The world has been stunned in recent weeks by the incredible performance of Aja Huang in a monumental five-game competition of the board game Go. Merely an amateur at the game in comparison to world champion Lee Sedol, Huang boldly challenged the Go master, placing $1 million at stake. If Huang was victorious, he would graciously donate the money to charity; if he was defeated, all the money would be given to Sedol. What an insanely audacious move by Aja Huang! How did he possibly expect to overcome the grandmaster at this extremely complex and intricate game?
In an unbelievable turn of events, Huang…
*tapping on le shoulder*
…4 wins and 1 loss…
*persistent tapping on le shoulder*
ARGH Yanhao what do you want?
*psst Huang didn’t actually play he was just acting as the physical avatar of an artificial intelligence machine he helped build called AlphaGo*
Wait, so Sedol was playing against Huang who was being mind-controlled by an AI? Ah, now that makes sense.
A few weeks ago I attended a lecture at a local university entitled Rise of the Algorithm – What Kind of Artificial Intelligence Have We Created?. It got me interested in the development of AI, especially with all the recent commotion regarding the Go games, so I thought I’d do a little bit of exploring.
Whilst the whole of Silicon Valley revels in the victory of Google’s AlphaGo, having created an artificial intelligence that has mastered one of our oldest and most complex games, in South Korea, the mood is different. Here the board game Go has been integrated into their culture and tradition for well over 2500 years, with many scholars studying the game full time in academies. AlphaGo, new to the game only a couple of years before, has effortlessly scrutinised the game thoroughly just by playing itself over and over again.
The nation is jarred as they watch their culture being firmly oppressed by technology, having previously been so naively confident in the ability of their hero, Lee Sedol, who predicted that he would sweep AlphaGo aside with ease.
Behind their shocked faces, there is another screen of amazement. Many South Koreans, expecting its moves to be mechanical and calculated, are in awe of AlphaGo’s beautiful playstyle which seems to exude an impression of intuition and creativity, rarely observed even in the most skillful and perceptive of players. However, although this certainly is a significant step forward for artificial intelligence, many are concerned about the technology breaching other forms of art and culture. Creativity is one of the few remaining defining traits of human nature, and we mustn’t let that be taken away from us. Is this something we should be worrying about? How does the cost-benefit analysis weigh out?
I’m sorry to cut this post so short as there is clearly a lot more to say on the matter, but I really should go and do some work now… stay tuned and perhaps there will be a second part explaining how AlphaGo works and what applications the AI could be used for in future. Maybe not, who knows? I’ll leave you with a vaguely relevant video to enjoy for the time being.