David R. Henderson  

Twitter vs. Flying Cars

Bastiat on Baseball... How U.S. Steel Helped Break Do...

You've probably heard that line from Peter Thiel:

We wanted flying cars, instead we got 140 characters.

That line expressed his disappointment about how technology has turned out.

Now, when rich libertarian entrepreneurs, especially ones I admire as I do Thiel, talk, I listen. I also evaluate.

So does Joshua Gans.

On flying cars, Gans writes:

Let's start with flying cars. We should question the premise of whether we want flying cars. We would like quicker and less constrained options from getting from A to B but flying around in a car, while it may be a representation of the future with such freedom, when you think about it, cannot really deliver that value, at least in cities. Indeed, the fact that flying cars do not exist where they might be of value -- in less dense areas -- leads us to question whether people want them at all. This is at least enough to give us pause on the issue.

And Twitter? Gans writes:
Twitter is not just 140 characters. It is a new communications network. The best example of this is how you can communicate with customer support of companies. When you have a problem with, say, an airline, it is often easier to tweet about it and wait for the airline to respond than to contact the airline's support number. Now this is currently an advantage from congestion but, in reality, you can see how it may be superior overall. First, complaints are limited to 140 characters. That is a feature, not a bug. That means if a company is getting lots of complaints about something, it can see it really quickly. Second, it is easier to find a way of communicating with a company via Twitter than searching for support numbers. It is just straightforward. Twitter is an address book with easy search. My point here is that Twitter is a new communications protocol and more than just social media. That puts it on a path to something more than just the trivial.

Comments and Sharing

COMMENTS (25 to date)
Handle writes:

What an apt coincidence that Twitter resembles fritter as in 'to fritter away'. We've made astounding progress in addictive diversions, but breakthrough enablers for the masses they ain't.

We seem to be living the 'compound interest' age of technological progress. Things are still getting better over time by the accumulation of small but steady refinements. But, save for the diversions mentioned above, you don't see a lot of dramatic existing-order-disruptors.

rapscallion writes:

We've had the tech to make cars that go up in the air since the 1930's. The problem is we're not likely to have the tech to make them come down safely or keep them from crashing into one another any time soon.

Douglass Holmes writes:

Almost all technological progress is an accumulation of small refinements. The internet is just a continuous accumulation of improvements in data transmission, storage, and processing technologies. Automobiles evolved from carriages, railroad, and other technologies (including bicycles). Airplanes would not have been practical if not for the internal combustion engines that the automotive industry made available.
Handle, can you provide us with an example of what you consider to be a dramatic, existing-order-disrupting technology?

David R. Henderson writes:

@Douglass Holmes,
Good point.
To answer your question to Handle, my best guess would be the wheel.

Jeffery Tucker explains that we might have had flying cars by now, but the government maintains a monopoly on air and ground transportation rules. http://mises.org/document/6528/

Jacob A. Geller writes:

I appreciate the pushback against people you agree with, but I'm not convinced by Josh's argument about the flying cars.

First, I'm not convinced that flying cars wouldn't help us get from Point A to Point B as Josh suggests. Maybe rush hour traffic would be completely eliminated (or at least become MUCH heavier before becoming real "traffic" again) if highways had a third dimension to them -- picture a stream of two, three, four, ten, or a hundred more highways hovering above the highways we currently have now. Right now it's mostly just air up there, but with flying cars every highway becomes many times more spacious. New highways could be made thinner, too, freeing up land land for other developments.

But set that argument aside for now, because to the extent that flying cars wouldn't help us get from Point A to Point B, a big reason why that would be true today is that there are currently no flying cars. If there were currently flying cars, the infrastructure needed to accommodate them (like parking garages on the 50th floors of skyscrapers) might materialize. Saying that flying cars wouldn't help much getting from A to B today, is a little like saying that automobiles wouldn't help much getting from A to B in 1900, on the grounds that there are no gas stations in 1900! It's technically true, but only in the short/medium run, and as the cost of the current technological iteration of "car" continues to fall, networks will build up, the benefits will rise, and the costs will fall even further. Hence, today we have gas stations all over the country, automobiles massively assist us in getting from Point A to Point B, and companies like Tesla motors may soon be building up huge new networks to accommodate electric cars. But you need the flying car (or the electric car, or the original car) to at least be invented , and somewhat cost-effective, before anyone will start building up the necessary network of infrastructure.

