Artificial intelligence just keeps getting better at beating us at games, and what game more lends itself to a computer being a cheating, blue-shelling bastard than Mario Kart? That’s the idea behind MariFlow, a Mario Kart-playing neural network programmed by Seth Bling based upon a neural program called TensorFlow.
The AI has already earned gold medals in the 50cc, Mushroom, and Flower Cups, but has only been able to silver in Star Cup. The video above shows four screens: three with MariFlow running and one with Seth’s dad playing. It was easy to guess which player was Seth’s dad because he was the only one driving like somebody’s dad who’s never played Mario Kart before.
The neural network is modeled after Seth’s play style and works by predicting what button presses Seth would make and mimicking his style of play at a rate of 15 predictions per second. It isn’t a normal feed forward neural network; it’s a recurrent neural network, meaning it retains some of the information it had in previous steps, allowing each “neuron” to act as a memory cell while simultaneously performing computations.
Seth trained the recurrent neural network through backpropagation using 15 hours of his own gameplay footage, in a similar human-mimicking method used by the nunchuck robot. (Okay, maybe the nunchuck robot wasn’t trained via backpropagation, but we still wanted to mention nunchuck robot, the robot that knows nunchaku.)
Unfortunately, 15 hours of gameplay footage wasn’t enough. The neural network would still encounter situations it couldn’t handle, with Lakitu fruitlessly urging it to turn to hell around. To better train the program, Seth set up interactive sessions in which the neural network would drive for a while, then Seth would take over the controller in situations where the network needed additional representative data.
The end result was a competent non-human Mario Kart player, with the details broken down in a Google doc and in the video above.
Strangely, the video didn’t include the part where (we assume) Mario says, “It’s-a me, the impending singularity.”