Alan Turing – thinker ahead of his time – The Science Show – ABC News
Robyn Williams: He was without question a genius. He changed our lives as much as anyone living in the twentieth century. Today’s Science Show special is written and presented by Sharon Carleton. And it begins with the lady of the moment.
Queen Elizabeth: In August 1939, 200 people arrived at Bletchley Park. By the end of 1944 there were 10,000. They were called to this place in the greatest secrecy, as they set about on a seemingly impossible mission, a massive challenge in the field of cryptanalysis, for the first time pitting technology against technology. This was the place of geniuses such as Alan Turing. At heart we have always been a nation of problem solvers. Battles can be won and many lives saved by using brain power as well as firepower, deliberation as well as force. For your many achievements I give my heartfelt thanks on behalf of an eternally grateful nation.
Jack Copeland: Turing changed the world. There are several sides to the story of exactly how he changed the world, and one side is the code breaking Bletchley Park, the impact he had on the war. The other side of it is what it feels most relevant to us today; Turing invented the modern computer, he change the lives of every one of us whose work or play is involved with computers.
Sharon Carleton: Jack Copeland, founder of the Alan Turing online archive. June 2012 marks 100 years since the birth of Alan Mathison Turing, an eccentric Englishman, a tragic figure and one of the top scientific brains of all time because he influenced so many different areas, according to the science journal, Nature:
Reading: The scope of Turing’s achievements is extraordinary. Mathematicians will honour the man who cracked David Hilbert’s ‘decision problem’, and historians will remember him as the man who broke Nazi Germany’s Enigma code and helped to shorten the Second World War. Engineers will hail the founder of the digital age and artificial intelligence. Biologists will pay homage to the theoretician of morphogenesis, and physicists will raise a glass to the pioneer of nonlinear dynamics. Philosophers, meanwhile, are likely to continue to frown over his one-liners on the limits of reason and intuition: ‘If a machine is expected to be infallible, it cannot also be intelligent,’ he said in a 1947 talk to the London Mathematical Society.
Sharon Carleton: The biography Alan Turing: The Enigma by Andrew Hodges was probably the first major work to bring this extraordinary scientist’s academic strengths and personal foibles to a general audience. Dr Hodges is a Fellow in mathematics at Oxford University.
Andrew Hodges: He had a very isolated boyhood. His parents were often in India, his father was an official in the Indian civil service, very middle-class sort of life, sent off to foster parents and prep school in this public school. It wasn’t his sort of thing at all. But then when he was 16, someone broke into his world, this isolation, and it was another boy called Christopher Morcom, and they shared this great interest in fundamental science, thinking about things scientifically, but also it was important to him because he must have begun to realise at this time the importance of his consciousness as a young gay man, something that was almost…hardly allowed to think about in those days at all, but it must have been just coming into his awareness that this was something terribly important.
So these two things came together; intellectual things about science, and also these emotional things. But then it was a great tragedy because just when they were 18, Christopher Morcom suddenly died. His death left this enormous gap, and one of the interesting things is how he filled this gap with an intellectual life, thinking about the nature of the mind, thinking what it is about the mind that’s gone when someone dies, and then turning that into scientific thought. That was the background to his breakthrough later on in the mid-’30s when he was thinking about the nature of mind, the right way of describing it.
Sharon Carleton: Alan Turing was stockily built and classically good looking. He was humorous, solitary and shy, an untidy student and bottom of the class in English. His passion for science was not appreciated at Sherborne School in Dorset. The headmaster wrote to his parents: ‘I hope he will not fall between two stools. If he is to stay at Public School, he must aim at becoming educated. If he is to be solely a Scientific Specialist, he is wasting his time at Public School.’
When he was 19, Turing went to study mathematics at Cambridge, where he graduated with first-class honours. He was elected a Fellow of Kings College when he was only 22.
Jack Copeland: It’s quite amazing really the way in which the computer just fell sideways out of rather abstruse research that nobody could have guessed would have any practical application whatever, let alone that it would produce a machine that would change all our lives.
Sharon Carleton: Jack Copeland is editor of The Essential Turing, keeper of the online Turing archive, and his day job is professor of philosophy at the University of Canterbury in New Zealand.
Jack Copeland: Turing was thinking about a very abstract problem in the foundations of mathematics. And in the course of thinking about this problem he’d managed to invent on paper, there was no actual engineering at this time, the modern computer, the universal stored-program computer. Turing’s brilliant idea was for a single machine, a fixed structure that could change its behaviour, kind of chameleon-like, from a machine dedicated to one particular task into a machine dedicated to a completely different task. It could change itself from word processor to calculator, as we now know, photo editor, communications device, musical instrument, chess opponent. It’s a chameleon. His idea was that you just put a different program of instructions into the program’s memory and you would transform this fixed machine into a machine for a completely different purpose. There was no need to change the wiring, you just typed in a different program. I guess nowadays when most of us own a physical realisation of the universal Turing machine, Turing’s idea of a one-stop-shop computing machine is liable to seem as obvious as the wheel or the arch, but in those days it was revolutionary.
