I'm wondering if strong players will learn to beat AlphaGo by playing unconventional openings, so as to confuse the neural net, and get the upper hand before th

I'm wondering if strong players will learn to beat AlphaGo by playing unconventional openings, so as to confuse the neural net, and get the upper hand before the MCTS can take over in the middle game. This could be good for the game of Go in general as top players are forced to find new ways to play. Just speculating as I am not a particularly strong player myself.