为什么物理AI是下一个前沿 | Applied Intuition

读过吗？

been Have you

那本书你看过吗？

read that book?

是啊，几年前我读过。

So I years ago, I have.

这可是史上最棒的书名之一，咱们得停下来感叹一下，约翰·德洛里安当时心里肯定在想：啥玩意儿？

It's one the great all time book titles, and we should just pause and say, John DeLoren was like, what?

他可是汽车界的超级天才。

He was like the super genius of the car industry.

他本来要当下一任通用汽车的总裁。

He was gonna be the next president of General

通用汽车的总裁，后来他自己创办了一家汽车公司，那家公司的故事在《回到未来》里出现过。

Of General Motors, and then later on, he started his own car company, which

对，就是《回到未来》里的那家。

was Which was in Back to the Future.

《回到未来》之后，整个项目因为各种原因最终垮了。

Back to the future, and then that whole thing collapsed for a variety of reasons.

但他确实是个传奇人物。

But, yeah, he was like a legend.

他是汽车行业中推动创新的核心人物之一。

He was like one of the main principal drivers of innovation in the car industry.

没错。

Exactly.

李·艾柯卡，就是这一类人。

Li Iacoco Bob Lots, this category.

对。

Yeah.

你得记住，他就在这个角落里。

And you gotta remember, is in this right corner.

抱歉，你能再说一遍这本书的名字吗？

Sorry, repeat the title of the book.

晴天时，看得见通用汽车。

On a Clear Day, You Can See General Motors.

那这本书为什么取这个名字？

And why was that the title of the book?

因为

Because

里面全是胡扯。

there's a lot of bullshit.

特别庞大又复杂。

Very large complex.

是的。

Yeah.

复杂得很。

Complex.

是的。

Yeah.

是的

Yeah.

简直就像一个国家一样。

It's like a it's like a nation state.

是的

Yeah.

我的意思是，真的。

It mean, really.

我的意思是，确实是。

I mean, is.

我觉得人们有时候会随便这么说。

Like, I think people would say that, like, sometimes almost, like, flippantly.

对。

Right.

但这些公司实际上就像是延伸体。

But these companies are, like, extension.

比如现代汽车其实就是国家的延伸。

Like, Hyundai is an extension of the state.

丰田也是国家的延伸。

Toyota is an extension of the state.

大众汽车的董事会成员，根本就是政府官员。

Volkswagen is literally, Volkswagen board members are members of the government.

没错。

Right.

所以这些公司都是国家的延伸，几乎每一个都是。以前有句老话，说对通用汽车有利的就是对美国有利。

So these are extensions of the state and and and almost every and there used to be old saying, what's good for General Motors is good for America.

没错。

Right.

你根本无法低估通用汽车在美国企业历史中的重要性。

And you cannot under state how important General Motors is the history of the American corporation.

斯隆写的《我在通用汽车的岁月》和《白领人的冒险》，如果你在管理大型工程团队，一定要读一读。

Sloan's My Years at General Motors, An Adventures of White Collar Man, if you run a large engineering organization, you should read that.

现代企业这种模式并不是凭空出现的，它有它的来龙去脉。

That is the the this like, do you this this thing that that we talk as a modern corporation didn't just emerge.

没错。

Right.

斯隆和凯特林——凯特林是工程主管——一起打造了这套体系，有层级、有副总裁，还有职能型和矩阵型的组织结构。

Sloan and Kettering create Kettering's the head of engineering, created this with this arc you know, with levels and vice presidents, and how do you do functional and matrix organizations.

这简直就是现代企业的源代码。

It's there really is like the source code.

后来约翰·德洛里安来了，他说他要当总裁，但对这家公司已经忍无可忍。

Comes along, John, know, comes on DeLoren, and he says He writes that he's gonna be president, and he's so fed up with a company.

但当时通用汽车其实经营得非常好。

But what was controversial was GM was doing really well at the time.

我们说通用汽车在《财富》百强里排第一，其实它包揽了第一、第二和第三名。

GM was like a When we say like GM was number one in Fortune 100, it was like number one, two, and three.

它无所不在，被公认为美国最顶尖的企业。

It was everything, and it was seen as the best company in America.

所以有人敢公开批评这家公司。

So somebody to openly criticize the company.

于是他写了一本书，辞职时满心不满，想揭露通用汽车管理上的种种问题。

And so he has a hope He writes this book as he quits out of out of how annoyed he was how General Motors was being read led.

书写完后，他冷静下来，觉得自己不该让这本书出版。

He writes his book, and then after he, like, sobers up, he's like, I don't want that book published.

之后他花了好几年时间，拼命阻止合著者出版这本书。

And so he fights for years for his coauthor not to publish the book.

但合著者还是把它出版了。

The coauthor still publishes.

这本书真实地揭示了一家大型企业的内部运作。

So it's a real true insight into a large corporation.

说来也巧，我最近才读到这本书，虽然我二十多年前就在通用工作过，对这家公司相当了解。

I'm incidentally just reading it out, even though I've, you know, worked at GM twenty some years ago and had, you know, know a lot about the company.

最让人震惊的是，这不只是通用的问题，大多数大型制造企业如今内部依然这么运作。

And what's shocking is it's not only about GM, most of the major manufacturers actually still operate that way on the inside.

所以我觉得，大家真正该明白的不是这些公司管理者笨。

And so the question isn't the the the point, I think, for everyone to take away isn't that these people who run these companies are stupid.

对。

Right.

他们不笨。

They're not stupid.

这就像你跟国防部做生意时的情况一样。

They're it's kinda like, you know, when you're selling to the department of war.

对。

Right.

人们会问，那你为什么这么做？

And people say, well, why are you doing that?

因为销售渠道决定了你的生意模式。

It's like, well, the distribution defines the business.

对。

Right.

也就是说，这个分销渠道的问题，可能这个数据有点过时了，但二十年前我在安全系统部门工作的时候，还记得通用汽车天天在我们耳边强调。

That like, so like, the distribution is this is a consumer product of of this this stat might be outdated, but when I worked in Safety Systems twenty years ago, I remember GM used to pound into your head.

在美国历史上最严重的五起消费者诉讼案中，有三起都跟汽车有关。

On the top five consumer lawsuits in in American history, three are automotive.

是的。

Right.

我们占了大多数。

We got the majority.

