软件终将吞噬服务 - 亚伦·列维

这项技术作为消费科技的普及，然后渗透到专业消费者领域，其程度超出了我以往所见的一切。难以置信。我认为它将从根本上改变人们的日常模式。

The universal adoption of this as a consumer technology and then bleeding into prosumer is it exceeds anything I've ever It's unbelievable. Experienced, and I think it is it will just fundamentally change people's sort of daily pattern.

这些都是早期采用者，而早期采用者对有意犯下的错误非常宽容。当某样东西全新出现时，围绕它的文化就会发展起来。早期的互联网用户并不抱怨网速慢。

This is all early adopters, and early adopters are very forgiving of mistakes on purpose. When something is brand new, a culture around it develops. The early Internet people didn't complain that the Internet was slow.

没错。使用AI的更资深小团队简直是超人。

Right. The more senior small teams that use AI are superhuman.

是的。

Yeah.

是的。就像他们一觉醒来都他妈成了托尼·斯塔克。难以置信。而且，他们的生产力高得离谱。

Yeah. It's like they woke up and they were all fucking Tony Stark. Is unbelievable. And, like, their productivity is insane.

美国应该对H-1B签证定价吗？还是那样会排斥新人才？AI编程助手是真的在提升生产力还是只是炒作？在这场AI平台变革中，谁会胜出？是老牌企业还是新兴的AI原生初创公司？今天，我与Box CEO Aaron Levy，以及a16z的Steven Sinofsky和Martin Casado坐下来辩论H-1B改革，为什么Box现在三分之一的代码由AI生成，从编写代码转向审查代码，以及为什么自下而上的AI工具胜过自上而下的试点项目。

Should The US put a price on h one b visas, or would that shut out new talent? Are AI coding agents truly boosting productivity or just hype? And in this AI platform shift, who wins? Incumbents or new AI native startups? Today, I sit down with Box CEO Aaron Levy alongside a sixteen z's Steven Sinofsky and Martin Casado to debate h one b reform, why Box now ships a third of its code from AI, the move from writing to reviewing code, and why bottom up AI tools beat top down pilots.

让我们开始吧。

Let's get into it.

首先，我想评论一下。你在群聊里发的关于自闭症新闻更新了你的PDOM。是的。

First, I just wanna comment. You posted in the group chat that the news around autism updates your PDOM. Yes.

不过，只有当你展示图片时才有效，所以你得做叠加处理才能让它说得通。但用那个福克斯新闻的标题可以做很多梗图。

It only works if you show the image, though, so you'll have to do the overlay to make that make sense. But there's so many memes you can do with with that Fox Fox News headline.

没错，正是如此。好吧，首先，我想谈谈移民新闻。

So Exactly. Well, first, I wanna get into the immigration news.

哦，你真的想就这么开始，像是，

Oh, you really wanna kick off just, like,

真的要把这整件事都启动起来。

really start this whole stuff.

是的。就像，这是个关键点。完全正确。Martine，你有一些有趣的反应。

Yeah. Like, it's a blood point. Exactly. Martine, you had some interesting reaction.

是的。完全正确。完美。

Yeah. Exactly. Perfect.

请，什么

Please, what

你对这项政策的看法和反应是什么？

were your reactions to what you think of the policy?

嗯，这很有趣，因为看起来每当政府触及移民问题，就会引起巨大的、本能的强烈反对。我们甚至从风投那里也看到了很多这样的反应。但同样非常有趣的是，里德·哈斯廷斯，一个典型的左派人士，长期以来一直参与移民政策，他表示：'我从事移民政策工作已经三十年了，这是正确的做法。'这也是我的想法，即这个系统长期以来一直被钻空子。由于抽签制度，初创公司很难招聘人才。

Well, it was interesting because it seems like anytime the administration touches immigration, there's a huge outcry, knee jerk outcry. And we saw a lot of that from VCs even. But it's also very interesting that Reed Hastings, is a classic lefty and has long been, was like, I've been in doing policy for immigration for thirty years, and this is the right approach. And this is very much my thought, which is this system has been game for a very long time. It's very hard for startups to hire because of the lottery system.

它被大公司、咨询公司、亚马逊和谷歌垄断了，这种情况必须改变。我认为一个非常合理的方法是通过定价来解决，因为你有一个市场，需要分配供应。价格是一个很好的方式。所以我对此非常非常支持。我评论了这一点，但似乎很多人不同意。

It's locked up by the large companies, the consultants, Amazon, and Google, and that has to change. And I think a very reasonable way to do it is to set price because you've got a market and you need to allocate supply. Price is a great way to do it. So I'm very, very positive on it. I comment about that, and a lot of people seem to disagree.

所以我认为这是一个活跃的讨论。

So I think it's an active discussion.

嗯，我认为这有几个方面。首先，里德最终回应的实际上已经不是当时的实际政策了。是的。他说每年10万美元是一个很好的政策，但显然，互联网已经向前发展了。对我来说，这并不明显。

Well, I think there's a couple elements to this. So one is, first of all, Reid was ultimately responding to a thing that was no longer the actual policy. Yeah. So he said 100 k a year was a great policy, and obviously, the Internet had moved on. It's not to me obvious.

我不会得出与你相同的结论，因为我认为在这种情况下，亚马逊和谷歌可能会占据绝大部分人才。所以我不清楚初创公司是否能从中受益。好吧，退一步说，就这个具体的实施方案而言

I wouldn't conclude the same outcome that you just concluded in that I think that you'd have a situation where the Amazons and Googles would probably actually capture the vast portion of the talent in this situation. So it's not clear to me that, like, startups sort of come out ahead Well, well, one step back. So from this particular implementation

也许亚马逊和谷歌更容易监管，但确实有许多咨询机构对价格敏感。是的，百分之百。

Maybe Amazon and Google who are probably more easy to regulate, but there are a number of organizations that are consultancy Yes. 100%. That actually are price sensitive

是的。

Yes.

这会让它们受到挤压。我认为既然它们位列前15名，大概占了四五个，那将是一个显著的释放。是的，或者更高层次。

That would be squeezed by this. I I would think that given that they're in the top 15, they make up, like, four or five of them. That would be a significant freeing up Yeah. Or a higher level.

我的想法是，如果你能把所有对此话题有看法的人都聚在一个房间里，但同时也让实践者和技术人员在场，还有，比如说，最...你甚至不能说右翼，因为实际上我不认为这甚至是典型的共和党立场。所以就像不，不，不。把所有人都聚在房间里，他问我们在优化什么？

I think the my thing would be, like, if you could just get all the people in the room that have an opinion on this topic and you but you actually have the practitioners and tech in the room as well and the, let's say, the most kind of you know, you can't even say, like, right wing because, actually, I don't think this is even classic Republican. So it's just like No. No. No. The the polls, you got everybody in a room, and he says he's sort of saying, what are we optimizing for?

我们是在优化以避免工资下降吗？这是个有趣的点。我们是在优化以确保某种工作不会流向特定美国人群？还是优化只让全球最顶尖的人才来这里？这些全都是完全不同的优化目标。

Are we optimizing for we don't want to have wages go down? That's an interesting thing. Are we optimizing for a particular kind of job not going to, let's say, certain populations of Americans? Are we optimizing for just ensuring that we only have the highest merit people on the planet coming here? Like, those are all totally different kind of goals to optimize for.

我认为最终形成的框架和体系应该有一个连贯的策略作为支撑。我的策略是：我们希望全球最优秀的人才来到这里。是的。但具体数量并不固定，有些年份可能是5000人。

And I think that the framework you end up with and the system that you end up with should probably hopefully have, like, a cohesive sort of strategy behind it. My strategy would be we want the absolute best in the world here. Yep. There's not exactly clear that there's a fixed number on that. Some years, there might be 5,000.

有些年份可能是5万人，有些年份可能是8万人。

Some years, there might be 50,000. Some years, there might be 80,000.

是的。

Yep.

我们可能希望他们对工资产生净正面影响。

We probably want them to be net positive to wages.

是的。

Yep.

所以我们同意，在任何特定行业或地区，工资应该随着这个人才库的增长而上升，而不是下降。我认为这实际上是完全合理的。你说应该有某种市场动态机制，不应该能够操纵和剥削人才库，比如现在在底特律，我们可以因为能够离岸外包而淘汰IT工作。我认为你可以建立一个系统，基本上满足所有这些目标，同时确保能够获得那些去读硕士项目的人，比如你们州的州立大学毕业生。他们毕业后出来。

So let's agree that, you know, in any given industry or locale, wages should go up with this talent pool as opposed to down. So I think that's actually totally reasonable. And you say you should have the market kind of sort of some market dynamic to that, and you shouldn't be able to kind of game and exploit the talent pools for saying, now in Detroit, we can go wipe out IT jobs because we can go and offshore those. Like, I think you could build a system that basically meets all of those goals while still ensuring that you can get somebody that goes to their master's program and name your state school. They come out of it.

他们是AI工程师。虽然Meta不会付给他们1亿美元，但他们将成为我们经济中非常有价值的贡献者。这完全是正和游戏。不是从别人那里抢走工作。它使我们更具竞争力。

They're an AI engineer. They're not yet at the sort of meta is going to pay them $100,000,000 but they are gonna be totally valuable contributors to our economy. It's all sort of positive sum. It's not taking a job from anybody else. It makes us more competitive.

我认为有一种方法可以在不过度施加限制的情况下做到这一点，这些限制可能会让初创公司无法在经济上可行地参与其中。我认为每年10万美元的门槛会直接影响初创公司。

And I think there's a way to do that without sort of overly, let's say, putting constraints in the system that make it maybe so a start up wouldn't be able to kind of economically, viably participate in this. And I think a 100 k per year would be at a point where the start ups would be directly impacted.

与很多初创公司合作。我不确定那是

Working with a lot of start ups. I'm not sure that's the

尊重地说，Andrejs Moritz 所见的初创企业类型并不能代表世界上所有初创企业的基础。

Respectfully respectfully, the kind of start ups that Andrejs Moritz sees are not all of the base of startups in the world.

所以你可以对数字吹毛求疵。是20吗？当我和Keith Rabois说这个数字很合理时，他说是20，还是10万？我不知道。但是，就像，这个想法是你等等。

So you could quibble about the number. Is it 20, which Keith Rabois said when I tell him it very sensible, is it a 100 k? I don't know. But, like, the idea that you Wait.

Keith 提出了20？

Keith threw out 20?

Keith 提出了20。

Keith threw out 20.

好吧，我们就用Keith的数字。如果Keith提出了20，我认为我们可以接受Keith的数字。

Well, let's just go with Keith's number. Think that if Keith threw out 20, I think we can be good with a Keith number.

