Fri June 29, 2012
NPR: Alan Turing turns 100
Originally published on Fri March 1, 2013 11:17 am
IRA FLATOW, HOST:
This is SCIENCE FRIDAY. I'm Ira Flatow. Your telephone is a computer, really. Your microwave, it's got a computer in it. Your television, it's got a computer there. Even, of course, your computer has a computer. Your iPhone, your cellphone. Everything - just about everything in electronics these days has a computer, and they all work the same way like a Turing machine. Decades before your PC, your Mac or your Commodore, Alan Turing was designing a machine which could calculate almost anything: a universal computer.
His machine made breaking the Nazi enigma code possible, saving millions of lives as it helped the Allies win World War II. But Turing wondered if a machine could calculate anything, could it calculate thought, and this idea gave birth to the field of artificial intelligence. But Alan Turing was gay, a taboo that led him to his arrest and chemical castration by the British government. He was found dead at the age of 41, ruled a suicide by the coroner.
Turing feared that his work would be forgotten and dismissed because, as he put it, Turing believes machines think, Turing lies with men, therefore machines do not think. Here with me to remember both a war hero and a scientist reviled in his time is David Leavitt, professor at the University of Florida in Gainesville, author of the book "The Man Who Knew Too Much: Alan Turing and the Origins of the Computer." Welcome to SCIENCE FRIDAY.
DAVID LEAVITT: Thank you.
FLATOW: Let's talk about his achievements. Tell us what made him most famous and his biggest achievements were.
LEAVITT: Well, it really depends on, I think, your own perspective. Turing really had four major achievements in my view. One was that he resolved a major problem in mathematical logic, the so-called decision problem. In order to do that, he invented a kind of hypothetical universal machine that was the avatar of the modern computer. He was responsible for the building and the design of the machine that broke the enigma code during World War II. And with some of his later papers, he really established the field of artificial intelligence.
So in some ways, the question what was his major achievement, it sort of depends on who you ask. I think for anyone to have done any of those things in a lifetime would have been extraordinary. To have done all four is mindboggling.
FLATOW: Yeah. What is a Turing machine?
LEAVITT: A Turing machine is a very simple hypothetical machine that is - initially, the idea of the Turing machine was that it would - was programmed to solve a particular algorithm. Turing then went on to prove that there could be what he called the universal machine that is to say a machine into which instructions could be fed by which that machine could resolve or solve or approach or try to solve any algorithm, an infinite number of algorithms. And he actually proved that such a machine could hypothetically exist. And so the Turing machine is that machine. It's that machine that can be turned into any other machine, and that was his huge breakthrough.
FLATOW: And there was also a test that he devised, was there not? About whether you could tell if somebody in a different room was a machine or a person.
LEAVITT: Yeah. This was one of Turing's most, I think, radical and interesting ideas. He posited that the way to determine whether a machine could think was whether that machine could convince a human being that it was thinking. And so he proposed this game, which he called the imitation game, which is actually now played annually at a sort of competition, where there'd be two human beings and a computer, and one of the human beings would be asking questions of the other human being and the computer. And based on the answers, he or she would have to decide which was the computer and which was the human being. If the computer could fool the interlocutor into thinking that it was human, then according to Turing, the machine could think.
FLATOW: Did he actually build the device that would fool people that way?
LEAVITT: He was kind of moving that way when he was in Manchester, when he was actually working on computers. Turing had a very, I think, restless imagination, and I think he was much more interested in the theoretical foundations of things. It seemed like whenever he was onto something, he would want to move to something else.
FLATOW: Yeah, yeah.
LEAVITT: So he was going on into biology toward the end of his life, into an entirely different area of inquiry.
FLATOW: 1-800-989-8255 is our number. Talking about the life and times of Alan Turing. And if you'd like to talk about it, we love to take your calls. You can also tweet us, @scifri, @-S-C-I-F-R-I, talking with David Leavitt, author of "The Man Who Knew Too Much: Alan Turing and the Origins of the Computer." "The Man Who Knew Too Much," what are you getting at with that title?