对吧？

Right?

所以，必须格外谨慎，这反而是一种幸运。

So it's a gift to be extremely careful.

公司里还有些奇怪的规定，不是简单的红黄绿，而是……比如紫色。

We had these, like, weird things, like, inside the company, you couldn't it wasn't red, yellow, green.

你总得当个解码员，你知道为什么吗？

It was, like, purple, or like, you'd always have to as decoder, because you know why?

因为一旦出事打官司，他们就会说：你们明明看到安全系统标了红色，还让它投入生产。

Because when they go to lawsuits, they're like, you let a safety system that was marked red go to production.

结果你得解释：不对，那是洋红色。

Was like, no, it was marked magenta.

是的。

Right.

没错。

Right.

你能想象这有多让人抓狂吗？

So, can you imagine how infuriating that is?

每次你都得琢磨：橙色到底代表什么？

Every time you're like, what does orange mean?

这还只是让你赶紧跳到你那个会议里，去搞清楚那个系统到底什么意思。

Does this mean I have to like so fast forward to your meeting that system.

很长一段时间以来，他们的口号一直是‘质量第一’。

Well, slogan for a very long time was that it was quality is job one.

对吧？

Right?

是的。

Yeah.

对。

Yeah.

还是安全才是第一要务？

Versus safety is job one?

是的。

Yeah.

没错。

Yeah.

正是如此。

Exactly.

安全才是汽车行业的重中之重。

Safety is And and and that's the one two punch of automotive.

是质量和安全。

It's quality and safety.

质量和安全。

Quality and safety.

尤其是质量，因为日本人彻底重塑了这一阶段，这背后有一整段不同的汽车发展史。

And quality really because because the Japanese really reset that that stage in in because that's a whole separate automotive history.

我们光谈汽车历史就能聊上一小时。

We could talk about automotive history for for an hour.

但关键在于，硅谷的公司正撞上一块坚不可摧的石头。

But the punchline is you have the Silicon Valley company meeting this immovable object.

现在就有一个平行世界在运转，哪怕是在通用汽车内部也是如此。

There is a parallel universe that cruises out there right now, even as a part of General Motors.

所以你必须结合公司所处的阶段、那一年工会谈判的具体情况来理解。

So I think you always have to take it into the context of where the company is, where union negotiations are happening literally that year.

如果你是工会，你会想：你们不能给我们赚十亿美元，却还在投资这件事。

And if you're the union, you're like, you can't make a billion dollars for us, but you're funding this thing

没错。

Right.

这可是在害人，而且太马虎了。

That's killing people, and it's sloppy.

我并不是断言一定就是这种情况，说得清楚点，这是一个多因素交织的问题。

So I'm not saying precisely that's what happened, to be very clear, but it's a multivariate problem.

我另一个观点是，我曾在两家公司都工作过，就是谷歌和通用汽车。

My other hot take is, I worked at both companies, right, Google and General Motors.

这两家公司其实非常相似，远比人们以为的要相似。

Those companies are way more similar than they are different.

比大家想象的还要相似得多。

Way, way more similar than they are.

其实，根本没必要让外人知道这些细节。

Literally, people don't need to know this.

谷歌的职级体系和通用汽车的职级体系完全一样。

The Google leveling system is the same as the General Motors leveling system.

我以前常这么说，你知道的，在谷歌会议里头，嘿。

And I I used to say I used say, you know, this inside of Google Meetings is like, hey.

其实我认识的一些通用汽车的工程师，比这儿的还要厉害，可别人听了一脸震惊，就好像我说教堂里没上帝一样。

Actually, some of the engineers I knew at General Motors are better than engineers here, and people would look at me like, I'm saying there's no god in church.

他们简直气坏了，说我这个来自底特律的金属弯管猴子，怎么敢这么说？

It's like they're like, how dare you, you metal bun bending monkey from Detroit?

我说，不是这样的。

I was like, no.

实际上，制造一台现代内燃机复杂得不得了。

Actually, like, making a modern combustion engine is extremely complex.

这可不是什么简单的事儿。

It's not just like, you know, it's it's it's not simple stuff.

所以我想说的重点是，这背后有好多因素在起作用。

And so the the the macro point, I think, is it's a bunch of things.

我觉得，安全永远是他们最优先考虑的事。

I think safety is always at the top of their top of their list.

我觉得我们确实招了不少来自Cruise的人。

I do think, you know, we've hired lots of cruise people.

我觉得他们处理政府那件事的方式有问题，遇到这种情况得讲究策略，可他们没把握好分寸。

I think the way they dealt with that specific issue with the government, you you gotta you gotta dance a particular way when that happens, and they just didn't dance exactly right.

这反而给了政府官员更多把柄。

And that just gives government bureaucrats more ammo Mhmm.

好让他们趁机发难。

To go after.

像通用汽车这样的大公司，你得小心行事，你们看过《好家伙》那部电影吗？

And you're a big target like General Motors, you gotta you know, it reminds you, you guys ever see that movie like Goodfellas?

记得《晨光之家》结尾那段吗？所有老派大佬都坐在法庭后头，悄无声息，事情就是这么发生的。

You know, there there that the the one of the last scenes of the House of the Rising Sun, you know, the all the old bosses go in the back of the courtroom, and they're like and, you know, that's what happened.

董事会当时就问，我们该怎么处理Cruise的事？

They're like the the the board was like, what do we do about Cruise?

我们到底能做点什么？

Like, what can we do?

我当时就想，凯尔这人挺不错。

I was like, Kyle's a good guy.

然后突然间，就想到《旭日东升》那部电影了。

But And then it like, Qasar The Rising Sun.

你懂的。

You know?

一群人冲进旧金山的办公室里。

People running through a San Francisco office.

开玩笑的。

Just kidding.

别把这拍成AI视频啊。

Don't make that an AI video.

只是想发条狠话给

Just gonna a mean text from

凯尔。

Kyle.

没错。

Exactly.

所以我觉得，应该存在一个能挺过来的宇宙，但确实不容易。

So, I think there is a universe that would've survived, but it's tough.

所以，Applied Intuition 做的很多事，就像你刚才说的，那种配合。

So, then a lot of what Applied Intuition does is kind of, as you said, like that dance.

就是怎么才能成为这些公司的优秀合作伙伴。

It's like how to be a great partner to these companies.

