14 DAY TRIAL // The second edition of the BotPrize contest, which aims at producing an artificial intelligence (AI) that cannot be distinguished from a live person in a virtual environment, has its roots deep in history, all the way to the 1950s. At the time, Alan Turing developed a test that held that a real AI was only achieved when a person couldn't tell if their “hidden” counterparts were machines or humans. In this edition of the BotPrize, participants had to adapt their AI so as to behave as humanly as possible within the confines of the video-game called Unreal Tournament 2004.
The goal of the competition was to produce an AI advanced enough, that judges would not be able to tell the difference between the human players and a computer-controlled one during gameplay. More than 15 teams joined the challenge this year, from countries including Brazil, Canada, the United States, Italy, Japan, the United Kingdom and Spain. The contest was held in Milan, Italy. One of the participants was Carlos III University of Madrid Information Technology Department Professor Raul Arrabales.
“As my research is focused on Artificial Intelligence, it seemed to me that the Turing Test adapted to video games was a good domain in which to empirically test our advances,” he says. None of the AI bots in this year's competition was able to deceive 80 percent of the judges. These results were in tune with those obtained at the first edition of the BotPrize, and also on the original Turing Test.
“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,” Arrabales adds, quoted by AlphaGalileo.
Regardless of whether it's for a robot or a video-game simulation, mimicking the complexity of human behavior is not a very easy task, as evidenced by the fact that no one has managed to do it until now. Different cognitive abilities need to be combined in a single algorithm, and the accomplishment has thus far eluded a large number of research teams working on AI around the world. “We look first at how the brain works, we try to understand it and later we try to imitate it in the machine. The problem is that we know a lot about the brain, but only at a relatively high level, the function that each concrete area of the brain carries out and how it is connected and is related to other brain areas,” the expert concludes.