Artificial intelligence put to the Turing Test
Loebner Prize founder on why he's backing AI
Published: 10 October 2008 11:41 GMT by Tim Ferguson
Show related articlesThis weekend could potentially be a historical one in the development of artificial intelligence (AI) as the tech world gears up for the latest annual Loebner Prize contest.
The prize was founded in 1990 by academic Hugh Loebner and this year's challenge - to be held at the University of Reading on Sunday 12 October - is the 18th annual contest.
silicon.com caught up with Loebner, who was recently voted into silicon.com's Agenda Setters 2008 list, to find out why he has been backing the prize for so many years and what the future might hold for AI.
A former academic, with a career that included stints as director of computing at the University of Maryland and a consultant to the State of Massachusetts Department of Employment Security, Loebner has little to do with AI research these days but remains a keen champion of its development.
Loebner's set up the prize as a result of a life-long interest in AI but also due to a desire to use technology to achieve a world where no human needs to work any longer.
Loebner has been backing the contest since the inaugural event in 1990 and remains firmly behind the cause. "I know it's kind of quaint nowadays not to give up easily. I guess I'm somewhat obsessive compulsive but I'm committed to my word," he told silicon.com.
The prize uses the Turing Test to determine whether a computer is able to demonstrate human-like intelligence during five minutes of questioning from a number of judges.
The prize is judged by a panel that includes a journalist, a professor of computer science, a psychologist and a philosopher.
The Turing Test originates from a 1950 paper written by computing pioneer and World War II code-breaking genius, Alan Turing, in which he suggested machines are capable of thought.
Turing came up with the idea that if a computer can respond in a way that is indistinguishable from a human - as examined in the Turing Test - it can be said to be intelligent.
Although the $100,000 Loebner Prize gold medal for the machine that can fool judges into thinking it's human hasn't yet been won, the device that most closely replicates a human response wins a $2,000 cash prize and a bronze medal each year.
Asked whether the Turing Test might ever be passed, Loebner said: "Oh I think so - I don't know about the timescale."
Although Loebner is convinced there will be a time when we have robots with artificial intelligence in our homes, he remains concerned such technology may not be distributed as fairly as it could be, much in the same way as global wealth is concentrated on a tiny percentage of individuals.
Of the prize's contribution to the AI cause, Loebner said: "I would like to see it advance artificial intelligence. I'm not sure it has particularly but that wasn't my criteria."
However, Loebner feels AI is already being applied in many areas including the "miraculous things" Google is doing with its search technology as well as the Darpa Urban Grand Challenge in which driverless cars race each other.
He also believes there could be much more advanced AI technology being developed behind closed doors. "There are possible examples of AI that aren't publically publicised that are being kept under wraps," he said.
Loebner was voted into silicon.com's Agenda Setters list for the first time this year. Among the Agenda Setters judges was the University of Reading's professor of cybernetics, Kevin Warwick, who is working with Loebner to run this year's prize.
For the results of this year's Loebner Prize and to find out if any machines passed the Turing Test, click here
Reader Comments (0)