Similarly, human-computer centaur teams are likely to be characterised by a constant tug of war between the humans and the computers,instead of settling down to a lifelong partnership. Teams made exclusively of humans - such as Sherlock Holmes and Dr Watson - usually develop permanent hierarchies and routines that last decades. But a human detective who teams up with IBM's Watson computer system (which became famous after winning the US TV quiz show Jeopardy in 2011) will find that every routine is an invitation for disruption, and every hierarchy an invitation for revolution. Yesterday's sidekick might morph into tomorrow’s superintendent, and all protocols and manuals will have to be rewritten every year.
同樣地，人類與計(jì)算機(jī)相結(jié)合的半人馬團(tuán)隊(duì)中，人類和計(jì)算機(jī)之間可能會(huì)發(fā)生一場(chǎng)持久戰(zhàn)，而不是建立終身伙伴關(guān)系。完全由人類組成的團(tuán)隊(duì)——如夏洛克·福爾摩斯和華生醫(yī)生——通常會(huì)發(fā)展出能維持幾十年的永久等級(jí)制度和慣例。但是，一個(gè)與IBM的華生計(jì)算機(jī)系統(tǒng)（因2011年在美國(guó)電視智力競(jìng)賽節(jié)目《危險(xiǎn)》中贏得大獎(jiǎng)而名聲大噪）合作的人類偵探會(huì)發(fā)現(xiàn)，每一道例行程序都是在招致干擾，每一個(gè)等級(jí)都可能引發(fā)革命。昨天的助手明天可能會(huì)搖身一變成為主管，所有的協(xié)議和手冊(cè)每年都要重寫。
同樣，“半人馬”組合很有可能變成一場(chǎng)人類與計(jì)算機(jī)之間不斷的拔河角力，而不是穩(wěn)定的終身伙伴關(guān)系。完全由人類組成的團(tuán)隊(duì)（比如福爾摩斯和華生），常常會(huì)形成長(zhǎng)期的階層和慣例，并能夠延續(xù)數(shù)十年。然而，如果偵探IBM的超級(jí)計(jì)算機(jī)系統(tǒng)“沃森”合作（該人工智能系統(tǒng) 2011年在電視益智搶答節(jié)目《危險(xiǎn)邊緣》（Jeopardy!）中獲勝，會(huì)發(fā)現(xiàn)所有的階層都可能被打破，所有的慣例也都可能被干擾。昨天的搭檔，明天可能就成了你的主管；所有的規(guī)章和守則也都必須每年重寫?！玖挚『辍?/p>
A closer look at the world of chess might indicate where things are heading in the long run. It is true that for several years after Deep Blue defeated Kasparov, human-computer cooperation flourished in chess. Yet in recent years computers have become so good at playing chess that their human collaborators lost their value, and might soon become utterly irrelevant.
仔細(xì)觀察國(guó)際象棋，可能會(huì)發(fā)現(xiàn)它的長(zhǎng)遠(yuǎn)發(fā)展方向。的確，在“深藍(lán)”擊敗卡斯帕羅夫后的幾年里，人機(jī)合作在國(guó)際象棋領(lǐng)域蓬勃發(fā)展。然而，近年來(lái)，計(jì)算機(jī)越來(lái)越擅長(zhǎng)下棋，人類合作者失去了價(jià)值，可能很快就會(huì)變得完全無(wú)關(guān)緊要。
仔細(xì)觀察國(guó)際棋壇的動(dòng)態(tài)，或許可以預(yù)估未來(lái)世界將走向何方。“深藍(lán)”戰(zhàn)勝卡斯帕羅夫之后的幾年間，人機(jī)合作是國(guó)際棋壇的熱門形式。但近幾年來(lái)，計(jì)算機(jī)已經(jīng)變得非常擅長(zhǎng)下棋，以至于人類合作者失去了他們的價(jià)值，而且可能很快就會(huì)變得完全無(wú)關(guān)緊要?！玖挚『辍? ??
