I've spent the last 20 years in Silicon Valley where I've worked as a operator running different companies, starting companies, and as an investor.

在過去的 20 年裡，我一直在硅谷工作，擔任過不同公司的運營者、創業者和投資者。

And I first ran a company that had about 120 people, it grew to 150, and then it shrank to 13 people in five rounds of layoffs.

我最初經營的公司大約有 120 人，後來發展到 150 人，又經過五輪裁員縮減到 13 人。

So I got laid off in the third round.

所以我在第三輪就被解僱了。

It was a telecom equipment company that was backed by Sequoia and Matrix.

這是一家電信設備公司，得到了紅杉和經緯的支持。

After that, I ended up joining Google and then left to start my first company, which is a data infrastructure company, which Twitter bought back in 2009.

之後，我加入了谷歌，然後離開谷歌創辦了我的第一家公司，這是一家數據基礎設施公司，Twitter 在 2009 年收購了這家公司。

At the time that Twitter bought my company, Twitter was about 90 people, and I helped scale it to 1,500 people over two and a half years.

在 Twitter 收購我的公司時，Twitter 大約只有 90 人，在我的幫助下，Twitter 在兩年半的時間裡發展到了 1500 人。

I started off running different product areas like Search, which again was very AI-centric, with Geo, and then moved over to effectively become a company fixer.

我一開始負責不同的產品領域，比如搜索，這也是非常以人工智能為中心的領域，還有 Geo，然後轉而成為公司的固定人員。

As an investor, I've backed about 40 companies worth a billion dollars or more each, 30-something at the seed or Series A.

作為一名投資者，我已經支持了約 40 家公司，每家公司的價值都在 10 億美元或以上，其中有 30 多家是種子公司或 A 輪公司。

So that's things like Airbnb, Airtable, Coinbase, Figma, Besto, Instacart, Stripe, Square, Dior, Rippling, Notion, and others.

這些公司包括 Airbnb、Airtable、Coinbase、Figma、Besto、Instacart、Stripe、Square、Dior、Rippling、Notion 等。

And then I also started investing in generative AI quite early, in part because I'd been doing AI and machine learning for so long.

我很早就開始投資生成式人工智能，部分原因是我從事人工智能和機器學習的時間太長了。

And so I was an early investor in companies like Perplexity, Hardy, Character, Mistral, Pica, and others.

是以，我是 Perplexity、Hardy、Character、Mistral、Pica 等公司的早期投資者。

I moved out for tech.

我搬出去是為了學技術。

I got my PhD at MIT.

我在麻省理工學院獲得了博士學位。

Before that, I'd studied mathematics and biology.

在此之前，我學過數學和生物學。

I moved out to Silicon Valley because I thought it was the best way to potentially have a about doing things that are useful.

我搬到硅谷，是因為我認為這是最好的方式，讓我有可能做一些有用的事情。

And so the reason I got a PhD, for example, was I thought initially that I'd be an academic and I'd really be focused on human disease and computational understanding of disease and other things like that.

是以，舉例來說，我獲得博士學位的原因是，我最初以為我會成為一名學者，真正專注於人類疾病和對疾病的計算理解，以及其他類似的東西。

I couldn't realize or felt that any one individual in the field of biology can have that big of an impact.

我無法意識到或感受到生物學領域的任何一個人都能產生如此大的影響。

I'd already done a background in mathematics for undergrad and had already done some CS-related things.

我在大學階段已經學習了數學，並做過一些與計算機科學有關的事情。

And so I moved out to Silicon Valley to work on companies with the hope that I could do something useful through the context of technology and company formation.

於是，我來到硅谷為公司工作，希望能通過技術和公司組建做些有用的事情。

Google in 2004 was growing extremely fast.

2004 年的谷歌發展極為迅速。

So when I joined, there was about 1500 or so people.

所以我加入時，大約有 1500 人左右。

It grew to 15,000 or so over three years, three and a half years.

