A celebration of the life and work of the pioneering computer scientist
See full coverage
Consider how Turing went about his proof. He took human computers as his model. There they sat at their desks, doing one simple and highly reliable step after another, checking their work, writing down the intermediate results instead of relying on their memories, consulting their recipes as often as they needed, turning what at first might appear a daunting task into a routine they could almost do in their sleep. Turing systematically broke down the simple steps into even simpler steps, removing all vestiges of discernment or comprehension. Did a human computer have difficulty telling the number 99999999999 from the number 9999999999? Then break down the perceptual problem of recognizing the number into simpler problems, distributing easier, stupider acts of discrimination over multiple steps. He thus prepared an inventory of basic building blocks from which to construct the universal algorithm that could execute any other algorithm. He showed how that algorithm would enable a (human) computer to compute any function, and noted that:
The behavior of the computer at any moment is determined by the symbols which he is observing and his "state of mind" at that moment. We may suppose that there is a bound B to the number of symbols or squares which the computer can observe at one moment. If he wishes to observe more, he must use successive observations. ... The operation actually performed is determined ... by the state of mind of the computer and the observed symbols. In particular, they determine the state of mind of the computer after the operation is carried out.
He then noted, calmly:
We may now construct a machine to do the work of this computer.
Right there we see the reduction of all possible computation to a mindless process. We can start with the simple building blocks Turing had isolated, and construct layer upon layer of more sophisticated computation, restoring, gradually, the intelligence Turing had so deftly laundered out of the practices of human computers.
But what about the genius of Turing, and of later, lesser programmers, whose own intelligent comprehension was manifestly the source of the designs that can knit Turing's mindless building blocks into useful competences? Doesn't this dependence just re-introduce the trickle-down perspective on intelligence, with Turing in the God role? No less a thinker than Roger Penrose has expressed skepticism about the possibility that artificial intelligence could be the fruit of nothing but mindless algorithmic processes.
I am a strong believer in the power of natural selection. But I do not see how natural selection, in itself, can evolve algorithms which could have the kind of conscious judgements of the validity of other algorithms that we seem to have.
He goes on to admit:
To my way of thinking there is still something mysterious about evolution, with its apparent 'groping' towards some future purpose. Things at least seem to organize themselves somewhat better than they 'ought' to, just on the basis of blind-chance evolution and natural selection.
Indeed, a single cascade of natural selection events, occurring over even billions of years, would seem unlikely to be able to create a string of zeroes and ones that, once read by a digital computer, would be an "algorithm" for "conscious judgments." But as Turing fully realized, there was nothing to prevent the process of evolution from copying itself on many scales, of mounting discernment and judgment. The recursive step that got the ball rolling -- designing a computer that could mimic any other computer -- could itself be reiterated, permitting specific computers to enhance their own powers by redesigning themselves, leaving their original designer far behind. Already in "Computing Machinery and Intelligence," his classic paper in Mind, 1950, he recognized that there was no contradiction in the concept of a (non-human) computer that could learn.
The idea of a learning machine may appear paradoxical to some readers. How can the rules of operation of the machine change? They should describe completely how the machine will react whatever its history might be, whatever changes it might undergo. The rules are thus quite time-invariant. This is quite true. The explanation of the paradox is that the rules which get changed in the learning process are of a rather less pretentious kind, claiming only an ephemeral validity. The reader may draw a parallel with the Constitution of the United States.
He saw clearly that all the versatility and self-modifiability of human thought -- learning and re-evaluation and, language and problem-solving, for instance -- could in principle be constructed out of these building blocks. Call this the bubble-up theory of mind, and contrast it with the various trickle-down theories of mind, by thinkers from René Descartes to John Searle (and including, notoriously, Kurt Gödel, whose proof was the inspiration for Turing's work) that start with human consciousness at its most reflective, and then are unable to unite such magical powers with the mere mechanisms of human bodies and brains.