Second, about the demand side -- yes, that flying cars do not exist should "give us pause," but press unpause for a moment and pretty quickly it starts to sound pretty odd that "flying cars don't exist because people don't want them."

Honestly, do you want a flying car? I do! I'd buy one for an extra, hmmm, $5,000.

Here Josh might be making the same mistake as the economist in the old joke about the 20-dollar bill on the sidewalk -- clearly there is no 20-dollar bill on the sidewalk, because if it were there then someone would have definitely picked it up already, so it can't be there; clearly there is no demand for flying cars, because if there were then someone would have definitely invented them and built up the necessary infrastructure already. What makes him so sure?

Look at it this way: if people didn't want flying cars, then flying cars wouldn't exist; but if people DID want flying cars, but the inventors just couldn't figure out how to invent a cost-effective flying car, then flying cars... wouldn't exist. The fact that they don't exist tells you nothing about the reason why.

Intuitively though, the latter explanation makes a lot more sense. I would love a flying car, personally, if it costed not too much more than a regular car, and I think a lot of people would too; it's just that, it's not so easy to make a cost-effective flying car like in the Jetsons, but it's totally technologically feasible to build Twitter. It's trivially easy, actually, to build Twitter, and almost impossible to build a cost-effective flying car. And I'm not saying that because I can see that nobody's built one, I'm saying that because *I* want one AND nobody's built one AND the laws of physics require a flying car to burn a pretty huge amount of fuel to stay afloat Jetsons-style.

So yeah, I can see how "Twitter instead of flying cars" technology might be a little bit disappointing, unless you were expecting the Great Stagnation.

Intuition and common sense often mislead, of course, but come on: I really want a flying car! How can I be alone in that...

Sol writes:

I wouldn't mind having a flying car, but I really want a low-cost, perfectly safe teleporter.

Handle writes:

I left a reply to Douglass Holmes, perfectly lucid, obviously human, no sales pitches, vulgar words or hyperlinks, but I was told that it was intercepted by the spam filter of all things!

How's that for progress in our time?

[Handle: I don't see any comments by you in the spam area, and I've gone through all of the hundreds of truly spam comments there now. Feel free to resubmit it. If you get an error message again, please email it to me at webmaster@econlib.org .--Econlib Ed.]

Eelco Hoogendoorn writes:

I agree with Jacob. It is quite rich to say there is no demand for more effective transportation. I could think of a use or two. But as an engineer myself, I simply have no idea how to build a practical and affordable flying car, despite standing on the shoulders of the wright brothers and henry ford. We need better power plants. We need better power storage. The internal combustion engine empowered the wright brothers, but it hasn't changed so much since to take flying from a military or highly centralized operation down to the personal level.

To answer douglas: the internal combustion engine was an excellent example of such a disruptive technology, like the steam engine before it. The IC was powerful enough to take transportation off the rails and into the hands of the average joe (and powerful enough to put airplanes in the sky). We havnt had any revolutions in transportation of that magnitude since. FYI, the IC has been around for a while.

Marc F Cheney writes:

Flying buses exist, and are wildly successful.

Flying cars exist, but are a niche market.

Finch writes:

We have flying cars. They are called "helicopters."

Helicopters are noisy, slow, dangerous, and require an expert pilot and expert maintenance, despite great economic pressure that they be otherwise. I'd love to have one, especially if the world was designed around more people having one, but it seems like it's just hard to make them cost-effective.

There are probably some regulatory issues holding them back. The FAA inhibits innovation, for example. And maybe the newfangled drone world we live in will make pilotless helicopters a reality, thereby lowering the costs and maybe increasing the safety.

But I think it's mostly energy arguments holding back flying cars. As with helicopters, you give up a lot when you can't rely on the ground to hold you up while you are stopped or coasting.

Eelco Hoogendoorn writes:

Marc: what you are saying is true. By the same measure, one might say that the middle ages enjoyed a globalized economy. Sure, the cost of shipping a few crates around the world was more than even most kings could afford, and would require a small army in sunken sailors as a sacrifice; but technically, everything was possible with a little persistence. Would you say the modern container ship was just an afterthought?

An affordable flying car is still as much a dream as a personal car was in the early 19th century. With a degree in aerospace engineering, I have a little appreciation as to why. Still, my 1950's self would have been profoundly disappointed by that realization, were he to extrapolate from the pace of technological advancement of the day.

