Tropes

www.voicechronicle.com_tech-and-science 2015 000033.txt

#Google's Algorithm Masters Video games Google has stated that they have developed the first computer program capable of learning a wide variety of tasks independently, in what has been hailed as a significant step towards true artificial intelligence. According to the engineers at Google, the program (or agent as its developers call it) learnt to play 49 different retro computer games, and came up with its own strategies for winning. They added that in the future, the same approach could be used to power self-driving cars, personal assistants in smartphones or conduct scientific research in fields from climate change to cosmology. The research was carried out by Deepmind which is a British company that was bought by Google last year for $500 million, whose stated aim is to build mart machines The team of engineers stated that a total of 49 classic Atari arcade games from the 1980 were utilized as part of the experiment. These games were used because they are simple enough for an AI system to learn quickly, but complex enough where winning was not trivial. They found that the agent performed at 75%of the level of a professional games tester or better on half of the games tested, which ranged from side-scrolling shooters to boxing to 3d car-racing. On some games such as Space Invaders, Pong and Breakout, the algorithm significantly outperformed humans, while on others it fared far worse. The team stated that this was mostly because the algorithm has no real memory. It is, therefore, unable to commit to long-term strategies that require planning. They also stated that with some of the games, the agent got stuck in a rut, where it had learnt one basic way to score a few points, but never really grasped the game overall objective. The team added that it is now trying to build in a memory component to the system and apply it to more realistic 3d computer games. Demis Hassabis founder of Deepmind, said, his is the first significant rung of the ladder towards proving a general learning system can work. It can work on a challenging task that even humans find difficult. It the very first baby step towards that grander goal, but an important one. n

< Back - Next >