in parking lots.
3) Accidents will be virtually eliminated, so insurance will be minimal, and paid for by the fleet owners.
4) Many cars will become incredibly small. Even Smart Cars and Cooper Minis have room for two passengers. In the 2040s, single-seat cars will be extremely common. They will likely be powered by batteries, rather than gasoline. This is because there will be no need for "range anxiety". That is, no car will come to you unless it knows it has enough battery life to get you to your destination and then to a charging station.
5) An intercity-trip (say Durham, NC to Charlotte, NC...a distance of about 150 miles) might involve getting into one car, then changing to minibus or large bus (or train), for the trip between a particular point in Durham and a particular point in Charlotte. Then there might be another switch to a car in Charlotte. The delay in these two changes of vehicles would be more than compensated for (compared to the current situation) by the fact that travel between Durham and Charlotte would be at well over 100 mph. The total time, door-to-door, would probably be under 2 hours.
6) Transportation will be much more electrified. Virtually all short-range trips will be by battery-power. Total U.S. gasoline usage would cut by more than half.
7) Parking lots would be eliminated. No parking lots at football and baseball stadiums. No parking lots at malls (which might not exist anyway). No parking lots at businesses.
8) Traffic congestion would be completely eliminated. Instead of traffic lights or stop signs, cars at intersections would simply pass within feet of one another at right angles, at full speed, because the computer controls between the cars would communicate which car crossed first. Speeds on freeways, even within cities, would be 70+ mph, at all times. Cars and buses would travel at these speeds within feet of one another.
9) Big-box stores like Walmart and Target would be eliminated. Grocery stores would be eliminated. Instead of going to the store, one would shop at a virtual store (on one's 80 inch organic light emitting diode TV), order what one wanted, and have it delivered by an automatic vehicle (which would coordinate with all other customer deliveries).
10) Home garages and commercial parking decks will be eliminated. Home garages will be converted to living space.
11) The total number of vehicles will be greatly reduced, because many vehicles will be running almost continuously (except when charging batteries). The number of mechanics will also be significantly reduced, because cars will run continuously, and get maintenance at more regular intervals.
It's going to be quite a change. (Hope I'm around to see it.)
My comments are quite a bit after this original blog article was posted, but here are my thoughts.
AI driven cars will change law-enforcement, especially pertaining to drinking and driving. Theoretically, one could be allowed to drink while they ride all they wish.
AI driven cars will allow people to use their PMD's and whatever our phones have evolved into by that time. No more worrying about texting and driving. Heck, the "driver" could watch a movie, TV show, or play video games the entire trip.
I do agree with Douglass's idea that many people (myself included) will still want more independence than pre-determined trips to places. We might not want to always travel via these fleet vehicles.
Also, something could be said as to the different kinds of fun some people enjoy in the backseats of their cars. Would that even be possible anymore with AI-driven utilitarian purposed cars? Part of the magic of independent driving is the freedom one has to do things spontaneously.
Conversely, what if one would want to just drive. I often will take drives to unwind and to think away from the home. I do not have a destination. I just drive. That appeal might not be available with pre-determined AI driven vehicles, or it would be substantially cut back. That is something I cannot say I support heartily.
Posted by: Beth | March 24, 2014 at 06:26 PM
Hi Chaeremon,
You write, "Computerized control of transportation always was and always is possible, wanted and in use. Yet this field of study is Operations Research and not artificial intelligence. Outside of your country, there are so many computerized transportation control systems, and each is patented..."
What I'm speaking of is computerized control of automobiles, trucks, and buses. Google now has cars that are completely driven by computers. There is a human sitting at the wheel, but the human does not touch the wheel, the accelerator or the brake:
http://www.nytimes.com/2010/10/10/science/10google.html?pagewanted=all&_r=0
http://www.pcmag.com/article2/0,2817,2402380,00.asp
The Google cars are not mass-produced, but most people expect mass-produced automobiles capable of fully computerized control (the driver never needs to touch the wheel, the brake, or the accelerator) within approximately one decade (i.e., before 2023):
http://www.forbes.com/sites/eco-nomics/2012/09/25/self-driving-cars-will-take-over-by-2040/
Posted by: Mark Bahner | May 21, 2013 at 12:42 PM
Hi Mark,
You write "artificial intelligence is going to have a huge impact on transportation"
Unfortunately artificial intelligence is just words on paper, and can never fulfil any promise. In particular, what you have not put in advance INTO the AI system will never, ever emerge sensationally (look here: invisible clothes, "it" thinks!). And every single case which was reported as success has, in real world scenarios, always suffered from the Fallacy of Small [episodic] Samples, so small that interpretation is dead easy for courted investors (and of course the media).