没错。

Exactly.

同时还得考虑到他们真实存在的困难和限制。

Bearing in mind their own real very real issues and constraints.

我觉得通用汽车当时也有。

I I think General Motors also had

商业模式这个话题。

the topic of business model.

对吧？

Right?

没错。

Right.

所以当时Cruise在推自动驾驶出租车，但通用汽车的主要利润来自私家车销售。

So you have Cruise was going after the robotaxi concept, but GM makes its profits from personal car ownership.

对。

Right.

这两者之间确实有点矛盾，我想这也是其中一部分原因。

And those things can be a bit odd, so I think that was also a bit of the Bit of the equation.

对。

Right.

好的。

Okay.

是的。

Yeah.

我觉得当时并不清楚。

And I think it wasn't clear.

顺便说一下，你其实是在2013年Y Combinator的活动上讲过话的，当时在场的每个人都有听。

I mean, by the way, you know, you, actually, all people, you spoke at y c at twenty in 2013.

我当时就在台下。

I was in the audience.

那时候我是合伙人。

I was a partner at the time.

你当时说了一句话，我觉得特别有意思，有种自我循环的感觉。

And you said something which I think is it's it's it's very, like, recursive here.

我们其实是在互相传递你自己的建议。

We're we're feeding each other your own advice.

这在新兴科技行业里是关键所在。

It's the key thing in in new techno in the new technology business.

实际上，大家多多少少都能摸索出技术的方向，尽管这依然很难。

Actually, everyone kind of figures out the technology, though that's still hard.

有时候，要做出这么复杂的东西真的很难。

It's so hard sometimes to build really complex things.

关键在于你什么时候、怎么把它推向市场。

It's when and how you deploy them into the market.

时机变得特别重要。

The when becomes really important.

你要是早了两年，那就完了。

You're two years early and you're doomed.

你要是晚了两年，竞争对手就太多了。

You're two years late, there's too many competitors.

你得正好踩在那个点上。

You have to hit it at the right spot.

我觉得这话说出来可能有点争议，但我真觉得Cruise的推进速度比Waymo快多了。

I think it's like I mean, controversial thing to say is like, I actually think Cruise, know, they were certainly moving at a much faster pace than Waymo.

他们起步晚了很多，可到最后被叫停的时候，俩人几乎并驾齐驱。

They started way behind, and you're talking about neck and neck when ultimately the plug was pulled.

所以，长远来看会发生什么，谁说得准呢？

So who knows what happens in the long term?

在同样的逻辑里，我们的假设其实是分销渠道。

Our hypothesis in that same equation is actually the distribution.

你让制造商来负责这部分。

You let the manufacturers do that.

我们现在在日本就运行着自动驾驶卡车。

Like, we run self driving trucks right now in Japan.

它们运送着商业货物。

They carry commercial loads.

车上虽然有安全员，但车是完全自主运行的。

They're they're you know, they're safety drivers there, but they're autonomously running.

但你根本看不出来，因为车身上标的是五十铃的品牌。

And but you won't know that because the brand is Isuzu.

对。

Mhmm.

那就是客户。

That's the customer.

我们选择和五十铃合作之所以这么好，是因为这家公司已经存在将近一百年了。

And and why it's so good for us to partner with Isuzu in that case is that company's been around for almost a hundred years.

对吧？

Right?

如果我没记错的话，他们是二战前就成立的老牌企业，而且非常熟悉政府流程。

If I'm if I'm not mistaken, a pre pre World War two company, and they are you know, they know the government.

他们有自己的测试场地。

They have test tracks.

他们深谙安全标准。

They know safety.

对自家的卡车了如指掌。

They know their own trucks very well.

所以我们提供智能系统并整合进他们的车辆时，这简直就是绝配。

So when we go and provide them with the intelligence and the integration into their physical machinery, that's a fantastic one two punch.

我觉得现在这个世界已经准备好接受人工智能在现实中的应用了，这很大程度上要归功于ChatGPT、Anthropic，还有这一系列的突破。

I think today, the world is ready to consume AI in the real world, and that's a lot because of ChatGPT and Anthropic and all these you know, everything that's happened.

现在没人再问，自动驾驶汽车到底是什么了。

So people are no longer like, what's a self driving car?

这都要归功于Waymo和特斯拉。

And there's because of Waymo and Tesla.

市场已经成熟，随时准备接纳了。

So the market is ready to consume.

我认为关键是要以最恰当的方式去迎合市场。

And I think you just have to meet the market in the way that the best way possible.

在我们看来，一直以来的做法都是去接触当前实际运营经济的那些人——不管是矿场运营商、国防部，还是制造商。

In our view, that has always been, you go through some of the people who run the economy right now, whether it's a mining operator, whether it's a department of war, whether it's manufacturers.

我们会在每个行业里，找到合适的合作伙伴一起合作。

And we work with, you know, within each vertical with the right partner.

但这跟特斯拉或Waymo那种垂直整合的模式，本质上是完全不同的。

But that's a fundamentally different view than a Tesla or a Waymo, which are gonna be vertical.

对。

Mhmm.

我们主打的是横向布局。

We're we're really playing the horizontal.

我们看待自己公司的方式，就像一家芯片制造商。

And I think the way we can always think we think about our companies, and we're kind of like a like chipmaker.

是的。

Mhmm.

我们说话做事，真的跟一家半导体公司一模一样，只不过我们不生产芯片罢了。但我们有设计定点，有长期稳定的合作关系，一旦进去，就很难被取代。

You know, we we we actually look and talk and walk a lot like a silicon company, except we obviously do, we don't make chips, but, you know, we're our we have design wins, and then we have really large long term relationships, and then once we're in, we're in, it's really hard to, you know, to to take us out.

所以你需要深厚的信赖。

So you need deep trust.

我们的合作伙伴对我们有着极深的信任，我们也非常了解他们的行业。

Our our partners have really a lot a lot of deep trust, and we know their markets really, really well.

詹森最清楚的就是他的客户。

The the things that Jensen knows is he knows his customers.

这就是为什么英伟达的成功不仅仅是因为他们掌握了非常复杂的技术。

That's why NVIDIA does well beyond the fact, obviously, they make a very complex technology.

那么这些传统的汽车公司正在如何为未来做准备？

So how are these legacy car companies preparing for the future?

