Sunday, November 14, 2010

Remembering Alan Turing

This article is about how a society treated a war hero, someone who also had a deep impact on multiple scientific fields.

During the World War II, Turing, a British citizen, was largely responsible for breaking the German Enigma encryption system. Though he did not have the glorious position of a General who leads his army to victory by using clever strategies, Turing's contributions to the war effort were no less. The Germans believed that Enigma was unbreakable and used it to encrypt sensitive communications. The work of Turing and his team gave the allied forces an edge as they were able to decrypt those communications and know in advance what the enemy was up to.

His other important contributions were in fields of mathematics, computer science and philosophy. So important were his works on computer science that he is usually referred to as the father of computer science.

But he was also gay. He lived in a time when there was no respect for a person's right to a private life and when society loved to play moral police. So when it was found that he was homosexual, he was charged guilty of that "crime" and was ordered to take hormones to kill that "abominable" instinct. After taking the medications for a while, Turing decided to kill himself. He injected cyanide into an apple and ate it. So it happened that society, prompted by religious morals, managed to kill one of the brightest humans to have ever walked on this Earth.

Of course, over the years the British society's attitude to homosexuals has changed a lot and their government even apologized for the way Turing was treated. But not all present day societies are that open minded and his story should serve as a warning of how misguided morals disrupt progress; progress meaning a better understanding of our Universe. It shows what people who think that they have the right to dictate another person's private life are capable of. They can drive a brilliant man to kill himself. Turing by no means was academically unproductive at the time of his death. Think of what more could he have contributed to humanity's knowledge had he lived a full life.

While most of Turing’s works are too esoteric for me to understand, his philosophical treatise on whether Turing Machines can mimic human intelligence is pretty accessible.

Turing on Artificial Intelligence

This is a brief summary of the 1950 paper by Turing titled “Computing machinery and intelligence”. The paper makes a strong argument for artificial intelligence, an uncomfortable notion which undermines the uniqueness of being human.

If the general public have heard of Turing, it is through his thought experiment which is known today as the Turing Test. Given a machine and a human and an observer, where the observer can communicate with the machine and the human only by typing on a keyboard, can the observer find out who is human and who is the machine? If she cannot, the machine can be considered to be as intelligent as the human and effectively there is no difference between the machine and the human.

The machine Turing refers to is the Universal Turing Machine. It is a digital computer and is universal in the sense that it can simulate other digital computers. That rids us of worrying about the specifics of how a particular computer is designed. As long as we know that it can simulate any other computer, we know that our machine is not missing any capability. The requirement that communication happen only by typing is to separate intelligence from other physical attributes which can bias the observer.

Turing then talks about various objections that can be raised against such a thinking machine. I’m going to talk about only the theological and mathematical objections. Theological, because the majority population subscribes to it. Mathematical, because if there are any logical contradictions in the thought experiment, it can never be realized.

The idea of a thinking machine is scandalous to theology, which views man as god’s special child who has an immortal soul that is the seat of intelligence. Even in eastern theologies which lack a personal god and and the idea of a man as child of god, subscribe to the idea of humans as the most exalted life form. For example, Karma hypothesis presupposes that superiority. Only a human can attain moksha, not other “lower” life forms.

Turing gives three arguments to counter the theological objection - humans are not that different from other animals, different religions have different ideas of a soul and finally, saying that machines can’t have a soul places restrictions on the omnipotence of god (Who are we to say that an all powerful god cannot endow souls to machines?). Today, it is even more easier to throw out the theological objection given that the ideas of god and soul themselves have no basis in reality.

The mathematical objection is a lot more interesting. It comes from Godel’s Incompleteness Theorem, which in Turing’s words, shows that in any sufficiently powerful logical system, statements can be formulated which can neither be proved nor disproved within the system, unless possibly the system itself is inconsistent.

Turing argues that this objection becomes baseless if the human mind itself is not free from inconsistencies. Neuroscience is providing a lot of evidence in that direction.

Turing’s paper can be said to have kick started the AI field. He showed that it is theoretically possible to build a machine which can think like a human. Though at this moment it is difficult to say if we can build such a machine. But there is a good chance that we will. Computers are getting more powerful and are beating humans in tasks (like chess) that were once thought the domain of human intelligence. They are also becoming good at pattern recognition, another field where humans excel. It will be interesting to see if they can display the full range of human intelligence and more importantly if they will have consciousness like us.