"It'll never work": a collection of failed predictions
lhup.edu
lhup.edu
IIRC, this is a misquote. The original was something like "All this talk of space travel is utter bilge. It would cost as much as a major war to put a man on the Moon." Which was more or less correct.
Anyway, for reference: http://en.wikipedia.org/wiki/Richard_van_der_Riet_Woolley
When a distinguished but elderly scientist states that something is possible, he is almost certainly right; when he states that something is impossible, he is probably wrong.
It is indeed possible for a particle to travel at a speed greater than the local phase velocity of light through a medium. This is what brings about Čerenkov radiation, which is that eerie blue glow associated with the core of nuclear reactors.
1) Prognosticating outside of one's expertise should raise a red flag.
2) Predictions should not rely only on scientific principles, but also involve economic analysis. (Larry Niven once wrote something to the effect: Anyone can write about teleport booths, but it takes a good author to come up with "flash crowds.")
3) Absolute laws are not always so simple to apply as it would first seem. (Example: the Halting Problem does not preclude useful bug checking software, just perfect bug checking software.)
Elder statesmen scientists aren't the only ones to make these mistakes. I posit that the common perception of reality is riddled with similar misapplication of fundamental knowledge. I also posit that there's huge profit potential hiding behind many such popular misapplications.
> When a distinguished but elderly scientist states that something is possible, he is almost certainly right; when he states that something is impossible, he is probably wrong.
Clarke's first law is an unnecessarily complicated version of:
It's probably possible.
kb
> Instead of a car in every garage, there will be a helicopter.... These 'copters' will be so safe and will cost so little to produce that small models will be made for teenage youngsters. These tiny 'copters, when school lets out, will fill the sky as the bicycles of our youth filled the prewar roads.
(“TFA”?! Really?!)
(See: http://www.paulgraham.com/say.html)
One famous example is Professor Joseph Le Conte's mistaken engineering analysis demonstrating the impossibility of flight. (See below.)
Another common example is the use of the Halting Problem to rule out the entire notion of tools for detecting bugs in software. (Yes, I've had professors tell me this flat-out.) The Halting Problem doesn't rule out such programs, it only demonstrates that they can't be perfect. To date, there are lots of tools that do an imperfect but still valuable job, to the point where people can even charge for them.
http://www.microquill.com/heapagent/ha_comp.htm
A major area where basic principles are misunderstood is in security. It's a truism that no security is perfect. However, it doesn't follow that no one is able to do online banking without instantly being hacked and robbed. Yet many who find the previous idea ridiculous also think that all "DRM doesn't work." That's simply not true. While it's true that all DRM can eventually be broken, it's not true that all of it instantly evaporates on contact with the internet. Breaking DRM involves a certain cost. If enough people are "willing to pay" the cost, then it will be broken. This is almost always true for big-budget hollywood movies. It's certainly not true for all digital content.
---- Professor Joseph Le Conte's mistaken engineering analysis
Put these three indisputable facts together:
One: There is a low limit of weight, certainly not much beyond 50 pounds, beyond which it is impossible for an animal to fly. Nature has reached this limit, and with her utmost effort has failed to pass it.
Two: The animal machine is far more effective than any we can hope to make.; therefore the limit of the weight of a successful flying machine can not be more than fifty pounds.
Three: The weight of any machine constructed for flying, including fuel and engineer, cannot be less than three or four hundred pounds. Is it not demonstrated that a true flying machine, self-raising, self-sustaining, self-propelling, is physically impossible?
— Joseph Le Conte, Professor of Natural History at the University of California, Popular Science Monthly, November 1888
But that efficiency is due to our remarkable performances on nothing but a few grams of low-grade fuel. It doesn't say anything about what would happen if you had some really high-grade fuel.
Sugar has about 18.8 joules per gram. Kerosene, which was pretty well known in 1888, has almost 46 000 joules per gram.
Perhaps we can extract some general principles about prediction from that, like: given a complex enough system, if one factor changes by two or more orders of magnitude, previously observed behavior is useless as a guide to the future. I'm fudging on 'complex enough', though. Trigonometry works just fine with triangles the size of pencils or skyscrapers, but predicting the scope of inventions doesn't seem to work well when one factor changes radically.
Exactly what does effective mean?
We still can't make ornithopters that approach the performance of birds. That is what his analysis actually shows. Our successful aircraft still don't have performance that resembles that of birds. Then again, no birds cruise at an altitude of 30,000 feet at nearly Mach 1. Our aircraft don't have to be bird-like to be economically useful, and this is indeed effective to us.