在三年或三年半的時間裡，它增加到了 15 000 人左右。

And so the company was doubling every, I don't know, six to 12 months.

是以，公司每隔 6 到 12 個月就會翻一番。

And so they added 13,000 people in three years, which is amazing.

是以，他們在三年內增加了 1.3 萬人，這是令人驚歎的。

And so it's very chaotic.

是以，這裡非常混亂。

And I think in a good way, but also it's quite chaotic.

我認為這是好事，但也很混亂。

Every six months or so, there'd be a big reorg where they change the organizational structure because it was growing so fast that if you had 1500 and then 3000 and then 6000 and then 12,000 people, it was a different organization every year.

每隔六個月左右，公司就會進行一次大的重組，改變組織結構，因為公司發展太快了，如果你有 1500 人，然後是 3000 人，然後是 6000 人，然後是 12000 人，每年的組織結構都不一樣。

And so that needed a different org structure.

是以，需要一個不同的組織結構。

And so suddenly the person that was working for you became your boss's boss and they'd switch divisions and things would change so rapidly.

於是，為你工作的人突然成了你上司的上司，他們交換了部門，事情變化得如此之快。

And when I joined, Larry Page had just gotten rid of all the middle managers.

我加入的時候，拉里-佩奇剛剛把所有的中層管理人員都辭退了。

And so every director of engineering had somewhere between 50 and 100 people working for them, which meant that there was a big gray market for talent.

是以，每個工程總監都有 50 到 100 名員工為其工作，這意味著存在著一個巨大的人才灰色市場。

A lot of managers didn't know what their employees were doing because they had too many employees working for them directly.

很多經理不知道員工在做什麼，因為他們有太多的員工直接為他們工作。

So they couldn't do enough one-on-ones.

是以，他們無法進行足夠的一對一訓練。

If you have a hundred people working for you, you can't see them that often individually.

如果有上百人為你工作，你不可能經常單獨見他們。

And so a lot of people lost track of who was working for them, which meant there was an enormous gray market for talent.

是以，很多人都不知道誰在為他們工作，這意味著有一個巨大的人才灰色市場。

And so one of the ways I helped start up a lot of the mobile efforts is I convinced people to work on mobile when their managers didn't know about it.

是以，我幫助啟動許多移動工作的方法之一，就是在人們的經理還不瞭解移動工作時，說服他們從事移動工作。

And so that was how a lot of the early mobile team got running.

是以，很多早期的移動團隊就是這樣開始運作的。

I think there's sort of personal lessons and there's professional lessons and maybe they're the same thing.

我認為，有個人的教訓，也有專業的教訓，也許它們是一回事。

I think one big lesson was that networks in Silicon Valley really matter and small groups of people who are really at different companies go on to achieve amazing things together.

我認為一個重要的教訓是，硅谷的人際網絡真的很重要，真正在不同公司工作的一小群人在一起會取得驚人的成就。

And so the famous example of that may be PayPal.

PayPal 就是一個著名的例子。

Peter Thiel started Palantir and Founders Fund and all these things.

彼得-蒂爾創辦了 Palantir 和 Founders Fund 等公司。

Keith Rupp always started multiple companies as well as became a prominent investor.

基思-魯普總是創辦多家公司，併成為著名的投資者。

Max Levchin started a firm.

Max Levchin 創辦了一家公司。

Elon Musk, of course, went and did so many things between Tesla and SpaceX and et cetera.

當然，埃隆-馬斯克在特斯拉和 SpaceX 等公司之間做了很多事情。

Reid Hoffman started LinkedIn.

裡德-霍夫曼創辦了 LinkedIn。

There's a small group of people that were at the core of PayPal that did all these things.

PayPal的核心團隊中有一小群人做了所有這些事情。

David Sachs was another person.

大衛-薩克斯是另一個人。

Similarly, the early Google network both went off to start a lot of different companies, Pinterest and Instagram and a variety of other things like that, but they also ended up as very prominent operators in Silicon Valley.