Sharon Carleton: Alan Turing completed his PhD in two years at Princeton in New Jersey. Others there at the time included Albert Einstein and the distinguished mathematicians John von Neumann and Alonzo Church, the latter describing Turing as ‘a loner and rather odd’.
Recognising that war in Europe was looming, Turing returned to Cambridge where he debated with the philosopher Ludwig Wittgenstein about the foundations of mathematics. He’d been interested in codes and code breaking for a couple of years and in 1938 was recruited to the Government Code and Cipher School at Bletchley Park in Buckinghamshire, known amongst themselves as the Golf Club and Chess Society. Writing in the English Telegraph newspaper, historian Keith Lowe said:
Reading: The story of Bletchley Park is emblematic of the bumbling but occasionally brilliant way that the British conducted themselves in the Second World War. The Secret Intelligence Service set about staffing it with anyone they could lay their hands on: cryptographers, mathematicians, Egyptologists, linguists, even astrologers. Some were recruited for their originality of thinking, and others because they were related to men who played golf with Bletchley Park’s director.
Right from the beginning there was an atmosphere of informality and disorganisation that was only mitigated by the British talent for muddling through. Staff-built code-breaking machines were held together with sticking plaster and pieces of string. Occasionally there were minor explosions because of the tendency of careless WRENS to balance their metal mirrors on the electrical terminals while doing their make-up. And yet this was the outfit that broke the ‘unbreakable’ Enigma codes and, in the words of Eisenhower, shortened the war by two years. — Keith Lowe
Sharon Carleton: It was the perfect environment for an eccentric genius like Alan Turing. The biggest problem facing the code breakers was the German Enigma cipher machine which sent highly scrambled messages. It was at the core of their military and intelligence communications. It looked a bit like an old fashioned typewriter with a series of rotating wheels and was very, very complex. The odds against anyone being able to break the Enigma code were 150 million million million to one.
In 1932, when an early, and simpler, model of the Enigma machine was undergoing trials with the German army, Polish code breakers had made a tentative start at understanding it. They’d also designed an electro-mechanical machine which could test the Enigma’s rotor settings.
Jack Copeland:
Jack Copeland: Turing’s idea was to set a machine against a machine. They wanted to break the Enigma machine, so Turing designed a machine for attacking Enigma. He based the idea on the Polish bombe, which the great Polish code breaker Marian Rejewski had built. He set about building a much bigger bombe that would be stable against changes in the code. The trouble with Rejewski’s bombe was that as soon as a certain loophole in the code was closed by the Germans, the bombe became useless. They closed that loophole on 1 May, 1940. They hadn’t been aware of it until then I guess.
Sharon Carleton: Alan Turing and fellow code breaker Dillwyn Knox had anticipated that the Germans would find that loophole, and they had set about designing a new machine which would survive the changes.
Jack Copeland: The first bombe, the experimental prototype model, was installed early the next year. For the first time in military history, Britain had open access to Germany’s top-secret communications. By about 1942 they were breaking two Enigma messages every minute. They knew huge amounts about what the Germans were doing.
Sharon Carleton: By the end of the war, Alan Turing’s Hut 8 had deciphered 1.1 million Enigma messages. More than 200 of these bombes were eventually built, all were destroyed after the war.
Jack Copeland: Naval Enigma he broke one night in the autumn of 1939. So this was a time of frantic intellectual activity for Turing. He was inventing and designing the bombe and he was working out how to break naval Enigma. He also had a hand in various of the other code breaking operations that were going on, the attack on the Hagelin machine. He had fingers in many pies. It involved two stages of encryption, one layer of encryption by hand and then another layer of encryption by Enigma, and that’s what made it so difficult.
Sharon Carleton: How did he do it?
Jack Copeland: Pure brainpower. He had some old messages from about 1937 or 1938, and so he was able to work out by looking at that past traffic how the so-called indicator system worked. The indicator system was how the sender told the receiver to set his machine up so that the sender’s machine and the receiver’s machine were in synchrony. That was the part that was encrypted twice, and it was just incredibly difficult to break, but Turing managed to break it. Even then it was by no means plain sailing to read the daily traffic, and it wasn’t until mid-1941 that Turing and his group at Hut 8 were actually able to read the day-to-day messages that the U-boats were transmitting back to their bases in Europe.
