http://news.bbc.co.uk/2/hi/uk_news/england/berkshire/7666246.stm

Test explores if robots can think

An experiment has been taking place in Berkshire to see if robots are capable of intelligent thought.

Scientists at the University of Reading tested five machines to see if they could pass themselves off as humans in text-based conversations with people.

The test was devised in 1950 by British mathematician Alan Turing, who said that if a machine was indistinguishable from a human, then it was “thinking”.

One robot, Elbot, came close on Sunday by reaching 5% below the pass mark.

No robot has ever passed the Turing Test, which requires the robot to fool 30% of its human interrogators.

During the experiment, five artificial conversational entities (ACEs) competed in a series of five-minute long, unrestricted conversational tests.

Artificial intelligence

The ACEs tried to pass themselves off as humans to the judges.

“During the tests, all of the ACEs managed to fool at least one of their human interrogators,” a University of Reading spokesman said.

The tests took place as part of the 18th Loebner Prize for artificial intelligence.

The annual competition awards prizes for the most human-like machine of those entered.

Top machines from around the world were “whittled down” to the five taking part in Sunday’s final, the university spokesman added.

Prof Kevin Warwick, who organised the tests, said: “Today’s results actually show a more complex story than a straight pass or fail by one machine.

“Where the machines were identified correctly by the human interrogators as machines, the conversational abilities of each machine were scored at 80% and 90%.

“This demonstrates how close machines are getting to reaching the milestone of communicating with us in a way in which we are comfortable.

“That eventual day will herald a new phase in our relationship with machines, bringing closer the time in which robots start to play an active role in our daily lives.”

This year’s winner was Elbot, despite failing the Turing Test. The programme’s developers were awarded a $3,000 (£1,760) prize.

Damn, I was going post this a few days ago but plum forgot.
Guess thats means I’m no AI chat program. Don’t got the memory for it.

That’s kind of impressive. We may have turing-capable AI in our lifetime..

Now, whether it will be more than just a good user interface, or a fun little chat program.. that’s another story.

DJ_Canada - 15 October 2008 07:09 AM

Damn, I was going post this a few days ago but plum forgot.
Guess thats means I’m no AI chat program. Don’t got the memory for it.

Well if you start talking to robots any time soon you better start geting a life…

silver caliber magnum - 17 October 2008 01:02 AM

DJ_Canada - 15 October 2008 07:09 AM

Damn, I was going post this a few days ago but plum forgot.
Guess thats means I’m no AI chat program. Don’t got the memory for it.

Well if you start talking to robots any time soon you better start geting a life…

Didn’t you know that DJ Canada is a robot?

Just because a computer can simulate a human well enough to fool another human doesn’t mean the computer is truly capable of independent thought or analysis.  The Turing Test was quickly shown to be inadequate for determining artificially intelligent or sentient computers.  What is has shown us is just how complicated the question really is, and made us question what being self-aware actually entails.

Charybdis - 17 October 2008 12:11 PM

... is just how complicated the question really is, and made us question what being self-aware actually entails.

*suddenly notices that his feet are rather potent at the moment*
*wishes he wasn’t so self-aware*

You could try walking aroung the office robot-style.  Maybe people will start to avoid you so it won’t be a problem.

Charybdis - 17 October 2008 12:27 PM

You could try walking aroung the office robot-style.  Maybe people will start to avoid you so it won’t be a problem.

I am alone in the office - the shoes came off and the odor assaulted me