同樣，早期的谷歌網絡都去創辦了很多不同的公司，如 Pinterest、Instagram 和其他類似的公司，但他們最終也成為了硅谷非常傑出的經營者。

So a lot of the COOs in Silicon Valley for the next generation of companies came out of Google.

是以，硅谷新一代公司的許多首席運營官都來自谷歌。

That's Sheryl Sandberg, the CEO of Facebook, Dennis Woodside, the CEO of Dropbox, the COOs of Stripe and Gusto and all sorts of companies came out of Google.

這就是 Facebook 首席執行官雪莉-桑德伯格（Sheryl Sandberg）、Dropbox 首席執行官丹尼斯-伍德賽德（Dennis Woodside）、Stripe 和 Gusto 的首席運營官以及從谷歌走出來的各種公司。

And so you similarly see this sort of dispersion of talent, but everybody knew each other.

是以，你同樣可以看到這種人才分散的現象，但每個人都互相認識。

And so it created a lot of collaboration across companies because of that.

正因為如此，公司之間產生了很多合作。

And so number one is just these small pockets of people who work together repeatedly really matter.

是以，第一條就是這些小範圍的人一起努力，反覆合作，這才是真正重要的。

That was one big takeaway.

這是一個很大的收穫。

I think the second takeaway was really the power of ambition.

我認為第二個啟示是野心的力量。

The main reason Larry Page or Sergey Brin would get upset with you or Eric Schmidt, they kind of ran the company together, would get upset with you is either number one, you weren't thinking big enough.

拉里-佩奇（Larry Page）或謝爾蓋-布林（Sergey Brin）或埃裡克-施密特（Eric Schmidt）會對你不滿的主要原因是：第一，你想得不夠遠大。

You'd go into a product review and you'd say, I have this idea for this hundred million dollar market.

你在做產品評論時會說，我對這個價值一億美元的市場有一個想法。

And they'd say, why is it in a $10 billion market?

他們會說，為什麼是在一個 100 億美元的市場？

Like literally they would say that, you know, they'd always push you on thinking bigger.

就像字面上的意思一樣，他們會說，你知道，他們總是鼓勵你想得更遠。

Or why are you happy with a million people using it?

或者說，為什麼你對一百萬人使用它感到高興？

Why isn't it a hundred million people using it?

為什麼沒有一億人使用？

And then number two is they pushed very hard on the innovation and what makes the thing that you're working on very special or different.

其次，他們非常注重創新，以及是什麼讓你所從事的工作非常特別或與眾不同。

It really forced some very clear product thinking and some very big thinking and ambitious thinking.

這確實迫使我們進行一些非常清晰的產品思考，以及一些非常宏大和雄心勃勃的思考。

And then the third is when you're an employee at a hyper-growth startup or a company, things will generally work out and you shouldn't worry too much about yourself.

第三，當你是一家高速成長的初創企業或公司的員工時，一般情況下事情都會迎刃而解，你不應該太擔心自己。

And if your boss suddenly, if your coworker suddenly becomes your boss's boss, then all these things shift.

如果你的老闆突然變成了你的同事，如果你的同事突然變成了你老闆的老闆，那麼所有這些事情都會發生變化。

A lot of people start to worry about themselves.

很多人開始擔心自己。

What about my career and what will happen to me?

我的職業生涯怎麼辦？

That the company's growing and changing so rapidly that there's lots of opportunity.

公司的發展和變化如此之快，機會很多。

And something that was sort of my third takeaway.

這也是我的第三個收穫。

I always wanted to start a company.

我一直想開一家公司。

And so going into Google, I thought someday I'll start a company.

所以在進入谷歌之前，我就想有一天我會開一家公司。

And I thought that was important to me.

我認為這對我很重要。

At the time there was no Y Combinator.

當時還沒有 Y Combinator。

And so you couldn't just apply and show up and get money and start working on things.