Thomas Boyle writes:

How the car flies - the physics - is quite important. Regulations are a big deal, too. Finally, flying is a remarkably difficult skill, and we don't yet have the ability to automate it.

If the car flies by using fixed aerodynamic surfaces that deflect air downward when moved at speed, then we have had flying cars for decades. They're called "personal airplanes". They have disadvantages, it's true: the propellers are quite dangerous; the wings require huge parking spaces and make the cars incompatible with city streets; they require long runways to start and stop; and they require light structures that are not crashworthy in traffic.

If the car flies by using rotary wings, again we've had flying cars for decades. They're called "helicopters". The rotors can be aligned with the vehicle, which makes it somewhat compatible with city streets and narrow parking spaces, but the vehicle is still very long in that configuration. Crashworthiness is an issue. However, the rotors are quite dangerous and produce enormous amounts of wind during takeoff and landing; they're also very loud. In practice, helicopters are limited to operation in and out of heliports, because of the danger, wind and noise.

If the car flies by using high-thrust reaction engines (jet engines or rockets) it will produce very high levels of wind directly below the vehicle - winds high enough to injure people and damage structures. It will also be unacceptably loud. In addition, it will be very fuel-inefficient. Flying cars like these have not seen any success at all.

If the car flies by levitating into the air at zero speed, without producing significant (or any) downwash, without requiring large rotors or wings, then it uses physical principles (such as antigravity) that, to date, have not been discovered. Note that this isn't a failure to translate known physics into engineering and a marketable product: it relies on physics not yet known (and that possibly do not exist).

Next, we have regulations. To pick just one simple one, in many places it is illegal to take off or land an aircraft, except at a designated airport. It sounded good at the time, but of course now we're stuck with it: a chicken and egg, where there will be no flying cars because it is illegal to operate them, and no demand to change the law because there are no flying cars. And then there are regulations on aircraft certification, pilot certification, etc., etc., etc., etc., etc,. etc., etc... And, don't forget crashworthiness regulations for cars, which are difficult for lightweight flying machines to meet.

And, finally, flying - the act of guiding a flying machine - is hard. Despite our dreams of flying, humans normally have quite a poor grasp of motion in 3D; and there is a wealth of airmanship to learn (e.g., most people know next to nothing about weather and navigation when there's no road to follow, and there is significant judgment involved in making go/no-go decisions in the air). Automation will eventually solve this one, but so far that has proven harder than it looks.

Mark Bahner writes:
We've had the tech to make cars that go up in the air since the 1930's. The problem is we're not likely to have the tech to make them come down safely or keep them from crashing into one another any time soon.

The human brain is capable of about 1 petaflop (1 quadrillion operations per second). Presently, the cost of a computer with 1 petaflop of capability is about a million dollars.

In 10 years, the cost will be $1000. And in 20 years, the cost will be $1.

My prediction is that autonomous personal aircraft ("flying cars") are about 10-20 years away. I place "flying cars" in quotation marks, because I don't think it will make sense to have a vehicle that can drive on roads and also have wings or helicopter blades to fly. So what will happen will be that people will get in an autonomous ground vehicle to take them to an "airport" (which will potentially simply be shopping mall parking lots, which will become obsolete) and then take off in their autonomous flying vehicles.

The key to flying cars is a computer capable of flying the vehicle...because humans simply aren't safe enough without expensive training. And even then, look at the list of celebrities who have died in small plane crashes.

Famous people who died in plane crashes

jpa writes:

For those who say we don't have the computer technology to pilot copters safely... see this


shecky writes:

Thiel is in a much better position to develop a flying car than almost anyone else. No excuse for slacking off.

Andrew writes:

I'll take driverless cars over flying cars.

Finch writes:


The issue with safe autonomous flight is not (at least not primarily) raw computing power. It's safe software. The software is a much harder problem than the hardware, and making the hardware better doesn't ease the problem much.


Would you bet your life on the software flying those things not crashing? On dealing with mechanical failure graciously? On recognizing a hazardous situations from context?

And flight control is not even the long pole in the tent. Power is the real problem. If we had itsey-bitsey nuclear reactors, maybe we'd have the ducted-fan flying cars from 50s-futurism.

