Why People and AI Make Good Business Partners | Shervin Khodabandeh | TED

譯者: Lilian Chiu 審譯者: Helen Chang

I've been working in AI for most of my career,

我的職涯大部分 都投入在人工智慧中，

helping companies build artificial intelligence capabilities

協助企業建構人工智慧的能力，

to improve their business,

改善它們的事業，

which is why I think what I'm about to tell you

這就是為什麼我認為 我接著要分享的內容

is quite shocking.

會蠻震撼的。

Every year, thousands of companies across the world

每年，

全世界有數以千計的企業

spend collectively tens of billions of dollars to build AI capabilities.

總共花費數百億美金

在建構人工智慧能力上。

But according to research my colleagues and I have done,

但根據我和我同事所做的研究，

only about 10 percent of these companies get any meaningful financial impact

只有大約 10% 的企業，

能讓它們的投資在財務面上 產生有意義的影響。

from their investments.

These 10 percent winners with AI have a secret.

這 10% 的人工智慧 贏家有一個秘密。

And their secret is not about fancy algorithms or sophisticated technology.

它們的秘密和很炫的演算法 或精密的科技無關，

It's something far more basic.

反而更基本許多，

It's how they get their people and AI to work together.

那就是如何讓它們的員工和人工智慧

一起工作。

Together, not against each other,

合作，不是對抗彼此，

not instead of each other.

不是取代彼此，

Together in a mutually beneficial relationship.

是在雙方都能受益的關係中合作。

Unfortunately, when most people think about AI,

不幸的是，大部分人 想到人工智慧時，

they think about the most extreme cases.

就會想到最極端的狀況。

That AI is here only to replace us

比如人工智慧只是要來取代我們，

or overtake our intelligence and make us unnecessary.

或者會追上我們的智慧， 讓我們無足輕重。

But what I'm saying

但我要說的，

is that we don't seem to quite appreciate the huge opportunity that exists

是我們似乎不太懂得感念

在這中間所存在的巨大機會，

in the middle ground,

where humans and AI come together

也就是讓人類和人工智慧共同達成

to achieve outcomes that neither one could do alone on their own.

單方面所無法達成的結果。

Consider the game of chess.

想想西洋棋。

You probably knew that AI today can beat any human grandmaster.

你可能知道，現今的 人工智慧可以打敗

任何人類大師。

But did you know that the combination of a human chess player and AI

但你可知道，人類棋手 和人工智慧的組合

can beat not only any human but also any machine.

不僅能打敗任何人類，

也能打敗任何機器？

The combination is much more powerful than the sum of its parts.

合體遠遠勝過當中的任何一部分。

In a perfect combination, AI will do what it does best,

在完美的組合中，

人工智慧會做它最擅長的，也就是

which is dealing with massive amounts of data and solving complex problems.

處理非常大量的資料 以及解決複雜的問題。

And humans do what we do best

而人類做人類最擅長的，

using our creativity, our judgment, our empathy, our ethics

運用我們的創意、判斷力、

同理心、倫理，

and our ability to compromise.

以及妥協的能力。

For several years,

數年來，

my colleagues and I have studied

我和同事

研究了數百間贏家級企業， 並且與它們合作，

and worked with hundreds of winning companies

who are successfully building these human-AI relationships.

這些企業都成功建立了 人類與人工智慧的關係。

And what we've seen is quite interesting.

我們的發現相當有趣。

First of all, these companies get five times more financial value

首先，比起只用人工智慧 來取代人類的企業，

than companies who use AI only to replace people.

這些贏家級企業的 財務價值有五倍之多。

Most importantly, they have a happier workforce.

更重要的是，它們的 整體職工都更快樂。

Their employees are more proud, more fulfilled,

它們的員工更驕傲、更滿足，

they collaborate better with each other, and they're more effective.

他們彼此合作得更佳，且效率更高。

Five times more value and a happier workforce.

價值是五倍，

整體職工還更快樂。

So the question is, how do these companies do it?