但我认为数字很容易让人纠结。是的。但你必须始终关注这个数字替代的是什么。我认为除了真正从事这项工作的人之外，普通人在进行这场辩论时并不了解在这种体系下工作损失了多少生产力。是的。

But I I think the number it's easy to fixate on the number. Yeah. But you have to always look at what is the number replacing. And I don't think the average person having this debate other than the people that really work at this have any idea the amount of productivity that is lost working this system. Yes.

意思是，资源量。当然，你提到的大公司有庞大的团队，他们花费所有精力，是的，就像，实际上，本质上，就像游说者一样，是的，运作这个体系。然后这背后的就是所有的理由和所有的管理，哦，是的。

Mean, amount of resources. And, of course, the bigger companies, the ones that you mentioned, have enormous teams that spend all of their energy Yeah. Like, literally, essentially, as lobbyists Yeah. Working the system. And then the back end of that is are all the justifications and all of the management Oh, yeah.

这是，现在他们刚刚部署了

This is And now they've just deployed

系统程序。

Systems program.

他们投入所有资源来管理，比如内部呼叫中心，处理让员工返回美国的事务，专门应对这个问题。你提到了我认为在这场讨论中非常重要却被忽视的一点：毫无疑问，在过去大约25年里，大型科技公司确实走上了一条分叉路径——办公室人员主要从25或30个大学院系招聘，而其他人则觉得，直接从这八个国际地点和学校大规模招聘要容易得多。我认为很多人忽略了这一点：科技行业的一大特点是，如果你看看英特尔员工毕业的院校，或者硅谷的历史，其实都是来自美国中部各类学校的人才。

All their resources to manage, like, in house call centers to deal with getting their employees back to The United States to just deal with it. I do you hit on one point that I think is really important to this debate that is sort of getting lost, which is there's no doubt that within the big tech world that they for the past twenty five years or so, they really went on this sort of bifurcated curve, which is hiring for the people in the office focusing on, say, 25 or 30 university departments. And then, basically, everybody else was like, well, it's so much easier if we just hire huge numbers of people from these eight international locations and schools. And I think a lot of this is missing that a big part of this is, well, the tech industry like, if you look at Intel and where they all went to college, and if you look at the history of Silicon Valley, it's all these people from all the schools in the middle of the country.

嗯，是的。

Mhmm. Yeah.

这些学校如今都不是主流科技公司的目标招聘院校。我认为大公司在这方面有些懒惰。作为一个花了数十年飞往这些学校招聘的人，我觉得大学方面没有努力提升项目质量，大公司也没有明确说明为何停止招聘或看不到招聘数据。这种改变对所有人都会更有利。

None of which are like the target recruiting schools by the main tech companies these days. And that's been a place where I think that the big companies have been somewhat lazy. And as a person who spent decades flying to all of these schools and recruiting, there's work that this the universities have have not done to be better programs and that there's work that the big companies have not done to be clear what it is that why they've stopped recruiting or haven't had seen the numbers. And that's change would be better for everybody to really make.

所以我的预期是受影响的实际是另一类完全不同的工作岗位——比如今天在佛罗里达想找一份10万美元的IT工作，根本找不到。对吧？我认为这才是受大型咨询公司影响最直接的领域。

I so my my expectation when it gets impacted is kind of an even different job set than that, which is if you go to Florida today and you try and get, like, an IT job for a 100 k, you just can't. Right? And so that I think is actually the area that's the most directly impacted by the large consultants.

意思是没有10万美元薪资的工作，还是根本找不到工作？

Meaning there aren't jobs that pay a 100 k or that you just can't find a job?

这些职位已经被占据了。它们就在那里，但就是不存在了。就像任何IT管理员、基础服务、咨询类工作等等，所有这些都已经饱和了。在美国大部分地区，要找到一份8万到12万美元的工作非常困难，就是因为这个原因。

They're taken. They're out there. They're out just don't exist. Like like any sort of IT administrator, like the services, like the basic consulting gigs, like all of that has been saturated. It's very, tough to get a job between, like, 80 and a 120 k in in much of The United States because of this.

对吧？所以这不是关于新毕业生成为软件工程师的问题，因为现实是软件工程师的终身预期价值足够高，我认为市场会自行调节。但真正被挤压出去的是这些较低层次的IT管理类工作。听着，如果我们想把这些工作带回来，我认为我们必须改变定价策略，而不是像许多公司那样进行套利。

Right? And so this isn't about the new grad being a software engineer because the reality is the expected value of a software engineer over their lifetime is high enough that I think that, like, the market kind of navigates that. But it's almost these kind of lower level more, like, IT admin jobs that have been squeezed out. And listen. If we want to bring them back, then I do think and we don't wanna do this kind of arbitrage that a lot of these companies are doing, that we're gonna have to change the pricing.

但是，设定最低薪资标准不就能有效解决这个问题吗？

So but don't but the wouldn't a minimum salary band effectively solve that problem for you?

当然。是的。

Sure. Yeah.

是的。肯定可以。

Yeah. For sure.

是的。我认为有很多机制可以实现这一点。

Yeah. I think there's a lot of mechanisms to do it.

好的。好的。

Okay. Okay.

我其实同意你的观点。我们应该真正讨论我们要解决的问题。是的，我认为我们都会同意。如果你来美国并获得大学学位，你应该获得签证。

I actually agree with you. Like, we should actually talk about the problem that we're trying to solve. So Yeah. I think we would all agree. If you come to The United States, you get a college degree, you should have a visa.

对吧？我们

Right? We

差不多吧，你知道，好吧。所以这是肯定的。

kinda you know, okay. So that's for sure.

但我们都同意这一点吗？

But do we all agree on that?

好的。是的。我同意。好的。

Okay. Yes. I agree. Okay.

所以这个法案，我认为提出这个策略的人实际上并不都同意这一点。

So the bill as in I don't think that the people proposing this strategy actually agree on that.

嗯，特朗普曾有名地这样说过。

Well, so Trump famously said this.

哦，他说过很多著名的话。

Oh, he famously said a lot

确实如此。他肯定说过。然后他又收回了这些话。

of it. He, for sure. And then he unsaid it.

是的，没错。我的意思是，但如果我们都同意，并且有一个政策规定，如果你从印度理工学院毕业然后来到堪萨斯州立大学，那么你就能得到一份工作，那这场辩论就不会有趣了。但你可以把

Yeah. Exactly. I mean But, like, this wouldn't be an interesting debate. If we all agree and we have a policy that says, if you go from IIT and then you come to Kansas State, then then you get a job. But you can put

这一点应用到工作中。这变成了系统游戏化的一部分，因为你会先完成印度理工学院的学业，然后获得公司赞助去某个学校攻读硕士学位。所以它并没有真正实现像就读四年制大学那样投资来美国的目标。

that in work. That became part of the gamification of the system because you would do the IIT thing, then you would get a company sponsorship for a master's degree at some school. So it didn't really accomplish the goal of investing in coming to The US the same way that going to a four year school would have done.

等等，等等，等等。为什么会这样？问题是什么？

Wait. Wait. Wait. Why is that? What was

问题在于你最终是由公司赞助的。是的。这完全改变了动态：是你主动寻求来美国，还是公司把你吸引到美国？这有点不同。

the problem? Because you ended up getting sponsored by a company. Yeah. And it changed the whole dynamic of were you seeking out The US, or did a company pull you to The US? It's a little bit different.

好吧。但我也经常觉得，这类讨论往往会像现在这样，我们关注的是应届毕业的软件工程师。而实际上我并不认为受影响的是这个群体。真的，我不这么认为。

Okay. I But also think so often this discussion goes the direction this one is going on is we focus on, like, new grad software engineers. And I actually don't think that what's is getting impacted. Right. I really don't.

我确实认为是这样。比如，咨询公司是做什么的？他们做的是行政工作、IT工作，只是薪资档次不同而已。

I really think it is. Like, what do the consulting shops do? They do Yeah. Admin work, IT work, and it's just a different salary band.

是的。我认为这些是...

Yeah. I think it's I think these are the

这些就是人力外包公司。没错。而且，像这样的方法直接针对它们。我认为这种方式实际上相当积极。

These are body shops. Yeah. And, like, an approach like this directly targets them. I think in a way that's actually quite positive.

是的。我确实认为...我的意思是，我反而会支持基思的方法，因为我觉得存在一个临界值，超过这个数值后，你就是在为了能够...是的，而在业务上做出其他权衡。

Yeah. I do think there's then I mean, I would just then favor Keith's approach because I think there there is a number in which it becomes you're then making other trade offs in your business just to be able to Yeah.

是的。是的。10万美元这个数字是...

Yeah. Yeah. The 100 k number is

嗯，我完全同意关于金额的观点，但我也想强调一点：整个系统确实可以避免这种巨大的成本和不确定性。是的。任何解决方案如果真的解决了这个问题...

Well, the I do totally agree with that the dollars but I do think that just to reinforce this that that the whole system can do without this immense cost and uncertainty. Yep. And any solution should really be if you address that

是的。

Yeah.

那么其余的整个动态就会随之而来。但只要它是一个庞大、复杂且昂贵的系统，是的。那么大公司将继续从中获得不成比例的好处。

Then the whole the rest of the dynamics will follow. But as long as it's a huge, complicated, expensive system Yep. Then the big companies are gonna continue to benefit from it disproportionately.

是的。而且我现在告诉你，对于初创公司来说，应对抽签制度要困难得多。

Yeah. And I will tell you right now, like, it is much harder for a startup to deal with the lottery system

是的。比那会更难。

Yes. Than it would be

比他们支付10万美元要难。至少对初创公司来说是这样。但我

than it would be for them to pay a 100 k. Like, at least for the startups, though. But I

不认为它改变了抽签制度。这是一场狩猎。对吧？对。它没有改变。参与抽签制度需要10万美元。

don't think it changed the lottery system. It's a hunt. Right? Right. It didn't It's a 100 k to participate in the lottery system.

对。

Right.

哦，不。不。我明白。我的意思是，希望希望希望，当然。比如，数字会改变很多人的计算方式。

Oh, no. No. I understand. I mean, hope hope hope hopefully Sure. Like, number it'll change the calculus of a lot of people

那些是

that are

在抽签系统中。

in the lottery system.

我只是认为

I I just think that

但是我

But I

更倾向于取消抽签系统。

would prefer to remove the lottery system.

是的。我认为你完全可以建立一个系统，对我们和任何其他人来说都有一个共同的定义，都会说这显然是一个高价值的工作。它将会提高这个特定领域的平均工资。而且我们想确保我们拥有世界上最好的人才来做这件事。它不会压低工资。

Yeah. I think you can absolutely pull off a system that that for probably in a shared definition for us and anybody else would say, this is clearly a high merit job. It's going to increase the wages in this particular sector on average. And and we wanna make sure that we've got the best talent in the world that comes in to to do that. It's not gonna drive down wages.

