Chess Variants

Chess has always been variant. Today there are thousands of mutations, transformations, and reorganizations of the classic game. I’m partial to those chess programs by Torus Games which allow you to play on boards stretched over imaginary topologies like Möbius strips and Klein bottles:

Alongside these programs you can find thousands of custom chess boards featuring strange situations and perplexing puzzles:


A History of Computer Chess

Understanding the history of computer chess as a field for measuring the intelligence of machines (and humans) is crucial to developing Neosentient Games. At first, citing the Mechanical Turk as a starting point for this history seems like a misnomer. The Mechanical Turk exhibits no autonomy and, in fact, isn’t strictly a computer. However, since the Neosentient will be invented through anthromorphic introspection, perhaps the idea of an augmented human (or augmented computer) is a good place to start developing Neosentient Games. After loosing to Big Blue in 1997, Gary Kasparov invented a chess variant for augmented play. Many chess masters create variants based on their intimate knowledge of the sport, adding or omitting pieces and extending the board, but Kasparov’s version of chess incorporated focused on augmenting the original game of chess by giving each human player a mechanical team mate. In Kasparov’s Augmented Chess, each human player could query their computer assistant for information. Is this the beginning of Neosentient Games? How can we develop this field towards the notion of the Neosentient?

Computer Vision Tools

There are many amazing tools for computer vision. Some helpful programming environments are Max/MSP/Jitter and PureData:

With Max/MSP/Jitter there are a couple libraries for tracking and interpreting video:

cv.jit by Jean-Mark Pelletier –
softVNS by David Rokeby –
Cyclops by Eric Singer –

You can also use different types of input, like the Kinect and Wii Remote. More on this later…

End Game and Game Over

Kasparov vs. Deep Blue

On February 10, 1996, Deep Blue became the first machine to win a chess game against a reigning world champion (Garry Kasparov) under regular time controls. After the loss, Kasparov said that he sometimes saw deep intelligence and creativity in the machine’s moves, suggesting that during the second game, human chess players had intervened on behalf of the machine, which would be a violation of the rules. Though Kasparov felt “deep intelligence and creativity,” this aesthetic of sentience was merely a bi-product of Deep Blue ability to evaluate over 200 million positions per second. Can we design a playful machine?

Artificial Artificial Intellegence

Amazon has recently launched a new service called the Amazon Mechanical Turk. They propose a casual labor system in which humans get payed to do simple tasks that are particularly difficult for computers to perform. You can view these “Human Intelligence Tasks” or HITs, and see how much money one might earn, here.

Some of the tasks include transcribing audio clips to text, finding urls containing specified information, labeling images with the appropriate tags, and even signing up for accounts on different websites. Amazon’s mechanical Turk is fascinating in the way it blends human and computer labor. The various logics required to complete each task are notable for being particularly human and requiring sentience. Amazon’s Mechanical Turk could function as a guide to certain pragmatic qualities of sentience.

On top of this, Amazon’s clear reference to the original Mechanical Turk is interesting and problematic. More soon on this…I’m thinking of both selling and buying some human labor as research for my project.