|Screenshot from 'Starcraft 2' gameplay / Korea Times|
By Ko Dong-hwan
With global eyes on the go match between artificial intelligence(AI) AlphaGo and Korea's professional go player Lee Sae-dol, a computer engineer predicted AI still needs more years of tuning to beat humans in the Starcraft computer strategy game.
According to reports, computer engineering professor Kim Kyung-joong from Sejong University, who developed AI for the Blizzard game in 2011, claimed the AI may surpass humans in unit control ― AI spools about 20,000 actions per minute (APM) while top players operate at around 200-300 APM. But such skill "isn't certainly all that it takes to win a Starcraft match," Kim said.
Rather, he said strategy was the most decisive key to winning the game, and Starcraft AI still was way behind humans in blueprinting strategies as well as making unplanned actions spontaneously.
|Lee Sae-dol, right, plays with Google's artificial intelligence program AlphaGo at the Four Seasons Hotel, Seoul. / Yonhap|
"No matter how well Starcraft AI may control a 12-unit group of Zealots ― an infantry unit from Protoss, one of the game's three species ― the computer has always been short of winning whenever humans countered with a stronger combination of different units or infiltrated AI's main base with a surprise attack like using a dropship ― a transport unit from Terran, another species of the game," Kim said.
Kim's research team had developed Starcraft AI "Xelnaga," Korea's first Starcraft AI. The AI finished top in preliminary games and third in main games at the Computational Intelligence and Games 2011 held by the Institute of Electrical and Electronics Engineers.
Starcraft AI has maintained a 50-60 percent winning rate since 2012 until last year in AIIDE (Artificial Intelligence and Interactive Digital Entertainment) Starcraft AI Competition.