亚马逊CEO安迪·贾西谈为何亚马逊投资2200亿美元于人工智能

即使今年投入了2200亿美元，我们仍然留下了大量的需求未能满足。

Even spending $220 billion this year, we're leaving a meaningful amount of demand on the table.

我们总是会预留一部分用于按需服务，但2026年的资源显然已经售罄。

We always leave a certain amount for on-demand, but '26 is obviously sold out.

2027年的资源基本上也售罄了。

'27 is largely sold out.

2028年的资源也有很大一部分被预订完了。

A good chunk of '28 is sold out.

金融市场在说："世界上没有足够的资金来资助这些项目。"

The financial markets are saying, "Oh, there's not enough money in the world to finance these things."

当今世界上我们做的每一件事，股市、银行业、医疗保健，都在使用数据中心。

Everything we do in the world today, stock market, banking, healthcare, they're all using data centers.

我们的任务就是想办法尽可能多地提升产能。

Our job is to try to figure out how to have as much capacity as we can.

我们有机会在更多的客户体验方面提供帮助，让他们的日常生活变得更美好、更轻松。

We have an opportunity to help across so many more customer experiences and make their lives better and easier every day.

那些大型实验室愿意为电力、冷却和芯片几乎付任何价格。

The big labs being willing to pay almost anything for power, cooling, chips.

嗯，也许我们应该收购那些老家伙，然后卖给他们那些新玩家，因为他们会不管多少都付给我们。

Well, maybe we should buy out the old guys and sell them to the new guys because they'll pay us whatever.

亚马逊预计今年将花费大约2200亿美元的资本支出。

Amazon expects to spend roughly $220 billion in CapEx this year.

而根据CEO安迪·贾西的说法，即使这样也不足以满足人工智能的需求。

And according to CEO Andy Jassy, even that isn't enough to meet AI demand.

在本期节目中，我和本·霍洛维茨与安迪坐下来，深入探讨了这一建设计划，他为什么相信AWS最终可能达到每年1万亿美元的业务规模，以及为什么推理成本仍然需要大幅下降。

In this episode, Ben Horowitz and I sit down with Andy to unpack that build-out, why he believes AWS could eventually become a $1 trillion annual run-rate business, and why inference costs still need to fall significantly.

我们还讨论了人工智能如何重塑亚马逊本身，从更小的团队更快地构建产品，到从头重新思考产品，以及安迪在网络安全、工作、机器人和太空方面看到了什么。

We also discuss how AI is reshaping Amazon itself, from smaller teams building faster to rethinking products from the ground up, and what Andy sees ahead for cybersecurity, work, robotics, and space.

安迪，欢迎来到播客。

Andy, welcome to the podcast.

很高兴你能来。

It's great to have you.

谢谢邀请我。

Thanks for having me.

你领导亚马逊已经五年了，作为CEO做了很多事。

So you're five years into running Amazon, and you've done a lot as CEO.

现在，资本支出2200亿美元，AWS增长了28%。

Right now, $220 billion in CapEx spend, AWS 28%.

你取得了不少成就。

There's a lot that you've achieved.

相比于2021年你正式接任CEO时，你对2026年亚马逊的雄心怎么看？

How do you think about the ambition in 2026 as it relates to Amazon relative to 2021, when you took over formally as CEO?

2021年感觉太遥远了。

2021 feels so long ago.

感觉像是——

It feels like-

八辈子了。

Eight lifetimes

……十年前的事了。

... a decade ago.

但你也知道，世界变化很快。

But as you know, the world changes fast.

跟2021年相比，我们整体营收大概增长了80%，AWS业务几乎翻了三倍。

And if you compare it to 2021, our overall top line's about 80% higher, AWS business almost three times bigger.

我们的广告收入大约增长到原来的2.5倍。

Our ads revenue is about two and a half times bigger.

所以业务确实增长了不少，我觉得实际上发生了好几件事。

So the business has just grown quite a bit, and I think there are a few things that have really happened.

我上任之后不久，也就是疫情快结束的时候，在零售业务上，我们遇到了一个真实的配送成本挑战，基本上我们花了25年建成的履约网络，规模翻了一倍。

I think pretty quickly after I started in the role, which was on the tail end of the pandemic, in our retail business, we ended up with a real cost to serve challenge where we had basically doubled the size of our fulfillment network that we'd built over 25 years.

我们在大约18个月内把网络规模翻了一倍，然后又在大约20个月内建成了跟UPS规模相当的最后一英里运输网络。

We doubled the size of it in about 18 months, and then we built a last-mile transportation network the size of UPS in about 20 months.

建得这么快，肯定会有各种地方不够优化，我们的配送成本和履约网络都不是理想状态，这带来了很多挑战和不安。我记得那时候，也就是我上任不到一年的时候，很多人说：“哎呀，这真是个糟糕的时机，运气真不好。”

And whenever you build that quickly, you end up with all sorts of things that you don't have optimized, and our cost to serve and our fulfillment network were not what we wanted, and it created a lot of challenges and angst, and a lot of people at that time, I remember it was just a year or so, a little less than a year into starting the job, people said, "Boy, what a terrible time to start, and what bad luck."

现在回头看，我觉得那真的是一个绝佳的起点，也是一种运气，因为它让我们能思考很多原本可能要过很多年才会考虑的事情，而这些想法对客户和业务都非常有价值。

And I think in retrospect, it was really a great time to start and a real blessing because it allowed us to consider a bunch of things that we probably wouldn't have considered for many years, but they were really good ideas for our customers and for the business.

其中最关键的一点是，在我们的配送网络中，我们取消了美国全国性的扁平化网络，改为区域化布局。

And the biggest of which, in our fulfillment network, ended up being that we regionalized a flat national network, a fulfillment network in the US.

我们把网络区域化之后，商品离客户更近了，这样一来，运输距离缩短，既能降低成本，又能更快地把商品送到客户手中。

We regionalized it, so we got items much closer to customers, which turned out to both lower our cost to serve because it was shorter transportation distances while getting items to customers much more quickly.

在我们为配送成本苦苦挣扎的那段时期，重新审视我们最核心的信念，反而激发了配送速度和零售业务的需求，这也是我们业务增长这么快的一个重要原因。

And that reevaluation of our most closely held beliefs during that time when we were struggling on cost to serve really unleashed speed of delivery and a lot of demand in our retail business, and a big piece of why we've grown as fast as we have there.

当然，接下来没多久，人工智能在舞台上进一步爆发了。

And of course, pretty quickly thereafter, just how AI exploded further on the scene.

其实人工智能并不是什么新鲜事物。

It's not like AI was new.

我们一直都在做机器学习，而且有过很多波人工智能浪潮，都曾被看作真正的转折点。

We have been all doing machine learning for a long time, and there've been lots of waves of AI that were supposed to be the real tipping points.

但到了2023年，生成式人工智能确实带来了巨大的不同。

But you get to 2023, and you really see with generative AI a huge difference.