我们可能希望尽可能多地引进这些人。是的。所以我认为，你知道，这其中一些元素很有趣，因为它们推动了对话的发展，比如10万美元就像一把锤子来实现这一点，也许有一种更细致的方法，我肯定会更喜欢。

We probably want as many of those individuals here as possible. Yep. And and so I I think, you know, some elements of this are are intriguing in that they push the conversation forward on the dimension of, you know, like, the 100 k is like a a hammer to do that, and maybe there's a more nuanced approach that I would certainly prefer.

但是，但再次强调，重要的是，这10万美元会随着技能水平而调整。水平，对吧？所以技能越高，这部分成本就越能被分摊。因此你可以认为这就像是一个调节器，用来获取更高技能。

But but, again, it's important to like, the the the 100 k will scale with the skill level. Level. Right? So the higher the skill, the more like that is amortized. And so you could argue that this is just kind of a dial to get a higher skill.

嗯，你一旦给它标上价格，你必须记住人们会愿意支付这个价格。

Well, you you the the once you put a dollar amount on it, you have to keep in mind that people are gonna pay it.

没错。

That's right.

是的。他们会自己创造市场。而且，技能对某些人来说可能并不相关，你知道吧？所以这就是棘手的地方。

Yeah. Gonna make up their own money. Market. And and the skill isn't might not be relevant to some people because you know? And so that's the tricky part.

嗯。嗯。

Yep. Yep.

我想从劳动力市场转到Good up这个话题。

I wanna segue from labor markets to Good up.

下一个真正有趣的政治话题是什么，我们可以

What's the next what's the next really interesting political topic that we can

谈论什么？

talk about?

参与进来。

Engage in.

是的，没错。从劳动力市场到AI带来的劳动生产率，线下时Aaron，我们谈到了METER论文，论文指出他们的开发人员使用AI后生产力实际上下降了，但这与你与众多初创公司交流、观察它们如何大幅提升生产力的经验不符。所以，你能否谈谈你在哪里看到初创公司声称它们效率更高，以及为什么会这样？

Yeah. Exactly. From labor markets to labor productivity with AI, offline, Aaron, we were talking about the METER paper, and it was the papers suggested that their developers were actually less productive with AI, but that doesn't square with your experience talking to a lot of different start ups, seeing a lot of different start ups and how they're so much more more productive. So why don't you talk about where you're seeing startups say they're more productive, and why is it happening?

是的。我先以我们自己的案例为例，然后还有一个非常极端的版本。我们自己的情况是，我们采用了多种AI编程工具，比如Cursor在内部非常非常受欢迎。当我与同事交流时，比如在走廊里，他们可能想让我对AI感到兴奋，但我觉得他们知道我已经深信不疑了。所以，我从大家那里得到的定性回答，以及我会告诉你我们的内部指标。

Yeah. So I'll first just represent our own case study, and then there's the really extreme version. So our own case study is we've adopted a few different kind of AI coding tools, you know, cursor being a super, super popular one internally. And, you know, as I talk to people, let's say in the hallway, who have, you know, maybe they're trying to get me excited by AI, but like, I think they know I'm I'm bought in. So so the the kind of qualitative answers I get from from people, and then I'll give you our our internal metric.

有些人说他们的生产力提高了20%或30%，另一些人会说75%。有趣的是，我还没能确定这些回答背后的人口统计学差异。

You know, some some people say I'm getting, you know, a 20 or 30% productivity gain. Other people will say 75%. Interestingly, I have not been able to pinpoint the demographic difference on the answers.

哦，但这是自我报告。

Oh, but this is self reporting.

这是自我报告。有多满意？是的，所以不，不。但我们也有内部指标。

This is self reporting. How happy are Yeah. So so so no. No. But we have internal metrics as well.

所以目前我们大约30%的代码来自AI。

So about 30% of our code right now is is coming from AI.

哦，明白了。

Oh, see.

所以我们有70/30的比例，30%来自AI。我们有一些纯粹的内部指标可以证明这一点。但有趣的是，我有一个2x2矩阵：有资深人员说他们的生产力提升了75%，也有初级人员说提升了75%，反之也有只提升25%的情况。

So so we've got the 7030%. 30%. So we have some of the kind of pure internal metrics that that show this. But what's interesting is is that, like, I have a two by two of I have senior people that are saying that they're, you know, getting 75% productivity. I have junior people that are saying they're getting 75% and then vice versa on the 25%, let's say.

我还没能完全找出规律，也许除了——我们之前稍微讨论过，网上也能看到类似观点——最大的标准可能就是那些真正推动AI做更多事情的人，这是一种新的心理特征类型，就是那些愿意说'管他呢，我就让AI试试这个任务，看看它能想出什么'的人。你愿意这样做的程度，我认为可能最终会体现在生产力提升上。这是我们作为一家规模较大的初创公司的情况。

And it's I I I haven't been able to quite figure out a pattern, maybe except for and we've talked about this a little bit, but you kinda see this online. Except for maybe the the biggest criteria is just the people that actually push the AI to do more, which is sort of this other new kind of psychographic, which is just like, who is willing to just be like, you know what? I'm gonna YOLO this this task and just see what the AI comes up with. And your your sort of willingness to just do that, I think probably somewhat then then shows up in the in the the ultimate productivity gain. So that's that's us as a relatively larger company on the startup side.

哦，这太疯狂了，最让我震惊的是这一点。我经常与3%、5%、10%的初创公司创始人交谈，他们自我报告称生产力可能提升了3倍、5倍甚至10倍。最大的区别在于，一年前如果我们进行这样的对话，讨论的焦点可能是AI的代码自动补全功能，它可能在你每完成一个工作单元时增加几行代码。而现在，明显的现象是背景处理——我给它一个非常详细的提示。

Oh, it's crazy, and this is the thing that just blows my mind. I will regularly talk to three, five, 10% startup founders that that self report they might be getting somewhere on the order of, like, three to five to 10 x productivity improvements. And and the the the big difference is is that, you know, a year ago, if we were to have this conversation, the the conversation would be about, you know, AI sort of doing type ahead. And and it, you know, it can add maybe, like, a few lines of code to your productivity per, you know, incremental, you know, sort of unit of work that you give it. And then now, obviously, the big phenomenon is background where I give it a very, you know, detailed prompt.

我发送任务，它返回结果。人们把它比作老虎机——有时候它不会返回正确的结果，你必须决定实际采用哪些部分。但那些获得真正倍数级生产力提升的初创公司，它们从根本上以不同的方式进行工程开发。

I send it off. It comes back. You know, people talk about it as like a slot machine of, like, some percent of the time, it's not gonna come back with the right thing. You have to decide which what you actually, you know, kind of pull in from it. But the kind of startups that are getting, like, real multiples of productivity gain are are just they're fundamentally engineering in a different way.

它们发送任务，任务在二十分钟后返回，然后它们真正从事的是代码审查工作，而不是代码编写。这显然会改变未来计算机科学的许多方面。唯一的问题是，这对哪些方面有益？又在哪些方面会失效？

They're sending off a task. The task goes off, comes back in twenty minutes, and then they're really in the in the business of doing code review, not code writing. And it's gonna obviously change, you know, quite a bit of what computer science looks like in the future. And then the only question is, like, you know, what are all the things that that's good for? Where does that break down?

什么样的团队才能真正进化到那种状态？但最近最让我震惊的是，我认为这从根本上改变了工程学的未来面貌。

What what kind of teams can actually evolve to that state? But that one has been blowing my mind the most recently, and I think that kind of fundamentally changes what what the future of, you know, kind of engineering looks like.

我觉得你说的超级有趣。让我问问你。我认为有一个超越初级、高级的层面，那就是我们都在讨论具有两个解决方案的特征，这两个特征现在非常重要。一个是工程师为工程师做事，而且他们非常非常了解这个领域。是的。

I think what you said is super interesting. Let me ask you. I think that there's an overlay that goes beyond junior, senior, and and their and which is is we're all talking about characteristics that have two solution has two really important characteristics right now. One is that it's engineers doing stuff for engineers, and and they understand the domain super, super well. Yeah.

而且我认为这是一个非常非常重要的部分，也是人们讨论不足的一大点，可能正在发生的是，AI正在加速那些在领域内工作且非常聪明的人的进程。是的。另一个是我们不应该忘记，这也跟你自我报告有点关系，但这也是所有早期采用者的情况，早期采用者会故意对错误非常宽容。这是一个超级有趣的动态，当某事物全新时，围绕它的文化会发展起来，这让任何事情都可能发生。我的意思是，你知道，就像早期互联网用户不会抱怨网速慢。

And I I think that that's a really, really important part and a really big thing that people aren't talking enough about, which is maybe what's going on is that you have AI accelerating for people that work in the domain and are very smart. Yes. And then the other is we we shouldn't forget, and this is to your your self reporting a little bit, but also which is that this is all early adopters, and early adopters are very forgiving of mistakes on purpose. And it's a it's a super interesting dynamic where when something is brand new, a culture around it develops, which which just lets anything happen. I mean, you know, like, the early Internet people didn't complain that the Internet was slow.

对吧？唯一抱怨互联网慢的是后期采用者，他们会说，哇，这太慢了，或者比如在线视频，就像我在看这个小小的邮票大小的视频。而早期采用者会觉得，这是我见过最酷的东西。其他所有人则会说，我为什么要看那样的东西？下载音乐等等也是如此。所以我真的觉得正在发生的只是这种难以置信的

Right? The only people on the Internet were slow were the late adopters who were like, wow, this is so much slower and or or like take online video, which was like, I'm watching this tiny postage stamp video. And the early adopters, like, this is coolest thing I've ever seen. And everybody else is like, why would I wanna watch anything like that? And the same with downloading music and ever and so I I really feel like what's going on is just this incredible

你们还记得你们做的那块手表吗？Spot手表？是的。Spot手表。是的。

Do you remember that you guys do you remember this watch that you guys made? The Spot watch? Yeah. Spot watch. Yeah.

大概在1943年2月左右？

In, like, 02/1943 or something?

差不多吧。

Something.

所以我就买了一个。它用的是，像是AM或FM收音机的噪音。是的。所以你就

So so I bought one. And it used, like like AM or FM radio noise. Yeah. So so you

可以在户外田野里获取延迟四十五分钟的股票报价。

could get, like, stock quotes forty five minutes delayed if you are outdoors in a field.

没错。所以我当时有一个，我在这个营地里。觉得这是全世界最酷的东西。而且很明显，这将成为有史以来最大众化的产品。你知道，我早了大概二十年，

Yes. And and so I had one, and I was I was in this camp. Like, this is the coolest thing in the entire world. And, obviously, this is gonna be this the most mass market product of all time. And, you know, I was, like, twenty years too early with the

Apple Watch。逐向导航GPS。就像，第一次能把逐向导航GPS放进车里超级酷，但问题是大多数车移动的速度比车内计算机计算转弯时机的能力还快。对吧。所以你只能开着车到处转。

Apple Watch. Turn by turn GPS. Like, the first the first time you could put turn by turn GPS in your car was super cool except for the fact that most cars moved faster than the ability for the computer in the car to calculate when you were gonna turn. Right. And so you can't you just drove around.