LEAVITT: That was a - that's a pretty coy title, obviously. I was referring to the Hitchcock film, "The Man Who Knew Too Much." But I meant it in two ways. On the one hand, he knew too much for his own good. I mean, to possess that quantity of knowledge, I think, could be a sort of overwhelming experience for anyone.
He also knew too much in the sense that he was perceived by the British government in the years after World War II as a security risk because of the quantity of classified information that he had in his possession. And, therefore, he was really hounded by the British police who were concerned that because he was gay, he would - he could be easily blackmailed or seduced into going over to the East.
Remember, this was the sort of epoch of Guy Burgess. So his knowledge, as that little syllogism that you quoted implies, he really felt that his knowledge was going to somehow be his undoing and that his homosexuality was going to interfere with getting the word out about the - what he believed what and he discovered.
FLATOW: Mm-hmm. He never kept his homosexuality.
LEAVITT: He was very open about it, but I would not say that he was particularly - he didn't really advertise it. His attitude was that it was perfectly normal and not a big deal. And so he behaved as if everyone else felt the same way, which was obviously a big mistake at that time.
FLATOW: Tell us - talk a little bit about that, that it was a crime to be a homosexual at that point - that period and to act out in your homosexuality.
LEAVITT: Yeah. There was a law on the books that was colloquially known as the blackmailer's charter because it criminalized, quote/unquote, "acts of gross, indecency between adult men in public or private." And it was under this law that Oscar Wilde was arrested and sent to prison. The law was obviously a terrible law and it could be - it was applied very selectively.
Turing had been having an affair with a young man in Manchester whom he suspected of having robbed him. So he reported the robbery to the police, who turned around and arrested him under the auspices or under the terms of the blackmailer's charter. He was then given a choice between serving prison time or undergoing a, quote/unquote, "cure" for his homosexuality, which consisted of massive doses of estrogen and amounted to what we would now call chemical castration. Even after this period when the so-called treatments had ended, the police followed him everywhere and made it very, very difficult for him to have anything like a normal life.
FLATOW: And he was found dead at a young age, right?
FLATOW: He was found dead in his room with clues like a mystery novel. There was a half-eaten apple next to his bed, and the coroner ruled that it was a suicide.
LEAVITT: Yes. It was - I think it's most widely believed now and - there are three possibilities. One is that he committed suicide. One is that his death was accidental. And the third is that it was set up, that he was actually killed and that it was contrived to look like a suicide. That third, I think, has pretty much now been dismissed.
The suicide idea I think grabbed people's imagination because Turing loved the film "Snow White and the Seven Dwarfs." And when he was at Cambridge, he would often quote from the evil queen's incantation, dip the apple in the brew. And so the idea that he'd committed suicide by eating an apple dipped in cyanide, you know, was incredibly terrifying, but also kind of, you know, created a - to led the whole mythology, including the myth that the apple and Apple computers was a reference to Turing.
It's really unclear. I don't think that an absolute answer can be given. However, my own feeling is that the argument that he died accidentally because he accidentally got cyanide on the apple that he went ahead and ate kind of reminds me of the old thing about Rose Mary Woods' argument when she was Nixon's secretary about how she erased the tapes.
FLATOW: Yeah, but - yeah.
LEAVITT: It seems kind of a - it seems at least as improbable as the suicide theory.
FLATOW: There are people who have said that the investigation of his death really was not carried out correctly, that the evidence was not looked at. And there have been people who have looked at the coroner's report in years since then and say that the evidence points to more of inhalation of the cyanide rather than ingestion by eating it.
LEAVITT: Yeah. You know, the science of sort of historical forensics is so fascinating, but it's also so frustrating. I also wrote a book about the mathematician Ramanujan, whose death was very mysterious. And then in subsequent years, people have proposed a lot of different theories.