On 7 December 2017 a critical milestone was reached, not when a computer defeated a human at chess—that's old news—but when Google's AlphaZero program defeated the Stockfish 8 program. Stockfish 8 was the world's computer chess champion for 2016. It had access to centuries of accumulated human experience in chess, as well as to decades of computer experience. It was able to calculate 70 million chess positions per second. In contrast, AlphaZero performed only 80,000 such calculations per second, and its human creators never taught it any chess strategies - not even standard openings. Rather, AlphaZero used the latest machine-learning principles to self-learn chess by playing against itself. Nevertheless, out of a hundred games the novice AlphaZero played against Stockfish, AlphaZero won twenty-eight and tied seventy-two. It didn't lose even once. Since AlphaZero learned nothing from any human, many of its winning moves and strategies seemed unconventional to human eyes. They may well be considered creative, if not downright genius.
2017年12月7日是一個(gè)關(guān)鍵的里程碑時(shí)刻，不是因?yàn)橛?jì)算機(jī)在國(guó)際象棋比賽中擊敗了人類（這已經(jīng)過(guò)時(shí)了），而是因?yàn)楣雀璧腁lphaZero程序擊敗了Stockfish 8程序。Stockfish 8是2016年計(jì)算機(jī)象棋世界冠軍。它獲得了人類幾個(gè)世紀(jì)以來(lái)積累的國(guó)際象棋經(jīng)驗(yàn)，以及數(shù)十年的計(jì)算機(jī)經(jīng)驗(yàn)，每秒能夠計(jì)算出7000萬(wàn)個(gè)象棋位置。相比之下，AlphaZero每秒只能進(jìn)行8萬(wàn)次這樣的計(jì)算，它的人類發(fā)明者從未教過(guò)它任何象棋策略，甚至連標(biāo)準(zhǔn)的開局策略都沒(méi)有教。相反，AlphaZero使用最新的機(jī)器學(xué)習(xí)原理，通過(guò)與自己較量來(lái)自學(xué)國(guó)際象棋。然而，在新手AlphaZero與Stockfish之間進(jìn)行的一百場(chǎng)比賽中，AlphaZero贏了二十八場(chǎng)，其余七十二場(chǎng)與Stockfish打成平局，一次也沒(méi)輸。由于AlphaZero從沒(méi)有從任何人身上學(xué)習(xí)，因此在人類看來(lái)，它的許多致勝的招數(shù)和策略都是不合常規(guī)的。即使這些招數(shù)和策略不是天才所為，也可以認(rèn)為是極具創(chuàng)造力的。
2017年12月7日，這是圍棋具有里程碑意義的一天，但這一天并不是計(jì)算機(jī)擊敗人腦（那已經(jīng)是舊聞了），而是谷歌的AlphaZero程序擊敗了Stockfish8程序。Stockfish8是2016年的全球計(jì)算機(jī)國(guó)際象棋冠軍，運(yùn)用的是幾百年來(lái)累積的人類國(guó)際象棋經(jīng)驗(yàn)，再加上幾十年的計(jì)算機(jī)象棋經(jīng)驗(yàn)，每秒計(jì)算7000萬(wàn)次走法。相較之下，AlphaZero每秒只計(jì)算8萬(wàn)次走法，而且寫程序的時(shí)候完全沒(méi)教它任何國(guó)際象棋規(guī)則，它連基本的起手下法都不會(huì)！AlphaZero完全是運(yùn)用最新的機(jī)器學(xué)習(xí)原理，不斷和自己下棋，就這樣自學(xué)了國(guó)際象棋。雖然如此，在AlphaZero與Stockfish8的100場(chǎng)比賽中，AlphaZero贏28場(chǎng)、平72場(chǎng)，完全未嘗敗績(jī)。AlphaZero完全沒(méi)向任何人類學(xué)習(xí)任何東西，許多獲勝走法和策略對(duì)人類來(lái)說(shuō)完全是打破常規(guī)的，可以說(shuō)是創(chuàng)意十足，甚至是天縱英才?！玖挚『辍?/p>