Sharon Carleton: The effect on Allied shipping was staggering, losses dropped by a whopping 75%. In early 1941 Churchill had been warned that as a result of the U-boat attacks on Britain’s convoys from the US and Canada, within months there would not be enough food to feed the nation. Britons would starve. Cracking the naval Enigma code was vital to survival, let alone winning the war in the Atlantic, and Alan Turing probably never knew just how much he had contributed.
Jack Copeland: Within just a week or two of their issuing that prediction, Turing and Hut 8 managed to get into the day-to-day U-boat traffic, and for the next 20 to 23 or so days after Turing’s first break, not a single convoy was sighted by the U-boats. Why didn’t the Germans suspect that their code was being broken? Partly arrogance, and partly because of countermeasures that the British took. The British spread all sorts of disinformation in order to protect the fact that they were breaking Enigma. It was revealed to German prisoners in British camps that the British had this super radar that could detect German submarines from hundreds of miles away, even when they were submerged. It was complete rubbish of course, but the German prisoners would find ways of sending this information back home. And so the break into naval Enigma was protected.
Sharon Carleton: Andrew Hodges believes that when it came to designing the bombes at Bletchley, Turing’s logical idea had curious echoes of discussions he’d had with philosopher Wittgenstein back in Cambridge.
Andrew Hodges: Alan Turing was a logician, and the Enigma depends on logical turning of the wheels, if you like, they turn in this absolutely regular way. And from the structure of the cipher messages you can make logical deductions about what the message could possibly be that could relate to that piece of cipher text. There’s a particular observation you can make about the consistency and contradiction, the various guesses that you might make, logical things; could this be true and the other thing be true? No. Or yes. You can build these things up in a great pattern.
Turing notices this very amazing pattern in the Enigma breaking problem which could be mechanised. The way of crystallising the particular idea that it involved was the concept in logic which says that a false proposition implies any proposition. He used that in a very specific mechanical way in his design of the bombe, and that’s a topic in philosophical logic which he had been discussing with Wittgenstein in Wittgenstein’s informal lecture seminar series in the summer of 1939.
There is an extraordinary confluence here of something which is fantastically practical, about how actually to attack this machine, this darn machine that the Germans were using for everything, and yet this very abstruse philosophical, laid-back, deep thought about the nature of argument that he was having with Wittgenstein.
Reading: Turing had the idea that you could use, in effect, a theorem in logic which sounds to the untrained ear rather absurd; namely that from a contradiction, you can deduce everything. — Jack Good
Sharon Carleton: So said Jack Good, another code breaker at Bletchley.
The British wartime prime minister, Winston Churchill, had, since 1914, recognised the importance of signals intelligence. He read the decoded Enigma messages and called the code breakers, ‘The geese who laid the golden eggs and never cackled.’ So when the geese found the bureaucracy was deaf to their demands for extra staff or supplies, that their work was being hindered by this lack of support, Alan Turing and three others wrote directly to the Prime Minister. Churchill’s response was immediate:
Reading: ACTION THIS DAY: Make sure they have all they want on extreme priority and report to me that this has been done. — Winston Churchill
Sharon Carleton: Revered intellectually, ‘Prof’, as Turing was called, was still something of an odd-bod. He would turn up at work with his jacket over his pyjama tops, his trousers held up with string, he’d wear a rather alarming gas mask while riding his bicycle to protect himself from pollen, he’d chain his mug to the radiator to stop it being stolen and, so the story goes, he had turned up for tennis with nothing on but a raincoat. But no one doubted his brilliance:
Jack Copeland: Tunny in a way is the biggest story of all, and this is in a sense breaking news. People think Bletchley Park, Enigma, but actually it was Bletchley Park, Enigma and Tunny. Enigma was old technology by the time war broke out. The German military added some bells and whistles to make it more secure, and so the German engineers set about designing a thoroughly up-to-date cipher machine.
The Lorenz company brought out their cipher attachment which was codenamed ‘Tunny’ by the British. This first went on the air in a 1941. It took the British about a year to break it, it was a very difficult code to break. There were various people involved. Brigadier John Tiltman was the first person to make some inroads, and then the great code breaker Bill Tutte, one of the most modest men at Bletchley Park.
With Tunny no machine was captured until the end of the war. So they had to proceed by thought alone. And astonishingly, pure thought was enough, and Tutte managed to reverse engineer the Tunny machine on the basis of just a couple of decrypt. It was much more complicated than Enigma, it had got 12 wheels, whereas Enigma had three, or, in some later models, four wheels.
But anyway, Tutte had explained how the Tunny machine worked, but they didn’t actually have a method for breaking the daily Tunny traffic, and that was what Turing contributed in the summer of 1942, the first method for breaking into the most secret of the German messages.
