Turing Revisited
by Brett Humphreys
As the twentieth century draws to a close, Alan Turing remains one of its most notable British gay atheists.
FRS OBE (1912-1954)
October 2000 marks the fiftieth anniversary of the publication of Turing’s classic paper Computing Machinery and Intelligence, now available in several places on the Web, in which he considered the question “Can machines think?". The paper was admired by the humanist philosopher and mathematician Bertrand Russell, who co-sponsored Turing’s election as a Fellow of the Royal Society the following year. Turing proposed to decide the question of machine intelligence by means of an “imitation game” (now known as the Turing Test), in which a human interrogator is allowed to address arbitrary questions to another human and a computer both concealed in a separate room. The interrogator’s objective is to deduce from the replies which is which. The computer’s objective is to deceive the interrogator into getting it wrong. Turing offered this bold prediction:
I believe that in about fifty years’ time it will be possible to programme computers, with a storage capacity of about 109 [bits], to make them play the imitation game so well that an average interrogator will not have more than 70 per cent chance of making the right identification after five minutes of questioning. The original question, “Can machines think?” I believe to be too meaningless to deserve discussion. Nevertheless I believe that at the end of the century the use of words and general educated opinion will have altered so much that one will be able to speak of machines thinking without expecting to be contradicted. I believe further that no useful purpose is served by concealing these beliefs. The popular view that scientists proceed inexorably from well-established fact to well-established fact, never being influenced by any unproved conjecture, is quite mistaken. Provided it is made clear which are proved facts and which are conjectures, no harm can result. Conjectures are of great importance since they suggest useful lines of research.
The particular significance of the Turing Test from a humanist perspective is its relevance to the dualist belief that humans (but not other species, at least according to Christian theology!) have some kind of soul or spirit capable of existing independently of the body. “Is a soul greater than the hum of its parts?”, as Douglas Hofstadter wittily put it. A machine passing the Turing Test would do much to help discredit the dualist view – although, as Turing himself wryly pointed out, a god who can confer souls on people should have no difficulty in giving one to a machine in such circumstances. However, while Turing’s forecast of the memory capacity of present-day computers has proved amazingly accurate, his prediction of their intellectual capacity has unfortunately turned out to be too optimistic.
Since 1991 an annual Turing Test contest has been sponsored by Hugh G. Loebner, a flamboyant United States philanthropist whose campaigns include decriminalisation of prostitution and publicising the fact that Olympic “gold” medals are actually made of gilded silver. Loebner offers a one-off Grand Prize of $100,000 plus a (real) gold medal for the first computer to pass the test, with an interim prize of $2,000 for the entrant that comes closest each year. The often entertaining transcripts of the winners’ dialogues with the judges tend to suggest that the judges are not trying very hard, but nonetheless it seems that Loebner’s $100,000 is pretty safe for the time being. Next year for the first time the Loebner Prize Contest comes to the United Kingdom, where the London Science Museum has agreed to host it for the next 44 years – unless the Grand Prize is won in the meantime.
The philosophical implications of the Turing Test have been the subject of lively controversy over the last half century, ranging from the scepticism of John Searle and his renowned Chinese Room to the support of Douglas Hofstadter and his 1981 Coffeehouse Conversation. Some of the debate can be found via the Turing Test Page maintained by Ayse Pinar Saygin, formerly at Bilkent University, Turkey, and now at the University of California, San Diego. An interesting, albeit slightly heavy, item on this site is the review paper Turing Test: 50 Years Later [PDF, 410 Kbytes] by Saygin and colleagues at Bilkent. For something a little lighter, take a look at The Automatic Confession Machine: A Catholic Turing Test, an amusing satire on Catholicism by Greg Garvey.
An important benefit of the Internet is its ability to open up easy access to information previously available only to people visiting specialist libraries or archives. A welcome development in this respect is the Turing Digital Archive, based at the University of Southampton. Since June 2000 this has provided web access to high quality digital scans of mainly unpublished personal papers from the Turing archive held at King’s College, Cambridge.
Another recent development is the online Turing Archive for the History of Computing set up by philosophers Jack Copeland and Diane Proudfoot at the University of Canterbury, New Zealand. Scheduled for completion by the end of this year, like the Turing Digital Archive it provides facsimile copies of original documents. One of them is the typescript of Alan Turing’s 1948 report entitled Intelligent Machinery, which remained unpublished for 20 years. As a precursor to Computing Machinery and Intelligence it gives an insight into his thinking leading up to the famous paper.
But the definitive site for information about Alan Turing remains the excellent Alan Turing Website maintained by Andrew Hodges, whose powerful biography Alan Turing: the Enigma, first published in 1983, is republished in a new edition in October 2000. The Alan Turing Website was the inspiration for the first of this series of Web Watch columns back in Spring 1997. The structure of the site remains essentially as it was then, but information is constantly added and updated. New pages since 1997 include Hodges’ oration given at the unveiling of the official Blue Plaque at Alan Turing’s birthplace in London on what would have been his 86th birthday, 23 June 1998.
Indeed it seems that memorials to Alan Turing are in vogue. In December 1999 an Alan Turing Award for Human Rights was established by Rank Outsiders, the UK support group for gay, lesbian and bisexual armed forces personnel. And the website of the sculptor Glyn Hughes has details of the Alan Turing Memorial project, whose aim is to install a life-size silicon-bronze statue of Alan Turing next year in Sackville Park, Manchester – the city where Turing programmed the world’s first working stored program electronic digital computer in 1948, and where he later made the fateful pick-up that was to lead to his prosecution and conviction for “Gross Indecency contrary to Section 11 of the Criminal Law Amendment Act, 1885” two years before his suicide at the age of 41.
The true gross indecency of that vicious piece of Victorian legislation is the blight it has caused to the life not only of Alan Turing but also of countless others over the last century and more. It’s good to observe that for the first time there is now a real prospect of that law being properly repealed. In July the Home Office published a consultation paper, Setting the Boundaries: Reforming the law on sex offences [PDF, 723 Kbytes], along with an abridged Summary Report and Recommendations [PDF, 136 Kbytes]. Both are available on the Home Office website for public comment by the end of February 2001. One of the more refreshing, if long overdue, recommendations is the simple proposition that “the criminal law should not treat people differently on the basis of their sexual orientation”. Setting the Boundaries sets out the first, and quite possibly the last, major review of the UK’s criminal law on sexual activities for many years. Let’s hope that a new millennium really does herald a new attitude.