他们是在更多地收购其他公司吗？

Are they making more acquisitions?

他们会怎么做呢？

They going to?

他们是自己研发，还是和你们合作？

Are they building and partnering with you?

他们打算怎么跟那些天生的科技公司竞争？

How are they going to compete with tech native companies?

这就像是在问，政府是怎么应对人工智能的。

It's like saying, like, how are governments dealing with AI?

这话题太广了，每个汽车厂商，比如本田、日产、丰田，这三家日本老牌车企，各自的策略都不一样。

It's it's such a broad topic, and each manufacturer, like, even, like, you take Honda, Nissan, Toyota, three Japanese manufacturers with, you know, long legacies, they all approach it very differently.

他们的策略大致横跨两个极端：要么自己研发，要么直接收购，中间还有各种混合方式。

They're roughly in a spectrum of we're gonna build to we're gonna buy, and other both extremes.

现在更多公司选择收购，因为它们试过自己搞，而我们一直都在这儿等着。

More than ever, we're gonna buy is the common answer, because they've been trying, and we've been there the whole time.

对于打算自研的客户，我们提供工具，也会简单介绍一下我们这次发布的新产品。

For the folks that are gonna build, we provide them tools, And we talk a little bit about our new product that we're announcing here.

而对于只想购买的客户，我们就直接卖给他们嵌入在设备里的智能系统。

And then on the ones that just want to buy, we sell them the actual intelligence that goes on the machines.

所以我们根据客户所处的阶段，提供相应的解决方案。

And so we meet the customer wherever they're ready in their journey.

更细致地说，实际情况是，每个产品都不一样。

The more nuanced version of that is, you know, the reality is, like, every every product is a different product.

你能投入多少芯片、多少钱在传感器上，客户愿意付多少，全看最终产品是什么——但从长远来看，所有这些设备最终都会实现完全自动驾驶。

And so the amount of silicon and amount of dollars you can put towards it, towards sensors, what the customer's willing to pay, all that depends on what actually gets in In the long horizon, all these things will be fully autonomous.

但中间的过渡阶段，就像当年的个人电脑一样，是逐步升级的，总有一天，没人会再关注笔记本的配置，甚至可能连手机的参数都没人关心了。

But the intermittent steps are very much what we saw in the PC, where you you have this slow step up to, one day, that'll be like, now nobody really looks at laptop specs and even maybe, frankly, your phone specs.

但从80年代中期到2002年、2005年左右，情况就不同了，那时候人们终于不再纠结于具体参数，转而全面转向笔记本电脑。

But that's not the case from basically '85 to 2002, 2005, where finally people stop actually specking at all, and and and then they're really moving to laptops.

但我觉得，这里也存在一个类似的二十年时间线。

But there's a similar kind of twenty year, I think, horizon there.

总的来说，当我们谈论机器变得智能时，本质上，机器就是由各种不同组件集成而成的。

Broadly, when we talk about machines and machines becoming intelligent, right, fundamentally, a machine is is a collection of these different components that are integrated.

对吧？

Right?

通常，最后做集成的那家公司会把自己的品牌贴上去，但其实有很多公司都在开发这些机器内部用到的技术。

And whoever does that final integration is oftentimes the company that puts their badge on it, the brand name, but many, many companies are building technology that goes into that machine.

所以我们现在有了很多可以嵌入机器的技术组件和平台，同时我们也销售那些核心的技术，供客户用来

And so we now have a bunch of technology components and platforms that can go into these machines, but we also sell the core technology that can be used

开发自己的产品。

to develop them as well.

如果你掀开一台推土机、联合收割机或者柴油卡车的引擎盖，你会发现里面用的是康明斯的发动机。

And if you look, by the way, under the hood of a of a dirt mover or, like, combine or diesel truck, they'll have Cummins engines in them.

但没人会说，因为这些人都买康明斯的发动机，所以卡特彼勒就不是家好公司。

But nobody says, well, because all these guys buy Cummins, this this means that they're, you know, whatever, Caterpillar is not a good company.

根本不是这么回事。

It's like, no.

这不过是他们采购的一个零部件而已。

That's just a component that they buy.

它们扮演的角色是不一样的。

They they had they have a different role.

所以当你去看任何一个垂直领域，都会发现背后是一张错综复杂的合作网络。

So if you when you look in the any of these verticals, there's it's just a complex web of folks.

这就是为什么我一直说，芯片这个类比其实特别贴切——这些公司自己都不造芯片，但全都用芯片。

That's why I always say, like, the the chip kind of analogy actually works quite effectively because some none of those companies make chips, but they all buy chips.

所以我觉得，这样理解确实很到位。

And so I think that that's the the it's a good way to think about it.

自动驾驶汽车，我们一直都在聊这个，记得最早大概是2005年左右，从DARPA大挑战赛开始的，之后谷歌很快就加入了这个项目。

So self driving cars, so, you know, we've all been talking about self driving cars for, like, I think the whole thing started, like, around, what, 2005 or something with the DARPA Grand Challenge originally, and so and then Google engaged on the program shortly after that.

是的

Yeah.

二十世纪末

Late double o's.

是的

Yeah.

二十世纪末

Late double o's.

差不多快二十年了，可能还差一点。

So almost 20 basically, around 20 little less than twenty years maybe.

过去二十年里，人们一直预测自动驾驶汽车马上就要实现了。

And there have been lots of predictions over the last twenty years of self driving cars are imminent at any moment.

所以坏消息是，今天大多数汽车仍然不是自动驾驶的。

So I guess the bad news is we're sitting here today, most cars are not self driving.

好消息是，现在确实已经有自动驾驶汽车了。

The good news is there are now self driving cars.

是的

Yeah.

现在越来越多的自动驾驶汽车在各地运行，像旧金山的人们已经觉得坐上这些车再平常不过了

And so the way more cars are driving all over, you know, in the places they're deployed, it's become you know, I think people in San Francisco I think, treat it now as routine that they get into

有些车，我觉得你可以把特斯拉和通用人工智能联系起来

some I think you can call I think Tesla Tesla's like the AGI thing.

没错

It's like Right.

你知道的

You know?

如果回看二十年前，我们现在看到的一切简直像是通用人工智能突然实现了

If we're talking twenty years ago, everything we're seeing right now is like a mind blowingly AGI.

