機器人革命不會殃及人類子孫

The fact that the Financial Times special series on robots has been framed around the question “Robots: friend or foe?” is fascinating to me. I understand that change is scary, and that the cyborg from The Terminator is really scary. But my reading of the evidence is that the rapid changes we are seeing in artificial intelligence, sensor development and many other fields are not sending us hurtling towards a future in which the machines become self-aware and take over. Worrying about that future is so misplaced that, as the AI pioneer Andrew Ng puts it, it is like worrying about overpopulation . . . on Mars.

英國《金融時報》關於機器人的特別系列報道一直圍繞着這樣一個問題：機器人是敵是友？這讓我很感興趣。我知道變化是可怕的，我也知道《終結者》(Terminator)中的半機器人確實可怕。但我讀到的證據表明，我們在人工智能(AI)、傳感器開發以及其他很多領域看到的快速變化，並沒有在把我們快速推向一個機器產生自我意識並接管一切的未來。擔心這種未來毫無意義，就像人工智能先驅吳恩達(Andrew Ng)所說的那樣，這就好像擔心……火星上人口過多。

A much more proximate and real threat, and hence a scarier one, is that of economic dislocation. Robots (a term I am using here as shorthand for the modern, rapidly expanding toolkit of digital technologies) are quickly learning lots of skills — everything from understanding natural language to diagnosing disease to driving cars — that used to be the preserve of human beings alone. These skilful machines are going to spread throughout the world’s economies in the years to come, and they are going to automate some people, perhaps many of them, out of their jobs.

一種更緊迫且更切實（因此也更可怕）的威脅是經濟可能出現混亂。機器人（我姑且使用這個詞代指快速擴張的現代數字技術工具）正快速學習大量曾經是人類專屬的技能——從理解自然語言到診斷疾病、再到駕駛汽車。未來數年，這些技術高超的機器將出現在全球各個經濟體中，他們將用自動化取代人工，致使一些（或許多）人失業。

Surely this means that robots are our foes? No, it absolutely does not. To believe otherwise is to reduce us humans to the status of mere labourers — and that is an offensively reductive view of our species.

這當然意味着機器人是我們的敵人嘍？不，完全不是這樣。認爲答案是肯定的，就是把人類貶低爲只不過是體力勞動者，這未免太小瞧我們這個物種了。

For one thing, many people are not workers at all; they are children or pensioners, the sick or infirm. For these groups, technological progress is a virtually unalloyed good. It will allow the elderly to lead more autonomous lives (think of a self-driving car that will let them visit friends and relatives, for example), while letting their families closely monitor them and intervene if they fall or become disorientated. It will let children learn what they want, at their own pace. It will let us tailor medical treatments to individuals, rather than simply doing what works best on average.

首先，很多人根本不是勞動者；他們是兒童或退休人士、病人或殘障人士。對於這些人而言，技術進步是近乎純粹的福音。它將讓老年人能過上更自立的生活（比如，設想一下無人駕駛汽車帶着他們探親訪友），同時讓他們的家人能密切監控他們的狀況，在他們摔倒或迷路時進行干預。它將讓孩子們能按照自己的進度學習他們想學的知識。它將讓患者能獲得最適合自己的醫學治療，而不是僅僅獲得通常而言最有效的治療。

I struggle to see how technological progress could be negative, on balance, for any of these groups. Yes, children can spend too much time staring at their smartphones, and parents should ensure their screen time is limited. But we survived the advent of television, a technology perfectly designed to turn us into zombies who cease to engage with each other in any way. I am confident, therefore, that the kids will be all right.

總的來說，我想不出對所有這些羣體而言，科技進步何以會是一件不好的事情。確實，兒童可能會花費太多時間盯着他們的智能手機，父母應控制他們玩手機的時間。但電視這種原本足以讓我們變成彼此不進行任何互動的殭屍的技術誕生了，我們安然無恙；因此，我相信，孩子們會沒事的。

But what about those of us who actually do work for a living? Even for us, robots are not our enemies — for two important reasons. First, they are not going to put us all out of work any time soon. There are still a lot of things that technology cannot do — from clearing a table to coaching a team to writing a novel. And even the biggest techno-optimists do not think that these things will be 100 per cent automated in the near future.

但我們這些真正在爲生計工作的人呢？即便對於我們，機器人也並非敵人，原因有兩個。首先，他們不會很快讓我們都失業。仍有很多事情是機器做不了的——從收拾桌子到教一個班的學生寫小說。就連對技術發展最樂觀的人也不認爲，這些工作將在不遠的未來實現100%自動化。

It is true that a lot of the jobs that technology cannot do today do not pay very well (largely because lots of people can do them, and this large supply of potential labour keeps wages down). The best way to address this challenge, I believe, is with a negative income tax that provides low-income workers with an earnings boost.

確實，機器現在做不了的很多工作都薪水不高（這主要是因爲可以做這些工作的人很多，大量的潛在勞動力供應讓薪資高不起來）。我認爲，解決這一挑戰的最佳辦法是將收入所得稅降至負值，爲低收入員工增加收入。

The second and much more important reason that robots are not our foes is that they make us richer overall. By increasing our capabilities and productivity, they create more bounty and abundance. We like to communicate, learn, entertain ourselves, travel and consume goods and services. Technological progress lets us do more of all of these things for a given amount of money (or, increasingly these days, for no money at all), and at higher levels of quality.

機器人並非敵人的第二個也是更爲重要的原因在於，他們讓我們總體變得更爲富有。通過提高我們的能力和生產率，他們創造了更富足的生活。我們喜歡交流、學習、娛樂、旅行以及消費產品和服務。技術進步將讓我們能用同樣的錢（乃至免費——按照如今的趨勢），以更高的質量做更多我們喜歡的事情。

It is true that the way most of us gain access to much of this bounty is by getting paid for our labour. It is also true that this “labour bargain” is becoming a tougher one for more and more people as their skills become less valuable, because of both globalisation and technological progress. We need to figure out how to deal with this situation This will be one of the most important policy arenas over the coming decades.

確實，我們多數人享受到這種富足的方式是通過獲取勞動報酬。此外，確實，對於越來越多的人而言，這種“便宜勞動力”正變得愈發難以消受，因爲由於全球化和技術進步，他們的技能變得不那麼有價值了。我們需要弄清楚如何應對這種情況。這將是未來數十年最重要的政策領域之一。

But we also need to keep in mind that this is a situation brought about by the fact that technology is letting us do and create much more with much less drudgery and toil. If we cannot figure out how to deal with this, and how to make sure that the fruits of robots’ labour are shared in a way that reflects our shared values and protects our most vulnerable, then shame on us. In that case, we will have met the real foe in that case, and it will be us.

但我們還需要記住，這種情況源於這一事實：技術讓我們得以費少得多的力氣創造多得多的價值。如果我們無法弄明白如何應對這種形勢，如何確保以反映我們共同價值觀、保護弱者的方式分享機器人的勞動成果，那麼我們應該感到羞愧。如果是那樣的話，我們已遭遇真正的敵人，那就是我們自己。