而我得说，那时候我们的AWS业务增长非常快。

And I would say our AWS business was growing really quickly at that point.

我们的功能最全面，运营表现和安全性也都更胜一筹，而且我觉得我们的合作伙伴生态也与众不同。

We have with the broadest functionality by fair bit and stronger operational performance and security, and I think a type of partner ecosystem that's different.

我们是绝对的领导者，而且我们觉得那块业务长期下来能带来几千亿美元的收入，大概十年以上吧。

We were the clear leader, and we felt like that business could be a few hundred billion dollars of revenue over a long period of time, figure over 10 years.

现在我们再看看AI带来的变化，我们所知道的每一个客户体验都会被推理和AI重塑。

Now we look at what's happening with AI and how every customer experience we know of is going to be reinvented with inference and AI.

我们其实认为AWS业务不仅增长很快，上个季度我们年同比增长了36.7%，年化收入达到1690亿美元。

We actually think the AWS business, not only is it growing really rapidly, last quarter, we grew 36.7% year over year on $169 billion annual revenue run rate business.

现在的机会真的非常惊人，而且如果你再看看我们新业务的种子，无论是团队在Prime Video业务上的进展，现在它已经是一个大型盈利业务，还是我们正在用Amazon Leo建设的低轨道卫星，或是Zoox的自动驾驶打车服务，还有我们的医疗健康业务，以及我们在机器人领域的探索。

Just the opportunities today are really striking, and I think if you layer on top of that the seeds that we have of new businesses, whether it's how the team has progressed our Prime Video business, which is now a large and profitable business, or the low Earth orbit satellite we're building with Amazon Leo, or our autonomous driving ride-hailing service with Zoox, or our healthcare business, or what we're doing in robotics.

我觉得我们有机会去帮助更多的客户，覆盖更多的客户体验，让他们的日常生活变得更好、更轻松。

We have an opportunity, I think, to help so many more customers across so many more customer experiences and make their lives better and easier every day.

而那个让客户的生活每天变得更好、更轻松的使命，确实把我们整个公司做的事情都串在了一起。

And that mission of trying to make customers' lives better and easier every day really unifies everything we're doing across the company.

接下来有几个问题想跟进一下。

So a couple of follow-ups on that.

首先，你提到建设配送网络速度太快，会让效率受到一些影响。

So one, you mentioned when you're building out the fulfillment network so fast that it's going to be a little inefficient.

当速度太快时，确实会出现一些混乱。

When you go that fast, there's some sloppiness to it.

然后你提到内部对此有些焦虑。

And then you mentioned there was angst around that internally.

从CEO的角度来看，你如何在不降低文化标准的前提下传达你的做法？

From a CEO perspective, how do you communicate what you're doing without lowering the cultural standard of...

亚马逊以卓越的运营效率、节俭等闻名，但你现在却好像在说“我们要建一个UPS”之类的。

Amazon's known for its excellence in operating efficiency, being frugal, all these kinds of things, but here you are going like, "We're going to build UPS" and like that.

那你怎么——

And so how-

……平衡这两者呢？

... do you reconcile those?

有意思的是，我们有一套很多人都知道的领导力原则。

Well, what's interesting is we have this set of leadership principles that a lot of people know about.

我们有这16条领导力原则，其中很多原则其实是为了……

We have these 16 leadership principles, and a lot of these leadership principles, they're supposed to really...

而且它们确实规范了我们作为团队和文化上的运作方式。

And they do govern the way we think about operating as a team and culturally.

但有几条原则之间是相互矛盾的，有时候人们对此感到困惑。

But there are several that are in tension with one another, and sometimes people are confused about that.

但现实是，这恰恰反映了真实世界的运作方式。

But the reality is it just reflects the way the real world works.

没有什么事情是完美无瑕的。

Nothing is really perfectly clean.

你总是在做取舍。

You're always making trade-offs.

所以回到当时我们决定那么激进、那么快地建设物流网络和最后一公里运输网络的那个时候，团队面临的问题是：我们知道如果这么快推进，肯定会非常困难。

So if you go back to that time when we made the decision to build out the fulfillment network and the last mile transportation network that aggressively, that quickly, the question the team was facing is we know if we go this fast, it's going to be really hard.

这可能比我们慢慢推进要更贵。

It'll probably be more expensive than if we went slower.

我们可能不会像慢慢推进那样把所有事情都做到最优，但客户和卖家确实需要它。

We will probably not get everything as optimally as we would if we went slower, but customers and sellers really need it.

所以我们是否真的会偏向于支持……记住，在疫情早期，所有实体店都关门了。

So are we going to actually shade on the side of enabling... Remember back then, in the early part of the pandemic, no physical stores were open.

如果你看看亚马逊在那段时间里在每个人生活中扮演的角色——

If you look at the role Amazon played in everybody's lives-

在疫情期间，它非常重要。

During the pandemic, it was very seminal.

所以我们不得不问自己：我们是选择快速建设，尽管可能不是最优，来服务那些试图维持自身业务的客户和卖家，还是不做？

And so we had to ask ourselves, are we going to build fast and maybe not as optimally as we can with the trade-off of serving customers and sellers who are trying to also have businesses that sustain themselves or not?

这是你必须做出的第一个决定。

That's the first decision you have to make.

我们做出的决定，几乎总是如此，是偏向客户。

And we decided as we almost always do, to shade on the side of customers.

然后很明显的是，我们在建设过程中，并不是所有计划中的事情都进展得很顺利。

And then it was fairly obvious that as we were building, it was obvious that not everything that we were planning was going perfectly.

我们的物流团队比公司其他大多数部门的人更经常在办公室，但还是有很多人需要参与大量的设计、规划和评估工作。

Our fulfillment team was more in the office than most of the rest of the company, but still a lot of people that have to participate in a lot of the designing and the planning and the assessing.

那时候，人们分散在各处，很多人在远程办公。

You had people all over the place and remote.

现在你需要在行动导向或速度和极高的标准之间找到一个平衡点，这意味着你要按照一直计划的成本结构、配送方式和设计来交付。

And now you're balancing the combination or the tension of bias for action or speed and relentlessly high standards, which means you deliver at the cost structure and the delivery and the design that you'd always planned for.

所以又一次，你必须做出取舍。

And so again, you have to make trade-offs.

我们做出了取舍，选择站在客户一边，快速行动，以便服务好卖家和消费者。

We made the trade-off to side with customers and to move quickly to be able to serve sellers and consumers.

我们当时以为会有一些后续的清理工作。

And we thought there'd be some amount of cleanup.

那我们现在知道地形会发展成现在这个样子吗？

Now, did we know exactly how the topography would work out in the way it did?

不是。而且有些东西我们没有优化，我也不确定我们事先没预料到它们不会被优化。

No. And there were certain things that we didn't have optimized that I'm not sure we foresaw wouldn't be optimized.

