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

翻译人员: Yip Yan Yeung 校对人员: Grace Man

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

我大部分的职业生涯 都在研究 AI 领域，

helping companies build artificial intelligence capabilities

帮助公司培养 人工智能（AI）的技术能力，

to improve their business,

改进业务，

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

这就是为什么我认为 我接下来要告诉你的事

is quite shocking.

十分令人震惊。

Every year, thousands of companies across the world

每年，全球成千上万家公司

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

都会在培养 AI 的技术能力上 花费上亿美元。

But according to research my colleagues and I have done,

但是根据我和我同事做的研究，

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

只有大约 10% 的企业 从这类投资中

from their investments.

收获了显著的财务改善。

These 10 percent winners with AI have a secret.

这 10% 的 AI 赢家 有它们的秘诀。

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

它们的秘诀与酷炫的算法 或者复杂的技术无关。

It's something far more basic.

这个秘诀简单得很。

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

就是如何让人与 AI 协作。

Together, not against each other,

通力合作，而不是针锋相对，

not instead of each other.

也不是取而代之。

Together in a mutually beneficial relationship.

是互利共赢。

Unfortunately, when most people think about AI,

不幸的是， 大多数人想到 AI 的时候，

they think about the most extreme cases.

会联想到最极端的例子。

That AI is here only to replace us

AI 只会取代我们，

or overtake our intelligence and make us unnecessary.

或者超越我们的智力， 让我们没有用武之地。

But what I'm saying

但是我要说的是

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

我们似乎并没有关注中间地带上

in the middle ground,

存在的丰富机会，

where humans and AI come together

在这里，人类和 AI 可以携手

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

达到二者缺一不可 才能创造的成果。

Consider the game of chess.

比如国际象棋。

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

你可能已经知道，现在的 AI 已经 可以赢过所有人类国际特级大师。

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

但是你知道 人类棋手和 AI 的组合

can beat not only any human but also any machine.

可以战胜的不仅是所有人类， 还有所有的机器吗？

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

这个组合比双方的总和 要厉害得多。

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

在一个完美的组合里， AI 会发挥所长——

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

处理海量数据， 解决复杂问题。

And humans do what we do best

人类也发挥其所长，

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

充分利用我们的创造力、判断力、 同理心、道德感

and our ability to compromise.

和妥协的能力。

For several years,

多年以来，

my colleagues and I have studied

我和我的同事们学习

and worked with hundreds of winning companies

并与数百家成功利用 AI 获利的公司合作，

who are successfully building these human-AI relationships.

它们已经成功地建立了 这种人类与 AI 的关系。

And what we've seen is quite interesting.

我们所见的非常有趣。

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

首先，这些公司 比那些仅仅只为了取代人类而

than companies who use AI only to replace people.

利用 AI 的公司多获取了 五倍的经济价值。

Most importantly, they have a happier workforce.

最重要的是， 它们的员工更快乐。

Their employees are more proud, more fulfilled,

它们的员工更自豪、 更有满足感，

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

他们更有效地合作，更加高效。

Five times more value and a happier workforce.

多五倍的价值， 更快乐的员工。

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

问题是，这些公司是怎么做到的？

How do they achieve these symbiotic human-AI relationships?

它们是如何达到这种 人类与 AI 的共生关系的？

I have some answers.

我有几个答案。

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

首先，它们不会用极端的情况——

only to replace humans.

取代人类，来看待 AI 。

Instead, they look deep inside their organizations

相反，它们深入研究组织内部，

and at the various roles their people play.

研究员工扮演的不同角色。

And they ask:

它们会问：

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

“AI 如何让我们的员工 更满足、更高效、

more amplified?

价值最大化？”

Let me give you an example.

我来举个例子。

Humana is a health care company here in the US.

休曼纳公司（Humana） 是美国的一家医疗服务公司。

It has pharmacy call centers where pharmacists work with patients

它有药房呼叫中心， 药剂师通过电话

over the phone.

与患者交流。

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

这个岗位需要相当的同理心 和人道主义精神。

Humana has developed an AI system

休曼纳研发了一个 AI 系统，

that listens to the pharmacists' conversation

收录药剂师与患者的对话内容，

and picks up emotional and tone signals

监测情感和语调的信号，

and then gives real-time suggestions to the pharmacists

给药剂师提供实时建议，

on how to improve the quality of that conversation.

