Turing Tales
So it looks like Alan Mathison Turing will have to wait for his pardon, perhaps as long as Gallileo waited, but Nature has done a nice job of collecting a special issue for the hundredth year from his birth. Turing raised so many profound questions in so many sciences. Here are just a few:
“The search for new techniques must be regarded as carried out by the human community as a whole, rather than individuals.” That sounds so obvious to us on the net; but the textbooks still relate science as though it leapfrogs from one genius to another. Which do we think is more important to progress in science–the community (like us) or the lone genius ahead of their time, like Mendel or McClintock? Was Turing himself the lone genius he denied existed?
The invention of the “universal computing machine” that could do what any other machine did by following a description of it. In other words, software. Are today’s computers in fact “universal” — can they do anything, given the right software? Or do their architectures limit what they can do?
Turing’s 1952 paper on “The chemical basis of morphogenesis” invented pattern formation in biological development. How does a whole human form out of a formless cell? The organism develops from a combination of “morphogens,” chemicals that induce tissue to differentiate in patterns; and random events, random fluctuations of molecules. Both principles have been discovered and propagated throughout biology. What key unsolved riddles in biology need a Turing today?
Fittingly, or ironically, the same Nature issue describes a study using magnetic resonance imaging (that giant magnet you put yourself into) to study the mechanics of how the brain “suspends disbelief” in fiction. Writers beware.
Comments are closed.
Ultraphyte 
As matters stand, a computer can do any fully intelectual activity that a human can, as long as we can fully define a set of rules for it to work by, and will wait for its processing time. This is why we can build a computer that is effectively a chess Grand Master, but there are still large numbers of other board games that computers just can’t play effectively.
Computers also can evolve by selection of random errors–achieving programs no human could create. Computers have been made to lie, learn, write new music, and teach invented language to other computers.
When computers evolve a limbic system, things will really get interesting.
What is the motivation/purpose of programming a computer to ‘lie’? I’m assuming that the lie is deliberate, i.e., the computer is ‘aware’ of the non-truth of the statement, and that this isn’t of a purported link to having an imagination.
The idea of deliberately causing a computer to “lie”, or “make mistakes,” is part of how you teach it to be creative. “Mistakes,” or doing something differently than normal, are part of creativity. Artists never paint exactly what they see; they introduce “lies.” Scientists don’t “lie,” but they sure do make mistakes. Just ask around my lab–one mistake after another this week (the week before break).
But sometimes the mistake leads to a new discovery. Just as in evolution, the rare beneficial mutation makes the new species.
Turing doesn’t deserve a pardon; he deserves an apology. A pardon suggests that he did something wrong rather than the government.