但如果你有好的指标，而且你的文化是讲真话的文化——这在我们文化中也是非常重要的一点——就是把数据摆到桌面上，把问题摆到桌面上，我们彼此直截了当。

But if you have good metrics and if you're a culture that's a truth-telling culture, which that is another really important part of our culture, is that we get the data on the table, we get the issues on the table, we're direct with one another.

我们不允许人们掩盖问题。

We don't allow people to paper over issues.

如果你有这样的文化，那么在指标里会很明显，在讨论中也会很明显。然后你只需想办法如何去改进，哪些是小的生产力问题，哪些是架构性问题。

If you have that type of culture, it's fairly obvious in the metrics, and it's fairly obvious in the discussion, and then you just have to go about figuring out how you want to improve it and which pieces are just kind of small productivity things and which are architectural pieces.

在获得那部分容量之后，我们回去真正重新设计了网络的关键部分。

We went back and really re-architected key parts of the network after we got that capacity installed.

那可能当时对那个团队来说，在他们的优先事项清单上排第八或第九，但后来变成了更高的优先级，以便真正实现速度和成本的结合。

That was probably at the time for that team was maybe eight or nine on their list of priorities, but became a much higher priority in a way to actually get the combination of speed and cost.

太有意思了。

Fascinating.

你好像在刚开始做AWS的时候，有点像是“建好了他们就会来”那种感觉。就是说，你在建设所有这些基础设施，然后你要把它们租出去。

You kind of like at the beginning when you started AWS, you like started AWS, it was kind of build it and they will come and that, okay, you're building all this infrastructure, and then you're going to rent it.

所以他们必须得来，不然你们就麻烦了。

And so they needed to come or you'd be in a lot of trouble.

但这在更大规模上，你和其他人，但我想主要是你，或者至少你肯定是最大的参与者之一，正在如此快速地建设数据中心和芯片产能，以至于金融市场都在说，'哦，全世界都拿不出足够的钱来资助这些东西。'

But this is that at a much bigger scale in that you and others, but I would say primarily you, or at least you're certainly one of the biggest players, are building data center and chip capacity so fast that the financial markets are saying things like, "Oh, there's not enough money in the world to finance these things."

然后我们正好看到了核心债券市场和股票市场出现的那些波动。

And then when we just kind of saw that skittishness in the core weave debt market and equity markets and so forth.

所以，鉴于需求如此旺盛，你们增长如此迅速，而且在供应链上还要争抢冷却设备、电力和芯片等等，你怎么知道建设规模多大才算过头呢？

And so how do you know how much is too much on that build-out, given the demand is so steep and you're growing so fast, and then there's a contention for that supply chain that you're kind of fighting for to get your cooling and power and chips and all that?

这是个非常好的问题，我希望有一个简单的公式能回答它。

It's a very good question, and I wish there was an equation that would make it simple.

那样我们团队就能省下很多时间。

It would save us a lot of time as a team.

我觉得在云计算的第一波浪潮中，我们刚创建AWS的时候，有几分'建好了他们自然会来'的味道，因为我们当时也不知道人们会不会用。

I would say in the first wave of cloud, when we started AWS, there was a little bit of we didn't know if people would use it, so build it and they will come.

但我们公司内部有很好的信号，那时候大概在2000年代初。

But we had very good signals inside the company, which at the time this was around the early part of the 2000s.

有些团队进展慢得让我们着急，尽管我们自己觉得相对已经算快了。

We had teams that were not moving nearly as quickly as we wanted them to, even though we felt like relatively we moved quickly.

但我们就是搞不懂为什么他们花那么长时间。

But for us, we just couldn't understand why it took so long.

结果发现，我们的产品团队把70%到80%的时间花在了基础设施上，而那些并不能让他们脱颖而出，只有大约20%的时间花在了真正能体现产品特色的部分上。

And it just turned out that our product teams were spending 70% to 80% of their time on the infrastructure pieces, which didn't differentiate them, and only about 20% of their time on what differentiated the product pieces.

我们意识到每个团队都在重复造轮子。

And we realized every team was doing it and reinventing the wheel.

所以我们最初创建AWS，就是为了让我们自己内部能更快地推进。

And so we originally conceived of AWS to allow ourselves to move more quickly internally.

我们想，如果我们能构建一套大规模、高可用、低成本、高灵活性的基础服务，比如计算、存储、数据库、内容分发、消息等等，那我们就能在消费业务上开发得更快。

And we figured if we built a set of high scale, high availability, low cost, highly flexible building block services like compute and storage and database and content distribution, messaging, and things like that, we would be able to build much quicker in our consumer business.

然后我们觉得，相对来说，我们做得还不错。

And then we just said we're pretty good, relatively speaking.

从技术上讲，可能其他公司也有同样的问题。

Technically, probably other companies have the same issues.

但即使在对外发布之前，我们内部就已经有很多团队感兴趣了，这让我们知道我们走对了路。

But even before we launched externally, we had so many teams internally interested that we knew we were on the right track.

我想我们谁也没预料到它在外部会发展得这么快。

I don't think any of us anticipated it would grow as fast as it did externally.

在人工智能方面，你知道，这些年一直有不同的风潮，人们说：“就是这个时候了。”

On the AI side As you know, there have been these different waves of people saying, "This is the time.

人工智能终于要火了。”

AI is finally here."

甚至可能在生成式人工智能风潮到来的六、七年前，人们就感觉人工智能的时代来了。

And even maybe six, seven years before the generative AI wave, people felt like it was AI's time.

我们构建了一大堆人工智能服务，这些服务在AWS这边确实是个不错的业务，但和现在看到的完全没法比。

We built a whole bunch of AI services that actually was a pretty good business for us on the AWS side, but nothing like what we're seeing right now.

在AWS早期，我们感觉自己增长得很快，也为此感到自豪，三年后我们的年运行率业务约为6000万美元。

In the early days of AWS, we felt like we were growing really fast, and we felt good about it, and we were about a $60 million annual run rate business after three years.

在这次生成式人工智能风潮的头三年里，光是AWS上的人工智能业务运行率就超过了250亿美元。

In the first three years of this generative AI wave, just our AI business in AWS is over $25 billion in run rate.

所以这要快好几百倍。

So it's several hundred times faster.

我们真的从没见过这样的情况。

We just haven't seen anything like it.

然后问题就变成了，你愿意投入多少资本支出去抓住这个机会？

And then the question becomes, how much CapEx do you want to invest to go after the opportunity?

我想告诉你的是，我们已经说过，今年，也就是2026年，我们预计资本支出大约会是2200亿美元。

And what I would tell you is, we've said this year, in 2026, we expect to spend about $220 billion of CapEx.

顺便说一句，我记得15年前在领导团队的早餐会上，我们简直不敢相信电信公司会花500亿美元在资本支出上。

By the way, I'll just tell you that I remember being at leadership team breakfast 15 years ago when we couldn't believe the telco company spent $50 billion in CapEx.

