Project NERO is a real-time strategy game in which the latest in AI research from the University of Texas Neuro-evolution Group has been imported into a commercial game engine.

NERO, Neuro-Evolving Robotic Operatives, is set in a post-apocalyptic world, where robots struggle over the relics of human civilization.

In the first phase of play, individual players deploy robots in a “sandbox” and train them to their tactics. In the second phase, players pit their robots in a battle against those trained by some other player, to see how well their training regimens prepared their robots for battle.

The robots in NERO use artificial neural networks for their “brains”, and they learn by means of neuroevolution. Neuroevolution is a genetic algorithm that rewards the agents that perform the best and punishing those that perform the worst. The rewards and punishments are specified by the players.
A bit sadistic.