By Griffin Schnitzer '18
In a historic match, Google pitted its AlphaGo computer system against the world champion of a complex strategy game. The contest in question, Go, is a board game similar to chess, however vastly more complex. Players must maneuver their pieces to control the board and, in doing so, “there are more possible positions on a Go board than atoms in the universe” (Wood). Due to this complexity, it is far from a simple task for a computer to play the game, much less win against an opponent who is “arguably the best player of the past decade” (Mundy). Lee Se-dol, the contestant, was so confident in human over machine abilities, that he stated it would be a huge victory for the computer if it “wins even one game” of the five game series (Jung-a). As it turned out, Google’s AlphaGo won the first three straight games of the series, before finishing it off with an overall 4-1 victory. The human contender was soundly defeated.
Beyond the outcome of this game, the results have much broader implications for the technology industry and for society as a whole. As far as the progression of artificial intelligence, this is a major step, as it goes beyond the innovations of IBM’s 1997 chess victory over a grandmaster with its Deep Blue system, a much less complex feat. The entire purpose for the creation of Deepmind, the Google company which built AlphaGo, is listed in a Google blog post as “to create general-purpose artificial intelligence.” The reason for designing AI, it says, is to use as a tool for solving real and complex problems that face humans – not just for beating us in board games. Much like the program was able to make Go moves that no humans could see, Google thinks it will be able to “find solutions that humans don’t necessarily see” in more practical situations.
Recently, Deepmind announced Deepmind Health, a mobile app that uses their advanced systems to analyze data and diagnose illnesses. This app, however, does not even use artificial intelligence in its analysis, but Google says it is “certainly” an area where they would be interested in applying it in the future (Novet). Additionally, in the realm of education, artificial intelligence can be applied for tutoring purposes, enabling new methods of learning. Some concerns have been raised, however, about how such developments may lead to lost jobs as artificial intelligence eliminates the need for humans in such positions. These predictions may be far off, and based in conjecture, as artificial intelligence has not yet been adapted to these functions.
Overall, the defeat of a human by Google’s AlphaGo will continue the trend of increasing artificial intelligence capabilities. The exact impact of this event may not be enormous, but it is a stepping-stone towards an outcome with far-reaching and powerful effects.
Sources:
Jung-a, Song. "Google DeepMind Takes on Go Grandmaster in AI Match." Financial Times. Nikkei, 8 Mar. 2016. Web.Mundy, Simon. "AlphaGo Conquers Korean Grandmaster Lee Se-dol." Financial Times. Nikkei, 15 Mar. 2016. Web.Novet, Jordan. "Google’s DeepMind AI Group Unveils Health Care ambitions." VentureBeat. 24 Feb. 2016. Web. 17 Mar. 2016.Wood, Geordie. "Google’s AI Is About to Battle a Go Champion-But This Is No Game." Wired.com. Conde Nast Digital. Web. 17 Mar. 2016.http://www.npr.org/sections/ed/2016/03/16/470011574/what-artificial-intelligence-could-mean-for-educationhttps://googleblog.blogspot.com/2016/03/what-we-learned-in-seoul-with-alphago.html Photo:http://airesearch.com/wp-content/uploads/2016/01/alpha_go-825x510.jpg