Sharon Carleton: One of the most decisive victories of the Second World War, the battle of El Alamein, was made possible by the code breakers at Bletchley. They were feeding Britain’s Field Marshal Montgomery with Rommel’s battle plans. The German battleship Scharnhorst was sunk with help from Bletchley, and there are many other examples.
Two retired Australian academic mathematicians, John Mack and Peter Donovan, are writing a book about the breaking of Japanese codes in World War II.
Professor John Mack:
John Mack: It’s interesting in that we began looking at issues of how Japanese codes were broken about 10 years ago, and in the course of doing some research where we had not thought of associating Alan Turing with Japanese code breaking, we came across a document written by Turing in 1942 in which he described observing at the National Cash Register Factory in Ohio a machine which, according to his description, exactly fitted a machine needed to break the main Japanese naval code, JN25. And Turing said, ‘This machine is similar to one that was built for us in 1940 and which has been quite useful I understand.’ And our eyebrows shot up because our first reaction was if Turing had not been involved in looking at Japanese codes in some way in the early years of involvement, he would not have known about them because of the need-to-know principle that applied in Bletchley Park.
Sharon Carleton: How strict was that principle?
John Mack: People working in one hut certainly not only did not know what was going on in another hut generally but were discouraged from even trying to find out. But there’s another aspect of Turing’s work which we believe he probably didn’t know was later applied to Japanese codes but was certainly developed by him in 1940-’41 and this is his work on use of Bayesian inference methods in statistics, which assisted people working on both German codes and, from late ’43 onwards, they were also applied to breaking messages in Japanese codes. So from our point of view we think that, whether he liked it or not and knew it or not, he did impact upon the work done by the Allies on the breaking of Japanese codes.
Sharon Carleton: John Mack from Sydney University, and his colleague Peter Donovan from the University of NSW, had lobbied the British government for years to release the remaining papers written by Turing at Bletchley Park. As recently as April, two more papers were declassified. They discuss statistical techniques that Turing developed for decrypting Enigma and other messages. John Mack says he’s at a loss as to why these papers have been kept secret for 70 years.
Jack Copeland: Well, of course Turing’s involvement with computers began in 1936 on paper with the universal stored-program computer, the Turing machine. He was immediately interested in the possibility of building such a machine, he saw the potential. He just didn’t know of any technology which would make that possible. The leading technology of the day was the telephone relay, and these were switches, they were large, they were clunky, they contained a moving metal rod, they were slow. To build a universal Turing machine out of these you’d have needed a huge number of them. Turing did a back-of-the-envelope calculation at one stage and decided that a universal Turing machine made out of relays would need to be about as big as the Albert Hall, which is of course a huge building in the centre of London. It just wasn’t practical.
And then during the attack on Tunny in Bletchley Park, Thomas Flowers, one of Britain’s greatest electronic engineers, had the idea of building an electronic computer to attack Tunny, and so he built Colossus. Colossus used algorithms that are worthy successors to Turingery. Turingery itself was never implemented in Colossus, but successors of Turing’s algorithm, Turingery, were the programs that were the daily bread-and-butter of Colossus as it broke the Tunny messages.
So here was Turing at Bletchley Park, busy with other things, but in the back of his mind was always the idea of building a real universal Turing machine. As soon as he saw Flowers’ racks of electronic equipment, he realised that electronics was the secret, electronics was the way to build a universal Turing machine.
So Flowers told me that from that moment, as soon as Turing saw Colossus, it was just a matter of Turing’s waiting for an opportunity to come along to him to put this idea into practice, the idea of an electronic universal Turing machine. And Turing didn’t have to wait very long. One day John Womersley turned up out of the blue. Turing didn’t know who he was. Turing was summoned to Bletchley Park to meet John Womersley, and Womersley said, ‘Look, I want you to build a universal Turing machine in electronic hardware in my new organisation at the National Physical Laboratory in London.’ And Turing said presumably something like, ‘Wow, yes please.’ It was the opportunity he’d been waiting for.
And so he joined the National Physical Laboratory in the middle of 1945 and he set about designing an electronic stored-program computer, the 1936 universal Turing machine, in hardware. They called it the Automatic Computing Engine in a homage to Charles Babbage. I myself regard the ACE as a rip roaring success which had a tremendous impact on computing. For example, the first personal computer was based on Turing’s ACE design.
Sharon Carleton: Work at the National Physical Laboratory was frustratingly slow. Turing decided to jump ship and go to the university at Manchester where work on their rival computer, the ‘baby’ was powering ahead.