是以，你不能只是提出申請，然後拿著錢就開始工作。

And there weren't a lot of online resources.

網上資源也不多。

And so it was really hard to actually find information about how to start a company.

是以，要找到如何創辦公司的資訊真的很難。

And that there weren't very many angel funds, operators with side funds, operators investing.

天使基金、擁有副基金的經營者、投資經營者並不多。

The venture ecosystem was very constrained.

創業生態系統非常有限。

And so there was a small number of firms that you could raise money from.

是以，你只能從少數幾家公司籌集資金。

And then there were some people pioneering, you know, angel funds or what were then called super angels.

還有一些人開創了天使基金或當時所謂的超級天使。

So that was Naval Robicont and Mike Maples and Michael Deering and a few other folks.

這就是海軍羅比康特、邁克-梅普爾斯、邁克爾-迪林和其他一些人。

But the reality was there just wasn't that much money either.

但現實情況是，也沒有那麼多錢。

And so it was hard to get information.

是以很難獲得資訊。

It was hard to get money and it was hard to get going.

錢不好掙，路不好走。

I think today it's actually dramatically easier to start a company and there's a lot more information about how to do it.

我認為，如今創辦一家公司實際上要容易得多，關於如何創辦公司的資訊也多得多。

But back then it was harder.

但那時更難。

So I went to Google in part because I thought it'd be a great place to transition, meet some people and maybe start something with them.

所以我去了谷歌，部分原因是我認為這是一個過渡的好地方，可以認識一些人，也許還能和他們一起做些什麼。

Yeah, we had two products.

是的，我們有兩種產品。

We started off with a site focused on location related information.

我們最初建立的網站主要提供與地點相關的資訊。

And then that morphed into an API focused on location information.

後來，它演變成了一個以位置資訊為重點的應用程序接口。

And we ended up eventually building one of the really early developer centric products.

我們最終打造了一款真正以開發者為中心的早期產品。

So at the time there wasn't a lot happening in terms of developer first products or product like growth or all these things that are very common now.

是以，在開發者首創產品、產品增長或所有這些現在非常普遍的事情方面，當時還沒有發生很多事情。

There are very few API centric products like Stripe or things like that.

像 Stripe 這樣以應用程序接口為中心的產品很少。

The only one I think that really existed and was taking off when we started was Twilio.

我認為，在我們開始創業時，唯一真正存在並正在起飛的是 Twilio。

And so building like a product as an API was kind of a radical concept in those days.

是以，將產品作為應用程序接口來構建，在當時是一種激進的概念。

And so we were one of the really early people building that.

是以，我們是較早建立該系統的公司之一。

And basically we for people to sort of interact with geodata in really rich ways, generate their own custom index and search over it, annotate location related data sets and things like that.

基本上，我們可以讓人們以非常豐富的方式與地理數據進行交互，生成自己的自定義索引並對其進行搜索，註釋與位置相關的數據集等。

So initially when we started the company, it was a team of ex-Google people, seven or eight employees.

是以，公司成立之初，我們的團隊只有七八名前谷歌員工。

We had all of them had worked at Google except for one person or something, or maybe two people.

我們所有的人都曾在谷歌工作過，除了一個人或其他什麼人，或許是兩個人。

And I think in hindsight, we should have had more diversity on the team early on in terms of where people came from because we did things the Google way, which we built very big infrastructure from day one that worked at enormous scale.

現在回想起來，我們應該在早期就在團隊中加入更多不同背景的人，因為我們是按照谷歌的方式做事的，我們從第一天起就建立了非常龐大的基礎架構，可以在巨大的規模上運行。

But if you don't have many users, it doesn't matter the skill that you have.

但是，如果你的用戶數量不多，那麼你所掌握的技能也就無關緊要了。

The positive of that is it made it very easy for us to do API related products and things like that because we designed it in a very modular way from day one.

這樣做的好處是，我們從一開始就採用模塊化設計，是以可以非常容易地開發與應用程序接口相關的產品等。

And the initial product that we had was more like a website where you could go and interact with location related information and then it morphed into the API.

