特别节目：麦肯锡在2026年消费电子展

这是麦肯锡播客，我们将帮助您理解当今世界最棘手的商业挑战。

This is the McKinsey podcast, where we help you make sense out of our world's toughest business challenges.

欢迎收听本节目。

Welcome to the show.

我是露西娅·拉希莉。

I'm Lucia Rahily.

我是罗贝塔·法萨拉。

And I'm Roberta Fasara.

今年，企业领导者在创新和消费电子领域应该关注什么？

What should business leaders focus on this year when it comes to innovation and consumer electronics?

本周麦肯锡参加了在拉斯维加斯举行的消费电子展，借此特别节目，我们将探讨这一问题。

We're gonna explore that question on this special edition of the McKinsey podcast because McKinsey was at the consumer electronics show in Las Vegas this week.

当然，所有讨论都围绕着人工智能能做什么展开。

And of course, all the talk was about what AI can do.

但麦肯锡全球主管合伙人鲍勃·斯特恩菲尔德表示，人类有三件事是唯一能够做到的。

But Bob Sternfelds, McKinsey's global managing partner says there are three things that humans can exclusively do.

那就是抱负、设定参数并创造性地思考。

And that is aspire, set parameters, and think creatively.

抱负。

Aspire.

设定正确的抱负。

Set the right aspiration.

这是人类独有的能力。

That's a uniquely human capability.

那么，你如何寻找那些关于树立抱负并让他人信服这一抱负的技能呢？

So how do you look for the skills about aspiring and getting others to believe in the aspiration?

人类的。

Human.

判断力。

Judgment.

那么，你如何根据公司价值观、社会规范等设定正确的参数和架构呢？

And so how do you set the right parameters, the architecture based on firm values, based on societal norms, whatever?

你如何培养能力，来确定正确的参数？

How do you build the skills to set what the right parameters are?

最后，真正的创造力。

And then finally, true creativity.

这些模型是推理模型。

The models are inference models.

下一个最可能的步骤，你如何思考那些非传统的方面？

The next most likely step, how do you think about orthogonal stuff?

它能让你重新审视自己在寻找人才时的一些假设。

It can take you back to challenging some of your assumptions on where you look for talent.

这意味着你毕业于哪所学校变得不那么重要了。

It actually means that where you went to school matters a lot less.

那么，你是否开始寻找原始的内在特质？

And so do you start looking for raw intrinsics?

你能扩大人才的基础吗？

Can you widen the base?

你能真正去看一下吗？以技术背景为例，不是你毕业于哪所大学，而是你的GitHub个人主页是什么样子？

Can you actually look at, let's take a tech background, not which university you graduated from, but what does your GitHub profile look like?

让我们真正关注内容本身，这是否意味着更多不同背景的人可以通过不同路径进入劳动力市场？

Let's actually get to the content, and could that actually start meaning that a wider set of people can enter the workforce with different pathways?

在CES上，麦肯锡高级合伙人比尔·威斯曼表示，人们正在热议人工智能在现实世界中的应用。

Also at CES, McKinsey senior partner Bill Wiseman says people were buzzing about the real world applications of AI.

那么我所说的这一点具体指的是什么？

So what am I talking about by that?

我看到一些量子公司与先进计算公司联手，探讨我们能共同应对哪些问题？

Saw some quantum companies coming together with advanced compute companies and saying, what can we jointly take on?

这涉及到你需要真实的物理化学和物理模拟来解决的问题。

And that's questions of where you need real physical chemistry and physics simulation to be able to solve problems.

分子相互作用，比如药物研发。

Molecular interactions, things like drug discovery.

今年这里就展示了这些内容。

That was on display here this year.

这是一个可以通过先进计算解决的实际问题。

That's a real problem you can solve with advanced computing.

另一个例子是一个控制手术机器人的大语言模型。

Another one was an LLM that was controlling a surgical robot.

这不是一个需要你指令它做手术的模型，而是通过训练让它自己学会如何做手术。

So that's not one that you instruct to do surgery, that's one that you train to do surgery and it figures out how to do it.

当时实际上正在进行一场针对假体的实时脊柱手术，而不是对真人。

And that was actually a kind of live spine surgery going on on a dummy, not a live person.

但看到有人真的能演示这一点，还是非常令人惊叹的。

But it was pretty amazing to see that someone was actually able to demonstrate that.

比尔还表示，当时还重点关注了降低计算成本。

Bill also says there was a focus on bringing down the cost of computing.

另一方面，今年人们也高度关注先进计算的成本问题。

On the flip side, say there was a lot of focus this year on cost of advanced compute.

但这并不是指芯片成本或数据中心成本。

And that wasn't about chip cost or data center cost.

那是指每个令牌的成本。

That was about cost per token.

而且有很多讨论集中在：是什么推动了每个令牌的高成本，以及如何以不同方式配置数据中心来影响服务水平，从而降低每个令牌的成本？

And there was a lot of instruction about, okay, what drives advanced per token and how can you configure a data center differently to affect service levels and therefore bring down cost per token?

因此，很明显，成本至关重要，我们需要降低这项技术的成本，才能继续推动先进计算的普及。

So just there's a clear focus that cost matters and that we we need to bring the cost of this down in order to continue to proliferate advanced compute.

