Cursor的崛起：工程师爱不释手的3亿美元年收入AI工具 | Michael Truell（联合创始人兼CEO）

Andrew是OneSchema的CEO，我们播客的长期赞助商之一。

Andrew is CEO of OneSchema, one of our longtime podcast sponsors.

欢迎你，Andrew。

Welcome, Andrew.

谢谢邀请，Lenny。

Thanks for having me, Lenny.

很高兴来到这里。

Great to be here.

那么OneSchema有什么新动态？

So what is new with OneSchema?

我知道你们和一些我喜欢的公司合作，比如Ramp、Vanta和Watershed。

I know that you work with some of my favorite companies like Ramp and Vanta and Watershed.

听说你们新推出了一款数据导入产品，可以自动化团队花费数小时手动导入、映射和整合CSV与Excel文件的工作。

I heard you guys launched a new data intake product that automates the hours of manual work that teams spend importing and mapping and integrating CSV and Excel files.

是的。

Yes.

我们刚刚发布了

So we just launched

这个

the

二点

two point

零版本的OneSchema文件馈送系统。

o of OneSchema FileFeeds.

我们从头开始用AI重建了它。

We have rebuilt it from the ground up with AI.

我们看到太多客户带着数据工程师团队来找我们，他们苦于清理杂乱电子表格所需的手动工作。

We saw so many customers coming to us with teams of data engineers that struggled with the manual work required to clean messy spreadsheets.

文件馈送2.0版本能让非技术团队仅通过简单提示就实现CSV和Excel文件转换流程的自动化。

Filefeeds two point o allows nontechnical teams to automate the process of transforming CSV and Excel files with just a simple prompt.

我们支持所有最棘手的文件集成方式，包括SFTP、S3甚至电子邮件。

We support all the trickiest file integrations, SFTP, s three, and even email.

我可以告诉你，如果我的团队必须构建这样的集成系统，那该有多好

I can tell you that if my team had to build integrations like this, how nice would it

能把这项任务从我们的路线图中移除，转而使用类似OneSchema这样的解决方案？

be to take this off our road map and instead use something like one schema?

完全同意，Lenny。

Absolutely, Lenny.

我们听过太多因单个错误记录导致系统中断的恐怖故事，无论是交易记录、员工档案还是采购订单，应有尽有。

We've heard so many horror stories of outages from even just a single bad record in transactions, employee files, purchase orders, you name it.

调试这些问题往往如同大海捞针。

Debugging these issues is often like finding a needle in a haystack.

OneSchema能阻止不良数据进入系统，并自动验证文件，生成包含所有问题文件的详细错误报告。

OneSchema stops any bad data from entering your system and automatically validates your files, generating error reports with the exact issues in all bad files.

要知道导入错误数据会给客户带来各种麻烦，并迅速失去他们的信任。

Know that importing incorrect data can cause all kinds of pain for your customers and quickly lose their trust.

Andrew，非常感谢你参加今天的对话。

Andrew, thank you so much for joining me.

如果你想了解更多，请访问1schema.co。

If you wanna learn more, head on over to 1schema.co.

网址是1schema.co。

That's 1schema.co.

在构建光标AI产品的过程中，你学到的最反直觉的事情是什么？

What is the most counterintuitive thing you've learned so far about building cursor building AI products?

我认为对我们来说最反直觉的一点——之前也稍微提到过——就是我们完全没预料到会自己开发模型。

I I think one thing that's been counterintuitive for us, hinted at at it a little bit before, But is we we definitely didn't expect to be doing any of our own model development when we started.

就像之前说的，当我们进入这个领域时，有些公司从一开始就专注于从头训练模型。

As mentioned, you know, when we when we got into this, there were companies that were immediately from the get go going and just focusing on kind of training model from scratch.

我们计算过训练GT4的成本，很清楚这不是我们能承担的事情。

And we had done the calculation for it to to to train GT four and just knew that that was not gonna be something we were gonna be able to do.

而且感觉把精力放在这个方向有点不对，因为市面上已经有很多优秀的模型了。

And also felt a little bit like focusing one's attention in the wrong area because there are lots of amazing models out there.

为什么要重复其他公司已经做过的工作呢？

And why do all this work to replicate kind of what other players have done?

尤其是在预训练方面，你知道的，就是从一个一无所知的神经网络开始，然后教会它整个互联网的内容。

Especially on the pre training side of things, you know, taking a neural network that knows nothing and then teaching it the whole Internet.

所以我们原本以为完全不会涉足这一领域。

And so we thought we were we weren't going be doing that at all.

从一开始我们就清楚地看到，现有模型其实可以为我们做很多它们目前没做的事，因为缺乏合适的工具。但实际上，我们现在做了大量模型开发工作。

And it seems clear to us from the start that the the existing models, there were lots of things that they could be doing for us that they weren't doing because, you know, there wasn't the right tool built for In fact, though, we do a ton of model development.

在内部，这是我们招聘工作的重点方向，并且已经组建了一支很棒的团队。

And internally, it's a it's a big focus for us on the hiring front and have assembled a fantastic team there.