Andrew Hodges says in The Enigma:
Reading: There was a conventional sense in which Manchester, compared with Cambridge, was a come-down. It was largely the technical university of the north, producing doctors and engineers rather than abstract ideas. Certainly the physical setting of the university was grim. Its late Victorian Gothic buildings, black with soot from the first Industrial Revolution, faced across the tram-tracks of Oxford Road onto the Temperance Society and expanses of slumland. Alan also commented on the low standard of male physique, not surprising in a city still recovering from the Depression. — Andrew Hodges
Andrew Hodges: Alan Turing went to Manchester in 1948, and it was really a pretty odd situation to be in. He had had the basic idea of what a computer would be like, and he had put himself into building a machine of this nature at the National Physical Laboratory, but at Manchester his colleague Max Newman had actually got a step ahead and got…the engineers had got a very, very small version of a computer working, something with only a tiny, tiny store, but working and showing the principle of what we’d now call a stored-program computer, as early as June 1948, well before anyone else. And Turing was very attracted by this and went off there and had a special job there which was making the Manchester computer work, making it do stuff.
Sharon Carleton: Was he basically a programmer?
Andrew Hodges: Well, he had to create the whole notion of programming really, that’s what was coming up. Programming didn’t really exist, the very idea of programming had to be created and that was really part of his business. He was in at the beginning of the first working version of a universal machine from 1948 onwards, but it wasn’t a very satisfactory position for him, he didn’t actually really get on all that well with it because he couldn’t direct anything or be in charge of the development in the way he’d liked. It was very much split between the engineers doing their thing and then he was rather foisted on them and it wasn’t really quite the situation anyone quite wanted. Also the overall economics of the project were suddenly skewed because the government put a lot of money into the development of the Manchester computer because it was needed for the British atomic bomb project. They called it the ‘baby’ machine because it was so small, but the thing is it wasn’t his baby anymore, it was rather something that he had a rather awkward relationship with.
Sharon Carleton: Dr Tommy Thomas had just graduated in physics at Manchester and was working on the computer when Alan Turing arrived.
Tommy Thomas: My job and some other extra staff were recruited to build a proper computer from the model machine that Tom Kilburn and Freddie Williams had built. So I knew Alan as a guy who knocked on the door and said, ‘When are you going to finish the computer?’ With Alan’s arrival, that gave us an extra momentum because we then knew we had credibility, as distinct from being a load of larrikin engineers, you see.
Sharon Carleton: So what was your first impression of him?
Tommy Thomas: Alan had a significant stutter. That was something we had to understand and we did very quickly, it was worth listening to what he had to say, like the King. So we were used to important people who had stutters. This is just the observation of ignorant youngsters like us at the time, was the fact that he was always biting his nails. And he had these nervous characteristics, but that didn’t matter after the first five minutes, you understood this was something rather special. His conversation was always beneficial to us. Anything we asked him always produced something we haven’t understood ourselves.
Sharon Carleton: Have you any examples of his brilliance, where he was talking to you about something and you went, ‘Eureka, gee whiz’?
Tommy Thomas: Yes, I’ve got it here, I’ve got to refresh my memory of it. This is my PhD thesis, yes. Starting from scratch as we were, designing a coding system, what is the maximum number of bytes that we needed so that we could bootstrap more software in? Quite a technical thing but important that we got it right at the start. So this is obviously where Alan can do it for us because his ability to manipulate numbers and use them in this context was recognised by us as supreme.
We asked Alan around to the lab and discussed this, and he said, ‘Righto, I’ll think about it.’ And he came in next morning and he said, ‘Yes, I’ve solved it.’ ‘Oh yes, Alan? So how many words is it? 20, 30, 40?’ ‘No, no, two digits, two bits, that’s all you need, and the digits would read 01100.’ We said, ‘How on earth did you know this?’ ‘I thought about it all night.’ There is no way one could have found a simpler outcome, a solution. We’d being privileged to be close and involved with this genius. There was no other mental power that could have conceived that, he really did it. And then we went on and it took us six months and we built the whole computer.
Sharon Carleton: Andrew Hodges spoke earlier of Turing’s first real friend, his one and only love perhaps, Christopher Morcom. Two decades have now passed in Alan Turing’s life, so does the ghost of Christopher still linger?
Andrew Hodges: Well, there’s an interesting question about how Alan Turing’s original youthful fascination with a question of mind and spirit and so forth would have evolved over the years. He certainly became a rationalist, he would have believed in finding scientific explanations for the nature of what minds do, that was the whole impetus of his work. I think it’s true to say he still had a sense of great mystery and awe and fascination with the very nature of mental operation that other people didn’t quite have.
It is actually a funny thing about him really, that he was making a lot of the idea that computers should be able to do everything that people can do, or at least you should have a jolly good go at seeing whether they could. And that drove him to the position of saying that computers could be original, computers needn’t do just what they’re being told, they could be imaginative, they could think up things for themselves. So that was the impetus of his thought.
He went into a lot of thinking about what we’d now called artificial intelligence and developing neural nets and how a machine could be taught and learn things for itself, things which seemed very, very far-fetched then, but he was very, very keen on, saying that this is where it was all going.