我們最初的產品更像是一個網站，在這裡你可以與位置相關資訊進行互動，後來它演變成了應用程序接口。

And I think the mistake we had with the first product, and all we really did was change how you display the information.

我認為我們在第一款產品上犯了一個錯誤，我們所做的只是改變了顯示信息的方式。

Is it programmatic through an API or is it on a website?

是通過應用程序接口進行編程，還是在網站上進行編程？

I think the mistake that we made was we were building for the past in some sense, right?

我認為我們犯的錯誤是，從某種意義上說，我們是在為過去而建設，對嗎？

We just come out of Google, the SEO world.

我們剛剛走出谷歌，走出搜索引擎優化的世界。

The second product that we had was more for the modern world of the future.

我們的第二款產品更適合未來的現代世界。

People are moving towards using APIs to integrate against their products.

人們越來越傾向於使用應用程序接口來集成自己的產品。

People are moving towards mobile.

人們正在轉向移動設備。

People are moving towards location services.

人們正在轉向位置服務。

And so I think a lot of mistakes that people make when they come out of a big company is they build for the era that the company that they're at was built in versus building for the future.

是以，我認為當人們從大公司出來時，他們會犯的很多錯誤是，他們為自己所在的公司所處的時代而建設，而不是為未來而建設。

And you always want to build a year or two into the future in terms of what's coming because you really want to take advantage of what is known as a why now statement.

你總是想在未來一兩年內預測未來，因為你真的想利用所謂的 "為什麼現在 "聲明。

What has changed that makes your company relevant today where it couldn't necessarily exist before?

是什麼改變了貴公司，讓它在以前不一定存在的今天變得如此重要？

And so the why now could be technology.

是以，現在的原因可能是技術。

The why now could be regulatory.

現在的原因可能是監管問題。

The why now could be something else.

現在的原因可能是別的。

So an example of a technology why now is Uber.

Uber 就是一個很好的例子。

Before you had the iPhone and effectively free lookups of GPS, it cost a dollar every time you do a GPS lookup for a mobile carrier.

在有 iPhone 和免費 GPS 查詢之前，每次為移動運營商進行 GPS 查詢都要花一美元。

And so when I was doing mobile at Google, all the carriers, Verizon and SoftBank and everybody else felt that nobody wanted to use location.

是以，當我在谷歌做移動業務時，所有運營商、Verizon 和軟銀以及其他所有人都認為，沒有人願意使用定位功能。

And the reason was they charged a dollar per lookup, which meant if you were doing an Uber style app, the trace would cost you $30 just to follow the car to your location.

原因是他們對每次查詢收取一美元的費用，這意味著如果你使用的是 Uber 式的應用程序，僅跟蹤汽車到你的位置就需要花費 30 美元。

So you couldn't do it.

所以你做不到。

When the iPhone came out, it broke open that hole that the carriers had, which suddenly meant things like Uber could exist.

當 iPhone 問世時，它打破了運營商的壟斷，這意味著 Uber 這樣的公司突然可以存在了。

You could push a button on your phone and a stranger in a car would show up and get in and feel okay with it.

你可以在手機上按下一個按鈕，就會有一個陌生人開車出現，然後上車，並覺得沒問題。

And then you'd see the wrap.

然後你就會看到包裝。

So that was a technology why now.

這就是為什麼現在要採用這項技術的原因。

Suddenly you had new mobile services.

突然之間，你有了新的移動服務。

A regulatory why now may be things like some SARA really grew on the backs of regulation, which forced you to have in-cab cameras.

現在的監管原因可能是一些 SARA 真的是在監管的支持下發展起來的，這迫使你必須安裝車內攝像頭。

If you're driving a truck across the country in the US, you need to be able to monitor the driver.

如果你在美國駕駛一輛卡車橫穿全國，你需要能夠監控司機。

And so it created an opening for their technology or their maybe regulatory shifts for healthcare or for other things that allow you then to build a company.