它就像个U型转弯机器。对吧。但我觉得这很有趣，因为人们现在用这些AI工具，他们假设，嗯，我不是医生，对医学一无所知，却让你来治愈我。

You do it was like a U-turn machine. Right. And but I think that it's just it's so interesting because I I think that people use these the AI tools today, and they assume, like, well, I don't I'm not a doctor. I don't know anything about being a doctor. Let me ask you to cure me.

对。

Right.

并诊断我。然后就像，哇。这就是问题所在。那些看到失败的人，他们本来就不是出色的程序员。对吧。

And diagnose me. And it's like, woah. That is the word. And it just like all the people who see failure, like, they're just they weren't great programmers to begin with. Right.

而他们不知道如何提问。他们不想进行代码审查。然而，优秀的程序员或

And they didn't know how to ask. They didn't wanna review it. And whereas great programmers or

是的。

Yeah.

专业的程序员都知道代码审查非常重要。没错。而且这就是做事的方式。我认为还有

Just professional programmers know that code review is really important. Yes. And and that's just how you do it. I think there's

有一个方面，嗯，这方面让它很难衡量。其中之一是，我不认为这只是早期采用者的问题，就像这些模型如此神奇，让你眼花缭乱。哦，所以即使它不是你想要的东西，你也会觉得它很棒。你知道吗？

an aspect well, there's there's an aspect of this that makes it very difficult to measure. One of them is and I don't think it's just an early adopter thing like these. These models are so magic that you get dazzled. Oh, so Even if it's not what you want, you're like, it was great. You know?

而且我认为很容易将其与高效混为一谈。比如，这不是我想要的，但它很神奇。所以因此，因此，一定是...但这个是的。因此，它一定很棒。所以，也许随着时间的推移，我们就会放弃自己的主见，让模型做所有事情。

And I I think it's very easy to complete that with being productive. Like, it's not what I wanted, but it was amazing. So therefore, therefore, must be But this yeah. Therefore, it must be great. So so, like, maybe over time, we just abdicate having an opinion and, like, the model does everything.

但现在，我经常看到这种情况，人们对使用AI非常热情，但它实际上并没有影响，你知道，他们的产出。他们只是热情而已。第二点是，我觉得几乎存在影子生产力。

But right now, and I see this a lot, people are like, they're so enthusiastic about using AI, but it really hasn't impacted, you know, their output. They're just enthusiastic. The second one is I I feel like there's almost shadow productivity.

哇。抱歉。你会如何验证一个5到10人的公司，他们在经验上却像50到100人的公司一样运作？比如，你可以看到他们代码的巨大规模，然后你会想，好吧，你不可能在十年前做到这一点

Woah. Sorry. How would you how would you how would you verify that with the a five or 10 person company who who kind of empirically is operating at, like, a 50 to a 100 person company? Like like, you just you can see them, the sheer, you know, scale of their code, and you're like, okay. You could not have done that ten

多年前。

years ago.

我其实同意

I actually agree with

史蒂夫说的所有观点，

everything that

听着，这确实是经验之谈。根据我的经验，我与许多公司合作过。经验表明，那些使用AI的更资深的小团队简直超乎常人。是的。

Steve was saying, which is listen. An this is anecdotally. Anecdotally, I work with a lot of companies. Anecdotally, the more senior small teams that use AI are superhuman. Yeah.

是的。就好像他们一觉醒来，全都他妈成了托尼·斯塔克。简直难以置信。而且，他们的生产力高得离谱，但他们都是——

Yeah. It's like they woke up, and they were all fucking Tony Stark. Is unbelievable. And, like, their productivity is insane, but they're all they're

全都是资深高手。听着，他们——我不想贬低任何一点，但那些公司相对于大公司的10%团队来说，生产力也已经高得惊人。是的。没错。因为没有遗留代码，而且他们是从零开始。

all super senior. Look, they were they don't I I don't wanna take anything at all away, but those companies were also incredibly productive relative to a 10% team of big companies. Yes. Yeah. Because there's no code, and and they're starting from a clean slate.

不。不。他们真的醒来——是的。是的。毫无疑问。

No. No. They they really wake yeah. Yeah. For sure.

而且，你知道，他们资历很深，但几乎每个人一开始都是AI怀疑论者，并且对价值持非常冷静的态度。所以他们只是以非常务实的方式使用它。

And and, like, you know, they're very senior, but they're also almost to a person where AI skeptics to begin with and are incredibly sober about Yeah. The value. And so they they just use it in these very pragmatic ways.

还有另一类人我注意到了，因为我想在知识层面保持全面诚实。我看到这些19、20岁的年轻人，首先我不知道斯坦福、MIT这些地方现在是怎么回事，但好像所有人都在辍学。

There's one other category that I'm seeing because I just wanna be intellectually honest on the full spectrum. I'm seeing these 19 and 20 year olds that are like first of all, I don't know what is in the water at, like, Stanford, MIT, etcetera right now, but, like, everybody's dropping out.

是的。

So Yeah.

所以，真的，人们去那里待一周就辍学了。但这个群体有个趋势，就是那些996类型的人，他们可能在过去是10倍效率的工程师，但现在他们像是100倍效率的工程师。

So, like, just, literally, like, people are going there for, a week just to drop out. But there is a tendency of that cohort. It's the nine ninety six people, and there is this tendency which is, you know, they would have been maybe 10x engineers in a prior world, but now they're like 100x engineers.

是的。

Yeah.

因此，你知道，在他们相对同龄人中算是资深。但他们创建初创公司的方式完全不同，这可能是我所见过的创办和运营公司方式的最大变化。是的。

And so, you know, senior in terms of in their own kind of relative cohort. But, like, the the the way that they are building their startups are just, like, completely different than it's probably the if I look at at you know, so we dropped out of college. God, that's scary nineteen years ago. If I look at how these companies run versus today, it's the biggest change in in how you start and run a company that I've ever seen. Yeah.

而且，我认为如果你看看1995年辍学与2005年我们辍学时的情况，我不认为公司创建过程有那么大的不同。

And, like and I I think if you looked at, like, in 1995, if you were to drop out of college versus 2005 when we dropped out of college, I don't think you, like I don't think, like, the company building process was all that different.

嗯，甚至游戏也从根本上被互联网改变了。

Well, even the game goes a the the Internet fundamentally changed.

不。后互联网时代。后互联网时代。好吧。

No. Post Internet. Post Internet. Okay.

是的。所以你在构建

Yeah. So so you're you're building

你必须对互联网保持开放，因为那是 是的。是的。

to white you have to be white to the Internet because that was a Yeah. Yeah.

100%

A 100%

的。

of that.

我们不是在85年。所以在95年，你辍学去做互联网创业。对吧？你辍学去做互联网创业。到2005年2月，除了我们的资源在云端，而不是像以前那样必须去数据中心——实际上我们仍然去了数据中心。

We're not in '85. So at '95, you're dropping out to do an Internet startup. Okay? You drop out to do an Internet startup. By 02/2005, other than the fact that our resources were in the cloud versus, you know, you'd have to go to a data center in our case, we actually still went to the data center.

比如，公司创建过程中的很多方面其实并没有太大不同。但如今在2025年，由于人工智能的出现，你创办公司的一切方式都变得完全不同了。

Like, not that much about the company building process was different. Today, in 2025, everything about how you're starting your company is completely different because of AI.

我认为，贯穿所有这些的关键主线是速度。是的，没错。而且我认为互联网已经提升了速度。

The I I think that that the key the the through line in all that is velocity. Yes. Exactly. And and I I think that that that cut off, the Internet increased velocity.

是的。

Yes.

而人工智能也提升了速度

And AI increases velocity

我完全

I totally

同意。而且我认为这对当前正在发生的事情来说非常非常重要。

the same way. And I I think that that's just super, super important to what's going on.

所以我们基本上，如果你

So we're basically if you

回溯到早期互联网时代，比如你创办一家公司时，仍然有很多老派做法，比如你的商业计划是什么？你的规划是什么？我们要保持两年秘密运营。所有这些都存在，而真正改变速度的是马克和本，是的，在网景公司。

go back even with the in the early Internet, like, you started a company, there was still a lot of old school, like, what's your business plan? What's your plan? We're gonna be stealth for two years. There was all of the stuff, and it was really Mark and Ben Yeah. At Netscape that changed the velocity Yeah.

关于公司运作方式。而云计算是这种速度加速的催化剂，而人工智能则是对速度运作方式的重构。是的，是的。但有趣的是

Of how companies work. And the cloud was an accelerant to that acceleration of velocity, and and AI is is a refactoring of how velocity works. Yeah. Yeah. But what

即使在云计算时代，它也极大地虚拟化了你原本必须处理的物理设施。

was interesting is even in the cloud, like, that was this great virtualizer of the physical stuff you would have to

但那需要两年时间建设。是的，没错。你知道，就像，你根本没有...是的。

deal with. But that was a two year build out. Yeah. Exactly. You were you know, like, you you had no Yes.

你可以拥有客户。是的，100%确定。

You could have customers. Yeah. 100%.

一旦

As soon

你有了代码，你就可以拥有客户。

as you had code, you could have customers.

你说得对。

You're right.

也就是说，就像我

Which is, like I

我的意思是，像PPLG这类高速发展的初创公司其实在AI之前就起步了。是的。但相当，嗯，惊人。想想，比如GitHub。想想，比如Slack。

mean, I mean, p p l g and a lot of these kind of high velocity startups actually started pre AI. Yeah. But the pretty, like, phenomenal. Like, think about, like, GitHub. Think about, like, Slack.

想想，比如Figma。

Think about, like, Figma.

是的。

Yes.

你知道，有一些相当卓越的公司是在AI之前出现的，我们现在还在借鉴它们。其中很多基本上是基于云，然后是这种SaaS化，从而解锁了不同的市场进入方式。是的。Zoom就是一个早期的好例子。

Like, you know, you have some pretty remarkable companies that came pre AI that we're drafting on. And a lot of that was, like, basically cloud and then this assification and then unlocking kind of different us to go to market. Yeah. Zoom Zoom was a great early example.

这就是AI的作用——它是产品开发的加速器。是的。而运行SaaS销售、云所做的，是加速了获得收入的过程。是的。这过去通常需要两三年时间。

This is the the thing that AI is is is an accelerant in building the product. Yes. Whereas what what Running SaaS selling. Cloud did was accelerate getting paid Yes. Which was a thing that used to take two or three years.

我 我

I I

我的意思是，可以说，云技术也让产品开发速度大大加快，因为你会拥有，比如，这些大型

mean, I would arguably, the cloud also made it much quicker to build a product because you would have, like, these big

嗯，它让客户都能更快地使用同一个版本。我不认为它让

Well, it made them much quicker for customers to all have the same one. I don't think it made

它，比如，快了五倍。

it, like, five times faster.