我们当时就说——

We're like-

对

Yes

“我只打算花那么多在资本支出上。”

... "I'm only spending that much on CapEx."

这些数字真的很惊人。

So these numbers are mind-blowing numbers.

即使今年在资本支出上投入2200亿美元，我们仍然有很多需求没有被满足。

And even spending $220 billion in CapEx this year, we're leaving a meaningful amount of demand on the table.

对我们有利的是，我们对客户需求有非常清晰的了解。

The thing that helps us here is that we have really good visibility into customer demand.

所以，如果你想想AWS的资本支出投资以及现金流的运作方式，主要有两个大的方面。

And so if you think about the CapEx investments in AWS and the way the cash cycle works, there's two macro areas.

一个是数据中心投资，另一个是服务器和网络设备投资。

There's the investment for data centers, and then there's the investment for servers and networking.

数据中心部分，你需要在服务器能够放入数据中心并开始产生收益之前大约两年进行投资。

The data center piece, you have to invest about two years in advance of when you can put servers in the data centers and start to monetize that.

好消息是，这种投资只需要做一次。

The good news is you only make that investment once.

这是一种使用寿命大约30年的资产。

It's a useful life asset that's about 30 years.

但你得提前几年就开始投入。

But you make that a couple of years in advance.

服务器和网络设备，具体取决于组件，通常只需要提前几个月。

The server and networking gear, depending on the component, it's only a few months in advance.

所以在你做出那个决定之前，你对客户需求有非常清晰的了解。

So you have very good visibility into customer demand before you pull that trigger.

如果事实证明你不需要填满那些数据中心，你就不会买服务器和网络设备。

And if it turns out you don't actually need the capacity to fill those data centers, you won't buy the server networking gear.

平均来说，他们大约三年就能回本。

That, on average, they tend to break even in about three years.

而且目前这些设备的使用寿命大约是五到六年，我们签的合同大多是五年期的。

And they have useful lives at this point of about five to six years, and most of the deals we're doing are about five-year deals.

所以当你以我们这样的速度同时建设这么多数据中心时，你必然会花费大量资本，而你的自由现金流在那些年份会面临压力，直到几年后进入后期阶段，所有数据中心都上线并开始产生收入，你就能收回服务器和网络投资的成本。

And so when you are building as fast as we're building with so many data centers simultaneously, you're going to spend a lot of capital, and your free cash flow is going to have headwinds in those years until you get to those later stages a few years down the road, where all those data centers are online, and they're all being monetized, and you're making back your money from the server and the network investment.

但由于现金流的运作方式，当你以如此快的速度增长时，初期年份会遇到阻力。

But because of the way the cash cycle works, it has headwinds in the early years when you're growing that fast.

说到数据中心，回到你刚开始做AWS的时候，那可能是世界上最不可能成为政治议题的事情。

So actually on that, data centers, going back when you started AWS, were probably the last thing in the world that you would believe would be a political issue.

但现在居然变成这样了。

But here we are.

尤其是在美国，挺有意思的，我相信你也遇到过这种情况，但全世界没有任何地方像美国政治那样反对数据中心，不管出于什么原因。

And particularly in the United States, it's funny, and I'm sure you've experienced this, but nobody anywhere in the world is as anti-data center as the kind of US politics for whatever reason.

那这事儿是怎么影响到你的——我想问，反对数据中心的观点有道理吗？

So how does that end up affecting this-- I guess, is there any merit to the anti-data center argument?

然后，这又是如何影响你的策略的，包括你在这里建了什么，以及在其它地区建什么？

And then, how does that kind of inform your strategy of what you built here, and what you build in the other regions?

嗯，我跟你分享一些我了解的情况，或者至少是我认为的情况。你想，我们今天做的所有事儿，股市、银行、医疗、娱乐、购物、教育、商业、打车。

Well, I'll just give you some of the context that I have, or at least I believe, which is, if you think about everything we do in the world today, stock market, banking, healthcare, entertainment, shopping, education, business, ride hailing.

我们现在做的每件事儿都是数字化的。

Everything we do is digitally driven.

它们都以非常简单的方式使用着数据中心。

They're all using data centers in a very simple way.

不，每个人都在说：我反对我自己。

No, everybody's, "I'm anti-me."

我再也不想用手机了。

I don't want to use my phone anymore.

而且他们都在用数据中心。

And they all use data centers.

这还是在AI之前的事。

And then that's before AI.

而AI是我们这个时代最大的技术变革，考虑到我们已经经历了互联网、云计算和移动技术，我觉得这已经说明了很多。

And AI is the biggest technology transformation of our time, which I think is saying something since we've lived through the internet and the cloud and mobile.

所以，现实是，为了支持我们刚才谈到的AI方面的快速增长，我们需要进行的那种基础设施建设的规模，自从1950年代我们建设高速公路和桥梁以来，世界上和这个国家里都没有出现过类似的情况。

And so, the reality is the type of infrastructure build-out we need to support the rapid growth we were just talking about on the AI side, there hasn't been anything like it in the world and in this country since maybe when we built out the highways and bridges in the 1950s.

我们正在经历的就是这种规模的建设。

That's the type of build-out we're going through.

这个国家，我觉得整个世界都算上，但我特别要说的美国，如果不去建设数据中心，后果是我们承担不起的，否则我们会发现像中国这样的国家会在这方面建立所有的领导地位和标准。

And the country, I think the world in general, but I'll say the US specifically, cannot afford not to build data centers, or we will find that another country like China will establish all the leadership and all the norms there.

而且我觉得我们任何人，无论是在国内还是全球范围内，都不会喜欢这种情况。

And I don't think any of us are going to like that, either domestically or globally.

所以我们必须要建数据中心。

So we have to build data centers.

那么，我是否认为我们整个行业可以有比现在更好的做法呢？

Now, do I think there are ways that we can be doing it better, all of us, as an industry than we have?

当然有。

Of course.

我觉得事情发展得很快，大家都在共同学习。

I think it's happening quickly, and I think everybody's learning together.

但目前有很多在我看来介于误解和谎言之间的说法在流传，因为这类问题很容易找到一个替罪羊，政治上，有时候投机者也喜欢找那个替罪羊来指责。

But there are a number of What I would consider somewhere between myths and mistruths right now being propagated, because it's one of those issues where you have an easy villain, and politically, and sometimes opportunists like to find that villain to point at.

我来给你们举几个例子。

But I'll give you examples.

高尔夫行业或者杏仁种植业消耗的水量比数据中心多得多。

The golf industry or the almond farming industry consume way more water than data centers do.

举个例子，高尔夫球场消耗的水量是我们所有数据中心用水量的200倍，而且它们两天用掉的水比我们数据中心一年用的还多。

Just as a data point, golf courses consume 200 times more water than all our data centers, and they consume more in two days than we do in a year in our data centers.

所以这是个很好的例子。

So that's a good example.

或者人们会说污染问题。

Or people talk about pollution.