I don't think there's a clear and absolute answer. It doesn't - the idea that he would commit suicide certainly does not seem improbable to me. He was prone to depression throughout his life. And while it's true that there were no warning signs, that he didn't leave a note, that he seemed to be doing fine, that's also true of a lot of people who commit suicide.
So I think we can't really reach an absolutely definitive conclusion. And in some ways, the argument that's going on now I think has as much to do with what people want Alan Turing to be, as with the question of who he actually was.
FLATOW: Mm-hmm. Let's see if I got a call in here. From Troy in Denver. Hi, Troy.
FLATOW: Hi there.
TROY: How are you doing?
TROY: Can you hear me?
FLATOW: Yes, go ahead.
TROY: Actually, I am just now leaving work and I couldn't believe you guys we're talking about Turing. I'm actually wearing a shirt right now that says, I failed the Turing test. We're (unintelligible) humor. But anyways, I'm calling because I actually didn't encounter Turing as a computer science major. I encountered him as a philosophy major, reading an argument called "The Chinese Room Argument," which hypothesizes a room in which you could put someone who only speaks English and then give them a book that has an infinite amount of responses to anything that could possibly be asked of the room, any instruction that could be given.
And the idea is you could give this guy Chinese - you could give this guy information in Chinese, and he could follow the instructions in the book and correct responses but have absolutely no idea what it was that he was actually doing. And in such a way, you could create a machine that could pass the Turing test without actually having to think, with no idea what it was doing.
FLATOW: Let me just remind everybody that this is SCIENCE FRIDAY from NPR. What do you think of that?
LEAVITT: You know, it's an interesting argument, that's John Searle's quite famous paper. And I think it is certainly one that's very worthy of consideration. Turing anticipated that argument and his position was that, essentially, well, yes, you can say to this computer, if you don't know that you're thinking, then you're not thinking. If you're not aware that you're not thinking, then you're not thinking.
But Turing said, you could also say that to me. I could say that to you. I could say, I don't have proof that when you give me answers to questions, you're actually going through a thought process and not just copying Chinese characters. Give me that proof. And Turing's point was that that's impossible. From his point of view as a mathematician, someone trained in mathematical proofs on a sort of reductio ad absurdum basis, if something functions in a way that suggests it's true, then it is true. This was also a very anti-Christian argument and that kind of negated the idea of the soul so - but he did anticipate that argument. And in his paper, he does, in a sense, refute it in advance.
FLATOW: But why do people like our last caller, why do they have to stumble on Alan Turing? And they - he's not a well-known computer guru.
LEAVITT: He - I think his reputation was if not suppressed, then sort of conveniently forgotten for many years because of the scandalous nature of his - what happened to him in the last years of his life. It was considered unseemly, and people didn't really want to talk about it. Turing was also the opposite of a self-promoter. He was very shy. He was very awkward. He didn't have any skills of putting himself across. And so, therefore, he didn't do much to ensure that he would - that his reputation would survive. I think the third element was the fact that so much of the work he did was classified, and, therefore, people actually couldn't get access to the information.
And it wasn't until Andrew Hodges' wrote his biography and Martin Davis, a professor at Berkeley, began doing a lot of research on the history of computers that Turing's - the incredible importance of Turing's role really came to light.
FLATOW: So eventually, the British government did apologize for what they did but...
LEAVITT: Apologized but not - but did not pardon him retroactively. Gordon Brown issued an apology. Subsequently, there was a motion in the House of Lords to issue a posthumous pardon, which was refused within the House of Lords. It was voted down on the grounds that even if what Turing - the crime he was convicted of would not be considered a crime now, it was considered a crime then. And, therefore, he was guilty because he knew it was a crime at the time. This is, to me, an absolutely outlandish and shameful position for the British government to take.
FLATOW: Leave it to the politicians again. Thank you very much, David.
FLATOW: David Leavitt...
LEAVITT: You're welcome.
FLATOW: David Leavitt, author of "The Man Who Knew Too Much: Alan Turing and the Origins of Computer." Transcript provided by NPR, Copyright NPR.