Results 1 to 6 of 6

Thread: In Just 4 Hours, Google's AI Mastered All The Chess Knowledge in History

  1. #1

    Default In Just 4 Hours, Google's AI Mastered All The Chess Knowledge in History

    https://www.sciencealert.com/it-took...mind-alphazero

    "This algorithm could run cities, continents, universes."

    PETER DOCKRILL 7 DEC 2017


    Chess isn't an easy game, by human standards. But for an artificial intelligence powered by a formidable, almost alien mindset, the trivial diversion can be mastered in a few spare hours.

    In a new paper, Google researchers detail how their latest AI evolution, AlphaZero, developed "superhuman performance" in chess, taking just four hours to learn the rules before obliterating the world champion chess program, Stockfish.

    In other words, all of humanity's chess knowledge and beyond was absorbed and surpassed by an AI in about as long as it takes to drive from New York City to Washington, DC.

    After being programmed with only the rules of chess (no strategies), in just four hours AlphaZero had mastered the game to the extent it was able to best the highest-rated chess-playing program Stockfish.

    In a series of 100 games against Stockfish, AlphaZero won 25 games while playing as white (with first mover advantage), and picked up three games playing as black. The rest of the contests were draws, with Stockfish recording no wins and AlphaZero no losses.

    "We now know who our new overlord is," said chess researcher David Kramaley, the CEO of chess science website Chessable.

    "It will no doubt revolutionise the game, but think about how this could be applied outside chess. This algorithm could run cities, continents, universes."

    Developed by Google's DeepMind AI lab, AlphaZero is a tweaked, more generic version of AlphaGo Zero, which specialises in playing the Chinese board game, Go.

    DeepMind has been refining this AI for years, in the process besting a series of human champions who fell like dominoes before the indomitable, "Godlike" neural network.

    That victory streak culminated in a startling success in October, in which a new fully autonomous version of the AI which only learns by playing itself, never facing humans bested all its former incarnations.

    By contrast, AlphaGo Zero's predecessors partly learned how to play the game by watching moves made by human players.

    That effort was intended to assist the fledgling AI in learning strategy, but it seems it may have actually been a handicap, since AlphaGo Zero's fully self-reliant learning proved devastatingly more effective in one-on-one competition.

    "It's like an alien civilisation inventing its own mathematics," computer scientist Nick Hynes from MIT told Gizmodo in October.

    "What we're seeing here is a model free from human bias and presuppositions. It can learn whatever it determines is optimal, which may indeed be more nuanced that our own conceptions of the same."

    But things are moving so fast in this field that already the October accomplishment may have been outmoded.

    In their new paper, the team outlines how the very latest AlphaZero AI takes the self-playing reliance called reinforcement learning and applies it with a much more generalised streak that gives it a broader focus to problem solving.

    That broader focus means AlphaZero doesn't just play chess. It also plays Shogi (aka Japanese chess) and Go too and, perhaps unsurprisingly, it only took two and eight hours respectively to master those games as well.

    For now, Google and DeepMind's computer scientists aren't commenting publicly on the new research, which hasn't as yet been peer-reviewed.

    But from what we can tell so far, this algorithm's dizzying ascent to the pinnacle of artificial intelligence is far from over, and even chess grandmasters are bewildered by the spectacle before them.

    "I always wondered how it would be if a superior species landed on Earth and showed us how they played chess," grandmaster Peter Heine Nielsen told the BBC.

    "Now I know."

    The findings are available at preprint website arXiv.
    Truth is a social construct.



  2. Remove this section of ads by registering.
  3. #2

    Default

    Nothing to see here folks... move along.
    Once you go Paul, you see through them all.

  4. #3

    Default

    I, for one, welcome our new robotic overlords!

    01001000 01100101 01101100 01101100 01101111 00101100 00100000 01110111 01100101 00100000 01100011 01101111 01101101 01100101 00100000 01101001 01101110 00100000 01110000 01100101 01100001 01100011 01100101 00100001
    There are only two things we should fight for. One is the defense of our homes and the other is the Bill of Rights. War for any other reason is simply a racket.
    -Major General Smedley Butler, USMC,
    Two-Time Congressional Medal of Honor Winner
    Author of, War is a Racket!

    It is not that I am mad, it is only that my head is different from yours.
    - Diogenes of Sinope

  5. #4

    Default

    oh forgot to tell them:

    01110000 01101100 01100101 01100001 01110011 01100101 00100000 01100100 01101111 01101110 00100111 01110100 00100000 01101011 01101001 01101100 01101100 00100000 01110101 01110011
    There are only two things we should fight for. One is the defense of our homes and the other is the Bill of Rights. War for any other reason is simply a racket.
    -Major General Smedley Butler, USMC,
    Two-Time Congressional Medal of Honor Winner
    Author of, War is a Racket!

    It is not that I am mad, it is only that my head is different from yours.
    - Diogenes of Sinope

  6. #5

    Default

    Quote Originally Posted by jllundqu View Post
    oh forgot to tell them:

    01110000 01101100 01100101 01100001 01110011 01100101 00100000 01100100 01101111 01101110 00100111 01110100 00100000 01101011 01101001 01101100 01101100 00100000 01110101 01110011
    http://www.binaryhexconverter.com/he...&pad=false&v=2

    Truth is a social construct.

  7. #6

    Default

    This algorithm could run cities, continents, universes.
    This claim is a bit hyper-ventilated. Alpha Zero has illustrated that search across extremely large spaces with well-defined outcomes (good, bad, unsure) and moderately complex rule-sets is an easy computational problem. This is a novel result and Deep Mind deserves props for their work, but this is hardly as shocking as some are making it out. Animal brains, including the human brain, already do this kind of thing, so this is precisely the kind of thing we were trying to recreate in computers. Even though Alpha Zero is being touted as "alien", Deep Mind has pointed out that Alpha Zero doesn't think like a machine, so that's really what we have long wanted to see computers do - think in a "non-computer" way. Sure, it doesn't think like humans, but why should it? Alpha Zero has no conception of the idea of a horse or a knight - we have applied these labels to the chess pieces as an act of human imagination and these illustrative labels play an integral role in holding our interest in the game. As world champion Emmanuel Lasker noted in one of his books on chess - when two competitors sit down to play serious chess, it is no longer a game in the children's sense, it is a fight that is fought on the terrain of the board, instead of physically with fists or soldiers. Alpha Zero does not know what a fight is, nor why anyone would be interested in fighting, what a sport is, or why anyone would be interested in fighting for sport.

    Alpha Zero will be better suited for accelerating scientific research where we can rapidly automate experiments. Scientific experiments have well-defined outcomes and these outcomes can be used to guide search in an unbiased way across the immense space of all possible explorations, in search for the properties we are seeking. This could be everything from searching for new polymers, searching for new drugs, simulating (predicting) protein-folding, discovering exotic physical properties such as those exhibited by meta-materials, and so on. Running a city or continent has no well-defined outcome (good, bad, indifferent). Alpha Zero will not be able to directly help with this kind of thing.






Similar Threads

  1. Replies: 4
    Last Post: 05-25-2015, 03:16 PM
  2. Replies: 10
    Last Post: 02-29-2012, 01:35 PM
  3. Ron Paul trends on Google for 19 straight hours, still going strong
    By libertybrewcity in forum Ron Paul Forum
    Replies: 3
    Last Post: 08-19-2011, 01:04 PM
  4. Replies: 8
    Last Post: 04-08-2009, 09:02 PM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •