Don't fear intelligent machines. Work with them | Garry Kasparov

譯者: Lilian Chiu 審譯者: Wilde Luo

This story begins in 1985,

這個故事開始於 1985 年，

when at age 22,

我在 22 歲時，

I became the World Chess Champion

打敗了卡爾波夫，

after beating Anatoly Karpov.

成為世界棋王。

Earlier that year,

那年年初，

I played what is called simultaneous exhibition

我參加了所謂的「車輪戰」，

against 32 of the world's best chess-playing machines

對抗 32 個世界最強博弈機器，

in Hamburg, Germany.

地點在德國漢堡。

I won all the games,

我大獲全勝。

and then it was not considered much of a surprise

在那時，我能同時戰勝 32 台電腦

that I could beat 32 computers at the same time.

並不讓人意外。

To me, that was the golden age.

對我而言，那是我的黃金年代。

(Laughter)

（笑聲）

Machines were weak,

在那時，機器特別弱小，

and my hair was strong.

與此相反，我的頭髮十分強韌。

(Laughter)

（笑聲）

Just 12 years later,

只不過經過了 12 年，

I was fighting for my life against just one computer

我卻得用盡吃奶的力氣 與單單一台電腦相博弈，

in a match

《新聞週刊》的封面將這場比賽稱為

called by the cover of "Newsweek"

"The Brain's Last Stand."

「人腦的最後一搏」。

No pressure.

不要有壓力。

(Laughter)

（笑聲）

From mythology to science fiction,

從神話到科幻小說，

human versus machine

人類對抗機器

has been often portrayed as a matter of life and death.

總是被描述成殊死搏鬥。

John Henry,

約翰亨利，

called the steel-driving man

被稱為「鋼鑽人」，

in the 19th century African American folk legend,

是十九世紀非裔美國人的民間傳奇。

was pitted in a race

在一場比賽中，他與氣鑽機競爭，

against a steam-powered hammer

鑽破山石，鑿出隧道。

bashing a tunnel through mountain rock.

John Henry's legend is a part of a long historical narrative

約翰亨利的傳奇 是歷史長流中的一部份，

pitting humanity versus technology.

代表著人性與技術的拼鬥。

And this competitive rhetoric is standard now.

而在今日，「人與機器之爭」 這種輪調已成常態。

We are in a race against the machines,

我們正與機器展開角逐，

in a fight or even in a war.

這是一場鬥爭，甚至一場戰爭。

Jobs are being killed off.

工作正在不斷減少。

People are being replaced as if they had vanished from the Earth.

人們被機器取代，彷彿已淪為配角。

It's enough to think that the movies like "The Terminator" or "The Matrix"

所有的一切足以讓人相信： 《魔鬼終結者》或《駭客任務》

are nonfiction.

並不是杞人憂天。

There are very few instances of an arena

極罕見到在一場比賽中

人類的血肉之軀和智慧

where the human body and mind can compete on equal terms

能夠與電腦或機器人平等競爭。

with a computer or a robot.

Actually, I wish there were a few more.

事實上我期待能有更多這樣的情況。

Instead,

但是，

it was my blessing and my curse

這是我的福氣也是詛咒，

to literally become the proverbial man

我成為眾所周知的名人，沒誇大，

in the man versus machine competition

人們至今仍然會談論

that everybody is still talking about.

這場人對抗機器的比賽。

In the most famous human-machine competition since John Henry,

在這場自約翰亨利以來 最出名的人類機器競賽中，

I played two matches

我參與了兩場比賽，

against the IBM supercomputer, Deep Blue.

對抗 IBM 的超級電腦「深藍」。

Nobody remembers that I won the first match --

沒有人記得我贏了第一場比賽。

(Laughter)

（笑聲）

(Applause)

（掌聲）

In Philadelphia, before losing the rematch the following year in New York.

在費城贏得了首場比賽。

而隔年在紐約第二次比賽，我輸了。

But I guess that's fair.

但我想那是公平的。

There is no day in history, special calendar entry

史上未曾有一天記錄著

for all the people who failed to climb Mt. Everest

那些沒能成功登頂聖母峰的人們，

before Sir Edmund Hillary and Tenzing Norgay

直到艾德蒙希拉蕊爵士和丹增諾蓋

made it to the top.

攻頂成功青史留名。

And in 1997, I was still the world champion

進到 1997 年我仍是世界冠軍，

when chess computers finally came of age.

那時我的機器對手終於成熟了。

I was Mt. Everest,

我是聖母峰，

and Deep Blue reached the summit.

而深藍攀上了山巔。

I should say of course, not that Deep Blue did it,

我應該說，當然，這不是深藍辦到的，

but its human creators --

是它的人類創造者辦到的：

Anantharaman, Campbell, Hoane, Hsu.

阿南蕯拉曼、坎柏、洪恩、許峰雄。

Hats off to them.

在此向他們致敬。

As always, machine's triumph was a human triumph,

一如既往，

機器的勝利其實是人類的勝利，

something we tend to forget when humans are surpassed by our own creations.

當我們人類被自身的創造物所超越時 我們常常會忘了這一點。

Deep Blue was victorious,

深藍的確獲勝了，

but was it intelligent?

但它是智慧的嗎？

No, no it wasn't,

不，它不是，

at least not in the way Alan Turing and other founders of computer science

至少不是艾倫圖靈及 其他計算機科學之父

had hoped.

所希望的那樣。

It turned out that chess could be crunched by brute force,

結果是，西洋棋臣服於蠻力之下，

once hardware got fast enough

只要硬體性能夠強勁，

and algorithms got smart enough.

且演算法夠精妙，就可以做到。

Although by the definition of the output,

不過從博弈中的表現來定義，

grandmaster-level chess,

西洋棋特級大師等級，

Deep Blue was intelligent.

深藍的確是智慧的。

But even at the incredible speed,

但即使性能如此之強，

200 million positions per second,

每秒計算兩億步棋，

Deep Blue's method

深藍的方法

provided little of the dreamed-of insight into the mysteries of human intelligence.

無法提供我們期待的洞見

來讓我們深索人類智能之謎。

Soon,

不用多久，

machines will be taxi drivers

機器將會擔任計程車駕駛、

and doctors and professors,

醫生、教授，

but will they be "intelligent?"

但它們會有「智慧」嗎？

I would rather leave these definitions

我更想把這些定義留給 哲學家和字典。

to the philosophers and to the dictionary.

What really matters is how we humans

重要的是，我們人類對於

feel about living and working with these machines.

跟這些機器一同居住、 工作有什麼感受。

When I first met Deep Blue in 1996 in February,

當我在 1996 年 2 月 初次遇見深藍時，

I had been the world champion for more than 10 years,

我保有世界冠軍的頭銜已超過十年，

and I had played 182 world championship games

曾參與過 182 場世界冠軍賽，

and hundreds of games against other top players in other competitions.

以及數以百計 和其他頂尖玩家交手的賽事。

I knew what to expect from my opponents

我可以預期我的對手，

and what to expect from myself.

我也可以預期我自己。

I was used to measure their moves

我習慣于酌量他們的每步棋，

and to gauge their emotional state

判斷他們的情緒狀態，

by watching their body language and looking into their eyes.

方法是觀察他們的肢體語言 以及注視著他們的眼睛。

And then I sat across the chessboard from Deep Blue.

接著，我隔著棋盤， 與深藍相對而坐。

I immediately sensed something new,

我立即有了新的感受，

something unsettling.

使人不安的感受。

You might experience a similar feeling

你可能也有過類似的感受，

the first time you ride in a driverless car

比如當你初次坐上 無人駕駛車輛時的感受，

or the first time your new computer manager issues an order at work.

或是聽到你的新任電腦經理 初次下達工作指令時的感受。

But when I sat at that first game,

但當我比賽第一局時

I couldn't be sure

我無法確定

what is this thing capable of.

這東西有什麼能耐。

Technology can advance in leaps, and IBM had invested heavily.

技術的進步有時會突飛猛進， 且 IBM 投資相當大。

I lost that game.

我輸了那一局。

And I couldn't help wondering,

我忍不住會想，

might it be invincible?

它有可能是無敵的嗎？

Was my beloved game of chess over?

我摯愛的西洋棋賽結束了嗎？

These were human doubts, human fears,

這些是人類的懷疑、人類的恐懼，

and the only thing I knew for sure

我只能確定一件事：

was that my opponent Deep Blue had no such worries at all.

