Placeholder Image

字幕列表 影片播放

  • Two players just played the Chinese board game Go.

    兩位選手剛剛結束如火如荼的中國圍棋比賽。

  • The game was between Lee Se-dol, the world's top Go player, and the computer program.

    比賽選手之一的李世乭是世界第一的圍棋選手,他對對手則是個電腦程式。

  • The computer won.

    電腦贏了。

  • And to understand why this is significant, you need to understand more about the ambitions of DeepMind, the British company owned by Google.

    要了解這件事情為甚麼如此備受矚目,你必須了解 DeepMind, 擁有 Google 的英國公司,的野心。

  • DeepMind built AlphaGo, the artificially intelligent system that just beat Mr. Se-dol at Go.

    DeepMind 一手打造了 AlphaGo, 這個於圍棋戰場打敗世乭先生的人工智慧。

  • The message we want to send with this game, with this AlphaGo system, is how flexible and powerful learning algorithms can be.

    透過這場比賽,透過AlphaGo,我們想要傳達的訊息是運演算法則能多有彈性,能多強大。

  • So far, computer engineers have been able to built machines that can beat humans at games like chess.

    目前為止,電腦工程師能夠開發出能在如下棋般的比賽中打敗人類的機器。

  • In 1997, IBM's Deep Blue super computer famously beat world champion Garry Kasparov.

    1997年,IBM 的 Deep Blue 超級電腦打敗了世界冠軍 Garry Kasparov,此舉舉世聞名。

  • With Deep Blue, program was built, a system that was able to analyze every possible outcome of every possible move.

    Deep Blue 的程式設計讓它可以分析每一個可能的結果,每一步棋。

  • But Go's a much much more complicated game than chess.

    但是圍棋可比西洋棋複雜多了。

  • There are more possible positions on a Go board than there are atoms in the universe.

    圍棋版上可能發生的棋步比宇宙中的原子還多。

  • This is too much information for even the most powerful super computer to process.

    這麼龐大的資訊量就連最厲害的超級電腦都無法負荷。

  • So DeepMind took a different approach:

    所以 DeepMind 轉了個彎:

  • It created an artificially intelligent system based on neural networks, much like the brain.

    它創造了立基於神經網路的人工智慧系統,像極了大腦。

  • It tries to learn and think like people.

    它試著像人類般思考、學習。

  • In the case of AlphaGo, it's played millions and millions of games against itself, learning from past mistakes and adapting each time.

    AlphaGo 這個案例,它跟自己比試了幾十萬次,從過去的錯誤中學習並精進。

  • All of this could have significance to the wider world beyond the Go board.

    這一切都可能應證於更寬廣的世界,而不是只限於棋盤上。

  • If DeepMind and other AI experts can create systems that think like us,

    如果 DeepMind 和其他人工智慧專家可以創造出能和我們一樣思考的系統,

  • it could have truly profound implications on the way we ordinary humans live our lives.

    它真的可以完全理解我們一般人類生活的方式。

  • Murad Ahmed, Financial Times.

    Murad Ahmed,金融時報。

Two players just played the Chinese board game Go.

兩位選手剛剛結束如火如荼的中國圍棋比賽。

字幕與單字

單字即點即查 點擊單字可以查詢單字解釋