Alas, someone who wants to argue against my statement is cordially invited to solve the following problem first, by computers (with or without artificial intelligence):
- http://www.claymath.org/millennium/P_vs_NP/
thereafter you will have my fullest attention. As an insider and professional, all that I have around me are disastrously failed AI projects and the ever growing list of newly discovered unsolvable AI problems. My prediction is, that Cold Fusion will be realized long, very long, before artificial intelligence is 1 ct more worth than just words on paper.
But I'm not here for bashing. Computerized control of transportation always was and always is possible, wanted and in use. Yet this field of study is Operations Research and not artificial intelligence. Outside of your country, there are so many computerized transportation control systems, and each is patented $$$. Asking for research funds and putting the project description into the invisible clothes of artificial intelligence is just the usual attempt of getting rich and famous fast. You have written about this psychopathic way of 'science' in use by so many climate "research" projects. Ah, and the amount of AI to get space station ISS up, running and supplied is zero, zip, nothing, nil.
Posted by: Chaeremon | May 15, 2013 at 08:27 AM
Certainly, early adopters of the computer driven cars would be in large cities where people are used to public transportation, so they are used to the idea of riding in a vehicle that doesn't belong to them. The other, obvious early adopters will be the elderly. Currently, when family or the community force an elderly person to give up driving, it means the end of independence for that person. If computer driven cars are available, allowing people to retain their independence when they give up driving, then we could quickly see a large portion of the aging population give up private car ownership. This could accelerate the process; when younger folks see the advantages of giving up control of their vehicles, they become more likely to give up theirs.
Posted by: Douglass Holmes | March 31, 2013 at 09:38 PM
Hi Douglass,
You write, "In order for these predictions to come true, there must be a demand for the changes. I believe that too many people WANT to drive their cars. When the weather gets nice, I'll be riding to work on my motorcycle, because I want to, not because I have to."
I think a significant majority of people will demand that incompetent human drivers get off the road, when they see how good computers are. I think it's possible that a lane on a few highways might be left for human drivers, or that human drivers might be allowed to drive on relatively deserted roads out in rural areas. But I don't think people will accept being slowed down by human drivers. For example, it takes me about 20 minutes to get to work, because of traffic lights and stop signs. If my computer-driven car could go through all those intersections without even slowing down, I wouldn't accept that I would have to stop, just because some selfish humans wanted to drive. Or imagine people commuting into NYC from CT or the suburbs. If they could go 80+ mph all the way to and from work, they'd never give that up for the sake of others who wanted to drive.
You also write, "However, your vision is really just a more efficient version of public transportation -..."
In some ways, yes. But in other ways, no. I still think a lot of people will be riding alone. It's just that their cars will be shrunk down significantly, and they won't own their cars. For instance, you mention your motorcycle. I'd ride in a tiny car (such as a Smart car) if I knew I would never crash. In a computer-driven-vehicle world, I would be certain that the car never would crash, so I wouldn't mind if it was barely bigger than my body (i.e. no protection in the front, rear, or sides).
Posted by: Mark Bahner | March 27, 2013 at 11:29 PM
In order for these predictions to come true, there must be a demand for the changes. I believe that too many people WANT to drive their cars. When the weather gets nice, I'll be riding to work on my motorcycle, because I want to, not because I have to. Also, too many people desire their own space, which an automobile provides. In fact, for many people, their automobile is more important than their home. Many men feel that their car is more "theirs" than what little space they have at home.
Granted, it is possible that private car ownership may become so expensive as to force people to use these computerized fleet cars. However, your vision is really just a more efficient version of public transportation - and even in cities where public transportation is pretty good, people still drive their own cars.
Posted by: Douglass Holmes | March 26, 2013 at 07:06 PM