So to go back to my other comment that predictions should involve economic analysis: our professor should have considered flight performance envelopes that would be commercially useful, not flight performance envelopes that resemble those of birds. (For example, he could have extrapolated something like the performance envelope of a balloon, but with twice the speed and better directional control.) If he had considered a few examples of those, he'd have quickly realized that using birds as a model wasn't a good idea for an analysis of human flight technology.
The real limit is our ability to burn huge amounts of it. Even a small plane burns over 1 lb / minute, which would be a shocking metabolism for an animal.
That puts Le Conte's error in even more perspective. Imagine if you said to him in 1888 "...but what if you had access to vast, cheap supplies of high-grade fuel which you burned off at a rate of 1 lb per minute?" He'd look at you like you were an idiot. Now your heavier-than-air craft is carrying several hundred pounds of fuel too, just to stay aloft for a few hours? And what do you do when you get where you are going?
Does this mean Le Conte's real failure isn't energy calculations, but predicting the availability of such fuel? In other words, it's all economics?
I had another comment about "toys" versus "tools for real work" in this thread. Maybe that's another defining element of "toys", the difficulty of maintenance and the lack of ubiquitous infrastructure. Only rich or eccentric people are going to take the trouble to provide infrastructure for their own machines, so it becomes by definition a "toy". Like planes or cars in the early 20th century.
Okay, is this a new heuristic? We look for devices which are fun or useful but impractical because right now, enthusiasts have to provide their own infrastructure at great cost?
This quote is in a way a direct refutation of the work done by many here in programming, and this perspective has its merits and flaws. But I don't see exactly how its a prediction of anything.
A few of those are included, like von Neumann's "nuclear power will make energy free!" prediction.
The future will surprise. It will surprise us by what is possible, and by what isn't.
In this case, it was just mental laziness. Yes, even John von Neumann can be guilty of this. E=mc^2 implies a freaking huge ratio in your favor. That makes nuclear energy seem of "virtually limitless" abundance for the same reason that nuclear bombs seem like "virtually unstoppable" weapons. Yet, simple ball bearings or a complex mechanism like an Orion rocket pusher plate can survive close proximity to a nuclear explosion with some judicious engineering. If John von Neumann started looking closely at nuclear energy in the broader context of our industrial and political infrastructure, he might have come to a different conclusion.
My conclusion is this: One of the best places for opportunities to hide is in the mistaken popular applications of basic principles and physical laws. This isn't to say that those principles and laws are incorrect, just that the way most people apply them either doesn't go far enough, or fails to also include an economic analysis.
This tells us how to look for opportunities hiding behind other common misapplications of principles. (Not necessarily ones involved in nuclear energy.)
1) Theological: It is an intervention in God's order, which makes nights dark...
2) Medical: It will be easier for people to be in the streets at night, afflicting them with colds...
3) Philosophical-moral: Morality deteriorates through street lighting. Artificial lighting drives out fear of the dark, which keeps the weak from sinning...
Raise your hand if you have a sleep disorder...
The jury is still out on that one.
Interpretations of quantum mechanics fall into a similar camp. There is the actual formal theory of quantum mechanics, which is clearly mathematics; but something like the Copenhagen Interpretation is also attempting to "describe" the universe, but is forced to do it extra-mathematically, by attaching interpretations (which can't be mathematically proven) to the formal results.
I suppose it also depends on what one means by "describe": does quantum mechanics fully "describe" the universe, or is it quantum mechanics plus an interpretation which describes the universe?
That's mathematically proven, so it must be true.
Also, mathematics may predict things which don't exist, or are impossible in nature.
As a former experimental physicist I can't help but notice how true this is. ;)
Somebody said "the Universe isn't just stranger than you imagine; its stranger than you Can imagine"
You're right though, it could mean a description of any complexity, something Mathematics is well capable of.
That's hardly a howler is it?! I see nothing suggesting that 1.5 tons is the endpoint or that we'll be using heavy computers in The Year 2000.
Will those people suggesting we'd one day have a supercomputer in our pockets be laughed at when we have microscopic omniputers floating around our then-useless brains?
IBM's Watson is powered by 90 of these (according to the wiki): http://www-03.ibm.com/systems/power/hardware/750/index.html
The datasheet says they weigh 120 pounds. 90 of those would be 10,800 pounds, or 5.4 tons. We're still a bit above the prediction, but I think Watson could be done in 1.5 tons of gear with some more development work focused on that.
I'd need to see the Popular Science article, but I doubt they were envisioning a personal computer. They might have been envisioning a community computer though. 1.5 tons would probably be enough for that in terms of a shared file/message board/connectivity supercomputer for a city block's worth of people.