这在产品质量方面也为我们带来了巨大提升。

And it's also been a big win on the the product quality side of things for us.

现在Cursor里每个神奇功能背后都或多或少有定制模型的参与。

And at this point, every magic moment in cursor involves a custom model in some way.

所以这确实既反直觉又令人意外。

And so that that was definitely counterintuitive and and surprising.

这是个渐进的过程，最初我们训练自有模型的应用场景确实不适合使用那些大型基础模型，后来这个模式在一个应用场景取得巨大成功后，就逐步拓展到其他领域了。

And it's it's been a gradual thing where, you know, there was an initial use case for for training our own model where it really didn't make sense to use any of the biggest foundation models that was incredibly successful, kind of moved to another use case that worked really well and had had been going from there.

进行这类模型开发的一个有益之处在于谨慎选择切入点，不试图重新发明轮子，也不把精力放在那些顶级基础模型已经非常出色的领域，而是聚焦于它们的弱点以及如何互补。

And one of the, you know, the helpful things in doing this sort of model development is picking your spots carefully, not trying to reinvent the wheel, not trying to focus on places maybe where the best foundation models are are excellent, but instead kind of focusing on their weaknesses and how you can complement them.

我想很多人听到你们拥有自己的模型会感到惊讶。

I think this is gonna be surprising to a lot of people hearing that you have your own models.

当人们谈论Cursor和这个领域的其他公司时，他们通常称之为GPT包装者。

When I you know, when people talk about Cursor and all the folks in the space, they would kind of call them GPT rappers.

他们只是基于ChatGPT或Sonnet进行开发。

They're just sitting on top of ChatGPT or Sonnet.

而你们说的是拥有自己的模型。

And what you're saying is that you have your own models.

谈谈背后的技术架构吧。

Talk about just like the stack behind the scenes.

是的。

Yeah.

当然。

Of course.

我们确实会以多种方式使用中国的大型模型。

So we definitely use, the biggest Chinese models at, you know, a bunch of different ways.

它们是实现Cursor用户体验的重要组成部分。

They're really important components of of bringing the cursor experience to people.

我们使用自有模型的场景。

The places where we use our own models.

有时是针对那些基础模型由于成本或速度原因根本无法满足的特定用例。

So sometimes it's to survey use case that a foundation model wouldn't be able to serve at all for cost or speed reasons.

这方面的一个例子就是自动补全功能。

And so one example of that is the auto complete side of things.

这对不编程的人来说可能有点难理解，但编程这种工作形式很特殊，有时通过观察你的操作就能完全预测接下来5到30分钟的工作内容。

And so this can be a little bit tricky for people who don't code to understand, but code is this weird form of work where sometimes really the next five, ten, twenty, thirty minutes of your work is entirely predictable from looking over your shoulder.

我想把这个与写作做个对比。

And I would contrast this with writing.

很多人都熟悉Gmail的自动补全功能，以及发短信或邮件时出现的各种补全形式。但由于仅凭之前写的内容通常很难预测后续内容，这些功能的帮助其实很有限。

So writing, you know, everyone or lots of people are familiar with, you know, Gmail's autocomplete and kind of the different forms of autocomplete that show up when you're trying to post text messages or emails or things like They can only be so helpful because often it's just really not clear what you're gonna be writing just by looking at what you've written before.

但在代码中，有时当你修改代码库的某部分时，很明显你还需要改动代码库的其他部分。

But in code, sometimes when you edit a part of a codebase, it's just you're gonna need to change things and then other other parts of a codebase.

因此Crusher的核心功能之一就是这种强化版的自动补全体验，它能预测你接下来要在多个文件、多个位置进行的操作。

And it's it's entirely clear how you're gonna need to change And so one core part of Crusher is this really souped up auto complete experience where you predict like the next set of things you're going be doing across multiple files across multiple places for a file.

要让模型擅长这种用例，首先速度很关键——这些模型必须非常快，要在300毫秒内给出补全建议。

And making models good at that use case, one, there's a speed component of those models need to be really fast and need to give you a completion within three hundred milliseconds.

此外还有成本考量——我们正在运行海量的模型计算。

There's also this cost component of we're running tons and tons and tons of molecules.

每次按键，我们都需要不断调整对你下一步操作的预测。

Every keystroke, we need to keep changing our prediction for what you're going do next.

这也是一个非常特殊的应用场景：你需要模型不仅擅长完成通用文本序列的下一个标记，更要擅长自动补全一系列代码差异——即分析代码库中的变更，并预测接下来可能修改的部分。

And then it's also this really specialty use case of you need models that are really good not at completing the next token of just to, like, a generic text sequence, but are really good at auto completing a series of diffs, You know, looking at what's changed within a code base and then breaking the next set of things that are going to change.

包括删除和新增的内容等等。

You know, both deleted and added and all of that.

我们发现专门针对这项任务训练模型取得了巨大成功。

And we we found a a ton of success in training models specifically for that task.

所以在这个领域，你看，并没有基础模型的参与。

So that's a place where, you know, no foundation models are involved.

这算是我们自己的特色。

It's kind of our own thing.

