[amoruso]

Here's the original research paper if you're interested.

http://arxiv.org/abs/1312.5602

I'll just quote their introduction instead of trying to summarize the paper:

"Our goal is to create a single neural network agent that is able to successfully learn to play as many of the games as possible. The network was not provided with any game-specific information or hand-designed visual features, and was not privy to the internal state of the emulator; it learned from nothing but the video input, the reward and terminal signals, and the set of possible actions—just as a human player would. Furthermore the network architecture and all hyperparameters used for training were kept constant across the games. So far the network has outperformed all previous RL algorithms on six of the seven games we have attempted and surpassed an expert human player on three of them."

[mturmon]

My gripe is with the post, not the paper. But you're right, the best way to figure out what's new is to go to the source.

The paper does a good job going over related work (section 3), beginning with the example I gave.