成本问题也同样在麦肯锡高级合伙人史蒂文·富克斯的考虑之中。

Cost was on McKinsey senior partner, Steven Fuchs' mind as well.

他说，尽管这些人工智能实验令人兴奋，但他没有听到足够多关于如何构建满足这一人工智能时代所需系统的讨论。

And he says that while these experiments in AI are exciting, he didn't hear enough conversation about how to create the systems required to meet this AI moment.

我们是在现有环境中工作的。

We are working in a brownfield environment.

每一家公司都有现有的工作流程和现有的人员配置。

Every single company has existing workflows, has an existing set of people.

我们该如何帮助整个体系实现转型？

How do we help the collective transition?

拥有使用场景、试点项目，以及推动思维都很棒，但真正实现规模化需要的是系统性变革。

It's great to have the use cases, it's great to have the pilots, it's great to have push the thinking, but really scaling it is a system change.

因此，很多讨论都集中在边缘案例上。

And so a lot of the conversations are around the edge case.

目前还没有围绕如何构建一个能够规模化释放价值的生态系统展开。

They're not yet around how do we really create an ecosystem that allows us to extract value at scale.

其中一个获得广泛关注的生态系统是涉及自动驾驶汽车的那一个。

One ecosystem that's gotten a lot of play is the one involving autonomous vehicles.

麦肯锡资深合伙人马丁·凯勒纳对今年CES上看到的诸多出行进步感到惊讶。

McKinsey senior partner Martin Kellner was amazed at all the mobility advancements he saw at CES this year.

我昨天到达时真的非常震惊。

I was really shocked when I arrived yesterday.

路上有这么多机器人出租车。

So many robotaxis on the roads.

它们更加果断，也更像人类，即使在恶劣天气条件下也能自如驾驶。

They are much more assertive, they are much more human like, driving around even in adverse weather conditions.

我现在预计这个行业将迈向商业部署和大规模运营，意味着更多的城市、更多的车辆上路，以及每辆车的出行次数增加。

I now expect the industry to moving to commercial deployment and large scale operations, meaning more cities, more cars on the road, but also more trips per car.

他说，行业领导者今年要想成功，需要做好三件事。

And he says industry leaders need to do three things to be successful this year.

第一件事是获得客户认可。

First one is getting customer traction.

第二件事是提高运营效率并控制成本。

Second one, being operational efficient and managing the cost.

第三件事是围绕那些自己不想做的领域建立合作伙伴关系。

And third one, also building partnerships around the things they don't wanna do themselves.

比尔·威斯曼对他在CES上看到的人形机器人印象深刻。

Bill Wiseman was impressed with the humanoid robots he saw at CES.

但他表示，明年这个时候，他对它们的期待会更高。

But he says this time next year, he expects even more from them.

我认为明年将是见证人形机器人大规模实际演示的一年。

I think next year is gonna be the year where we see real demonstrations of humanoid robotics at scale.

每个人都有一台机器人。

Everybody had a robot.

这些机器人移动得非常慢。

Those robots were moving very slow.

它们在做一些有趣的事情，但看起来很无聊。

They were doing interesting things, but they were boring to watch.

你意识到，你看到的几乎所有关于这些机器人的视频都快进了大约10倍。

And you realize that almost all the videos you see of those things are sped up probably by a factor of 10.

另一个令人印象深刻的是机器人的执行器和一些部件。

The other thing that was impressive was the actuators and some of the elements of a robot.

这些是驱动机器人关节的电机，无论是手指、肘部、肩膀还是腿部，它们都能产生巨大的力量和扭矩，而这正是让300到400磅的类人机器人成为现实所需要的。

Those are the motors that move joints in robots, whether it's fingers or elbows or shoulders or legs, they're gonna be able to generate a lot of force and torque, which is what you need for 300, 400 pound Cubanoid robots to become real.

我看到的最后一件事也非常有趣。

And the final thing I saw a lot of, it was really interesting.

我认为明年我们将在系统中看到触觉传感器。

I think we're gonna see in systems next year are tactile sensors.

确保机器人知道如何触碰而不损坏或掉落物品。

Making sure robots know how to touch without breaking and without dropping.

触觉传感器不仅存在于手上，还要分布在身体的不同部位，以免犯错。

And tactile sensors, not just in the hands, but in different places in the body so you don't wind up making mistakes.

今年我看到的另一件大量讨论的事情是安全性。

And that's the one thing that I did see a lot of this year, a lot of talk about safety.

在机器人领域，没有人真正解决了安全问题。

Nobody's solved the problem of safety when it comes to robotics.

明年这也将继续成为关注焦点。

That's gonna be on display next year as well.

非常感谢您收听麦肯锡播客。

Thanks so much for listening to the McKinsey podcast.

我是露西娅·拉希利。

I'm Lucia Rahili.

我是罗贝塔·法萨罗。

And Roberta Fasaro.

关注我们

Find us

请访问 mckinsey.com。

on mckinsey.com.

我们很快会发布本集的文本稿。

We'll have a transcript of this episode up shortly.

并下载麦肯锡洞察应用，您可以在其中找到本播客及其他每日更新的实用内容。

And download the McKinsey insights app where you can find this podcast and other helpful content updated daily.

如果您喜欢这个节目，我们非常希望您

If you enjoy the show, we'd love for you

能给我们打分并留下评论。

to leave a rating and a review.

两周后见。

We'll see you in two weeks.