所以，問題是， 這些企業怎麼做到的？

How do they achieve these symbiotic human-AI relationships?

它們如何達成這種人類 與人工智慧共生的關係？

I have some answers.

我有一些答案。

First of all, they don't think of AI in the most extreme case

首先，它們不偏頗地認定人工智慧

only to replace humans.

只用來取代人類。

Instead, they look deep inside their organizations

反之，它們更深入探究自己的組織，

and at the various roles their people play.

以及其人員所扮演的角色。

And they ask:

它們還問：人工智慧 如何能讓我們的人

How can AI make our people more fulfilled, more effective,

更滿足，更有效，

more amplified?

更能發揮？

Let me give you an example.

讓我舉個例子。

Humana is a health care company here in the US.

哈門那是一間美國的健康照護公司。

It has pharmacy call centers where pharmacists work with patients

該公司有藥局客服中心， 讓藥師透過電話服務病人。

over the phone.

It's a job that requires a fair amount of empathy and humanity.

這份工作需要相當的同理心和愛心。

Humana has developed an AI system

哈門那開發了一套人工智慧系統，

that listens to the pharmacists' conversation

系統會聽藥師的談話，

and picks up emotional and tone signals

找出情緒和語調上的信號，

and then gives real-time suggestions to the pharmacists

接著即時提供藥師建議，

on how to improve the quality of that conversation.

說明如何改善那次談話的品質。

For example, it might say “Slow down” or “Pause”

比如，它可能會建議

「說慢一點」，或「停一下」，

or "Hey, consider how the other person is feeling right now."

或「嘿，想想對方現在的感受。」

All to improve the quality of that conversation.

全都是在改善那次談話的品質。

I'm pretty sure my wife would buy me one of these if she could,

我很確定，如果可以的話， 我老婆會買這個系統給我，

just to help me in some of my conversations with her.

以協助改善我與她的對話。（笑聲）

(Laughter)

Turns out the pharmacists like it quite a lot, too.

結果發現，藥師也很喜歡這個系統。

They're more effective in their jobs,

他們不只把工作做得更有效率，

but they also learn something about themselves,

還更了解自己，

their own behaviors and biases.

認識自己的行為和偏見。

The result has been more effective pharmacists

結果就是，藥師更有效率，

and much higher customer satisfaction scores.

客人給的滿意度分數也更高許多。

Now, this is just one example of many possibilities where human AI collaborate.

這只是一個例子，

還有很多人類與人工智慧 合作的可能性。

In this example, AI was a recommender.

在這個例子中，人工智慧是建議者。

It didn't replace the human or make any decisions of its own.

它沒有取代人，也沒有自己做決定。

It simply made suggestions,

它就只是提出建議而已，

and it was up to the person to decide and act.

仍然由人類來決定和採取行動。

And at the heart of it is a feedback loop,

它的核心，是回饋的迴圈，

which, by the way, is very critical for any human-AI relationship.

對任何人類與人工智慧的關係， 這種迴圈都很重要。

By that I mean that in this example,

我的意思是，在這個例子中，首先，

first AI had to learn from humans the qualities that would make up a good

人工智慧得要先向人類學習

好的談話或不好的談話有什麼特性。

or not so good conversation.

And then over time, as AI built more intelligence,

隨時間，當人工智慧 建構了更多智慧，

it would be able to make suggestions,

它就能做出建議，

but it would be up to the person to decide and act.

但還是由人類來下決定和採取行動。

And if they didn't agree with the recommendation

如果人類不認同人工智慧的建議，

because it might have not made sense to them,

人類可能會覺得建議不見得合理。

they didn't have to.

In which case AI might learn something and adapt for the future.

此時，人工智慧就可以學起來， 做為未來的參考。

It's basically open, frequent, two-way communication,

基本上，這是開放、 頻繁、雙向的溝通。

like any couples therapist will tell you,

就像任何婚姻治療師都會講的，

is very important for any good relationship.