Hadur writes:

Almost every American now has a hand-held device through which they can access the sum of human knowledge. It's not a flying car, but it's pretty amazing. Even if we mostly just use them to look at pictures of cats and/or send naked pictures to our mistresses.

Arthur_500 writes:

Twitter is a new communications network that competes with other communications for eyeball space and time. As a user I need to determine if the information I am receiving from tweets is better information than what I get from e-mails, on-line chats, human interactions, letters, even other sites like that F-place. What sort of individual is sitting around sending out tweets every five minutes?

One great hotelier was known to have said the Right Customer is always right. I would paraphrase this to be the right source of information is most important not just every source of information

Boy I'd like one of those flying cars around 5:00 this evening...

Mark Bahner writes:
Boy I'd like one of those flying cars around 5:00 this evening...

Suppose you had a computer-driven car that did 80+ mph on an urban highway, a few feet from the car behind it? And never stopped at stoplight? And safely averaged ~20 mph above most posted speed limits around town?

Would you still need the flying car? Wouldn't that computer-driven car be fast enough?

Mark Bahner writes:
Would you bet your life on the software flying those things not crashing? On dealing with mechanical failure graciously? On recognizing a hazardous situations from context?

Absolutely. Much more than human beings. A computer can learn and remember perfectly the exact conditions that led to every crash of every airplane that ever crashed.

And the software can evaluate every engine parameter, every air parameter, and every geographic location parameter (including the locations of all nearby planes) several times per second.

And the software never thinks about the big fight it just had with its spouse, or its sick child, or anything like that.

That's why very soon (within a couple decades) every single fighter jet and bomber in the U.S. military fleet will be fully computer-driven.

James Oswald writes:

Any invention you can say what it is and has not been implemented is probably a bad idea. That they don't exist tells you what a bad idea they are. Flying cars would be noisy, fuel inefficient, and horribly dangerous.

Mark Bahner writes:
Any invention you can say what it is and has not been implemented is probably a bad idea. That they don't exist tells you what a bad idea they are.

Artificial heart? ;-) Space elevator? ;-)

Flying cars would be noisy, fuel inefficient, and horribly dangerous.

I agree that something that drives on the streets and also flies is not likely to "fly." Where I don't agree is that it seems very likely to me that aircraft similar to the V22 Osprey (i.e., vertical takeoff and landing (VTOL), capacity up to about 25 passengers, maximum speed 300 mph) will become very common…when planes become completely piloted by computers.

Here's a good letter (isn't the Internet cool?!) that contains some thoughts similar to mine. It points out:

1) There are 20,000 (!!!!) general aviation airports in the United States (Wikipedia estimates 5,200…but that’s still ~100 per state),

2) The V22 Osprey is tremendously expensive, but might be significantly reduced in cost if rotary engines replaced the turbines,

3) VTOL aircraft similar to the Osprey can get 85 passenger miles per gallon.


I agree that they would be noisy (I think neighbors of general aviation airports are going to be in for very unpleasant surprises in the coming decades!). But let's suppose that "only" 2,000 general aviation airports were sufficiently isolated from neighbors and in areas sufficiently desiring of the money that a fairly busy airport would provide. That would still mean 2,000 “new” passenger airports in the U.S., or 40 “new” passenger airports per state.

I don't agree that they would be "fuel inefficient" (see point 3 from the letter).

I also don’t agree they would be “horribly dangerous.” I agree they would be “horribly dangerous” with human pilots…but with computer pilots, I think they would be pretty safe. Specifically, with computer pilots, your concern about their use as a tool by terrorists would be essentially eliminated, because the terrorists would not be able to fly the plane. And the fact that computer hardware/software would continue to improve would make them more and more safe.

Per a U.S. DOT website, less than 1 percent of trips on U.S. highways are 100 miles or more, but they represent 15% of the passenger miles traveled every year. I can imagine a time when many or most of those trips are instead taken from nearby general aviation airports, via computer-piloted planes. I think this will also greatly increase the number of trips of 100 miles or more that will be taken each year.

Himanshu Sanguri writes:

This apple to apple comparison is not rational. A competitive world is always free to expand and choose the directions to move. No authority had ever abated the research for flying cars at the cost of Twitter. It is all about creativity, demand, supply and availability of talent and skills. Moreover, the premises that Twitter is not a public good invention is false.

Comments for this entry have been closed
Return to top