Communications of the ACM
In Search of Machines That Play at Being Human
This year, 15 teams from Brazil, Canada, the United States, Italy, Japan, the United Kingdom, and Spain, among others, participated in the BotPrize contest, which applies the Turing Test to video games. In the competition, a judge started a game against two players, one human and one artificial. After 15 minutes of play, the judge had to identify which player was a human and which was a program. None of the computer programs, or bots, entered in this year's contest was able to deceive 80 percent of the judges.
"In our case, we didn't have enough time to program a good bot, since I am still in the process of migrating the control architecture that I use in real robots to the bots in the Unreal Tournament 2004, so we didn't place among the first five, but I will try again next year with a more advanced bot which implements the abilities of prediction of the opponent," says Carlos III University of Madrid professor Raul Arrabales.
Replicating human behavior in any environment is complex because it becomes necessary to combine different cognitive capabilities, Arrabales says. The problem is that much is known about the brain but only on a relatively high level, which prevents the creation of artificial neuron networks that mimic human ones with enough detail.
The finalists in the competition were able to convince at least one of the judges that their bot was human.
From Carlos III University of Madrid (Spain)
View Full Article
Abstracts Copyright © 2009 Information Inc., Bethesda, Maryland, USA 
No entries found