這種溝通有助於發展好的關係。

Now the key word here is relationship.

這裡的關鍵字是「關係」。

Think about your own personal relationships with other people.

想想你自己和其他人的關係。

You don't have the same kind of relationship with your accountant

你和你的會計師、 你的老闆，或你的另一半

or your boss or your spouse, do you?

都有不同的關係，對吧？

Well, I certainly hope not.

但願是不同的啦。

And just like that,

同樣的，

the right relationship between human and AI in a company

在企業中，人類 與人工智慧的理想關係

is not a one-size-fits-all.

並非一體適用。

So in the case of Humana, AI was a recommender

在哈門那的例子中，

人工智慧是建議者，

and a human was decision-maker and actor.

而人類是決策者和行動者。

In some other examples, AI might be an evaluator

在其他的例子中，

人工智慧可能是評估者，

where a human comes up with ideas or scenarios,

由人類提出點子或情境，

and AI evaluates the complex implications and tradeoffs of those ideas

人工智慧來評估

那些想法的複雜牽扯及權衡，

and makes it easy for humans to decide the best course of action.

讓人類更容易決定 哪種行動步驟最理想。

In some other examples, AI might take a more creative role.

還有一些例子，

人工智慧的角色可能需要更多創意。

It could be an illuminator where it can take a complex problem

人工智慧可扮演照明者，

它可以接受複雜的問題，

and come up with potential solutions to that problem

並提出可能解決那個問題的方案，

and illuminate some options

照明一些人類可能看不見的選項。

that might have been impossible for humans to see.

Let me give you another example.

讓我再舉個例子。

During the COVID pandemic,

在新冠肺炎疫情期間，

if you walked into a retail or grocery store,

如果你走進零售店或雜貨店，

you saw that many retailers were struggling.

會看到許多零售商經營得很辛苦。

它們的架上空無一物，

Their shelves were empty,

their suppliers were not able to fulfill the orders,

它們的供應商無法履約，

and with all the uncertainties of the pandemic,

因為疫情中的不確定性，

they simply had no idea how many people would be walking into what stores,

它們完全不知道

會有多少人走進什麼商店，

demanding what products.

想買什麼產品。

Now, to put this in perspective,

讓我說明一下，

this is a problem that's already quite hard when things are normal.

在正常狀況下， 這個問題就相當難解了。

Retailers have to predict demand

零售商必須要預測

幾千個地點每日

for tens of thousands of products across thousands of locations

對數萬種產品的需求， 還會牽扯到數千家供應商，

and thousands of suppliers every day

to manage and optimize their inventory.

以管理庫存並將庫存最佳化。

Add to that the uncertainties of COVID and the global supply chain disruptions,

此外，還有

新冠肺炎的不確定性 以及全球供應鏈中斷，

and this became 100 times more difficult.

難度提高了一百倍。

And many retailers were simply paralyzed.

許多零售商就癱瘓了。

But there were a few who had built strong foundations with AI

但有少數幾間

建立了強力的基礎，用的是人工智慧

and the human-AI feedback loop that we talked about.

及我們剛才談到的人類 與人工智慧間的回饋迴圈。

And these guys were able to navigate all this uncertainty

這些零售商在這種不確定性當中就能

much better than others.

比其他零售商更清楚方向。

They used AI to analyze tens of billions of data points

它們用人工智慧來分析 數百億筆資料，

on consumer behavior and global supply chain disruptions

資料內容包括消費者行為、 全球供應鏈中斷、

and local government closures and mandates

當地政府的關閉及命令、

and traffic on highways

還有高速公路及海運 通路的交通流量，

and ocean freight lanes and many, many other factors

以及許多其他因素，

and get a pretty good handle on what consumers in each unique area

就能好好掌握 在每個特定的地區中消費者

wanted the most,

最想要什麼、

what would have been feasible,

什麼做法可行，

and for items that were not available,

至於無法取得的品項，

what substitutions could be made.

