Her name is Yahli.

她叫做雅里

She is nine months old.

她現在九個月大

Her mom is a doctor, and her dad is a lawyer.

她的媽媽是個醫生，爸爸則是一名律師

By the time Yahli goes to college, the jobs her parents do are gonna look dramatically different.

等到她上大學時，

In 2013, researchers at Oxford University did a study on the future of work.

她爸媽所做的工作，看起來會跟現在非常不同

They concluded that almost one in every two jobs have a high risk of being automated by machines.

在2013年，牛津大學的研究人員對未來的工作進行調查

Now, machine learning is the technology that's responsible for most of this disruption.

他們得出結論，未來每兩個工作中，就有一個面臨被機器自動化接管的高風險

It's the most powerful branch of artificial intelligence.

機器學習這項科技是導致大部分工作斷裂的主因

It allows machines to learn from data, and mimic some of the things that humans can do.

它是人工智慧最強大的一個分支

Now, my company, Kaggle, operates on the cutting edge of machine learning.

它使機器可以從數據中學習，

We bring together hundreds of thousands of experts to solve important problems for industry and academia.

並模仿人類可以做的一些事情

So this gives us a unique perspective on what machines can do, what they can't do, and what jobs they might automate or threaten.

我的公司「Kaggle」，運營著尖端機器學習科技

So machine learning started making its way into industry in the early 90s.

我們集合數十萬名這類的專家，

And it started with relatively simple tasks.

要來解決工業和學術界的重要問題

It started with things like assessing credit risk from loan applications, sorting the mail by reading handwritten characters from zip codes.

這給了我們一個獨特的觀點，瞭解什麼是機器可以做，

Over the past few years, we have made dramatic breakthroughs.

什麼是它們不能做的，

Machine learning is now capable of far, far more complex tasks.

以及哪些工作可能會被自動化或威脅

In 2012, Kaggle challenged its community to build an algorithm that could grade high-school essays.

機器學習在90年代初開始進入工業界

The winning algorithms were able to match the grades given by human teachers.

它始於相對簡單的工作

Last year, we issued an even more difficult challenge.

它始於像評估貸款申請者的信用風險

Can you take images of the eye and diagnose an eye disease called diabetic retinopathy?

以及透過辨識手寫郵遞區號，將郵件進行分類

Again, the winning algorithms were able to match the diagnoses given by human ophthalmologists.

過去幾年中，我們取得重大突破

Now, given the right data, machines are gonna outperform humans at tasks like this:

機器學習現已能勝任較以往更為複雜的任務

A teacher might read 10,000 essays over a 40-year career.

在2012年，Kaggle 挑戰其社群

An ophthalmologist might see 50,000 eyes.

建立一個可以為高中程度的論文評分的演算法

A machine can read millions of essays or see millions of eyes within minutes.

當時獲勝的演算法所產生的評分已經

We have no chance of competing against machines on frequent, high-volume tasks.

和人類教師的評分不相上下

But there are things we can do that machines can't do.

去年，我們發出一個更困難的挑戰

Where machines have made very little progress is in tackling novel situations.

你能透過拍攝眼睛成像診斷出「糖尿病性視網膜病變」這種眼疾嗎？

They can't handle things they haven't seen many times before.

再一次，獲勝的演算法能夠完全匹敵

The fundamental limitation of machine learning is that it needs to learn from large volumes of past data.

由人類眼科醫生所做出的診斷

Now, humans don't.

目前，只要賦予正確的數據，

We have the ability to connect seemingly disparate threads to solve problems we've never seen before.

在這種任務上，機器的表現將勝過人類

Percy Spencer was a physicist working on radar during World War II.

在40年的職業生涯中，一個老師可能讀過一萬篇文章

When he noticed the magnetron was melting his chocolate bar, he was able to connect his understanding of electromagnetic radiation with his knowledge of cooking in order to invent...

一名眼科醫生可能看過50,000隻眼睛

Any guesses? The microwave oven.

但一台機器可以讀過數百萬篇文章，或者看過數百萬隻眼睛

Now, this is a particularly remarkable example of creativity.

在數分鐘內

But this sort of cross-pollination happens for each of us in small ways thousands of times per day.

我們完全沒有機會與機器競爭

Machine cannot compete with us when it comes to tackling novel situations, and this puts a fundamental limit on the human tasks that machines will automate.

針對這種頻繁、大批量的工作

So what does this mean for the future of work?

但還是有些東西是我們可以做，機器不能做的

The future state of any single job lies in the answer to a single question:

機器在處理未曾遭遇的新情況這方面的進步很慢

To what extent is that job reducible to frequent, high-volume tasks, and to what extent does it involve tackling novel situations?

他們無法處理以前沒有見過幾次的東西

On frequent, high-volume tasks, machines are getting smarter and smarter.

這正是機器學習的基本限制，

Today they grade essays. They diagnose certain diseases.

它需要從過去的大量數據中加以學習

Over coming years, they're gonna conduct our audits, and they're gonna read boilerplate from legal contracts.

人類則不必

Accountants and lawyers are still needed.

我們有能力連結看似不同的個別線頭，

They're gonna be needed for complex tax structuring, for path-breaking litigation.

解決我們從未見過的問題

But machines will shrink their ranks and make these jobs harder to come by.

波西‧史賓塞是二次世界大戰期間，從事雷達研究的物理學家，

Now, as mentioned, machines are not making progress on novel situations.

他注意到磁控管會融化他的巧克力棒

The copy behind a marketing campaign needs to grab consumers' attention. It has to stand out from the crowd.

他能連結接自己對電磁輻射的理解

Business strategy means finding gaps in the market, things that nobody else is doing.

和烹飪的知識

It will be humans that are creating the copy behind our marketing campaigns, and it will be humans that are developing our business strategy.

而發明了... 猜到了嗎？ 微波爐

So Yahli, whatever you decide to do, let every day bring you a new challenge.

現在，這項發明成為特別引人注目的創造性案例

If it does, then you will stay ahead of the machines.

但是，這類「異花授粉」的情況，在我們週圍每天都以小的方式出現數千次

Thank you.

機器無法與人競爭，