我的對手深藍完全沒有這些擔憂。

(Laughter)

（笑聲）

I fought back

在受到這令我身心交瘁的 一擊後，我還擊回去，

after this devastating blow

to win the first match,

贏得了第一場比賽，

but the writing was on the wall.

但凶兆已然顯現。

I eventually lost to the machine

我最終敗給了機器，

but I didn't suffer the fate of John Henry

但我沒有步上約翰亨利的後塵。

who won but died with his hammer in his hand.

他在勝利中與世長辭， 手中還拿著錘子。

[John Henry Died with a Hammer in His Hand Palmer C. Hayden]

（繪畫《約翰亨利握着錘子離世》， Palmer C. Hayden）

[The Museum of African American Art, Los Angeles]

（洛杉磯，非裔美國人藝術博物館）

It turned out that the world of chess

結果，西洋棋的世界

still wanted to have a human chess champion.

仍然想要有個人類棋王。

And even today,

即使現今，

when a free chess app on the latest mobile phone

當免費的手機西洋棋 app ，

is stronger than Deep Blue,

比深藍還強，

people are still playing chess,

人們仍然在下西洋棋，

even more than ever before.

甚至比以前更愛下棋。

Doomsayers predicted that nobody would touch the game

災難預言者預測沒人會再碰西洋棋，

that could be conquered by the machine,

因為西洋棋已被機器給征服，

and they were wrong, proven wrong,

但證據證明他們錯了，

but doomsaying has always been a popular pastime

但災難預言一直是種熱門消遣，

when it comes to technology.

特別是在與科技有關時。

What I learned from my own experience

我從我自己的經驗學到的是

is that we must face our fears

我們必須要面對恐懼

if we want to get the most out of our technology,

以讓科技發揮得淋漓盡致；

and we must conquer those fears

我們必須要克服恐懼，

if we want to get the best out of our humanity.

以讓人性充分發揮。

While licking my wounds,

當我在舔傷口時，

I got a lot of inspiration

我得到很多靈感，

from my battles against Deep Blue.

來自我和深藍的對戰。

As the old Russian saying goes, if you can't beat them, join them.

有句很老的俄國諺語， 如果你不能打敗他們，就加入他們。

Then I thought,

接著，我想，

what if I could play with a computer --

如果我能和一台電腦一起 ……

together with a computer at my side, combining our strengths,

有一台電腦作為盟友， 將我們倆的強項結合，

human intuition plus machine's calculation,

人類直覺加上機器計算能力，

human strategy, machine tactics,

人類策略，機器戰術，

human experience, machine's memory.

人類經驗，機器記憶。

Could it be the perfect game ever played?

那是否會是史上最完美的比賽？

My idea came to life

我的想法實現了。

in 1998 under the name of Advanced Chess

在 1998 年 以「進階西洋棋」之名，

when I played this human-plus-machine competition against another elite player.

我參與了人類搭配機器的比賽 對抗另一位精英玩家。

But in this first experiment,

但在這第一次實驗中，

we both failed to combine human and machine skills effectively.

我們都失敗了，沒能有效 結合人類與機器的技巧。

Advanced Chess found its home on the internet,

「進階西洋棋」在網路上找到它的家，

and in 2005, a so-called freestyle chess tournament

2005 年，所謂的 「自由式西洋棋巡迴賽」

produced a revelation.

帶來了一個啟示。

A team of grandmasters and top machines participated,

參與者包括大師級玩家以及頂尖機器，

but the winners were not grandmasters,

但贏家不是大師，

not a supercomputer.

也不是超級電腦。

The winners were a pair of amateur American chess players

贏家是一對業餘的美國西洋棋玩家，

operating three ordinary PCs at the same time.

他們同時操作三台普通個人電腦。

Their skill of coaching their machines

他們訓練機器所使用的技巧

effectively counteracted the superior chess knowledge

有效地戰勝了大師級對手的 優秀西洋棋經驗

of their grandmaster opponents

以及其他人更強大的計算能力。

and much greater computational power of others.

And I reached this formulation.

我得到了這個組合公式：

A weak human player plus a machine

一個能力不強的人類玩家加一台機器

plus a better process is superior

再加上一個較優的操作流程，

勝過單單一台非常強大的機器，

to a very powerful machine alone,

but more remarkably, is superior to a strong human player

更棒的是還勝過 一個厲害的人類玩家加上機器，

plus machine

以及一個次等的操作流程。

and an inferior process.