可以用什麼取代。

But AI alone without the human touch wouldn't work either.

但光有人工智慧， 少了人類獨有的特性

也行不通。

There were ethical and economic tradeoffs that had to be considered.

倫理和經濟之間的權衡也得要考量。

For example, deciding to bring in a product

比如，決定要進一種產品，

that didn't have a good margin for the retailer

零售商的利潤不大，

but would really help support the local community

但能在地方社區需要的時候，

at their time of need.

提供協助支持。

After all, AI couldn't quite understand

畢竟，人工智慧不是很能 了解一些人類獨有的行為，

the uniquely human behavior of panic-buying toilet paper

比如恐慌性搶購廁所衛生紙，

or tens of gallons of liquor,

或者買數加侖的酒精

only to be used as hand sanitizer.

卻只用來做手部清潔。

It was the combination that was the key.

組合，才是關鍵。

And the winning companies know this.

贏家級的企業知道這一點。

They also know that inside their companies,

它們也知道在自己內部，

there's literally hundreds of these opportunities for human-AI combination,

真的有數百個機會可以用到 人類與人工智慧的組合，

and they actively identify and pursue them.

它們會主動找出 這些機會並投入去做。

They think of AI as much more broadly a means to replace people.

它們用更廣的視野去看人工智慧，

不只把它視為取代人類的手段。

They look inside their organizations

它們在自己內部檢視，

and re-imagine how the biggest challenges and opportunities of their company

重新想像

它們最大的難題和機會

can be addressed

如何能用人工智慧 和人類的組合來處理。

by the combination of human and AI.

And they put in place the right combination for each unique situation.

它們針對每個獨特的情況， 安排適合的組合。

Whether it's the recommender or the evaluator

不論人工智慧扮演的是建議者、

評估者、照明者、

or the illuminator or optimizer or many, many other ones.

優化者，或許多其他角色。

They build and evolve the feedback loops that we talked about.

它們會建構並改進我們 剛才談到的回饋迴圈。

And finally and most importantly, they don't just throw technology at it.

最後，也最重要的是，

它們不是就只把技術投下去。

In fact, this has been the biggest pitfall of companies

事實上，這一直都是

在人工智慧上做投資 卻無法回收的那些企業

who don't get their return from their AI investments.

最容易犯的錯誤。

If they overinvest in technology

它們在科技上投資過多，

expecting a piece of tech to solve all their problems.

預期靠科技就能解決 它們所有的問題。

But there is no silver bullet.

但萬靈丹並不存在。

Technology and automation can only go so far,

科技和自動化也有它們的極限，

and for every one automation opportunity inside a company,

在企業中，每有一個自動化的機會，

there's literally ten for collaboration.

就會有十個合作的機會。

But collaboration's hard.

但合作很困難。

It requires a new mindset

合作需要新的心態，

and doing things differently than how we've always done it.

且要用和我們平時 不同的做法來做事。

And the winning companies know this, too,

而贏家級企業也知道這一點，

which is why they don't just invest in technology,

這就是為什麼它們不僅投資科技，

but so much more on human factors,

還在人相關的因素上投資更多，

on their people, on training and reskilling

投資在人身上、訓練他們、 教他們新技能，並重新想像

and reimagining how their people and AI work together in new ways.

它們的員工和人工智慧 如何能用新方式合作。

Inside these companies, it's not just machines replacing humans.

這些企業並不只是以機器取代人類，

It's machines and humans working together,

而是機器和人合作，

learning from each other.

彼此互相學習。

And when that happens,

實現這一點時，

the organization's overall rate of learning increases,

組織的整體學習率會增加，

which in turn makes the company much more agile,

反過來，就會讓企業 更靈活也更有彈性，

much more resilient,

ready to adapt and take on any challenge.

準備好適應並面對任何挑戰。

It is the human touch that will bring the best out of AI.

是人，讓人工智慧發揮到最好。

Thank you.

謝謝。

（掌聲）

(Applause)