没错

Right.

主持人一直在移动。

The host keeps moving.

特斯拉的东西真的很厉害。

The Tesla stuff's amazing.

你可以看看好多厂商，比如蓝巡航、超级巡航、宝马、沃尔沃的驾驶辅助系统，都挺出色的。

You can look at a bunch of manufacturers, Blue Cruise, Super Cruise, BMW, Volvo's Pilot, they're all quite impressive systems.

但它们还不是完全自动驾驶。

They're not full self driving.

对。

Right.

不过呢，

But, yeah.

这其实已经是完全自动驾驶了，加上远程监控那套系统。

Well, it's it's full self driving x whatever remote monitoring is happening.

是的。

Yeah.

我们家在洛杉矶有个房子，你们可能还记得，洛杉矶之前发生过一场大火。

The Tesla we have we have a home in in in Los Angeles, and you you guys may recall there was a there was a large fire in Los Angeles.

是的

Yeah.

对

Yeah.

而且在那之前，加州的电网就已经撑不住了。

And then the power then then the and then the California power grid was buckling even before that.

所以特斯拉Cybertruck的一个优点就是它其实是个很不错的移动电池，确实如此。

So it actually turns out among the things CyberTrucks are good at is they're they're very good batteries Yeah.

可以用来给家里供电。

For powering your house.

没错

Yeah.

所以我们干脆把Cybertruck当作家里的备用电源。

And so literally, have Teprotrux as our backup battery for the house.

去年FSD更新的时候，我记不清具体是哪个版本了，但那次更新后，很多人觉得它真的有了质的飞跃。

And as of last year, whenever the FSD release, I forget the exact one, but there was one where it, like, at least a lot of people thought it, like, really turned the corner.

14。

14.

是的。

Yeah.

我昨天跟一个人聊了，他开的是Model Y，让车子全程自动开完1号公路，从大苏尔那段路一路开下来。

And, like, that thing drives you I talked to somebody yesterday talked to somebody yesterday who has a model y who let the let the thing let the thing do the full route all the way up Highway 1 through Big Sur.

是的。

Yeah.

我觉得现在系统脱手的平均里程特别高，好像已经到几千英里了。

Think mean disengage like, the meantime and, like, miles per disengagement are are really high, I think miles is, like, in the thousands.

这真的非常厉害。

Which is very impressive.

是啊，我知道。

Yeah, I know.

没开过1号公路大苏尔那段的人，可能不知道那条路有多紧张。

For people haven't driven the Highway 1 Big Sur, that's a stressful drive.

他说感觉特别棒。

He said it was great.

要不是这事太离谱了，我根本不会跟他聊。

Anyway, I wouldn't have been talking to him had it not been crazy.

直接就冲下悬崖了。

Would have gone right cliff.

因为他把方向盘给拆了。

Because he unbolted the steering wheel.

所以一松手就

So Right off the

冲下悬崖了。

cliff.

然后你知道吗，特斯拉的无人出租车已经开始在街上出现了。

So and then, you know, Tesla's rolling out their robotaxi, you know, that is is starting to starting to show up in the in the wild.

一方面，这些东西确实已经出现了。

And so so on the one hand, those exist.

但另一方面，你知道，99.999999%的车都还不是自动驾驶的，要说另外一种，可能就只有自动驾驶卡车了。

On the other hand, you know, 99.999999 percent of cars are are not self driving, and then I would say maybe just one other would be the self driving trucks.

媒体老是反复炒作卡车自动驾驶会让人失业，说所有卡车司机都要下岗了。

There's been this recurring kind of panic in the press of the trucks become self driving, and the employment, all these truck drivers gonna be out of a job.

可今天坐在这儿想想，路上真有那种没配安全员的自动驾驶卡车吗？

And sitting here today, I don't know, are there any trucks on the road that are self driving that don't have at least a safety driver in the truck?

我觉得答案大概还是没有。

And I think the answer is probably still Yeah.

数量还是极少。

Still very few.

所以咱们

So let's

咱们来分清楚吧，刚才提到的其实有好几个不同问题。

let's split let's split the there's multiple points that we brought up here.

第一个是关于私人拥有的车辆。

One is on the, let's say, personally owned vehicles.

对。

Right.

那为什么它们还没普及呢？

And why are they not more ubiquitous?

一部分原因是制造商在推广技术方面做得不够好。

The part of that is the manufacturers are not good at deploying technology.

一部分原因是他们想确保安全，但最主要的原因还是成本、成本、成本。

Part of that is they're they're they wanna be safety conscious, but most of it is cost cost cost.

你在中国看到的情况是，中国的电动车生态很不一样，主要是因为他们根本不看重利润。

The the what you're seeing in China, which is a China's kind of a different EV ecosystem mainly because they don't care about profits.

当一个行业根本不考虑盈利时，整个行业的逻辑就全变了。

When you're talking about business that doesn't care about profits, it changes the entire calculus of the entire industry that that doesn't care.

但你现在看到的是 L2+ 系统。

But what you're seeing is you're seeing l two plus plus systems.

所以我们可以把整个自动驾驶的讨论简化成一个问题：方向盘后面有没有人？

So we can simplify the entire self driving conversation to, is there a driver behind the steering wheel?

对。

Right.

所以方向盘后面还是有司机，但像特斯拉这种车基本上能开到任何地方。

So this is a driver behind the steering wheel still there, but generally, like Tesla drives everywhere.

成本还不到一千美元。

They're like sub $1,000.

对。

Right.

这里面涉及芯片、传感器、整体方案和软件，全都非常激进。

There's an aggressive that's chip, sensors, the package, the software, everything.

我们预计，一旦成本降到五百左右，汽车厂商就会免费提供这套系统。

We, you know, anticipate that there there's a very aggressive once you get to, like, 500, the automotive OEMs will actually subsidize it for free.

他们会直接送给你。

They'll just give it to you.

导航系统当年就是这么发展的。

This happened in nav systems.

你们还记得吧，以前导航系统可是个大卖点。

If you guys remember, nav systems used to be a big thing.

那时候得花三四千块买导航，结果突然之间就免费了，还成了标配。

You pay $4, 3,500 to get a nav system, and then suddenly it became free, and it just became default.

我觉得这事也会在自动驾驶上重演。

I think that'll happen.