是以，這就為他們的技術或醫療保健或其他方面的監管轉變創造了機會，讓你可以建立一家公司。

And so often you want to kind of ask, what is different now?

是以，你經常會問，現在有什麼不同？

And therefore, why is what I'm doing relevant for today?

是以，為什麼我所做的與今天有關？

Maybe the why now for Anduril would be the shift towards drones and machine vision as in new capability, right?

對 Anduril 而言，現在的原因可能是無人機和機器視覺作為新能力的轉變，對嗎？

It's a technology shift.

這是一種技術轉變。

So something Anduril could exist.

所以，安杜里爾可能存在。

So we basically had started building the API centric product and we were talking with Twitter as a customer.

是以，我們基本上已經開始構建以 API 為中心的產品，並將 Twitter 作為我們的客戶。

And one of the folks on the Twitter team sent me an email.

推特團隊的一位成員給我發了一封電子郵件。

I think it was Ryan Sarver that basically said, why don't you come in for a constraints of being outside of Twitter removed and let's brainstorm the product that you could build for Twitter.

我想基本上是瑞安-薩弗說的，你為什麼不跳出 Twitter 的束縛，讓我們集思廣益，為 Twitter 打造一款產品。

And it was meant to be a vendor discussion, but we could tell that they were kind of thinking about buying us because they wanted to incorporate the product into their platform and ecosystem.

這本來是一次供應商討論，但我們可以看出，他們在考慮收購我們，因為他們想把產品納入他們的平臺和生態系統。

And at the time, Twitter had a massive developer ecosystem.

當時，Twitter 擁有一個龐大的開發者生態系統。

It had just opened up API access for everyone to access tweets and other information.

它剛剛開放了 API 訪問，讓每個人都能訪問推文和其他資訊。

And they thought that we could be an ancillary service that would reside on top of Twitter and would be something every developer could use, right?

他們認為，我們可以成為 Twitter 上的一項輔助服務，讓每個開發者都能使用，對嗎？

Which was a very compelling vision.

這是一個非常引人注目的願景。

So we met with Ryan and started talking through that.

於是，我們與瑞安會面，開始討論這個問題。

And then we met with Ev and other members of the team at the time.

然後，我們會見了埃夫和當時團隊的其他成員。

Ev Williams was still the CEO of the company.

Ev Williams 依然是公司的首席執行官。

And at the time that we started talking with them, there were about 50 people.

在我們開始與他們交談時，大約有 50 人。

And by the time they bought us, which is about three months later, there were 90 people, right?

到他們收購我們的時候，也就是大約三個月之後，已經有 90 人了，對吧？

So they almost doubled in three months.

是以，它們在三個月內幾乎翻了一番。

So they're going very fast.

是以，它們的速度非常快。

Basically, we thought about it after they made an offer and we decided it'd be a very good home for what we were doing for three reasons.

基本上，在他們出價之後，我們考慮了一下，我們認為這是一個非常適合我們工作的房子，原因有三個。

Number one is they had this massive developer ecosystem.

第一，他們擁有龐大的開發者生態系統。

So we thought that we could reach more users faster and have more impact.

是以，我們認為我們可以更快地接觸到更多用戶，產生更大的影響。

Second, we thought it was a great place for our engineers to go because Twitter was a small team still and we thought all of them could have really amazing careers there.

其次，我們認為這是工程師們的好去處，因為 Twitter 還是一個小團隊，我們認為他們都能在那裡獲得非常出色的職業發展。

Third, we thought financially it was a good outcome because at the time Twitter was worth a billion dollars.

第三，我們認為這在經濟上是個好結果，因為當時 Twitter 的市值高達 10 億美元。

And we thought, oh, it's going to be at least a $10 billion company.

我們想，哦，這至少會是一家市值 100 億美元的公司。

And so any purchase price we have is magnified by 10 times.