实际情况是，当你注册新站点并配备备用发电机时，你需要登记并展示如果这些发电机全天满负荷运行所需的排放量。

What happens is when you register new sites and you have backup generators, you have to register and show the amount if you ran all of those generators full-time at full capacity.

但当然，这种情况从来不会发生。

But of course, that never happens.

备用发电机一年只运行0.1%的时间，也就是大约10小时用于维护。

You only run backup generators 0.1 of 1% of the year, like 10 hours a year for maintenance.

或者你谈到能源价格。

Or you talk about the energy rates.

如果你看看我们大型数据中心分布的不同地区，有些地方的能源价格上涨了，有些地方的则下降了。

If you look in the different geographies in which we have a big data center presence, some of the energy rates have gone up, some of the energy rates have gone down.

但真正的问题是，作为一个国家，我们在电网基础设施上的投资没有达到应有的水平，未来肯定会有很大的增长需求。

But the real issue is that as a country, we have not invested in the infrastructure of our grids the way we've needed to, like there's going to be meaningful growth.

所以，总有什么东西会推动这种投资，而恰好人工智能因为发展得太快、太猛，就成了那个催化剂。

And so something was going to push it over to investment, and it just turns out that AI, because it's growing so rapidly and so significantly, is the thing that pushed it over.

但是，当你真正把容量建起来之后，你会发现，在大多数地方，电价会趋于平稳甚至下降，就像印第安纳州、佐治亚州或密西西比州那样。

But building it out, of course, what you're going to find is once you actually get that capacity in most of the geographies, they're going to end up, the rates will flatline or go down as they have in Indiana or in Georgia or Mississippi.

所以，我们其实可以做得更好。

And so, there are things that we can do that are better.

还有一点大家容易忽略的是，在这些地区，大型数据中心运营商缴纳了大量的税款。

The other thing people forget about is in these geographies, the big data center providers are paying a lot of tax.

比如在印第安纳州的圣约瑟夫县，我们已经缴纳了大约30亿美元的税款。

So if you look in St. Joseph County in Indiana, we've paid about $3 billion of tax.

这些钱最终会落到社区手里，他们可以用在真正对社区有用的地方。

The communities get that money, and they get to spend it on things that really matter for those communities.

所以，社区其实也从这里面受益了。

So the communities get something out of it, too.

我们推出了像“共同建设”这样的项目，刚刚又承诺在未来五年内向我们设有数据中心社区投入另外十亿美元。

We've built programs like Built Together that we just committed to another billion dollars in the communities in which we have data center presences over the next five years.

所以，我认为我们必须继续与社区合作。

And so, I think we have to keep working on it with the communities.

这必须是公私合作。

It has to be a public and private partnership.

但建设对于国家的未来至关重要。

But the build-out is so important to the future of the country.

想象一下，如果我们没有修建高速公路和桥梁，国家的经济繁荣和能力会是什么样子。

Imagine if we hadn't built out highways and bridges and what the economic prosperity and the capabilities of the country would be.

我们必须正确看待这件事，并且必须成为我们建设所在社区的合作伙伴。

And we have to keep it in perspective, and we have to be partners in the communities in which we build.

让我们再回到资本支出这个话题。

Let's go back to the CapEx for a second.

那2200亿美元，也就是每天6亿美元，这么说只是为了让你有个概念。

The $220 billion, that's $600 million a day, just to put it in perspective.

我很好奇你现在怎么看AI的需求。

I'm curious how you think about AI demand right now.

之前在台下我们聊过，你把需求比作杠铃，也许你可以具体说说当前AI需求的情况，放在什么样的背景下看。

Offline, we were talking about how you see it as a barbell, and maybe you can situate or contextualize the state of AI demand at the moment.

嗯，首先一个重要背景就是我之前说的，简直难以置信AI发展得这么快、规模这么大。

Well, the first context piece is what I said earlier, which is just I can't believe how big and fast AI has grown.

它的增长速度比AWS起步时快了好几百倍，我们AWS在AI收入上，头三年就达到了250亿美元的年化运行率。

It's a few hundred times faster than the beginning of AWS in the first three years, $25 billion run rate for us just in AI revenue in AWS in the first three years.

所以规模确实非常大，但你如果看需求的结构，确实是杠铃型的。

So it's really large, and yet it really is barbelled if you look at how demand is.

我想说的是，杠铃的一端是AI实验室，他们在消耗巨量的算力，这是我以前从未见过的。

And what I would say is on one end of the barbell are the AI labs who are consuming gobs and gobs of compute, unlike anything I've seen.

还有就是少数几个爆款应用，比如ChatGPT、Cloud Code之类的。

And then a few runaway apps, ChatGPT, Cloud Code, things like that.

杠铃的另一端是企业，他们主要在利用AI来削减成本和提升生产力，比如自动化客服、反欺诈，或者业务流程自动化。

On the other end of the barbell are enterprises who are largely using AI for cost avoidance and productivity, so automating customer service or fraud or business process automation.

在杠铃中间的是企业目前的生产工作负载，其中许多已经在使用AWS和推理服务，但我得说，绝大多数企业还没有普遍采用。

In the middle of the barbell are the current production workloads of enterprises, many of which are using AWS and using inference today, but I would say the overwhelming majority of which still aren't in any pervasive way.

我认为这将是AI使用中绝对规模最大的部分。

And that I think is going to be the largest absolute segment of AI usage.

而且我觉得你会看到这一天的到来。

And I think you're going to see that.

我很难预测这会发生得有多快。

It's hard for me to predict the rate at which that'll happen.

每年都在更大程度上推进。

It's happening to a greater extent every year.

但如果把时间拉长到五年，五到七年，再看看那时候的情况，我觉得不会等那么久。

But I think if you zoom out five years, five to seven years, and you look at what it'll be like, and I don't think it's going to wait for five to seven years.

这种事从来不会一蹴而就。

It never happens in one fell swoop.

但只要放眼五到七年，我觉得相当多，甚至大多数现存的企业应用都会普遍使用推理技术。

But if you just zoom out five to seven years, I think that quite a few, if not most, of the enterprise applications that we use and that exist today are going to be using inference in a pervasive way.

所以我们面前还有很大的增长空间。

And so there is a lot of growth in front of us.

那你怎么看这个问题？

And how do you think about it?

因为到2027年，数据中心、芯片、内存这些资源短缺的问题就会来临，现在大家都在建设，但到年底未必能准备好，而需求却一点都没减少。

Because there's this impending data center, chip, memory, everything shortage that's upon us in really 2027, just because everybody's building now, but it's not going to be ready necessarily by the end of the year, and demand's not slowing down at all.

那你面临这样的情况怎么看？大实验室愿意为电力、冷却、芯片和环境付出几乎任何代价。

And so how do you think about it when faced with the big labs being willing to pay maybe anything, or almost anything for power, cooling, chips, the environment.

而另一边，你还有那些合作了很久的老客户，他们按正常价格来运行他们的业务。

And then you've got these kind of all been around a long time customers who are paying their regular rates to run their businesses.