有个挺有意思的现象：要给部分车型加装这套系统要花一笔钱，但要是全系标配，额外增加的成本其实很小。

The there's a weird thing, which is like, actually getting into a subset of your cars costs x dollars, and to get into all the cars costs x plus just a small incremental amount.

对。

Mhmm.

这些成本都是固定的，不管你要生产多少辆车，生产线怎么安排，认证测试、各种合规流程，都得一次性搞定。

These are just a fixed cost and the way that how many vehicles and the the way the assembly line comes and the way you have homologation, all these testing regimes, all this stuff.

所以我觉得会是这样：先等一阵子，然后一下子全铺开了。

So I think you'll have wait, wait, wait, and then a lot.

没错。

Yep.

每个汽车制造商，无一例外，即便是最便宜的品牌，都在研发全自动驾驶的竞品。

Every single OEM, without exception, even the lowest dollar OEMs are working on an FSD competitor.

是的。

Yeah.

所以这事儿迟早会成。

So it it it'll it'll come.

不过，要理解私人拥有的自动驾驶汽车，可以拿手机的发展做个类比。

But it'll just like, you know, with a good analogy to think about self driving in the personally owned ecosystem is mobile phones.

对。

Right.

以前有卫星电话，然后是高通那种大砖头手机，再后来是摩托罗拉的翻盖机，从九十年代末到两千年代末，大家一直在问：手机什么时候才能普及？

We had we had the the satellite phones, then we had the Qualcomm, you know, brick phones, then we had the Motorola razors, and from, you know, the late nineties to the late two, you know, double zeros, the review was like, when's mobile gonna come?

突然之间就爆发了，从iPhone发布那年算起，短短四年，Uber、Instagram、WhatsApp、Snapchat全冒出来了。

There was a huge, and then it it comes, and we we go by by o seven from the iPhone launch, like, it's like four years when you get Uber, Instagram, WhatsApp, Snapchat.

对。

Right.

这些才是真正的杀手级应用。

Those are the killer applications.

所以我觉得，自动驾驶也会经历类似的阶段——大家等着等着，突然间就普及到每辆车上了。

So I think there's a very, very similar kind of wait, wait, wait, and then it's just basically ubiquitous in every vehicle.

没错。

Right.

如果你问我具体什么时候，我觉得28年量产，29年启动生产，到29、30年，到了30年代初，成本就会降到几乎免费。

If you had to ask me for what that number is, 28 SOP, 29 start up production, '29, '30, and then by the early thirties, it'll start becoming very cheap to free.

到了30年代初，你买新车时，根本就默认它是自动驾驶的了。

Pretty much routinely by by the early thirties, you would just you buy a car, and you just assume it's self driving.

完全正确。

Exactly.

是的。

Yeah.

或者更准确地说，车上会配备L2+级别的驾驶辅助系统。

Or it has the driver in seat l two plus plus system being very specific.

如今的特斯拉Cybertruck车主已经拥有这些功能了，

Cyber truck Tesla owners have today,

具体是什么？

what Exactly.

将来会成为默认配置。

Will be in be be default.

那么问题来了，为什么我们到处看不到Waymo的车呢？

So then the question then the the the, you know, the the other side of this is, why don't we have a bunch of Waymo's everywhere?

具体来说，Waymo用的是不同的技术。

Specifically, Waymo has a different technology.

这里就不细说了，但特斯拉和很多中国公司采用的都是这种端到端的架构。

And without getting into the nuances here, but Tesla and many of the Chinese and Applied were very much in this end to end model architecture.

这是一种全新的自动驾驶实现方式。

This is a new way of doing self driving.

Waymo，说得直白点，不是这种模式。

Waymo, for the lack of a better word, is not that.

这并不意味着他们没学过。

It's not doesn't mean they're not learned.

只是他们不是用一个端到端的系统而已。

It doesn't it's just just not one end to end system.

他们用的不是一个整体的模型。

It's not one monolithic model.

他们的做法有一个特点，就是得依赖高精地图，本质上是一种地理围栏的思路。

One of the proclivities of their their approach is it does depend on HD maps, therefore, is a geofencing concept.

我觉得Waymo正在努力消除这个瓶颈，好让业务能更快地扩展到更多地方，但现实还没到那一步。

I think Waymo's trying hard to remove that bottleneck so they can expand geographically faster, but the reality of today isn't there.

另外，Waymo原本是从Alphabet的研究机构出来的，研究人员做事时根本没考虑商业限制。

The other thing is when you have researchers, which they which Waymo really was coming out of an alphabet research organization, they didn't put commercial constraints.

所以他们的传感器都是定制的，价格很贵。

So the sensors are bespoke and expensive.

还有车和里面的计算设备，根本谈不上经济可行。

The cars and the then the compute that are in there, they're they're just not economically feasible.

他们试了很多方法想降低成本，但说实话，从一个便宜的东西往上加功能，比从一个过度设计的东西下手删减、做到极致便宜，要容易多了。

And they've tried a lot to get that down, but it's kinda like it's a lot easier to go from something that's really cheap and make it more, you know, more feature full than something that's overbuilt and then trying to trim and make it really, really cheap.

这正是争论的核心。

And that's the big debate.

到底谁会先达成目标？

Who's gonna get there first?

是特斯拉的全自动驾驶，还是Waymo的成本优势和广泛覆盖？

Tesla with full self driving or Waymo with cost and geographic ubiquity?

但你知道我们没在争论什么吗？

But you know what we're not debating about?

自动驾驶会不会实现？

Is it gonna happen?

对吧。

Right.

你知道我们没在争论什么吗？

You know what we're not debating about?

我的意思是，是不是还得有什么重大的技术突破才行？

Like, is is is there's a big technical breakthrough that needs to happen.

这些都不是问题。

None of those things.

现在很明显，自动驾驶已经进入工程阶段了，就是得一点一点地把每英里的成本压下来。

So now we're clearly in the engineering side of self driving, which is just this grind down to, like, dollar per mile efficiency.

一旦成本降下来了，你觉得会怎样？

And the moment that it's cheap, guess what?

所有汽车厂商都很聪明。

All the OEMs are smart.

他们一看到合适就会立刻采用。

They'll just they just adopt it.

厂商们并不是因为觉得消费者不想要，或者不懂这项技术才犹豫。

It's not that OEMs are resistant because they don't think consumers want it, or they don't understand the technology.