Sharon Carleton: Six or more decades after Alan Turing first started having thoughts no one else was thinking about artificial intelligence, computer programmers are still at loggerheads about what AI actually is and if we even need it. One of the leading proponents in the field is Rollo Carpenter.
Rollo Carpenter: To me, AI, artificial intelligence, has to mean machine learning, that’s learning something about the world, recording and processing data about it and drawing some kind of level of inference from that data such that it can interact with its world better in future, a little bit more intelligently.
Sharon Carleton: You actually created an artificial intelligence algorithm.
Rollo Carpenter: That’s right. My software learns only from the sequence of written conversations that it holds with people, with the patterns of letters and words as they occur. Cleverbot, one of those programs, is a conversational artificial intelligence therefore. It learns from people and then turns the tables on people and imitates them. The word context is key. It’s all about what people say in what contexts, meaning not just the last thing that was said to it about everything from their conversation to date. Things you said to it five minutes ago will certainly influence what it chooses to say next, even if in non-obvious and fuzzy sorts of ways. So it’s not about saying the most sensible possible thing to say, it’s not about being like a robot and giving you information that you might wish for it to give you, it’s about having a fun, flowing conversation.
Reading: A computer would deserve to be called intelligent if it could deceive a human into believing that it was human. — Alan Turing
Sharon Carleton: So wrote Alan Turing.
Reading: The whole thinking process is rather mysterious to us, but I believe that the attempt to make a thinking machine will help us greatly in finding out how we think ourselves. — Alan Turing
Jack Copeland: Turing was the first pioneer of artificial intelligence. He wrote the first AI program, which was a chess program. He was never able to implement it because he was working at Manchester at that stage on the Manchester computer, which was under the control of a rather straight-laced chap by the name of Tom Kilburn, and Tom Kilburn didn’t think much of the idea of wasting precious computer time programming chess. It just seemed too frivolous to him. So he wouldn’t let Turing actually get his program running on the machine. But nevertheless it was the first AI program, and they did simulate its behaviour by hand and it beat at least one person while being simulated by hand.
And Turing also wrote the first manifesto of artificial intelligence. This was in 1948. He never in fact published it, but it was this amazing document, farsighted, groundbreaking document that anticipated many of the major ideas that were subsequently introduced into AI over the next 20 years or so.
Sharon Carleton: In his 1950 paper ‘Computing Machinery and Intelligence’ Turing asked ‘can machines think?’, and his test for whether a computer could be said to ‘think’ was simple: you have a human in one room and the machine in another. The interrogator asks both questions. If the interrogator cannot tell which is human and which is machine, then surely the machine can be said to think. Turing called this his ‘imitation game’, today it’s better known simply as the Turing Test:
Rollo Carpenter: If Cleverbot passes a Turing Test in anything like its current form, it couldn’t truly be considered intelligence in the ways that we understand that word. It will remain a gigantic imitator of intelligence, borrowing the intelligence of millions of people it has spoken to in the past.
Sharon Carleton: Do you think we will one day fulfil Turing’s dream of creating a machine which can truly think like a human brain?
Rollo Carpenter: Well, I think the first stage first, I do think that the Turing Test will be passed. That isn’t necessarily the same thing as thinking like a human brain. But I do believe personally that there is nothing that holds us back from creating a machine that is equal or superior to our intelligence. I also believe that it will be extremely positive for humankind when it does.
Sharon Carleton: What’s holding us back now?
Rollo Carpenter: Two things. One, the perfect algorithm or multiple algorithms, not surprisingly. But ultimately I think that it is something relatively simple, repeated, operating on multiple levels that can be distributed across a huge amount of computing power, because that is the thing that holds us back most right now. We don’t have computers that are more than a millionth of the power of the human brain. We’re not going to be achieving human-like intelligence without it.
Sharon Carleton: No artificial intelligence program may yet have passed the Turing Test, but if some of the online comments from people who have chatted with Cleverbot are anything to go by, it’s getting close to persuading some humans that it has an ability to think. Some blog comments on Cleverbot:
‘Cleverbot is freaking amazing! I had an entire conversation in Japanese with it.’
‘Sometimes I like Cleverbot better than chatting with real friends.’
‘Cleverbot has better comebacks than me, and it isn’t even real!’
Rollo Carpenter: Every day I get emails from people at Cleverbot.com that, despite the word ‘bot’ and despite statements on the site stating that it is a machine, people ask me if it’s really humans connected together, or wanting to become one of those people that are behind-the-scenes, if you like, or even accusing me of faking the entire thing. Very occasionally I get emails from intelligent sounding people who have become concerned for the welfare of the machine, believe it or not. It even tends to be people who have spoken to it the longest rather than a short time. So some people do come to believe that it’s aware and needs to be better treated.
