At one time, when computers had small capacity, it was a challenge for us to ask if we can program the machine to compete with a member of our species.
Today, when capacity is readily and cheaply called to hand, this is not how to measure progress, it is a cheap marketing ploy. We do not, for example, ask workers producing motor vehicles to compete with a factory full of heavy robots - and if we did the result would be obvious, the machines would win.
But this is what Google have done in their unfair competition with Lee Se-dol and the previously neglected game of GO. Se-dol was not playing against a Commodore 64, then the win would have been impressive. He was competing against a machine of very significant power utilization and computational capacity. He was a single worker competing against a factory of machines.
Shame on you Google (John and Peter) for not making the nature of the competition clear and allowing the community to mislead itself into believing that the machine is truly intelligent. It is certainly clever programming and orchestration of combinatorial results by the engineers, for that the Deep Mind team deserves applause. The win was certainly artificial but it was not an example of intelligence.
At best it was an example of admirable and effective automation - a proxy for the intelligence with which the machine was imbued.
I would still like to know a comparison of the amount of power actually used by Deep Mind and Lee Se-dol, so that we may gain a real estimate of how distant machines are from the capacity of our species.
How could the game be played and be fair? Match power utilization. Instead of allowing elapsed time to dominate and treating the machine and player as equal, allowing the machine to burn megawatts of power against the 15Joules per second of the man, make them truly equal and allow them only to use the same amount of power to play.