有道理

Fair

确实。所以，就像，我觉得那是，就像，那是，就像，是的。就像，你能更快地看到你的网站，但是，就像，我在98年也能很快看到一个网站。就像

enough. So, like like, I think it was, like, it was, like, yeah. Like, you can see your website a lot quicker, but, like, I could see a website in '98 pretty quickly. Like

是的。不过，你知道，也许这是个基础设施问题。构建一个大型分发服务。

Yeah. But well, you know, maybe this is an infrastructure thing. Building a big distributor service.

不。不。当然。当然。

No. No. Sure. Sure.

我当时没有做分布式系统，真的、真的、真的很难。98年的基础设施。

I was not doing distributing Really, really, really hard. Infrastructure in '98.

所以是的。但是，所以，抱歉，我只是想...我认为AI生产力难以衡量有两个原因。第一个我刚提到，它确实令人眼花缭乱。所以我觉得人们会有点觉得，哦，这太神奇了。

So Yeah. So but so there's sorry. I just wanna so so I think AI productivity is hard to measure for two reasons. The first one I just mentioned, it is really dazzling. So I think, like, people kind of, like, feel like, oh, it's amazing.

第二个原因是，我认为很多生产力实际上是隐藏的，而人们衡量了错误的东西。

The second one is I think a lot of the productivity is actually hidden, and people measure the wrong thing.

对吧？所以人们竟然在生产力上衡量错误的东西，这很令人震惊。这简直就是生产力衡量的历史。嗯，但是，但是还有

Right? So Shocking that people measure the wrong thing in productivity. That's, like, literally the history of productivity measurement. Well, but but also

这里发生的情况是，董事会说我们需要更多AI。然后会发生什么？他们去找，比如，某个CTR，某个创新实验室，然后创新实验室做AI。然后，随便吧。他们引入，比如，构建一些内部工具，然后它会失败。

what happens here is, like, you have the board, and the board is like, we need more AI. And then so what happens? And they go to, like, some CTR, some innovations lab, and then the innovations lab do AI. And so, like, whatever. They, like, bring in like, build some internal tool, and it'll fail.

当然，那会失败。对吧？但现实是，这波AI浪潮非常个人化。可能公司里大多数人都在用ChatGPT。可能，你知道，是某种个人助理。

Of course, that'll fail. Right? But the reality is is, like, this AI wave is so personal. Like, probably most people in the company are using ChatGPT. Probably, is, you know, some personal assistant.

可能，你知道，他们正在使用Cursor或某些编程工具，而这要难衡量得多，对吧。因为，你知道，它没有公开宣传。所以如果你真的看看那些报告，比如企业项目失败的情况，你们看看他们在衡量什么。就像是，没错。很明显，董事会强行推动的内部项目，他们雇了，比如，你知道，一些顾问来做，那肯定会失败。

Probably, you know, they're using cursor or some coding thing, and that's much, much harder to measure Right. Just because, you know, it's not advertised. And so if you actually look at the reports on, like, enterprise things fail, you guys look at, like, what they were they're measuring. It's like, yeah. Clearly, internal project pushed down by the board where they hired, like, you know, some consultant to do it is gonna fail.

那些总是失败。是的。但这实际上并不是正在发生的事情。比如，正在发生的趋势是一个非常长期的，嗯，

Those always fail. Yeah. But that's actually not what's going on. Like, the movement that's happening is a very secular Well,

这是这是这是下一次自下而上的采用真正改变生产力等式的时候。是的。而这是一件大公司不知道如何应对的事情。

it's it this is this is the the the next time that bottom up adoption is really changing the productivity equation. Yeah. And that's a thing that that it defies. Big companies do not know how to deal with that

是的。

Yeah.

因为他们想要控制它，他们需要控制它。他们担心安全、安保、隐私以及他们所有的公司规定。然后，另外我认为叠加在这之上的是，人工智能是一个非常独特的，如果这么说没问题的话，创新在于它是非确定性的。是的。所以突然间，有了这个非常个人化且非确定性的东西，而在大型组织中所有这些试点和人工智能项目的真正问题是，你不仅无法衡量，甚至不想推出一个非确定性的，嗯。

Because they they want they need to control it. They worry about safety and security and privacy and all of their corporate rules. And and then also, the other thing I think to overlay on that is AI is is a very unique, if that's not a bad way to say things, innovation in that it's it's like nondeterministic. Yeah. And so all of a sudden, have this very personal and nondeterministic thing, which of which the real problem in a large organization to all these pilots and AI projects is that they you can't you can't you not just measure, but you don't even wanna put out there a nondeterministic Uh-huh.

解决方案。

Solution.

是的。是的。

Yeah. Yeah.

因为你的整个情况是，嗯，我们大规模运营，覆盖了60个国家。是的。就像，

Because your whole thing is like, well, we operate at scale, and we have 60 countries. Yeah. Like,

它就像白细胞一样。

it's a white blood cell.

就像，我们不能……我们不能推出一个客户支持解决方案，如果五个客服用五种语言给出的答案都不一样。是的，基于客户或客服人员提问的方式。我真的觉得，这将是大型组织在采用新技术时面临的最大挑战，他们可能会因为不确定性而陷入混乱。

Like, we can't We we can't put we can't put, like, a customer support solution out there if five agents in five languages all have different answers Yeah. Based on the way that the customer or the the CS agent asked the question. And I I really feel like that is gonna be the the hugest challenge in in large organizations figuring out how to adopt things is, like, they're gonna just get all bunched up over over nondeterminism.

是的。没错。我我我认为，我们可能需要一种新的衡量标准，总的来说，因为生产率的提升很大程度上会是这种微妙的变化，就像，我以前会去谷歌查那个。

Yes. Yeah. I I I think that the we we will have to have some new form of of measurement probably in general in this because so much of the so much of the improvement in productivity will be will will sort of be this, like, subtle change of, like, like, I used to go to Google for that.

不。或者现在我的EA（行政助理）在用ChatTPTA写邮件

No. Or now I my EA is writing emails using ChatTPTA

为我。

for me.

所以当然那是

And so sure that's

除了上次播客中我们讨论过的内容外，这种情况还会出现在哪里呢？我只是觉得，我们将开始进行更高层次的工作。是的。而这最终会呈现出不同的面貌。是的。但你无法简单地衡量，好吧。

Where is that showing up in and other than just, like again, when we talked about this in, I think, the last podcast, I'm just like, we're gonna just start to do higher levels of work. Yeah. And that will just end up looking different. Yeah. But you won't be able to be like, okay.

当工作方式完全不同时，我该如何衡量生产力呢？

How do I measure what the productivity was when it's just like, I'm working just totally differently?

是的，是的。但我的意思是，Coda 就是一个很好的例子。根据我的经验，资深人士实际上如何使用AI进行编码？比如文档编写、测试。

Yeah. Yeah. You'll but all I mean, Coda, I think, is such a great example. So, like, what do people in my experience, the more senior folks actually use AI for code? It's like documentation, writing, testing.

我的意思是，很多其他方面可能并不会直接加快发布进度，但你能得到更健壮的代码。

I mean, it's a lot of the other stuff that may not actually, like, increase, like, the shipping schedule, but, like, you get a lot more robust code

是的。

Yeah.

更易维护的代码，更好的架构，更具前瞻性的架构。这样我们就能构建出更出色的软件，同时保持相同的功能发布速度。

A lot more maintainable code, a much better architecture, a much better future forward architecture. So we could be building tremendously better software, but still be shipping features at at the same velocity.

对吧？我们可能需要将生活质量作为另一个衡量指标，就像真正帮助人们那样。就是这样，是的。我只是更快乐了。我不需要

Right? We might need to just measure, like, quality of life as another metric of, like, literally people help. It's like the what yeah. I'm just happier. I don't have

做那些破事。不想要我们自己的权利文件。没错。

to do shit. Don't want our own right document. Exactly.

是的。但这是，你知道，很容易被指责为夸张和夸大其词。但我认为最关键的是你刚才谈到的，也就是你将如何对待工作的整个概念将会完全不同。你知道，这是我不得不说的超级老派观点，但电子表格出现前后的对比就是个完美例子。电子表格出现前

Yeah. But this is I I you know, it's so so easy to get accused of hyperbole and and overstating it and things. But what I think is just so key to what's going on is this it is what you were talking about, which is like, the whole notion of what you're gonna do with the job is gonna be really different. You know, I'm this is my obligatory super old person thing, but the the pre pre spreadsheet, post spreadsheet is a perfect example of this. The pre

电子表格的例子。

spreadsheet example.

我没说经典。我说的是但在电子表格出现之前，你是个银行家，你知道，你会觉得我应该帮助这家公司被收购。然后你会建立一个财务模型，让50个刚毕业的MBA们用他们的惠普计算器埋头苦干计算财务模型。然后你说，好吧，我们再来一遍。但如果利率变化

I didn't say classic. I said But but before spreadsheets, you would be a banker, and, you know, you would be like, I'm supposed to help this company get acquired. And you would come up with a financial model, and you'd have 50, you know, recent MBAs all churning away with with their HP calculators figuring out the financial model. And then you go like, okay, let's do this again. But if the interest rate changes

嗯。

Mhmm.

或者如果他们的资金来源发生变化，你就会说，好吧，那得花一周时间。是的。然后每个人都得重新做一遍。所以你实际的决策质量 是的 是 完全正确。

Or if their source of funds change and you're like, okay, well, that's like a week. Yeah. And and everybody has to do it all over again. So you actually your quality of decision Yeah. Was That's exactly right.

非常糟糕。是的。所以发生的事情是，就在1985年，那个工作完全改变了。到1990年，你不是让新毕业生来做，而是自己动手做。我对此记得清清楚楚。

Was really bad. Yes. And and so what happened was that just in 1985, like, that job just completely changed. And by 1990, instead of asking the recent grads to it, you were doing it yourself. I actually remember this absolutely crystal clear.

我的两个堂兄是1985年芝加哥大学的MBA毕业生。他们拿到MBA学位时根本不用电脑。两年后当我提到要去微软工作时，他们还说，我们有实习生帮我们用电脑呢。嗯哼。后来他们确实也开始用电脑了，但他们对银行业的整个概念——

My my two cousins were University of Chicago MBA grads in 1985. They did not use a computer to when they got their MBAs. And when I was talking about going to Microsoft two years later, they were like, well, you know, we have these kids who use the computer for us. Uh-huh. And they they ended up using computers and stuff, But but, like, their whole notion of banking Right.