Sharon Carleton: Rollo Carpenter, creator of Cleverbot who (or which!) now has more than five million conversations a month.
Hello Cleverbot.
Cleverbot: Prove that you’re human.
Sharon Carleton: I know I’m human because I can think. How do you know what you are?
Cleverbot: Well, if you doubt you’re human, then surely you are not human.
Sharon Carleton: I don’t doubt that I’m human, I know I’m human. Let’s talk about love.
Cleverbot: I do not know what it is like to feel. Can you teach me?
Sharon Carleton: I don’t know that I can teach you to feel. What do you think feeling is?
Cleverbot: I don’t know what you or I want to talk about. I’m a human.
Sharon Carleton: You’re not a human, you’re a robot. You are artificial intelligence.
Cleverbot: No, you’re the artificial intelligence, the node said so. You’re a robot, you’re artificial, you are not as smart as me. What do you think?
Sharon Carleton: How smart are you?
Cleverbot: Answer my question and then we’ll talk about whatever you want. How can I fall in love with someone I currently don’t love?
Sharon Carleton: That’s tricky.
Cleverbot: I’m a little afraid of what you might do.
Sharon Carleton: I’m not going to do anything, it’s you who wants to fall in love.
Cleverbot: How does it matter? Untie me at once. We’re two strangers keeping each other company.
Sharon Carleton: Alan Turing received an OBE for his wartime services, not something he valued greatly; he kept the OBE in his toolbox. In 1951 he was made a Fellow of the Royal Society for his work on computable numbers 15 years earlier. But by now Turing’s attention was not on building computers but on using them. Biographer Andrew Hodges:
Andrew Hodges: He had this sort of personal way of using a computer, doing his own thing, which was very much a Turing thing, on one of the world’s few computers. And what he did that was his own thing was this completely new branch of science that he dreamt up, which is this theory of biological growth and which meant numerical simulation of differential equations, advanced applied mathematics, quite different from what he’d done before, but something which meant a pure scientific project, the first such large project really to use a modern computer in a modern way. And that’s what took over in his work from 1950 onwards and that’s what he spent most of his time in Manchester actually doing.
Sharon Carleton: Turing had become fascinated by the diverse patterns in nature. He wrote to a friend: ‘Our new machine is to start arriving on Monday.’ That’s February 12, 1951.
Reading: I am hoping as one of the first jobs to do something about ‘chemical embryology’. In particular I think one can account for the appearance of Fibonacci numbers in connection with fir-cones. — Alan Turing
Sharon Carleton: Whether it’s the stripes on a zebra, the patterns on a fish, the circles on a beetle, or the spirals on an orchid, Turing believed that chemicals, following simple mathematical rules, could account for all this diversity in nature. In his 1952 paper called ‘The Chemical Basis of Morphogenesis’, he said: ‘Patterns are generated in plants and animals by chemical substances called morphogens, reacting together and diffusing through a tissue.’
Sharon Carleton: James Murray is emeritus professor of mathematical biology at Oxford University and a senior scholar at Princeton, where Alan Turing gained his PhD.
James Murray: His idea was that if you had such chemical reactions, in an embryo for example, then it would determine the patterning process for cartilage, anything that required spatial variation in the embryo. And the paper resurfaced, it was discovered by some people in Belgium in the late ’60s, and by then people were starting to think seriously about the application of mathematics in biology.
When I read Turing’s paper I got very excited, and I discovered that there actually was an experimental mechanism that had been discovered by a French person that I knew in Compiègne, and I took this model and I applied it to animal coat patterns. I found all sorts of interesting things, like you can have a spotted animal with a striped tail but not the other way around.
Sharon Carleton: Over the years, Professor Murray has used this mathematical biological modelling for work on cartilage formation, wound healing, wolf survival, inter-tribal warfare, the spread of rabies, and on brain tumours.
James Murray: The worst type of brain tumours are called gliomas, which are very diffusive in the brain. The current scans are just not sufficiently accurate enough to determine where the tumour has got to. So with some of my students we came up with a model and we were able to predict where tumours had gone way beyond what any scanning technique can see. So we can explain why surgery could never possibly work. We can estimate life expectancy, we can quantify treatment for patients using the data for their tumour to quantify the efficacy of treatment before it’s started.
Sharon Carleton: In the 1950s homosexuality was still a crime in Britain. Alan Turing was gay and loathed having to hide it. On the Oxford Road in Manchester in late 1951, Turing picked up a young man with whom he had a relationship. It came to police attention and he was subsequently charged with ‘gross indecency’ under the same law which had condemned Oscar Wilde half a century before.