This convinced me that we would need

這讓我意識到，我們需要

better interfaces to help us coach our machines

更好的界面來協助我們訓練機器，

towards more useful intelligence.

導出更有用的智慧。

Human plus machine isn't the future,

人類加上機器不在未來

it's the present.

而是現在。

Everybody that's used online translation

大家都用線上翻譯，

to get the gist of a news article from a foreign newspaper,

來了解外國報紙文章的要點，

knowing its far from perfect.

即使知道這翻譯遠非完美。

Then we use our human experience

接著，我們用我們的人類經驗

to make sense out of that,

從翻譯中理出意義來，

and then the machine learns from our corrections.

接著機器從我們的修正中深入學習。

This model is spreading and investing in medical diagnosis, security analysis.

這個模型被廣泛應用在 醫療診斷和安全性分析方面。

The machine crunches data,

機器計算大量的數據，

calculates probabilities,

計算可能性，

gets 80 percent of the way, 90 percent,

完成了 80%、90% 的工作，

making it easier for analysis

讓人類這一方更容易進行分析和決策。

and decision-making of the human party.

But you are not going to send your kids

但你送孩子上學時

to school in a self-driving car with 90 percent accuracy,

不會使用只 90% 精確的自駕車，

even with 99 percent.

甚至 99% 也不行。

So we need a leap forward

所以我們需要更進一步，

to add a few more crucial decimal places.

再加上幾位重要的小數。

Twenty years after my match with Deep Blue,

在我與深藍的第二場比賽 20 年之後，

second match,

this sensational "The Brain's Last Stand" headline

這個聳動的「人腦的最後一搏」標題

has become commonplace

變得很常見，

as intelligent machines

因為智慧機器

move

滲入了每個領域， 且幾乎每天都是如此。

in every sector, seemingly every day.

But unlike in the past,

但不像過去，

when machines replaced

過去機器是取代

farm animals, manual labor,

農場動物、手工勞力，

now they are coming after people with college degrees

現在它們要與擁有大學文憑 和政治影響力的人們競爭。

and political influence.

And as someone who fought machines and lost,

身為一個曾和機器對戰 並敗下陣來的人，

I am here to tell you this is excellent, excellent news.

我在此告訴大家，這是大好消息。

Eventually, every profession

最終，每個行業

will have to feel these pressures

都會感受到這些壓力，

or else it will mean humanity has ceased to make progress.

不然就意味著人性已經停止進步了。

We don't

我們沒得選

get to choose

when and where technological progress stops.

科技進展何時何地會停止。

We cannot

我們不能放慢腳步。

slow down.

In fact,

事實上，

we have to speed up.

我們要加快腳步。

Our technology excels at removing

我們的科技擅長

排除人生中的困難和不確定性，

difficulties and uncertainties from our lives,

and so we must seek out

所以我們必須要找到

ever more difficult,

更困難、

ever more uncertain challenges.

更不確定的挑戰。

Machines have

機器擅長計算。

calculations.

We have understanding.

我們強在思維。

Machines have instructions.

機器有指令，

We have purpose.

我們有目的。

Machines have

機器有客觀性，

objectivity.

We have passion.

我們有熱情。

We should not worry about what our machines can do today.

我們不用擔心 現今的機器能做到什麼，

Instead, we should worry about what they still cannot do today,

而應該要擔心 現今它們仍做不到什麼。

because we will need the help of the new, intelligent machines

因為我們將需要新的智慧機器

to turn our grandest dreams into reality.

來協助我們實現偉大的夢想。

And if we fail,

如果我們失敗，

if we fail, it's not because our machines are too intelligent,

如果我們失敗，那不會是 因為我們的機器太聰明，

or not intelligent enough.

或不夠聰明。

If we fail, it's because we grew complacent

如果我們失敗， 那是因為我們變得自滿，

and limited our ambitions.

限制了我們的野心。

Our humanity is not defined by any skill,

我們的人性並不是 由任何技能來定義的，

like swinging a hammer or even playing chess.

比如揮舞錘子或甚至下西洋棋。

There's one thing only a human can do.

有一件事，只有人類才能做到，

That's dream.

就是夢想。

So let us dream big.

所以，讓我們立下宏願。

Thank you.

謝謝大家。

(Applause)

（鼓掌）