他们只是想要一个能让他们维持低利润率、大规模生产的价格区间。

It's because they want a price envelope, which allows them to keep their thin razors, thin margins, and at a scale Right.

目前已经在一百多个国家部署了V1版本。

Which is deployed across a 100 plus countries in v one.

没错。

Right.

如果你只是做小规模部署，那情况就完全不同了。

And so if you're just doing a small deployment, it's very different.

我觉得最后一点要说的是，买斯巴鲁或铃木的车主，和买特斯拉的车主，对品牌的期待截然不同。

And I think and and that was the last thing I would say is the buyer of a Subaru or a buyer of a Suzuki have very different brand expectations than a buyer of a Tesla.

对的。

Mhmm.

所以还得考虑消费者的年龄，以及他们对未来的预期或对现状的理解。

And so including the age of the consumer and what they think will happen or or or what happen.

这也是原因之一。

So that that also the reason.

所以如果你是铃木，你会觉得：我的客户根本不需要这些东西，那我干嘛非得硬塞进车里呢？

So if you're Suzuki, you're like, my buyer's like not don't want this stuff, so I'm not gonna jam it into the car.

并不是因为他们技术上不行。

It's not because they're not like technically competent.

这是另一个领域的问题。

This is a different area.

你觉得什么时候会成为常态？

When do think when do you think it'll be routine?

比如说，在美国最大的200个城市里，你出门时理所当然地觉得自动驾驶出租车随时会来接你，这种场景什么时候才会普遍？

Let's say the 200 biggest American cities, like, it wouldn't be routine to walk outside, and you're you're just taking for granted that a robotaxi can come pick you up.

现在是26年了。

It's twenty six now.

我敢肯定，到30年肯定就普及了。

I mean, certainly by thirty.

明白了。

Alright.

好的。

Okay.

是的

Yeah.

肯定到2030年了。

Certainly by thirty.

我觉得关键变量其实是，Waymo自己会说，每个城市的成本收益已经能跑通了。

And I I would say the the big, like, variable there really is, like because you what Waymo will say is that the dollars and cents per city already work.

对。

Mhmm.

可既然有家公司的资金几乎无限，为什么还没在200个城市铺开呢？

And it's like, well, a company that has basically unlimited capital, why are they not already in 200 cities?

但你再看他们的扩张计划，节奏其实挺猛的，感觉真能冲到那个目标。

But then you see their launch schedule is pretty aggressive, and you're like, that that can that that can get there.

所以如果他们敢这么拼，我觉得2028年就能实现。

So maybe because being aggressive, I would say twenty eight.

是的

Yeah.

好的。

Okay.

大概两年吧。

Like, two years.

我觉得30年能投入使用，但普及可能要到32年、33年左右。

I I would say I would say available in 30, but routine in maybe, like, '32, 30 '3.

没错，因为需要扩大规模，提升产量。

Sure, because there's scale up, there's a volume.

而且如果你住在洛杉矶，五年前我去洛杉矶，人们还会问：‘Applied Intuition是什么？’

And also, if you live in LA, so, five years ago, I'd go to LA, people would be like, What's Applied Intuition?

他们根本不知道自动驾驶是什么。

I don't know what self driving cars are.

但最近几年，现在人人都知道自动驾驶了，有些人甚至听说过Applied Intuition，因为他们从其他车企那里听说过。

In the last couple years, now they all know self driving, and some of them even know Applied Intuition because they know from the other manufacturers.

我觉得再过两到四年，每个人都会知道这件事。

I think you fast forward another two to four years, everybody knows it.

现在是这样。

Now Mhmm.

那是不是所有人都在用Waymo的服务？

Does that mean everyone's taking Waymo's exclusively?

没错。

Right.

实际上，答案是否定的。

That answer is no, actually.

如果你看数据，情况就非常非常明显了，作为Uber，你肯定得担心。

Now there is a huge, huge if you look at the numbers, if you're Uber, you gotta be scared.

他们正在不断蚕食网约车市场。

I mean, they're just eating into into ride sharing.

没错。

Right.

是的。

Yeah.

但要实现完全普及，那得便宜到极致才行。

But to get a 100% ubiquity, I mean, that's that's that's another it has to be extremely cheap.

没错。

Good.

那长途货运呢？

And what about long haul trucking?

长途货运？那是另一回事，我们刚才说的是客运。

So long haul so that's what we so that's the passenger side.

长途货运的经济模式和商业模式完全不一样。

The long haul trucking, completely different economics, completely different business model.

现在有不少公司，我觉得至少有五家以上，已经在跑中美之间的长途货运卡车，车上都有司机运货。

There are many companies right now, I would say probably north of five, that are running long haul trucks with drivers carrying loads between America and China.

你要是看中国那边，可能已经进入两位数了。

If you see that China is probably getting into double digits.

对。

Mhmm.

所以它确实存在。

So it's there.

但你之所以不知道，也没把它放在心上，是因为它不是面向消费者的产品。

But the reason you don't know it and the reason it's not top of mind is it's not a consumer product.

没错。

Right.

而且跟Waymo和特斯拉那边不一样，投资者愿意给它们一些市值上的溢价，算是为它们声称的卡车运输业务潜力买单。

And unlike on the Waymo and Tesla side, where investors are willing to essentially give you, you know, some market cap, you know, adjustment for the potential of they say the trucking business is like, you know, made

买车是凭感情，你

You buy a car with your heartstrings, you

买

buy

卡车得靠算账。

a truck with a calculator.

对，这就是笔算账的生意。

Yeah, it's a calculator business.

所以这完全是算账的问题。

And so it's like pure dollars and cents.

因此，作为自动驾驶卡车的提供方，如果你是全栈做的话——有些公司就是这样，但我们不是——你就得每跑一英里都清楚地告诉客户，我能帮你省多少钱。

And so I think you, as the provider of self driving trucks, if you're doing the whole thing, like some of the companies are, which we're not, you have to show every mile, I'm gonna save you this many dollars.

这就像是——

And it's like,

当然，当然，当然——

for sure, for sure, for

对，因为买家并不专业，他们就想：我本来就有一支司机团队。

sure, because the buyer's unsophisticated, And they're just like, well, I I already got a staff.

他们能开车。

They can drive.