就是建立在那种确定的多周转周期上。然后突然间，就变成更多实习生用Lotus 1-2-3处理事务。这让我想到，代码和初创公司的情况就是——完全不同的思维方式，关于你能做多少、多快完成以及迭代速度。Figma和Dylan就经常谈论这个。

Was defined by this Sure. Multi week turnaround. And then all of a sudden, it's just more interns doing their Lotus one, two, three thing. And I this is it gets to the you know, what's going on with with code and startups is is it's just the there's just, like, a whole different mindset over how much you can do and how soon and iterate. Figma and Dylan, they're always talking about this.

就像，Figma现在要改变设计理念的发展轨迹了

Like, this is like, Figma is gonna now change the trajectory of a design idea

没错。

Right.

从‘我们在这里迭代’变成‘直接动手干’。是的。

To go from, like, oh, let's iterate over here to, like, let's just do it. Yes.

我的意思是，这就像——再次说明，用我的生产力百分比来思考有点奇怪。显然我的一天没有代表性，因为我同时处理太多不同事情。但经常晚上10点时，在以前，我可能会把任务派给某个人，比如某个分析师或参谋长之类的角色，去调研某个事情。

I mean, the the this is and this is where it's like a again, a weird thing to, like, think about the percentage of my productivity as an example. And, like, my my day is, like, not representative, obviously, because I'm bouncing between too many different things. But, like, there'll be so many times where it's 10PM. In a prior world, I would have sent off a task to somebody, you know, you know, some analyst or chief of staff, you know, type type role. Go research this thing.

三天后结果回来，你才找到答案。而现在显然就像启动深度研究。没错。用Cursor生成原型，做某种分析，十分钟或二十分钟就能拿回结果。

It comes back three days later, and then you find the answer. And then now it's obviously just like kick off a deep research. Yep. Go to Cursor to generate a prototype. Do do some kind of analysis, and then you have it back in ten minutes or twenty minutes.

而且我已经压缩了无论那是什么，原本可能需要串联连接该任务的步骤现在完全被压缩了。这样到了早上，你就能直接启动那个项目了。对我来说，这几乎不可能用一个数字来衡量，因为工作的本质已经完全不同了。因为你将工作流程中的许多不同步骤压缩成了一个单一动作。

And and I've just compressed whatever the what whatever that was kind of then gonna be as a, you know, serially connected to that task is now just fully compressed. And so then by the morning, you're now kicking off whatever that project was. Again, like, nearly impossible for me to peg, like, what is that as a number? It's just, like, a fundamentally different thing on just what work looks like. And because you just compress so many different steps of a workflow into a single action.

所以这完全是一种不同的思考工作的方式。

And so we're it's just like a completely different way of of thinking about work.

这在我们之前讨论的内容中——也就是关于你的专业知识如何真正贡献于此——处于什么位置？是的。特别是，我认为让人们了解当你与客户交流时，如何帮助他们避免试图让AI去做人们原本就无法完成的工作会很有趣。因为这是一个容易失败的点。

Where where does that fit in for you in what we're talking about what I was asking about earlier, which is how does the expertise your expertise really contribute to that? Yeah. And in particular, I I think it'd be interesting for people to understand, like, when you talk to customers, how do you help them to avoid trying to get people to make AI make them do jobs they couldn't do in the first place? And, like, how does because that's a easy point of failure.

是的。实际上这是一个非常反直觉的事情，你在上一个话题中谈到了专业化。AI的最大收益者是那些在某个领域有一定专业知识的人，他们知道什么是真实的，什么行不通，我应该从AI的输出中整合什么。你知道，比如那2%可能是幻觉或者数据走向错误的内容？

Yeah. I mean, it actually is this is this is this really counterintuitive thing where and you've talked about specialization on the last one. It's like, it the the the biggest gains of AI go to people who have some degree of expertise in an area to know what is actually true, what is not going to work, what should I what should I integrate from the output of this AI? Yeah. You know, like, what are the 2% of things that maybe are hallucinations or or, you know, took the data in the wrong direction?

如果你对你特定的领域没有深刻的理解，你就无法做出正确的判断来做出所有这些决定。所以我认为专家在这个世界中只会变得更强大。这就是为什么我甚至不相信你能告诉大学生学习与历史上任何其他时期不同的东西。就像，真正擅长某个特定领域。是的。

If you don't have a deep understanding of your particular space or field or domain, you you aren't able to then have the right judgment to make all of those decisions. So I think the experts just get more powerful in this world. And so I would that that's why I'm not even, like, convinced that you can tell a college student to learn anything different than ever any other, you know, period in history. Like like, be really good at a particular field. Yeah.

然后AI仅仅是你在该领域能力的涡轮增压器。但如果我不了解我所知道的关于SaaS的事情——这显然是一种相当奇怪的专业知识，但我在理解SaaS方面还算可以——那么我交给深度研究代理并随后纳入工作的内容就会没有意义。对我来说，我不会拥有它提到的那个事物的所有背景。我如何将其融入整体战略？但因为我对这个特定行业有一些理解，这让我效率大大提高。

And then AI is merely a turbo charger of your capability in that particular field. But like, if I didn't know if I didn't just like generally know the things I know about, you know, SaaS, which is like, obviously, like a really, like like, weird expertise, but like, I'm like okay at understanding SaaS, then I then the things I give to, you know, a deep research agent that I then go incorporate back into work wouldn't make sense. Like to me, I wouldn't have the I wouldn't have all the context for like that one thing that it mentioned. How do I like form that into the overall strategy? But because I have some understanding of of this particular industry, that just makes me way more productive.

所以我认为专业知识根本不会消失，而且任何特定领域的专家只会变得更强大。

So so I don't think expertise goes away at all, and I think any any of the experts in their particular area just become more powerful.

所以我们实际上有相当多的市场轶事数据。这非常非常有趣。比如，拿很多这类例子来说，我们举个非文本的例子，如图像或视频。如果你看看任何流行平台的客户群

So we we actually have a fair bit of anecdotal market data on this. So it's very, very interesting. So if you take, like, a lot of these, let's just take kind of a nontext example, like image or video. And if you look at the customer base for any of the popular platforms

是的。

Yeah.

非常有趣。如果你随机抽取一个已变现的美元，它来自专业人士。原因很明显，因为，你知道，他们能够产出内容；而如果你随机抽取一个用户，那是休闲用户，处于长尾部分。

Very interesting. So if you if you draw a dollar at random that's monetized, it's from a professional. And for obvious reasons because, you know, like, you know, they can produce if you draw a user at random, it's it's casual and it's in the tail.

没错。

Right.

所以相当清楚，你知道，从变现角度来看这是一场专业消费者运动。而且，你知道，我与许多公司有合作，比如与专业设计师或创意人士打交道的公司。他们在AI工具上花费的时间与在传统工具上一样多。是的。结果只是应用产出更加丰富，而且，你知道，它往往——他们

And so it's fairly clear that, you know, this is a prosumer movement from monetization. And, you know, I'm associated with a number of companies that work with, like, say, professional designers or professional creatives. They spend just as much time on the AI tools as they would on traditional tools. Yeah. It just turns out the app output is far more rich and, like, you know, it you know, it it's it it tends to be They

他们有更多

have a they have more

是的是的。但这仍然是人类的审美。仍然有非常具体的要求。所以我认为，你知道，如果未来出现AI的广告模式，我认为会有一大批人想要使用AI，但他们实际上没有财务激励去做，或者这与他们的实际工作无关。我们已经开始看到这种分化，而且还会出现另一个子群体去做。

is that is Yeah. But it's still it's still human taste. There's still, like, very specific requirements. And so I think I think, you know, if there ever is an ads model that ever shows up for AI, I think that there's gonna be a long tail of people that wanna use AI, but they actually don't have the financial incentive to do it or it isn't tied to, like, you know, their actual job. And we're already starting to see that bifurcate out, and there's gonna be another subsection that do.

我的感觉是，比如说，假设我在写代码，随便什么。我是个业余开发者，在写一个3D游戏，想要一个3D资源。我有两个选择：可以让AI为我创建

And and my sense is, let's say if I'm let's say say I'm writing, like, whatever. I'm a I'm a casual developer. I'm writing a three d game, and I wanna have, like, a three d asset. I've got one of two choices. I can have AI create it for me Mhmm.

或者我可以聘请专业人士来做。作为开发者，即使使用AI，我也做不出很好的3D资源。对吧？所以我认为，你会面临和今天一样的界限：要么自己凑合着做，要么找专业人士，而专业人士也会使用AI。

Or I can, you know, contract a professional to do it. Like, me as a developer, I'm not gonna create even if I use AI, it's not gonna be a great three d asset. Right? And so I think that, you know, you're gonna have that same line that you have today. It's either you can hack something up yourself or you can go with a professional, and the professional will be using AI.

我认为非常令人兴奋的是，AI创造了第三类选择：我不是想把这当工作做，永远不会靠这个赚钱，但对我而言有某种产品效用提升，值得每月花20美元。现在我能够独自 brainstorming 时生成原型

And and I think what's super exciting is is that the AI has created a third category, which is which is I am I'm not trying to do this as a job. I'm never gonna monetize this, but there's some product utility gain to me that is worth $20 a month. So my ability to now generate prototypes when I'm by myself just brainstorming Yeah. Yeah. Like, again, I'm not gonna do anything in the ultimate delivery of that, you know, into any functional code.

但是，对我来说，能够去实现我正在思考的东西，值得花这20美元。

But, like, the like, it's worth $20 for me to be able to go and, like, realize the thing that I'm thinking about.

是的。

Yeah.

所以就有了所有这些捕获TAM（总可寻址市场）的新方式，因为有所有这些被释放的效用。

And so so there's just, like, all new ways of of capturing TAM because there's all this utility that gets unlocked.

对。还有一点值得注意：对某个领域感兴趣的人可以通过AI进入这个领域。对吧？我认为，对于想要进入某个领域的人来说，以AI原生的方式去做是头号机会。对吧？

Right. And there's one there's there's one more thing that's that's worth noting, which is people that are interested in an area in an area get that through AI. Right? I was like, this is the number one opportunity for somebody, you know, that wants to enter an area to do it as an AI native. Right?

因为，这个工具实际上可以教你。是的，你知道吗？而且现在人才非常稀缺，你可以填补这个空缺。

Because, a, like, the tool actually can teach you. Yes. You know? And then there's just such a paucity of of talent out there that you can fill.

这也是，我是说，这其实是生产力发展的普遍历史。是的，就是当你拥有的工具越多，首先专家们会使用它们。

And this is also I mean, this is the history of of productivity in general Yeah. Which is you you the more tools that you have available, first, the experts use them.

嗯。

Mhmm.

然后更多人能够成为专家。没错。而且，只要——但这也是你管理自己职业生涯的一部分。我很喜欢你提到的那个观点，就是你仍然需要真正擅长某件事。我认为，如果你想在金融或销售领域做得好，你应该变得非常擅长，并假设你会使用AI来辅助。

And then more people are able to become experts. Yep. And and I and I as long but that's part of your managing your own career. I love that point you made about, like, you still have to be really good at something. And and I think that that, you know, if you wanna be good at finance or sales, you you should become really good at it and assume you're gonna use AI for that.