会不会有业务上的压力，想着“嗯，我们是不是该把老客户那边的资源让给新客户，因为他们愿意付任何价格。”

Is there any business pressure to go, "Well, maybe we should buy out the old guys and sell them to the new guys because they'll pay us whatever.

对于老客户，我不能把价格翻倍。

And the old guys, I can't double their prices.

那不公平。”

That wouldn't be fair."

或者你是怎么看待这个问题的？

Or how do you think about that?

因为这是一个很...

Because it's such a-

...奇怪的情况，你有一些合作了20年的老客户，他们做的事情，正如你所说的，可能是在做推理，然后又有这些新客户，他们似乎永远不满足。

... weird thing in that you've got people you've had in there for 20 years doing a thing that's ... kind of, as you say, probably going to inference, and then you've got these new guys who are insatiable.

他们什么都愿意买。

They'll buy anything.

这个问题问得好。

It's a good question.

我们并不是完全按照你刚才说的那种方式来思考这个问题，但我完全理解你的意思。

We don't think of it quite the way you're talking about, but I can completely understand.

你想啊，现在数据中心和电力就那么多。

If you think about it, there's only so much data center and power today.

内存也就那么多，而且我不认为这种短缺会在2027年结束。

There's only so much memory, and I don't think that that shortage ends in 2027.

我觉得，回到资本投资这个话题，其中一个原因是，为什么我认为我们可以投更多而不会浪费资本，因为26年显然现在已经卖完了。

I believe, one of the reasons, coming back to the capital investment, why I think we could invest a lot more and it wouldn't be a bad use of capital is that, 26 is obviously at this point sold out.

27年也基本卖完了。

27 is largely sold out.

我们总会留一部分给我们的推理服务中的按需使用，比如Bedrock或EC2，但大部分大块头都已经卖光了。

We always leave a certain amount for on-demand in our inference service, Bedrock or EC2, but largely sold out all the big chunks.

28年的一大部分也卖出去了。

A good chunk of 28 is sold out.

需求实在太大了。

There's so much demand.

如果你想想AI业务和我们所谓的核心业务，也就是非AI业务，现在它们都在以很快的速度增长。

If you think about the AI business and what we call the core business, which think of core as non-AI, they're both growing at really rapid rates right now.

而且，如果你只有一定量的内存，你就得决定把多少内存分配给HBM，给AI加速器，而不是给DRAM和核心业务。

And, if you have only a certain amount of memory, you have to decide how much of the memory you want to allocate to HBM, to the AI accelerators versus to DRAM, and to your core.

所以这些都是你必须做出的实际权衡，但我想说，我们在这个问题上的出发点是以客户为中心。

So these are real trade-offs you have to make, but I would say that where we start on this issue is being customer-focused.

对我们来说，我们在AWS和客户之间有着深厚的合作和伙伴关系。

And for us, we have deep relationships and partnerships with our customers in AWS.

同样，在疫情期间，不像我们的竞争对手，他们我觉得是想在季度内尽量优化，以在客户成本优化的大环境下尽可能多地提高收入，而我们采取了不同的方式。

And in the same way, during the pandemic, unlike our competitors, who I think tried to figure out how to optimize within quarters to drive revenue as much as possible in a world where companies were going to cost optimize, we took a different approach.

我们主动让我们的团队去帮助客户找到成本优化的方式，尽管我们清楚这会在短期和中期内减少我们的收入。

We proactively had our field go out and help customers figure out how they can cost optimize, even though we knew it would diminish our revenue, short and medium term.

但我们会帮助我们的客户取得成功，因为我们坚信，如果我们长期帮助客户，我们的业务也会随之发展。

But we're going to help our customers be successful because our very strong belief is if we help customers over a long period of time, our business will come along with it.

在核心业务上也是如此。

And that's the same thing here on the core.

我想你可以通过计算，认为由于供不应求，客户愿意支付更多，挤掉我们的核心客户，但他们是在AWS上运营业务的。

I guess you could do calculations and decide that you have customers, because of the scarcity and their willingness to pay, would pay more to squeeze out your core customers, but they're running their businesses on top of AWS.

他们运行着所有的核心应用。

They're running all their core applications.

我们不会挤压我们的客户，我们会确保他们能够扩展。

We're not going to squeeze our customers, and we're going to make sure that they can scale.

然后我们的任务就是尽可能多地增加容量，不仅能满足核心客户，也能让所有想在AI方面使用我们服务的客户都能如愿以偿，同时还要确保对我们来说在经济上也是可持续的。

And then our job is to try to figure out how to have as much capacity as we can to allow not just our core customers, but all of our customers who want to use us on the AI side, and people want to use us on the AI side, to be able to do so, and do so in a way where it's sustainable economically for us as well.

所以，我希望我能告诉你，我认为供应限制只是2027年才会出现的问题。

And so, I wish I could tell you that I thought that the supply constraints would be just a 2027 issue.

但我认为，未来几年我们会在不同的组件上看到这些问题。

I think we'll see them in different components for several years.

我觉得我们可能比其他任何人都在增加容量，因为我们非常热衷于继续保持我们在基础设施和AI领域的显著领先地位，就像我们过去一直做的那样。

I think that we're probably putting more capacity in than anybody else because we are very passionate about continuing to be the significant leader in the infrastructure and AI space like we've been.

但正如我所说，我们今年大概要花费2200亿美元，而且我们还可以花更多，这不会是资本的浪费，从长期和中期来看都是如此，但即使那样，我认为我们可能仍然无法满足所有的需求。

But as I said, we're spending $220 billion or so this year, and we could spend a lot more, and it wouldn't be a bad use of capital, and now both medium to long term, and I think we still probably wouldn't satisfy all the demand.

现在的需求简直太旺盛了。

There is just an incredible demand right now.

说到这个，你在2015年做了一件大事，收购了Annapurna Labs。那时候你可能并没有想到，也许你想到了，但当时并没有巨大的AI需求，因为2015年AI需求确实不大。

So kind of speaking of that, you did something back in 2015, when you bought Annapurna Labs, which was probably before, you probably weren't really thinking, maybe you were, but about giant AI demand in 2015, because there just wasn't much AI demand in 2015.

那么，是什么驱动了你做出那个决定？而在这个新世界里，这个决定又是如何回报你的呢？

So what drove that decision, and then how has it kind of paid off in this new world?

我觉得Annapurna团队是我有幸合作过的最有才华的团队之一。

I think the Annapurna team is one of the most talented teams and set of people I've ever had a chance to work with.

这也是亚马逊历史上最成功的收购之一，甚至可以说是最成功的。

And it's one of the very best acquisitions we've done at Amazon in the history of the company, if not the best.

我们确实预见到了一些东西，但我们收购Annapurna并不是专门为了AI芯片。

We foresaw some pieces, but we did not make the Annapurna purchase for AI chips per se.