而且他们根本不想换。

And it's like and they're like, they're just not inclined.

我们现在选择在日本做卡车业务，这绝不是偶然。

Now where we're playing in Japan, it's not random that we're doing trucking in Japan.

现在劳动力严重短缺，人口结构也在急剧恶化，你知道的，这种情况。

There's a massive labor shortage today, there's an imploding demographic, you know, situation.

所以几乎每个行业都有需求，这也是我们选择这个市场全力发展的原因。

And so there's a demand from almost every sector, and that's why we've we've picked that market to to to really grow.

但我觉得，你甚至可以看看更冷门的领域，比如采石场，你知道的，就是运水泥、运土的地方。

But I think, like, you can take, like, even more obscure, like, what will all queries, you know, literally, like, where you you know, you're you're moving cement, you're moving you're moving dirt.

不是‘queries’，是 Q-U-A-R-R-Y，采石场。

Not queries, q u e a r s Q U A R R Y.

对，采石场，就是采石场。

Yeah, Quarries, Quarries.

那些地方都是石头、岩石。

Those are rock, stone.

岩石和石头，那些地方什么时候需要？

Rock, stone, When are those?

我可以告诉你，那些拥有并运营采石场的人，现在就想要这个。

I can tell you, the people who own those things, and run those things, want it today.

所以说，根本就是根本赶不上需求，我们根本没法生产得这么快。

So it's literally, then you don't have a point, we can't make this stuff fast enough.

没错。

Right.

不过有个宏观层面的事，大家都不提，但我觉得这些事迟早都会发生。

The macro point, though, that people don't talk about, I think all this stuff's gonna happen.

很快就会发生。

It'll happen pretty soon.

用不了多久就会发生。

It'll happen fairly soon.

但宏观上，从政策和经济、政治经济的角度来看，AI现在面临的巨大阻力，是因为我们这些行业的人担心自己的工作会怎样，而风投那边，我敢说你们所有的同事都特别紧张。

But the macro point that in legislation, and kind of the kind of economics, the political economy of this conversation is AI is really you see, you have this big pushback in digital AI, because we're the accounts that are like, don't know what's gonna happen to my job, and VCs, I'm sure all of your associates are very scared.

但在我们这个领域里

But, like, in our universe

他们在赌，到底需不需要我们。

They're betting whether they need us.

是的

Yeah.

对

Yeah.

没错

Yeah.

在我们的领域里，情况恰恰相反。

In our in our universe, it's the other way around.

你瞧，我会遇到一些运营人员，他们直接说：只要你能做到，我们什么都给你。

It's like you they're literally, I'll I'll I'll meet these, you know, operators, and they're like, we'll give you everything.

只要你能做成这事，我们全力支持。

Like, if you can do this, we'll give you everything.

所以关键就看我们能不能够积极主动、全力以赴地推进了。

So then it's just up to us to, like, get there as as, you know, aggressively and physically

长期以来，卡车运输行业不知为何总能引发媒体的恐慌，总觉得会带来毁灭性的失业潮。

Well, there well, there you know, the the share for a long time has been for trucking, for some reason, triggers the at least the press's imagination on, like, you know, sort of apocalyptic levels of job loss.

比如，会不会有

Like, do do will there

这完全不对。

Well, that's it's so wrong.

你接着说。

Go ahead.

卡车司机根本不够用，你猜怎么着？

There's not there's not enough truck drivers, and guess what?

没人愿意当卡车司机。

Nobody wants to freaking be a truck driver.

为什么呢？

Why is that?

说说原因。

Explain that.

因为这工作太差了。

Because it's a terrible job.

就是说，因为。

It's like Because.

你啊，你就是觉得。

You're you're like you're like

顺便说一下，我小时候住的那个镇子，最主要的地标就是个卡车休息站，所以我也明白。

By the way, grew up the main feature of the town I grew up was a truck stop, so I Yeah.

这就是原因。

It's the answer.

为什么开卡车这份工作越来越没人干了？

Why is truck driving down

开卡车这事，你问我，你知道怎么回事吗？

the truck driving down You're asking me you know what?

这就像在跟我的孩子讲道理，他问：‘为什么我不能把手放在炉子上？’

This is like talking to my kid, who's like, Well, why can't I put my hand on the stove?

我就说：‘因为会烫伤手。’

It's like, Because it's gonna burn your hand.

就是这样，可为什么呢？

It's like, But why?

这是第三个为什么了，来吧，老兄，咱们干吧。

It's like, the third why, it's like, Come on, buddy, let's do this.

这可是个艰难的办法。

That's it on a hard way.

对，咱们开始吧。

Yeah, let's do

那什么最难？

So, what's hard

我开玩笑的，大家都知道，我没做过那事。

I'm kidding, just to make sure everybody knows, I did not do that

在这边。

on this

这边。

side.

对，对。

Yes, yes.

那到底难在哪里呢？为什么开卡车是个辛苦活，为什么年轻人长大后都不想干这行？

So what's hard why is being a truck driver a difficult job, or why would kids not want to do it when they grow up?

我打个比方，这个例子特别清楚，你一听就懂。

So, let me use a parallel analogy, which is very clear, and then you can Sure.

人们总说，现在没人愿意干活了。

You know, people will say, like, nobody wants to work anymore.

接着又说，你看麦当劳招不到人，空位一大堆。

Then they say, well, you know, McDonald's has all these job openings.

不对。

No.

其实呢，以前在麦当劳上班的人，现在都去跑外卖、开网约车了。

Actually, what it is is those people that used to work at McDonald's now DoorDash and Uber.

没错。

Right.

因为对他们来说更划算。

Because it's better for them.

更自由。

Better.

因为他们可以自己安排上下班时间，没人管着，也不用一直站着，接单间隙还能刷刷手机，这才是关键。

Because they can open, they can start their hours, end their hours, and they don't have to there's no boss, and they don't have to stand on their feet, and they can surf their phone in between, you know, orders, and they don't like, that's the reason.

这可不是随便发生的。

It's not random.

市场自有它的逻辑。

The market is efficient.

对啊。

Right.

那开长途卡车的呢？为什么没人愿意连续四到八天都离家在外呢？

And so in the truck driving example, why does somebody not wanna be away from their family for four to eight days in a row Yep.

跑长途货运太辛苦了。

Doing long haul trucking?