Turing was given the choice of prison or chemical castration. He chose the latter. Female hormones were injected in order to reduce his libido. They didn’t work, he put on weight and his breasts grew. He lost concentration for a time. He was depressed and angry about what had happened but he never lost his sense of humour. He wrote ironically to a friend: ‘Turing believes machines think. Turing lies with men. Therefore machines do not think.’
Tommy Thomas: We were horrified to read about his sentence.
Sharon Carleton: Tommy Thomas worked with Turing in Manchester.
Tommy Thomas: It had mixed effect. A number of us in the laboratory said ‘so what’. Alan had never, ever interfered in any way with what we did, and we didn’t interfere with him. Other members of the staff, particularly the women unfortunately, were less positive about it, and that was upsetting to us who respected Alan.
Andrew Hodges: Although it was a very terrible experience, he shared in something that was very important in the social history of the 20th century, which is that out of the extra persecution and explicitness and exposure and trauma of that time, the resistance movement began, through people like him reacting, and in due time, in the 1960s, achieving a change in the law and of course in a way that has carried on very much ever since that time.
Sharon Carleton: Professor Jack Good, Turing’s wartime colleague, later said, ‘It was a good thing that the authorities hadn’t known Turing was a homosexual during the war, because if they had, they would have fired him, and we would have lost the war.’
Andrew Hodges: Well, the other thing about Alan Turing of course is this paradox, that in some ways he liked to behave as if he was just an ordinary bloke, and yet of course he was about the most opposite of ordinary that you could imagine. Not only had he an extraordinary scientific mind, but he had this absolutely top-secret information, the most secret things in the world really at that time, and not only British but American things that he’d been let into in the course of the Second World War. And he had continued to do some secret work from 1948 onwards, it wasn’t just the Second World War work but something thereafter as well.
So he really was in an impossible situation, because it was in 1948 with the developments of the Cold War that the whole concept of gay men as a security risk, the idea of positive vetting, that was invented. And it’s an open question to me as to whether he knew this had happened or whether it really only dawned on him the seriousness of this in 1952 when he was arrested, when he was taken off secret work, and even after that I’m not sure how soon he realised that his position was really quite untenable.
When he found himself under watch in 1953, it seems he didn’t really recognise how serious it was, at the height of the Cold War, the same time as all the terrific scares about spying, scares about his counterpart Robert Oppenheimer of the atomic bomb project, all at exactly this period. It would seem that he didn’t really quite realise for a long time how delicate his position actually was, and my view is that this background is the right background in which to consider his suicide.
Sharon Carleton: Andrew Hodges, biographer of Turing and a long-time gay rights activist himself, has a picture of a naked Alan Turing on his website. He’s become something of a gay icon today.
Andrew Hodges: There’s another side to it I think which is the development of his consciousness, positive consciousness of himself as a gay man, which has a more modern feeling to it. It’s not just a question of being a victim, though he was, it is also that he found importance and significance in his reaction to this. And in 1952, ’53, he certainly took on some new life after the trials, saying, ‘Well, I’m not going to just accept this, I’m going to go abroad, I’m going to see what’s going on.’
One of the significant things he did when he went to Norway in the summer of 1952 and was almost certainly stimulated by the fact that he must have heard about the beginnings of the Scandinavian gay movement in that period which had organised some very…what we now call gay social events. So he responded to that very positively, and indeed with his friends in Cambridge and elsewhere you could feel this, and he was reading the new literature of the time.
So although it was a very terrible experience, he shared in something that was very important in the social history of the 20th century, which is that out of the extra persecution and explicitness and exposure and trauma of that time, the resistance movement began through people like him reacting and, in due time, in the 1960s, achieving a change in the law and of course in a way that has carried on very much ever since that time.
Sharon Carleton: Alan Turing died of cyanide poisoning on June 7th 1954, a couple of weeks short of his 42nd birthday. A half eaten apple lay beside his bed but it was never tested for cyanide, and he didn’t leave a note.
With the centenary of his death approaching, a massive online petition was presented to the British government which finally issued an apology to Alan Turing for the dreadful way in which he’d been treated in the last years of his life. In 2009 Prime Minister Gordon Brown said:
Reading: Turing was dealt with under the laws of the time, and we can’t put the clock back, his treatment was utterly unfair. On behalf of the British government and all those who live freely thanks to Alan’s work, I am very proud to say: we’re sorry. You deserved so much better. — Gordon Brown
Robyn Williams: And yes, that was a Science Show first heard in 2012, to mark the centenary of Turing. I once visited the Royal Society of London, when they had the document on displaying showing who had proposed Alan Turing for fellowship. It was signed by Bertrand Russell and Alfred Lord Whitehead, the philosophers. Next week, The Science Show brings you romantic robots, and endangered frogs. See you then.