金融時報:大數據造就老大哥？

大數據造就老大哥？

令人炫目的“大數據”技術有著無限的潛力改變世界，它能讓普通人也過上《唐頓莊園》里的貴族生活——輕松得到量身定做的個性化服務。但是，斯諾登泄密事件也深刻反映了大數據可能帶來的對個人隱私的侵犯。FT專欄作家John Gapper認為，必須警惕政府有關部門和網絡公司通過大數據正在積聚著的巨大權力。難怪，斯諾登事件爆出后，《1984》的銷量猛增。

測試中可能遇到的詞匯和知識：

Big Brother “老大哥”，喬治·奧威爾的名著《1984》中的經典形象，一個令人感到窒息和恐怖的、控制一切的“大洋國領袖”。"Big Brother is watching"是書中一句名言，指個人言行隨時被官方監(jiān)視著。

artificial intelligence 人工智能，即AI

Taj Mahal 泰姬陵

Uttar Pradesh 印度的北方邦

Brick Lane 磚塊街/布里克巷，在東倫敦，現在是孟加拉移民的社區(qū)。

Big data has to show that it’s not like Big Brother(977 words)

Sales of George Orwell’s Nineteen Eighty-Four have risen since Edward Snowden revealed how the National Security Agency of the US gains access to telephone records and data from technology companies. So far, if people do not exactly love Big Brother, they are prepared to accept some invasion of their privacy in return for security.

What about “big data”? Companies that hold rapidly expanding amounts of personal information are using new kinds of data analysis and artificial intelligence to shape products and services, and to predict what customers will want. Larry Page, Google’s chief executive, describes his ideal form of technology as “a really smart assistant doing things for you so you don’t have to think about it”.

The vision of living in a virtual Downton Abbey, with a computer to plan your day, suggest the best route to travel, the films you might want to watch and the best flight to catch – even to book it for you – has an allure. We are all pressed for time and want an easy life. Instead of being bombarded with information and forced to choose, it’s nice to get personal service.

But just as the NSA disclosures have taken people by surprise, although it has existed for 60 years, I doubt whether many grasp either the size of the data trail they create daily, or the advances in technology that are permitting a select group of big data enterprises to exploit it. The technology is evolving so quickly that what was unthinkable two years ago is routine.

“It is both a wonderful and scary future. Companies with huge amounts of data will know more about you than yourself. They will be able to predict what you might do next,” says Kai-Fu Lee, a Beijing-based investor and the former head of Google in China.

In a column last week I compared Google to General Electric in the late 19th century – an innovative industrial enterprise riding a wave of new technology. The flip side of that is that Google, Amazon, Microsoft and other technology giants are amassing powers that need to be controlled carefully.

The NSA and big data companies put their databases and computing power to different uses – one to identify spies and terrorists, and the others to match services to users. They have in common the use of very large databases and techniques such as pattern recognition and network analysis.

At the advanced end, this shades into artificial intelligence of the kind that, for example, intuits what you meant to search for even when you misspell the key words; can translate speech into another language in real time (as Microsoft demonstrated in China last year); or learns to recognise a photograph of a cat by viewing thousands of images.

The ability of computers to learn in a similar manner to humans is known as “deep learning” and it is notable that Google has hired several pioneers in the field, including the scientist and author Ray Kurzweil. Among the technology transfer offered by the NSA to private US companies are “cutting-edge machine learning technologies”.

Such software can infer a lot from scraps of information, provided that it has enough of them, as shown by the NSA’s effort to analyse phone call metadata from Verizon (and perhaps other operators). President Barack Obama assured Americans that “no one is listening to your phone calls”, but this alone is a trove.

A study by Latanya Sweeney, a professor at Harvard University, found that 87 per cent of people can be identified simply by knowing their age, gender and postcode, if these are cross-checked against public databases. That is typical of the data collected by social networks and internet companies.

The extraordinary power of big data companies comes from being able to combine the personal data of customers with observations about them, from which products they buy to where (as measured by global positioning satellite data from mobile phones) they are. That produces a set of “inferred data” about what they probably want.

If I search on an Android phone for “Taj Mahal” while standing in India, for example, Google will prioritise results for the shrine in Uttar Pradesh. If I do the same in Brick Lane, east London, it will suggest local Bangladeshi restaurants. How long before it offers to book a restaurant based on how I rated others as I walk around a foreign city at dusk?

At one level, I would be pleased if it did (as long as it was a good one) since it would save me doing the work myself. At another, as a World Economic Forum report on personal data put it: “Inferred data can feel like an all-knowing Big Brother watching the security camera.”

One of the concerns that springs from this is that big data companies with such software are very difficult to compete with. The more data that I and other users provide them with, the better they are at predicting what we want. The machine brain becomes cleverer with use.

Another is trust. Social networks have been poor at protecting users’ data, and they hold only a fraction of the information on people’s behaviour, habits and intentions on the new generation of services. It is no wonder that the NSA turns to them – it has computing power and they have swaths of material.

A third is ownership. We each have rights over our own information, but what happens when it gets mixed up with that of others and combined into a vast database of intentions? If I change my mind, how can it be unscrambled?

Above all, we don’t know what this technology means because we are only at the beginning of the era of big data. There are plenty of aspects to admire but it will take some time to love.

請根據你所讀到的文章內容，完成以下自測題目：

1.About artificial intelligence, which of the following is not correct?

A. It is based on data analysis.

B. Computers will be able to predict what customers want.

C. It was an idea first brought up by Google’s cofounder Larry Page.

D. "A really smart assistant doing things for you so you don’t have to think about it."

答案(1)

2.There are technology giants that "are amassing powers that need to be controlled carefully", which company is not among them?

A. Google.

B. Amazon.

C. Facebook.

D. Verizon.

E. Cisco.

答案(2)

3.What best explains the concern that big data companies are "very difficult to compete with"?

A. They usually have ponopoly power on the market.

B. The more data they have, the better they can do.

C. They are very rich and have armies of genius engineers.

D. They have already forged powerful lobby groups.

答案(3)

4.Which of the following is significantly different from the other three?

A. "network analysis"

B. "deep learning"

C. "cross-checking"

D. "inferred data"

答案(4)

* * *

(1) 答案：C.It was an idea first brought up by Google’s cofounder Larry Page.

解釋：人工智能的概念很早就有了，AB都正確，D是拉里·佩奇說的。

(2) 答案：E.Cisco.

解釋：從文中不難得知，那些可以得到大量用戶數據的、可以進行大數據業(yè)務的公司，正在積聚著巨大的權力，這樣的公司有搜索引擎、社交網絡、電信公司、網上商城等。ABCD都是典型。而思科主營電信硬件設備，不大可能擁有這樣的數據。

(3) 答案：B.The more data they have, the better they can do.

解釋：ACD在某種程度上是有道理的，但是最根本的原因還是大數據的“自然壟斷”性——類似水電燃氣等基礎設施，用的人多了，再多的人就只能用它們的。

(4) 答案：C."cross-checking"

解釋：ABD都是大數據技術專有的術語，而C相互檢驗和反復核實則是早就有的技術，至少信息化社會之前就有了。