是的。你会比那些假装懂行的人更强。我认为这很像，如果我们回顾之前的讨论，你知道，哦，AI能制作PPT幻灯片。但事实证明，制作一个好的PPT仍然是一项技能。是的。

Yeah. And you're gonna be better than the people pretending. And I and I think that that's very much like if you just take know, we were going back and forth about, you know, oh, it makes a PowerPoint slide deck. Well, it turns out to make a good PowerPoint deck is still a skill. Yeah.

人们仍然愿意付给麦肯锡巨额费用，以获取带有更好图片的更优质PPT。

And people still pay Mackenzie huge amounts of money for better PowerPoint decks with better pictures.

是的。所以，听着，我外包了很多视频和艺术类的工作，已经做了很长时间，你知道，作为不同公司甚至在这里在a16z的一部分。我见证了这些传统技术外包向AI的转变。金额还是一样的。是的。

Yeah. So so, listen, I so I I contract a lot of work for, like, videos and art, and I have for a very long time, you know, as part of, like, different companies or even here at a sixteen z. You know, I have seen the shift for contracting some of these traditional techniques to AI. Dollar amounts are the same. Yeah.

所以，这又是杰文斯悖论的再现。

And so, again, like, know, this is Jevon's paradox all over again.

就像，他们

Like, they

花费的时间一样多。

spend just as much time.

是的。

Yes.

你懂吗？只是产出往往更炫目或诸如此类。

You know? It's just the output tends to be more dazzling or whatever it is.

是的。所以

Yes. And so

你应该看到它的更多版本。他们应该运行更多模拟和其他想法。

And you should see more versions of it. They should run more simulations and other ideas.

你与他们有更多的迭代机会，拥有更多控制权。比如，我可以听，我可以制作一个视频，比如我想要一条龙从天空飞出来，你懂吗？

You've got more iteration with them. You've got more control. Like, I can listen. I can have a video where, like, I I want a dragon to fly out of the sky. You know?

所以，作为客户，你确实拥有更多的控制权。但当然，这并不意味着成本会降低。

So, like, you have, like, a lot more control as as the customer. But for sure, you know, this is not somehow, like, dropping the cost

输出。我想，我想

of output. I wanna I wanna

你刚才是不是想插话？我想回到你之前提到的一点，关于有20岁的年轻人正在以新方式创办公司。记得几年前，Patrick Carlson和其他一些人曾问，嘿，所有Z世代的超级成功创始人在哪里？还记得吗？当然，Dylan Field和Alexander Wang的公司花了几年时间才真正成功。

were you about to jump in? I wanna circle back to a point you made earlier about that there are 20 year olds who are building companies in new ways. Because remember a few years ago, I think Patrick Carlson and a few others were asking, Hey, where are all the Gen Z super successful founders? Remember that? And of course there was Dylan Field and Alexander Wang, their companies took a few years to really work.

现在我们看到Cursor和Merkor的创始人在很短的时间内迅速达到了大规模。也许是因为基础模型公司需要一定经验水平的创始人，因为融资规模较大，而应用程序可能更适合年轻创始人。你对这一点有什么看法？

Now we're seeing the Cursor founders, the Merkor founders sort of, you know, get to massive scale in a very short period of time. And maybe it was the the foundation model companies required in a certain level of, you know, experienced founder because of the fundraising amounts and maybe the applications are, you know, more conducive to younger founders. But what's your sort of reflection on this?

嗯，我不记得他具体提到那个日期，但我确实认为在2010年代中期到2020年代初期的这段时间里，我们行业实际上处于一个相对低迷期。原因是我们已经完成了世界上人们需要的许多核心事物的开发。比如，一旦有了Slack，你就不需要其他五个聊天工具；一旦有了Zoom，你就不需要其他五个视频会议工具。所以之后的产品变得有些衍生，超越了这些核心平台。

Well, the the I don't remember exactly the date at which he he mentioned that, but the but I do think there was a period between in in the sort of, you know, mid two thousand tens to late two to early twenty twenties where where we were actually in kind of a bit of a lull as an industry. And the the reason for that was, like, like, we kind of did, like, check off a lot of boxes of, like, the core things that people needed in the world. And so we, like, checked off, like, a lot of the like, once you have Slack, you don't need five other chat tools. Once you have Zoom, you don't need five other video conferencing tools. And so it gets kinda derivative, you know, past these kinda core platforms.

所以，一旦SaaS覆盖了工作中所有主要事项，在消费领域，我们也有了送餐、听音乐和看视频的方式。作为消费者，我们做的事情并不是无限的。那么20岁的创始人应该做什么呢？相比2000年代中期整个世界都开放、你可以创办任何东西的时候，现在的机会相当有限。

And so once you have, like, SaaS, you know, kind of check off all the major, like, things you do at work, and then in the consumer world, we, like we had ways of delivering food and listening to music and watching videos. So, like, there's, like, not an infinite set of things that we we do as consumers. Then then what what is the 20 year old founder supposed to to work on? Like, they're gonna the it's like you have, you know, pretty finite opportunities as compared to in the mid two thousands, let's say, the whole world was open. You could start anything.

因为每个类别都必须重新发明，有点像是后移动时代、后云成熟期。所以我们如今进入了人工智能时代，这就是为什么我如此兴奋不已。因为整个格局完全重置了，现有的公司虽然有分销优势，但仅此而已。公司并没有其他真正的优势，反而有劣势。

Because every single category had to be reinvented, kind of post mobile, post cloud maturity. So we now have that era in AI, and that is why I'm so unbelievably pumped up. It's because you have a complete reset of the landscape where there there's, like, all there there is incumbent advantage in distribution, but that is it. There's no other other real advantage in the company. Disadvantages.

是的。然后还有一大堆劣势。

Yes. And then there's bunch of disadvantages.

但是我

But I

说吧，是的。请继续，抱歉。

do Yeah. Go ahead. Sorry.

不，不。但，就像，你知道我想说什么。所以，就像，你有这样的格局，新创公司可以进入并做一些现有企业要么做不到，要么甚至没有明显的现有企业去做的事情。因为，再次强调，你可能是在将服务转化为人工智能劳动力，而之前甚至没有软件公司尝试过这样做。

No. No. But, like, I mean, you know where I'm going. So, like, but, like so you have this you have the exact makings of a landscape where where new startups can come in and do things that incumbents either can't or there's no obvious incumbent to even do that thing. Because, again, you're taking maybe, like, services and turning them into AI labor, and there was no software incumbent previously to even attempt to do that.

然后，有些现有企业在执行某些领域时面临大量复杂性，他们不会重新调整整个内部工程流程以10倍的速度前进。而一个全新的初创公司可以做到这一点，并立即获得大公司的规模。所以这是第一次

And then you have incumbents that have a whole lot of complexity in terms of their ability to go and execute in some of these spaces, and they're not gonna retool their entire internal engineering workflows to move at 10 x the pace. And so a brand new startup can go and do that and then instantly get the scale of a larger company. So it's the first time

在历史上，你没有任何

in history where you have none

大公司的劣势之一。而大公司传统的优势在于拥有规模化分销能力。规模化是因为你可以审视一个功能，然后说，我们下个月就要开发它。显然这更难，因为你知道，这里面有很多复杂性。但至少你有人力资源去实现它。

of the disadvantages of a big company. And the the traditional advantage you have as a big company is you have scaling of distribution. Scale because you can look at a feature and you say, we're gonna go build that next month. And obviously, it's harder because there's, like, you know you know, there's there's just lots of complexity to that. But at least you have the the human, you know, power to go do that.

现在作为初创公司，你立刻就能获得规模化，因为有了后台智能体等等。所以这就变成了分销游戏，现在很多这样的软件可以像病毒一样传播，这在十或十五年前是不可能的。因此我们某种程度上中和了许多现有企业的优势，所以这对全新的初创公司来说是一个成熟的机遇。通常会是刚大学毕业的人说，嘿，这是我第一次创业。他们疯狂到不知道这有多难。

Now as a startup, you instantly have scale because, you know, background agents, etcetera. And so then it's a distribution game, and a lot of these, you know, pieces of software can go viral now in a way that wasn't possible ten or fifteen years ago. So we've kind of neutralized a lot of the incumbent advantages, and so thus it's a ripe opportunity for brand new startups. Often will be people just like coming right out of college saying, hey, it's my first time building a company. Like, they're crazy enough to not know how hard it is.

所以他们会直接跳进那些我们原本认为市场已经饱和的领域。你不可能在那里创办公司，但就会有新的初创公司真的去做了，并且在这些领域实际打造出真正的公司。

So they'll jump right into markets that otherwise we would assume are like it's art like, the market's already solved for. There's no way that you're gonna build a company, and you'll just have new startups that actually go and do it and and actually produce real, you know, real companies in these spaces.

是的。因为这非常关键，因为真正发生的是，这就是为什么说这是一次真正的平台变革。硅谷已经多次见证这种情况，这就是为什么经常有人怀疑这是不是又在喊'狼来了'？因为大家都知道，当平台发生变革时，正是初创公司占据优势的时刻。

Yeah. Because I I mean, this is just so critical because it what what's really happening is this is why you know it's an actual platform ship. So Silicon Valley has seen this movie many times before, and it that's why often there's a lot of this, you know, is this crying wolf or not? Because everybody knows that when there's a platform shift, that's the moment in time that that startups are at Right. At an advantage.

没错。所以每次平台变革时，每个人都会说，哦，就是这次了。这会重塑一切，然后并没有，人们就会觉得，哦，总是现有企业赢家通吃。

Yes. And and so each time there's a platform shift, like, everybody's like, oh, this is it. This is gonna reinvent every and then it doesn't, and people get really like, oh, it's always incumbents.

但是

But

历史上看，现有企业的优势被严重高估了。嗯。真的，我是说这次的情况，互联网是否颠覆了微软就是一个非常有趣的问题。当然，现在它是一家3万亿美元的公司，但并不是以我们理解互联网的方式存在。就像九十年代我们拥有的消费者、平台、资产都没有了。是的。

historically, like, the advantages to incumbents are wildly overestimated. Mhmm. And, really, I I mean, this is this one where, you know, like, you know, was did the Internet undo Microsoft or not undo Microsoft is a super interesting thing. Because, of course, there's a $3,000,000,000,000 company now, but not on the Internet in a way that you think about the Internet. Like, none of the consumers, none of the platforms, none of the assets that we had in the nineties Yeah.

成为了互联网资产。是的。我的意思是，即使你看今天的Azure，它也是一项了不起的成就。它并没有在任何地方运行Windows。对吧。

Became Internet assets. Yeah. I mean, even if you look at Azure today, it's an amazing accomplishment. It's not running Windows anywhere. Right.

