I think that's what the AlphaGo team did - they trained their agent against itself, and it learned new moves not explicitly programmed in! With an evaluation f

I think that's what the AlphaGo team did - they trained their agent against itself, and it learned new moves not explicitly programmed in! With an evaluation function just saying ahead / not ahead.