UPDATE: AlphaGo defeats Lee Se-dol once again in the 2nd match.
What can only be described as an historic victory for the A.I. program, Lee Se-dol’s only response was:
“I was very surprised. I didn’t expect to lose. [But] I didn’t think AlphaGo would play the game in such a perfect manner.”
– Lee Se-dol
It isn’t entirely over for Lee Se-dol, however, as there are four more matches to be had against AlphaGo. As the fictional detective Sherlock Holmes would say, “The game is afoot.” The two will spar once again today at 1PM KST (11PM EST). Who will come out victorious is unclear, but it will most assuredly be a game you don’t want to miss.
Unlike chess, the abstract board game Go requires a great deal of strategy that can’t entirely be coldly calculated. This is what makes AlphaGo’s victory so significant. While it isn’t necessarily evidence of an A.I. program “thinking” as we know it in human terms, it certainly lays down the groundwork of machines going beyond its original programming to adapt to situations occurring in real-time. Like IBM’s Watson, AlphaGo has shown to be able to make very difficult decisions without too great of an effort and come out victorious as a result.
What this means for future A.I. isn’t entirely clear, but if this recent victory for AlphaGo tells us anything it’s that these programs are only going to become all the more intelligent at an exponential rate. Beyond the board games, future A.I. will not only reach the intelligence capacity of that of an average human, but will likely far surpass it soon thereafter. According to the futurist Ray Kurzweil, who is currently working for Google, this moment in time will be what is known as the Technological Singularity. When it will happen is anybody’s guess, but considering AlphaGo’s victory in a game of Go occurred ten years ahead of human predictions, it might just happen sooner than we think.