告诉他们如何提高对话的质量。

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

比如，它会提示 “慢一点”或者“停一下”，

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

或者“嘿，想想对方 现在是什么感受。”

All to improve the quality of that conversation.

这些提示都是为了提高对话的质量。

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

我很确定如果 我太太买得到这个系统，

她肯定会给我买一个，

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

让我更好地和她对话。

(Laughter)

（笑声）

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

结果是药剂师们 也很喜欢这个系统。

They're more effective in their jobs,

他们工作更高效，

but they also learn something about themselves,

同时也更了解自己，

their own behaviors and biases.

了解自己的行为和偏见。

The result has been more effective pharmacists

这个系统带来了 更高效的药剂师、

and much higher customer satisfaction scores.

更高的客户满意度。

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

这只是人类和 AI 合作的 众多例子中的一个。

In this example, AI was a recommender.

在这个例子中， AI 是一个建议者。

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

它没有取代人类， 或者自主做任何决定。

It simply made suggestions,

它只是提建议，

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

由人类决定如何 做出决定，采取行动。

And at the heart of it is a feedback loop,

它的核心是一个反馈回路，

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

对人类与 AI 的关系至关重要。

By that I mean that in this example,

我的意思是，在这个例子中，

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

AI 首先需要向人类学习 优秀和不怎么样的对话

or not so good conversation.

有什么样的特质。

And then over time, as AI built more intelligence,

随着时间流逝， AI 积累了更多智能，

it would be able to make suggestions,

它就能够提出建议，

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

但是由人类决定 如何做决定和执行。

And if they didn't agree with the recommendation

如果人类无法赞同这个建议，

because it might have not made sense to them,

也许这个建议没什么道理，

they didn't have to.

这个人就可以选择不采纳。

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

无论如何选择，AI 都可以 学到一些东西并为以后调整。

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

这种公开、频繁、双向的交流，

like any couples therapist will tell you,

婚姻治疗师也会这么告诉你，

is very important for any good relationship.

对任何良好的关系都非常重要。

Now the key word here is relationship.

此处的关键词是“关系”。

Think about your own personal relationships with other people.

想想你和其他人的私人关系。

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

你和你的会计师、

or your boss or your spouse, do you?

你的老板或者配偶 都不会有同样的关系吧？

Well, I certainly hope not.

希望没有。

And just like that,

就像这样，

the right relationship between human and AI in a company

公司里 人类和 AI 之间的正确关系

is not a one-size-fits-all.

并没有“均码”。

So in the case of Humana, AI was a recommender

以休曼纳为例， AI 是一个建议者，

and a human was decision-maker and actor.

人类是决策者和执行者。

In some other examples, AI might be an evaluator

在其他的例子中， AI 可以是一个评审员，

where a human comes up with ideas or scenarios,

人类提出想法或者场景，

and AI evaluates the complex implications and tradeoffs of those ideas

AI 评估这些想法的 复杂影响和优劣，

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

让人类更容易做出最佳决定。

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

还有别的例子， AI 扮演了更有创意的角色。

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

它可以指点迷津， 针对复杂的问题

and come up with potential solutions to that problem

提出可能的解决方案，

and illuminate some options

指出一些人类可能

that might have been impossible for humans to see.

无法察觉的选项。

Let me give you another example.

我再举一个例子。

During the COVID pandemic,

新冠疫情期间，

if you walked into a retail or grocery store,

如果你走进一家 零售商店或者超市，

you saw that many retailers were struggling.

你会发现很多零售商都处境艰难。

Their shelves were empty,

他们的货架空空，

their suppliers were not able to fulfill the orders,

供应商无法完成订单，

and with all the uncertainties of the pandemic,

疫情带来的不确定性

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

让他们对有多少人会前来购物、

demanding what products.

购买什么商品毫无头绪。

Now, to put this in perspective,

我们来仔细看一下这个问题，

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

就算一切正常的时候， 这都是一个很难的问题。

Retailers have to predict demand

零售商每天都需要预测

for tens of thousands of products across thousands of locations

来自成千上万的地点和供应商的

and thousands of suppliers every day

数以万计的商品需求，

to manage and optimize their inventory.