当你的规模达到一定程度，你就能接触到超大客户群体的行为模式，你会意识到，如果能垂直整合，自己构建某些能力，就能为客户创造更多价值，同时也能让业务本身的经济效益更好。

When you get to a certain scale, you start to have access and visibility to patterns of customers at a very large scale that you realize if you can vertically integrate and you could build certain capabilities yourself, you'll be able to provide more value for customers and better economics for the business itself.

所以Annapurna最初对我们来说，主要是为了核心的EC2服务器。

So what Annapurna started as for us was just on the kind of core EC2 server.

我们觉得服务器上有太多资源被虚拟化、网络、安全和存储占用了。

We felt like we were using too much of the server for virtualization and for networking and for security and for storage.

我们希望能够将服务器的几乎全部算力都提供给客户使用。

And we wanted to be able to use nearly 100% of the server for compute for customers.

所以在收购之前，我们就已经和Annapurna合作开发这款网卡，把那些组件从服务器上卸载到网卡上，这样就能尽可能多地把服务器资源用于客户的计算任务。

And so we had been working with Annapurna before we acquired them on this NIC that offloaded those components to the NIC to allow us to have as much of the server for compute as possible.

我们真的很喜欢这个团队，也很喜欢我们一起合作的那个产品。

And we really liked the team, and we really liked the product that we collaborated with them on together.

然后我们意识到，这个产品最终几乎会出现在每一台 EC2 服务器上，而且事实证明，它现在已经出现在我们所有的服务器里了。

And we realized that was going to end up being in virtually every EC2 server, and it's turned out to be in every server we have.

所以我们当初收购它，是因为它将成为我们计算能力的核心部分，而且因为这个团队在芯片制造方面非常出色，我们觉得他们可能还能造出其他类型的芯片。

So we originally bought it because that was going to be a core capability of our compute capabilities, and because the team was so talented at building chips, we felt like they could probably build other chips too.

那时候，我们和一家公司有着非常深入的合作关系，直到现在也依然如此，那就是英特尔。

And at that time, we had this very deep partnership, which we still have a deep partnership with this company, with Intel.

但当时英特尔几乎垄断了 CPU 市场，当你有一个几乎完全占据市场份额的合作伙伴时，他们通常不会想着如何提升性价比。

But Intel had virtually all of the CPU share, and when you have a partner who has virtually all the share, they're not often thinking about how to drive price performance.

他们更多考虑的是增加一些新功能，好让自己拿到更多市场份额，赚更多的钱。

They're trying to think about other features that allow them to get more share and drive more revenue.

说实话，这完全可以理解。

By the way, it's totally understandable.

这些就是市场运作的一些典型规律。

These are some of the dynamics of how markets work.

这就是创新者困境出现的一部分原因。

It's part of how the innovator's dilemma comes up.

所以我们一直想在性价比上施压，但很难引起他们的注意，不过我们有客户想越来越多地把本地工作负载搬到云端，他们需要更好的性价比。

And so, we had tried to push on price performance, and it was hard for us to get attention there, but we had customers who wanted to move more and more of their on-premises workloads to the cloud, and they wanted better price performance.

所以我们接下来让Annapurna团队去做的就是打造一款CPU芯片。

And so the next thing we turned that Annapurna team loose on was to build a CPU chip.

他们造出了Graviton，如今我们前1000家EC2客户中有99%都在普遍使用它。

And they built Graviton, which 99% of our top 1,000 EC2 customers use it very pervasively today.

它比其他X86处理器性价比高约40%。

It's about 40% more price performant than other X86 processors.

所以这对我们的客户和业务都产生了巨大的影响。

And so that has been hugely impactful for our customers and for our business.

几年前，我们在AI领域看到了同样的情况，那里有一个占主导地位的提供商，非常好，而且我们和他们有深厚的合作关系，现在依然有，并且在可预见的未来也会一直保持。

And a few years ago, we saw that same movie in the AI space where you really had one dominant provider, who's very good, and we had a deep partnership with, we still have a deep partnership with, we will for as long as I can foresee.

但他们的市场很大，正在构建很多不同的产品和优先事项，我们清楚客户需要更好的性价比。

But they have giant markets and lots of different products they're building and priorities, and we knew our customers needed better price performance.

这就是为什么我们在Annapurna内部单独组建了一个团队，来打造AI加速器，也就是Trainium，它的性价比也比其他同类处理器高大约30%到40%。

And that's why we spun out a separate team in Annapurna to build an AI accelerator, which we call Trainium, which is also about 30% to 40% more price performant than the other comparable processors.

我觉得，当我们谈到之前说的那个橄榄形中间部分的时候，如果你问是什么能让更多现有的企业生产应用去做推理，改变那个中间部分的格局，有几个因素。

And I think when you talk about that middle of the barbell we were just talking about, if you ask what are the things that will change that middle of the barbell to have more of those current enterprise production apps using inference, there are a couple things.

一部分是那些擅长做后训练和微调的人才，以及那些知道怎么排优先级、知道该做什么、怎么做的组织和团队能力。

Some of it is some of the talent that's comfortable doing post-training and fine-tuning and some of the talent and organizational muscle that knows how to prioritize what to go after and how to do it.

但很大一部分是推理的成本。

But a big, big chunk of it is the cost of inference.

在过去几年里，推理成本已经下降了不少，但在我看来，它还是贵了一个数量级，无法像理想中那样被广泛应用。

The cost of inference has come down quite a bit in the last few years, but it's still an order of magnitude too expensive, in my opinion, for it to be used as pervasively as possible.

因为如果你让企业把现在用CPU的应用重写，改成用AI加速器，如果成本差距很大的话，他们需要从AI加速器中获得的额外收益就得非常高，才愿意去重写那些应用。

Because if you're asking enterprises to rewrite their applications that use CPU today with AI accelerators, if the difference in those costs is very significant, the amount of incremental benefit they have to get from having the AI accelerators on there is a much higher bar to want to rewrite those apps.

但如果你能降低推理成本，那个门槛就低多了，而且所有应用在加了AI加速器和推理之后，都会变得更好、更聪明、能力更强。

But if you can take the cost of inference down, that bar is much lower, and all applications are going to be better and smarter and more capable with AI accelerators and inference.

所以，让客户能广泛使用AI的一个非常重要的部分，就是降低推理成本，而我们的芯片业务在其中扮演了重要角色。

So that is a really important part for customers to being able to use AI pervasively is driving the cost of inference down, and our chips business is a big piece of that.

我可以告诉你，在我们自己的公司，这事儿可能花很多钱。

I can tell you from our own firm that it can get very expensive.

去年，我们有一个员工花了200万美元干了不少好事，不过——

We had one employee spend $2 million in the last calendar year doing good work, but

……

...

我当时就觉得，好吧，这个推理成本可没那么便宜。

I was like, "Okay, this inference is not that cheap."

所以呢，你做的其中一件事是投资，特别是在Anthropic和OpenAI上，投入非常大，我觉得。

So, one of the things that you've done is you've made some investments, particularly in Anthropic and OpenAI, very large investments, I think.

