Computer Learns Board Games From Two-Minute Clips, Beats Humans Right After - perryhastannow

As of right now, we're officially one footfall nigher to Skynet. Like the computer antagonist, computer scientist Lukasz Kaiser's machine learning software package (PDF) is capable of learning at accelerated speeds. Unlike everyone's ducky Cyberdine Systems mistake, this one doesn't need military-score hardware–it just needs a laptop with a 4GB RAM, a 2.13GHz Intel L9600 mainframe, and only one CPU meat.

In his latterly published newspaper, Kaiser defined how a system guided by a decision-devising engine of sorts can learn how to play competently games with only a minimal amount of background information.

This is where things get a little technical, so bear with us: Kaiser states that while data processor scientists have done a great sum of work in regards to computerized physical object realization and visual scene interpretations, "only a few systems with the mental ability for erudition higher-level concepts has been presented as yet." Accordant to Kaiser, our computers are jolly good at deriving sequences of high-level symbolic information from television streams, but we nonetheless take in a long way to turn when it comes to learning from it.

He argues that a more than nuanced access using relative structures and four-fold logic systems is ameliorate suited for learning from optical data in comparison to the standard practice of utilizing formulas and singular logic systems. "These 2 fundamental changes countenance United States of America to certify a system that–wise only roughly rows, columns, diagonals and differentiating pieces–learns games corresponding Connect Quartet, Gomoku, Pawns or Breakthrough, each one from a few intuitive video demonstrations, together around 2 minutes in duration."

Is this where we set forth preparing for the rise of the machines? Not quite yet. Kaiser still needs to figure out how to baffle the system to solve problems requiring "hierarchical, structured learning or a form of probabilistic formulas." Until then, we'atomic number 75 safe. After that, it's anyone's game.

[Łukasz Kaiser (PDF) via Gizmodo / Photo: _e.t on Flickr (Mil Away-SA 2.0)]

Cassandra Khaw is an entry-stratum audiophile, a street dancer, a person who writes about video games for a living, and someone WHO spends too a great deal time on Twitter.

Like this? You might also enjoy…

MIT Develops an Muscularity-Harvesting Chip That You Can Shake and Bake
Give Maine Three Minutes and I'll Teach You How to KO'd-Skrillex Skrillex With Barcodes
Robotic Legs Can Walk Care a Human, Could Teach Hoi polloi to Walk Again

Get more GeekTech: Twitter – Facebook – RSS | Topple us off