管理并优化仓储。

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

加上新冠带来的不确定性 和全球供应链冲击，

and this became 100 times more difficult.

难度增加了百倍。

And many retailers were simply paralyzed.

许多零售商直接瘫痪。

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

但是有些零售商之前 就打下了 AI 的坚实基础，

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

建立了我之前谈到的 人类与 AI 之间的反馈回路。

And these guys were able to navigate all this uncertainty

这些公司比其他公司更好地

much better than others.

在不确定性中站稳脚跟。

They used AI to analyze tens of billions of data points

它们利用 AI 分析数亿数据点，

on consumer behavior and global supply chain disruptions

分析消费者行为、 全球供应链冲击、

and local government closures and mandates

当地政府关闭情况和指令、

and traffic on highways

高速公路交通、

and ocean freight lanes and many, many other factors

海运航线和许多别的因素，

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

很好地掌握了 每一个领域的消费者

wanted the most,

最想要什么，

what would have been feasible,

怎么样最可行，

and for items that were not available,

缺货的商品

what substitutions could be made.

有什么替代品。

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

但是，单单靠 AI， 没有人类的影响也无法成功。

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

还需要考虑道德和经济影响。

For example, deciding to bring in a product

比如，决定要不要引进一款

that didn't have a good margin for the retailer

对零售商没有什么利润，

but would really help support the local community

但是可以为本地的社区

at their time of need.

在它们需要的时候 提供支持的产品。

After all, AI couldn't quite understand

AI 毕竟并不能完全理解

the uniquely human behavior of panic-buying toilet paper

人类疯抢厕纸或者

or tens of gallons of liquor,

只是用作洗手液的大量酒水

only to be used as hand sanitizer.

这样的独特行为。

It was the combination that was the key.

组合在一起才是关键。

And the winning companies know this.

那些获胜的公司懂得这点。

They also know that inside their companies,

它们也知道在公司内部，

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

存在着成百上千 留给人类与 AI 组合的机会，

and they actively identify and pursue them.

它们主动地发掘并追寻这些机会。

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

它们不会把 AI 看作取代人类的 方式，而是思维更加开阔。

They look inside their organizations

它们观察企业内部，

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

重新考虑公司最大的挑战和机遇

can be addressed

应该如何利用

by the combination of human and AI.

人类与 AI 的组合解决。

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

它们会为每一种具体的情况 采取最佳的组合。

Whether it's the recommender or the evaluator

无论是建议者、评审员、

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

引路人、优化器还是其他角色。

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

它们建立并改进了 我们之前谈到的反馈回路。

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

最后一点，也是最重要的一点， 它们不会一门心思只搞技术。

In fact, this has been the biggest pitfall of companies

实际上有很多公司都遭遇了危机，

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

它们在 AI 上的投资 没有得到回报。

If they overinvest in technology

它们过度投资技术，

expecting a piece of tech to solve all their problems.

指望一项技术可以解决所有问题。

But there is no silver bullet.

并没有什么一招鲜。

Technology and automation can only go so far,

科技和自动化止步于此，

and for every one automation opportunity inside a company,

但是公司里每一个自动化的机会，

there's literally ten for collaboration.

都对应着十个合作的机会。

But collaboration's hard.

但是合作很难。

It requires a new mindset

这需要一种崭新的思维，

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

抛开传统的方式， 采取与众不同的行动。

And the winning companies know this, too,

获胜的公司也懂这些，

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

这就是为什么 它们不会只投资于技术，

but so much more on human factors,

还会投资于人力，

on their people, on training and reskilling

于员工，于培训和技能提升，

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

于重新考虑人类和 AI 协作的全新方式。

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

在这些公司里， 机器没有要取代人类。

It's machines and humans working together,

机器会和人类一起工作，

learning from each other.

互相学习。

And when that happens,

这时，

the organization's overall rate of learning increases,

这个公司的总体学习率提升了，

which in turn makes the company much more agile,

从而让公司变得更加敏捷、

much more resilient,

更加能屈能伸，

ready to adapt and take on any challenge.

适应性更强， 应对各种挑战。

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

人类的介入能让 AI 物尽其用。

Thank you.

谢谢。

(Applause)

（掌声）