But one quotation that I'm really taking issue with is the one about Sir Walter Scott dismissing public gas lighting. Far from that, Scott was actually a dedicated promoter of gas lighting in its early years, as evidenced by his tenure as the First Chairman of the Edinburgh Gas Street Lighting Company:
http://www.scotlandmag.com/magazine/issue30/12007614.html
But when was a little bit of historical research an obstacle in the way of feeling good at the expense of people who lived hundreds of years ago?
Whenever I see that quote I picture Martin Luther as the "Get A BRAIN! MORANS" guy.
I wonder if this sounded as dumb back then as it does now? What about things that do not move under their own power, yet move nonetheless, such as waves and wind? What about flowers that close every night and open in every morning? What about the fact that they knew almost nothing about the earth deeper than the tiniest scratch on its surface -- perhaps there are muscles of a sort under there?
What about the "fact," as they would have seen it, that God can make anything move however he wants, and that he might have a special mechanism that makes planets and stars move but does not apply the same way to things on Earth? And did he really believe that the moon and stars have limbs and muscles?
I think this illustrates how people can get away with any possible idiocy as long as they are on the right side of an issue. This is why scientists care about the timing of publication: they look a lot more competent if they publish their results at a time when everyone will tend to believe they are correct, instead of at a time when everyone will pick their paper apart looking for flaws (or simply assume the flaws are there.)
> Animals, which move, have limbs and muscles; the earth has no limbs or muscles, therefore it does not move. It is angels who make Saturn, Jupiter, the sun, etc., turn round. If the earth revolves, it must also have an angel in the centre to set it in motion; but only devils live there; it would therefore be a devil who would impart motion to the earth....
> The planets, the sun, the fixed stars, all belong to one species - namely, that of stars. It seems, therefore, to be a grievous wrong to place the earth, which is a sink of impurity, among these heavenly bodies, which are pure and divine things.''
A sink of impurity. There's a data point that has lasted through the ages.
I wonder, then, why he didn't find it plausible that it is devils or fallen angels who move the earth?
Nonetheless, we can certainly grant him an implicit "Only living things (animals) can move (by themselves)." Doesn't make it any less wrong, but I doubt that it's just a logical fallacy.
I'm still amazed that people spend so much time staring at plywood boxes.
A toy is something which fascinates the mind in some way. An "interesting toy" suggests something that has a scope of operation within some small realm. I'll posit that Legos are an "interesting toy" whereas a toy car is "just a toy".
Substituted, this is now "X are things which have great possibilities within a smaller realm of operation, but are not suitable for Y".
The problem should now be obvious. The next question is to ask what it would take to scale the X's realm up. If it's economically feasible, it will happen.
The use of "toy" is the rhetorical trick here, because it implies the thing in question has permanently limited scope, and it might not be. (There's also the implied put-down, that those who find them worthwhile are childish.)
So what things today are "interesting toys" but not suitable for "real work"? Mobile devices? Social networks?
Today the premier example would be mobile devices and mobile OS tablets. They have enormous potential, even the potential to replace traditional PCs, but today they are often regarded as toys.
Isn't the below an over-optimistic failed prediction?
a few decades hence, energy may be free—just like the unmetered air.... John von Neumann
Automobiles will start to decline almost as soon as the last shot is fired in World War II. The name of Igor Sikorsky will be as well known as Henry Ford's, for his helicopter will all but replace the horseless carriage as the new means of popular transportation. Instead of a car in every garage, there will be a helicopter.... These 'copters' will be so safe and will cost so little to produce that small models will be made for teenage youngsters. These tiny 'copters, when school lets out, will fill the sky as the bicycles of our youth filled the prewar roads.
I think this is the reason we don't have flying cars. We could build them. We could do a lot of things. But nobody really wants them, given the danger they would pose. (Energy constraints are probably another reason, but secondary. There are flying car prototypes that get MPG numbers comparable to SUVs.)
And yes, we want them. Many times faster than ground transport, safely usable by any person young or old, no congestion, frees up the ground for other uses. Feasibiliy aside, it would be undeniably utopian.
If I had to bet it would be on the "driverless" car.
So if the biggest problem here is safety (introduced to some extent by irresponsible drivers), maybe flying cars will be possible once we can create automated car control?
Once these robot cars are reliable enough, they will look like a much better alternative to human drivers to me: quicker reaction, possibility to have 360 degrees field of view at all times, impossible to DUI etc.
Let's keep an eye on DARPA's Grand Challenge and Google robot car attempts.
On a second thought, isn't bash.org exactly that?
"I confess that in 1901 I said to my brother Orville that man would not fly for fifty years. Two years later we ourselves made flights. This demonstration of my impotence as a prophet gave me such a shock that ever since I have distrusted myself and avoided all predictions."
I wonder how he explained birds?