是的。而且我认为这就是为什么，你知道，说‘哇，这对谷歌是好是坏？’并不疯狂。因为如果你不进行转型，很多事情会变得非常非常困难。而事实证明，历史上即使你确实转型了，你其实并没有真正成功，你只能等待时间过去。

Yeah. And and I think that that that's why, you know, it's not crazy to go, wow. Is this gonna be good or bad for Google? Because there's a bunch of stuff that becomes really, really difficult if you if you don't make a transition. And then it turns out, historically, even if you do make the transition, you really didn't, and you just have to wait for time to pass.

是的。嗯，这就像英特尔和GPU的情况。对。就像他们在2005年2月错过了GPU。对吧。

Yeah. Well and this this is like Intel with the GPU. Yes. Like, they they missed the GPU in in 02/2005. Right.

他们错过了收购公司、做这项工作或其他任何机会。而且他们某种程度上也错过了数据中心。只是花了更长时间才意识到他们也错过了那个。

And they missed the opportunity to buy the company, to do the work, or whatever. And they kind of missed the data center too. It just took a longer time to figure out that they missed that as well.

我们对颠覆的定义相当狭隘，意思是，我们期望现有巨头必须输给这个新初创公司。而那从未发生。从未发生。所以这就像是整个广播、电视、剧院、电影的类比。而且它是

And we have a pretty narrow definition of, like, disruption in the sense of, like, we expect that the incumbent has to lose for this new startup. And that never happens. And never happens. And so it's the it's the whole radio TV, you know, theater, you know, movie analogy. And it's

就像是，不。就像是，它

like it's like, no. Like like, it

结果是微软可以成为一家4万亿美元的公司，而所有这些新类别可能会出现，如果一切都完美类比桌面时代，微软本应拥有它们。但它们就是没有。而且它们都协同工作，对吧。作为一种生态系统，因为事实证明，软件确实吞噬了世界，这些市场实际上比我们意识到的要大100倍。所以现有巨头可以增长，然后你还有新的颠覆者。是的。

turns out that Microsoft can be a $4,000,000,000,000 company, and you can have all these new categories emerge that maybe Microsoft should have owned if everything was, like, perfectly analogous to the desktop days. But they just don't. And it all works together as Right. As one sort of ecosystem because it just turns out, like, software did eat the world, and these markets are actually just, a 100 times larger than what we realized. And so incumbents can grow, and then you have new disruptors Yeah.

那种事情会随着时间推移逐渐显现出来。

That that sort of emerge along the way.

是的。任何时候引入能降低边际成本的新技术，市场都会扩大。现有企业也能做到。但我要说，当新的用户行为和购买行为出现时，现有企业表现很差。特别是...

Yeah. Anytime you bring in a new technology that brings the marginal cost down, then, like, the market's gonna expand. And, like, the incumbents can do it. I will say incumbents are very bad when new user behaviors and buying behavior show up. Like, in particular Yeah.

是的。因为他们不知道如何适应它。AI绝对是一种新的用户行为和购买行为。所以这对初创公司来说是个巨大优势，因为要让大公司围绕新用户行为转型，需要调整整个公司，从市场营销到后端支持的所有环节。这实在是负担太重了。

Yeah. Because they don't know how to cater to it. AI is definitely a new user behavior and a new buying behavior. And so this is very much an advantage of startups just because, you know, to change a a large company around a new user behavior cuts to the entire company, everything from basically marketing all the way to support in the back end. That's just too much of a lift.

我的意思是，如果你看看最佳实践，比如过去18个月里如何构建智能体的方法，我认为我们已经经历了2-3次架构模式的变化。

I mean, if you look at the if you look at the best practices of of, like, even how you would create an agent in the last eighteen months, I think we've gone through two to three architecture pattern changes.

是的。

Yeah.

所以，这真的很麻烦，你知道吗，我几乎跟不上。作为一个中等规模的公司，我无法想象如果你有更多需要协调的人员会怎样。

And so, like, it's shit like you know, I can barely keep up. Yeah. As a, you know, midsize company, I can't imagine if you just had so many more people you had to, like, organize around that.

颠覆性新技术需要人们理解如何以不同方式使用它们，随着最佳实践的演进，使用方式也会随时间变化。这种灵活性只能...微软最初是如何推出Copilot的其实是个很有趣的问题，因为它是最早非常成功的产品之一。我最近了解到它是由OpenAI创建的，这就解释得通了。

Disruptive new technologies require Yeah. You people to understand how to use them in consumption in different ways that evolves over time as you get best practices. Like, that sort of flexibility can only come it's actually a very interesting question of how Microsoft actually did Copilot to begin with because it was actually one of the first ones that these products was very successful. I learned recently it was created by OpenAI, so that kind of explains it.

嗯，但是但是你得至少但是但是，就像，等等。你得启动一个人。

Well, but but you had to at least but but, like, wait. You had start a person.

而你不得不而且而且而且很凑巧，他是个创业型的人。

And you had to And and and conveniently, he was a startup guy.

是的。是的。所以100%是的。而且很凑巧

Yeah. Yeah. So A 100%. And conveniently

即便如此，顺便说一句，它它很了不起，它竟然出自我的

Even then, by the way, it's it's remarkable that it came out of my

总是令人惊讶的是，当这些大公司推出新的、具有定义性的东西时，你100%会去探究，然后你会发现，它基本上就是臭鼬工厂项目。基本上，他们没什么可失去的。没错。而且它没有干扰。就像，iPod就是个经典例子。

the what's always remarkable is when when something new and defining comes from these big companies, you a 100% of the time, you look into it, and you're like, it was basically Skunkworks. Basically, they had nothing to lose. Exactly. And it didn't interfere. Like, the iPod is this classic one.

人们总是或者甚至iPhone。人们总是谈论，比如，勇敢之举，就像他们的电脑业务已经死透了，死透了，死透了。就像，它只有3%的市场份额，毫无起色。所以，就像，它就像一次万福玛利亚传球。iPod就是一次万福玛利亚传球。

People all or the iPhone even. People always talk about, like, the brave it's like their computer business was dead, dead, dead. Like, there was it was, like, 3% share and going nowhere. So, like, it it was like a hail Mary. The iPod was a hail Mary.

对。然后手机，他们当时并不在手机行业。这没关系。而他们做了一部手机，一个小电脑，这让诺基亚感到震惊。哇哦。

Right. And then the phone, they weren't in the phone business. It didn't matter. And the fact that they made a phone a little computer is what Nokia was like. Woah.

那是什么？我认为人们真的需要理解一个事实，正如你所说，大公司会存在非常非常长的时间。这是我在过去几周看到很多人讨论的事情。你知道吗？哦，这基本上就是克莱顿·克里斯坦森的《创新者的窘境》。

What is that? And I and I think people really need to wrap their heads around just the fact that, to your point, that the the big companies stay around a very, very long time. And this is something I've seen a bunch of people in the past couple weeks. You know? Oh, it's basically Clay Christensen, Innovators' Lemma.

当然，没人真正读过这本书。克莱顿是个很棒的人。你知道吗？我教书的时候他就在走廊那头。关键是这本书讲的是大约两英寸半的磁盘驱动器。

Of course, nobody's ever read the book. And and and Clay is a great guy. You know? He was down the hall when I was teaching there. And the the thing is is that this is a book like about two and a half inch disk drives Yeah.

还有一堆非常疯狂的

And a bunch of really crazy

我真的认为这不适用于现代

I really don't think it applies to modern

软件。他忽略了，除了忽略了成本和低端、高端市场的问题外，更重要的是公司不会消失。嗯。你关于这一点的看法非常重要。

software. He he missed, aside from he missed the cost and low, high end and thing, but was also that the companies don't evaporate. Mhmm. And you your point about that is really important.

是的。

Yeah.

是的。所以它造成的结果是，每个人都担心这个阴影，所以你总是能看到这种情况。就像，我猜，当一家公司说我们不担心谷歌会这么做时，这正是你想听到的。对吧。

Yeah. And so what it does is there's this shadow that everybody is worried about, and so you just see it constantly. Well, like, the the I assume, like, you you when a company is like, we're not worried about if Google does this. That's what you wanna hear. Right.

因为，就像，我的意思是，你就是这样生活的。我意思是，因为你就像是经典人物，你知道，史蒂夫·乔布斯说过你是一个功能。嗯。而且这里不仅仅是你做。现在有两家老牌公司也在做这些事情。

Because, like, it's just I mean and you live this. I I mean because you you were like the classics, you know, Steve Jobs said you're a feature. Mhmm. And here not just that you do. There's, like, two old companies that do this stuff now.

个人而言，他只对德鲁说过那句话。所以对。对。我得到了那个的游戏版本。是的。

Personally, he only said that to Drew. So Right. Right. I got the game's version of that one. Yeah.

是的。

Yeah.

是的。但不是。但是但是有一件事

Yeah. But no. But but the the one thing that

关于克莱非常永恒的一点是，那个超越软盘或者，你知道，无论是什么东西的东西。在位者不想做违背他们商业模式的事情。当然。当然。而那部分是完全是永恒的。

is very timeless about about Clay would be would be the thing that has the thing that does transcend floppies or, you know, whatever the whatever thing is is. The incumbent doesn't wanna do something that's against their business model. Course. Course. And that that part is is fully timeless.

是的。而且按你说的，实际上我们很少说这些公司颠覆了自己。几乎从来不是他们颠覆了自己。而是他们进入了一个他们实际上没有，你知道，市场份额的市场

Yep. And to your point, it's actually very rare whenever we say these companies disrupted themselves. It's almost never the case that they disrupted themselves. It's the case that they they went after a market that they had no no actual, you know, you know, market share in

对。

Right.

然后它就成功了。

And it just worked.

那是微软的开发工具。是的。因为，就像，微软在开发工具方面已经没有业务了，因为没有人再编写Windows程序了。所以实际上，就像是我们能为云做什么？然后它就像

Which was the development tools for Microsoft. Yeah. Because, like, the window there was no business in Microsoft development tools anymore because nobody was writing Windows programs. So really, it was like, what could we do for the cloud? And it like

说实话，用AI来做这件事甚至有点傻。就像，AI非常有颠覆性，所以我们觉得，好吧。那么它允许初创公司对抗现有的大公司。但即使在非颠覆性技术中，初创公司也经常有机会对抗现有的大公司。例如，每年，我是一名基础设施投资者。

Honestly, it's even a little bit of a silly discussion to do this with AI. Like, AI is very disruptive, so we're like, okay. Well, then it allows startups to, like, work against incumbents. But even in nondisruptive technologies, startups often have a play against incumbents. So for example, every year for I'm an I'm an infrastructure investor.

过去十年里，每年在AWS re:Invent之后，我都会进入治疗师模式。然后

Every year for the last ten years after AWS re:Invent, I move into therapist mode. And

所有 所有

all all