好像给Anthropic投了180亿美元，给OpenAI投了500亿美元，差不多这数吧。

Like 18 billion in Anthropic and 50 billion in OpenAI or something like that.

那是什么让你想去做这些投资的呢？

And so what got you to want to do that?

这些投资，还有Nvidia类似的投资，让华尔街闹翻了天，因为这让他们想起了互联网泡沫时期那种糟糕的日子。

And that and similar investments from Nvidia and so forth got Wall Street up in a tizzy because it reminded them of the bad old days of the dot-com boom and so forth.

你是怎么考虑的，特别是因为他们也是客户，投资他们对你来说为什么合理？

How did you think about that, particularly since they are also customers, and why did that make sense for you to invest?

这两家公司团队都非常出色，公司也非常出色。

Both of those companies are really impressive teams and really impressive companies.

他们业务增长的速度，至少从营收来看，我们以前真的没见过这样的。

And the rate at which they've grown their businesses, at least in the top line, I don't know if we've ever seen anything like it.

真的是——

It's really-

太惊人了。

Amazing

——难以置信。

... unbelievable.

所以对我们来说，这是这些团队、他们构建的技术和模型，以及他们开发的一些应用，加上我们对AI变革性的信念，以及我们相信长期来看这两家公司会在这个领域成为最成功的两家公司，这些因素结合在一起。

And so to us, it's a combination of who these teams are, the technology and the models that they've built, and some of the applications they've built, our belief in how transformative AI is, the fact that we believe these are going to be two of the most successful companies in this space over a long period of time.

不是每家都能赢，但在这个领域会有很多非常成功的公司。

And not everyone's going to be a winner, but there are going to be a lot of very successful companies in this space.

但我们相信它们会是其中两家。

But we believe these will be two of them.

而且我们合作关系非常紧密。

And then the fact that we partner so closely.

如果先从Anthropic说起，我们合作时间最长，有很多客户在我们的推理服务Bedrock上运行Anthropic。

I would say that if I start with Anthropic, where we've had the longer partnership, we have so many customers that run Anthropic in Bedrock in our inference service.

我认为客户在Bedrock上运行Anthropic效果最好。

I think customers are most successful running Anthropic in Bedrock.

我们在芯片方面也有深入合作，他们基于Trainium开发已有一段时间，这对Anthropic来说非常好。

We have had a deep partnership with them on chips, where they have been building on top of Trainium for a while, and I think it's been very good for Anthropic.

这对我们的Trainium团队也很好。

I think it's been very good for our Trainium team.

我们双方都从彼此身上学到了很多。

I think we've both learned quite a bit from each other.

所以我们与它们有着深入的合作，无论是我们对AI未来的信念、这些公司的信念，还是合作关系，都让我们觉得通过投资来加深合作是合理的。

So we have a deep partnership with them, where both our belief in the future of AI, the belief of those companies, and the partnership, it made sense for us to make investments as part of those partnerships.

在OpenAI这边，我们有很多客户想同时用OpenAI和AWS，但因为他们之前的商业安排，一直没能实现。

And then on the OpenAI side, we just had so many customers who wanted to use OpenAI and AWS, and they weren't able to do so because of their prior business arrangements that they had.

而且很长一段时间，我觉得两家公司彼此都很尊重，也很想合作。

And for a long time, I think the companies had a lot of mutual respect and desire to work together.

所以这也是一样的情况，通过一连串的对话，经过几个月的时间，终于实现了在Bedrock上运行OpenAI。

And so this is another one of those where over a series of conversations and months The ability to be able to run OpenAI in Bedrock.

我们一起打造了一个独特的产品，就是那个有状态运行时环境，我觉得以后所有智能体都会这样构建。

We built a unique product together in the stateful runtime environment, which is how I think all agents are going to be built moving forward.

你必须要处理状态。

You're going to need state.

跟Anthropic一样，他们也承诺多年、多吉瓦的Trainium使用量，我们相信这家公司会在这个领域长期做得非常成功。

Like Anthropic, they made a multi-year, multi-gigawatt commitment to Trainium, and we just believe that that is going to be one of the most successful companies in the space over a long period of time.

所有这些都让我们觉得，加深合作伙伴关系并进行投资是合理的。

All those made it make sense for us to have deeper partnerships and investments.

接下来我们聊聊怎么用AI来打造公司。

Switching over to how you use AI to build a company.

所以，我记得你说过，Bedrock团队在76天内用六名工程师重建了你们的推理引擎，而以前这需要40个人花一年时间。

So, I think you said the Bedrock team rebuilt your inference engine with six engineers in 76 days, and that would've taken 40 people a year.

那确实很不一样。

So that's pretty different.

而且亚马逊以两个比萨团队而闻名。

And Amazon is famous for the two pizza teams.

那么，基于当前的技术，你们现在对组织设计有什么想法？

So what is your current organizational design thinking, given the technology?

因为这确实会带来很大变化，这些年来我们变得非常专门化，有设计师、产品经理、前端工程师、后端工程师等等，各种角色。

Because it does seem to really change things, and we got super specialized over the years with designers and product managers and front-end engineers and back-end engineers and this and that and the other.

而现在突然之间，每个人都成了全能型人才。

And now all of a sudden, everybody's a Renaissance person.

那在组织设计方面，你们是怎么考虑的呢？

And so how do you think about that in terms of org design?

是的，我同意你所说的。

Yeah, I agree with what you're saying.

我觉得你提到的那个Bedrock经历，对这里的每个人来说都非常有启发。

I think that the Bedrock experience that you alluded to was very eye-opening for everybody here.

我们花了很多时间在内部分享这个故事，来激励大家，提醒大家什么是可能的，因为随着时间的推移，会发生这样的情况——我在29年前创办了这家公司。

And we spent a bunch of time sharing that story internally to inspire people and remind people what's possible, because over time, what ends up happening-- I started the company 29 years ago.

我最初是助理产品经理，我当时大概负责五个产品吧。

I started as an assistant product manager, and I was the product manager for, I don't know, five products.

现在，当公司成功壮大，产品也越来越大的时候，就会出现这种情况：一个产品经理只负责一个产品，而且只管理其中的一小部分。

Today, when companies get successful and they get bigger and the products get bigger, you end up where a product manager works on one product and they just own one slice of it.

比如只负责定价。

Like they own pricing.

每个人都变得更专业化，大家都接受这样的做法，认为如果你在打造一个大型成功的东西，就必须这样做，其实不是这样。

Everybody becomes more specialized, and everybody accepts that that's the way you have to do business if you're building something big and successful, and it's not.

尤其是在你看到AI的发展时，绝对不是这样。

And it's definitely not when you look at what's happening with AI.

就在过去的一周里，我们在消费者业务中围绕智能体开展了某些工作。

And even just this past, if I look at the last week, we're working on some things around agents in our consumer business.