The Kludge That Won
It’s occurred repeatedly in our history. It’s occurred over and over during the evolution of life.
Evolution has dozens and dozens of examples of some structure evolving for one purpose, but then, once it’s there, being used for some other purpose. Quite often, the new ‘unintended’ purpose ends up being even more useful than the original. Because most biological structures have two or even three uses, such boot-strapping development is ubiquitous. And it doesn’t just occur on the level of our eyes and hands, it also occurs on a chemical level.
The development of new proteins is approximately the major function of biological evolution and again, as soon as one of those enzymes finds a new purpose, it will get used for biological advantage. And dual purposes can occur at the chemical as easily as the mechanical level.
But I want to describe some ideological kludges.
The major lesson is that nobody should ever be afraid to consider some crazy idea. Crumpling up a piece of paper to throw out the latest attempt is nearly free. And the more people with that attitude, the sooner we’ll see another one.
“Black Swans” can arise from such Kludges quite easily.
Max Planck tried to fix his math problem with a kludge.
The problem was that everyone knew what kind of power spectrum is created by a hot body. It’s got a little bit of power in the high frequencies, most of the power in the mid range and a decent chunk on the low end. We KNOW what kinds of radiation is produced, and at what levels. The difficulty was predicting it.
Raylegh-Jeans could predict one end of that curve just fine, but the high frequency prediction was for an infinite amount of energy, which couldn’t be correct. Another formula could predict the other end. Between them, scientists could get by, but it was obviously a problem.
Planck decided it wasn’t Somebody Else’s Problem and set himself the task of getting both sides of that curve from a single formula.
He introduced what even he considered a complete kludge, an unrealistic condition that he thought might help with the math, and then he figured he could go back and deal with the goofy conceptual issue once he had the formula.
Only one problem: he couldn’t take out the kludge. And the ‘kludge’ turned out to be one of the most important insights in the entire history of science:
Atoms can only vibrate at particular frequencies.
Claude Shannon began working, in the 1940s, on questions of communication.
Specifically, he developed a mathematical theory, which we now call ‘information theory‘ by which to predict how much redundancy a signal would need to succeed at sending a message despite a noisy or intermittent connection.
To start applying measurements to such things, Shannon began by asking “what is information?”
What he decided is that information is the amount of surprise a message generates – if someone tells you a message you already know, it’s not informative. If someone tells you the sun will rise tomorrow, they’ve wasted their breath, because you already knew that. If the sun wasn’t going to rise tomorrow, hearing about it would be outrageously surprising, and something approximately every person would want to know, it’s that informative.
You must be incapable of predicting the future of a message for it to truly be considered informative. Orderly messages are very predictable, and hence, not very informative. The more random a message seems, the more information it contains.
Shannon’s kludge…is that information is entropy.
He didn’t mean physical entropy – he thought he was merely borrowing the word as a metaphor.
What’s crazy is that since all of science is about using information about observations to predict future observations, the entire structure of science, and hence our model of the universe, could be completely stood on it’s head by this concept because science doesn’t actually require the stuff to actually exist. The material might all be an illusion like the Matrix or a hologram or some other strange possibility. And science will continue to be about using past observations to predict future ones. String Theory is, in part, the ultimate logical conclusion of Shannon’s theory about information.
“Maxwell’s demon” is a thought experiment named after James Clerk Maxwell, for illustrating the point: we can’t actually sort things without using energy. Information and energy aren’t distinct.
The role of information in the evolution of the universe is now considered so critical, Stephen Hawking concluded that black holes don’t exist because like energy, information can’t be created or destroyed, and given the choice, decided black holes were the mistake.
More recent observations suggest that gravitational waves store information in the structure of spacetime itself and can provide a mechanism for information to escape black holes.
Claude Shannon’s might just be the most impressive Kludge That Won:
Information = entropy
John Napier had a math problem, too.
The problem was that people had pencils and paper but no calculators or computers, but to figure out things like tide tables and star charts for navigation required working out things like 1048.234 x 407.841.
Even worse, some trigonometric calculations were done with series, meaning fancy formulas full of multiple terms that could produce a good answer, but the calculations could become enormous. And the larger they were, the longer they took, the easier it was to make a mistake, and the harder it was to check.
So Napier thought of a kludge. He wrote a book. A book with 57 pages of explanation and 90 pages with thousands of numbers. A book that reported which exponent would go on (1/2) in order to produce any other number.
Multiplying two giant numbers takes a decent amount of time. But since multiplying numbers means adding the exponents, Napier’s Big Idea was to produce a lookup table, by which people could convert their ugly complicated numbers requiring multiplication or division into less ugly easy problems requiring only addition or subtraction.
His lookup table ran from 1 two-millionth to 2 million.
Navigators, astronomers and architects could look up the exponents and add, instead of multiplying giant numbers. Once they had the new exponent, they could find the number they sought by a reverse lookup.
We now call that method ‘logarithms’ and they are unique in the history of science because unlike every other thing ever invented by people, Napier’s Logarithms were universally hailed as a Good Thing, obviously capable of saving outrageous amounts of time and effort and mistakes for some of the most important functions serving positive purposes here on Earth.
The other reason logarithms are such an important innovation is because most of our major sensory systems respond logarithmically, which is why the decibel, stellar magnitude, luminosity, and pH scales are all logarithmic.