我们在应用中没有过多标注或品牌宣传，但这确实是Crusher的核心功能部分。

We don't have a lot of labeling or branding about this in the app, but that, you know, cores of, you know, powers a very core part of Crusher.

另外，我们使用模型的另一类场景是辅助SONNET、Gemini或GPT这类系统。

And then, you know, another set of places where we're using our models are to to help things like SONNET or Gemini or GPT.

这些模型既作用于那些大型模型的输入端，也作用于输出端。

And those sit both on the input of those big models and on the output.

在输入侧，这些模型会遍历整个代码库，试图找出需要展示给大型模型的部分。

On the input side of things, those models are searching throughout a code base, trying to figure out the parts of a code base to show to one of these big models.

你可以把它想象成专门为代码库定制的迷你谷歌搜索，用于找出需要展示给大型模型的相关代码片段。

You can kind of think about this as like a mini Google search that's specifically built for finding the, you know, relevant parts of a codebase to show one of these big models.

而在输出侧，我们会处理这些模型针对代码库建议的修改草图。

And then on the output side of things, you know, we take the sketches of the changes that these models are suggesting you make with that code base.

然后，我们还有一些模型负责填充细节，比如那些最智能的模型完成高层次思考的部分。

And then, you know, we have models that then kind of fill in the details of like, you know, the high level thinking is done by these smartest models.

它们会花费一些token来完成这项工作。

They spend a few tokens on doing that.

接着，这些更小、更专业且速度极快的模型，配合一些推理技巧，将这些高层次的变更转化为完整的代码差异。

And then these smaller specialty, incredibly fast models coupled with some inference tricks then take those high level, changes and turn them actually into full code diffs.

因此，在需要专业任务的场景中，这对提升质量非常有帮助。

And so it's been super helpful for, pushing on on quality in places where you need a specialty task.

同时，这对提升速度也极为有益，而速度对我们产品品质来说同样是一个至关重要的维度。

And it's been super helpful for pushing on speed, which is such an important dimension of product quality for us, too.

这太有趣了。

This is so interesting.

我刚刚在播客中采访了OpenAI的首席产品官Kevin Wheel，他将这称为模型组合。

I just had Kevin Wheel on the podcast, CPO of OpenAI, and he calls this the ensemble of models.

他们的工作方式也是如此，利用每个模型的最佳特性，正如你提到的，还能享受使用更廉价模型的成本优势。

That's the same way they work to use the best feature of each one and to your point, the cost advantages of using cheaper models.

这些开放的其他模型，是基于Llama之类的开源模型吗？

These open these other models, are they based on like Lama and things like that?

就是你们直接接入并构建的开源模型？

Just open source models that you guys plug into and build on?

是的。

Yeah.

我们在这个领域的工作方式非常务实，不想重复造轮子。

So, we try try to be very pragmatic about the place that we're going to do this work and we don't want to reinvent the wheel.

所以我们会从现有的最佳预训练模型开始，通常是开源模型。

So starting from the the very best, you know, pre trained models that exist out there, often open source ones.

有时也会与那些不公开权重的大模型提供商合作。

You know, sometimes in collaboration with these big model providers that that don't share their weights out into the world.

因为我们不太在意能否逐行读取权重矩阵来获得特定输出，我们更关注的是训练和后期调校这些模型的能力。

Because the thing we care about less is, you know, the ability to read line by line, you know, the matrix of weights that then, you know, go to give you give you a certain output and it's we just care about the ability to kind of to to train these things, to post train them.

总的来说，确实主要是开源模型，有时也会与闭源提供商合作进行调优。

And so by and large by and large, yes, open source models, you know, sometimes working with the the closed source providers too to tune things.

这引发了一个许多AI创始人和投资者经常思考的讨论，那就是AI领域的护城河和防御性。

This leads to a discussion that a lot of AI founders always think about in investors, which is moats and defensibility in AI.

所以感觉定制模型是这个领域的一个护城河。

So it feels like one is custom models is is a moat in the space.

考虑到如你所说，不断有其他人入局试图抢占你的市场份额，你如何看待这个领域的长期防御性？

How do you just think about long term defensibility in the space knowing there's other folks, as you said, launching constantly trying to take you trying to eat your lunch?

我认为，确实存在一些方法可以建立惯性和传统壁垒。

I think that, you know, there are ways to to build in inertia and, you know, traditional notes.

但总的来说，我们处在一个必须持续努力打造最佳产品的领域，这个行业里的每个人都是如此。

But I, you know, I think by and large, we're in a space where, you know, it is incumbent on us to continue to try to build the best thing and and and everyone in this industry.

而且我真的认为这个行业的天花板如此之高，无论你建立了什么样的壁垒，都可能被后来者超越。

And I, you know, I truly just think that the ceiling is so high that kind of no matter what, you know, entrenchment you build, you can be leapfrogged.

我认为这类似于与过去的传统软件市场、传统企业市场有些不同的市场形态。

And I think that this resembles markets that are maybe a little bit different from normal software markets, normal enterprise markets of the past.

让我想到的一个例子是1999年的搜索引擎市场。

You know, I think one that comes to mind is is the market for search engines at the 1999.

或者说，在九十年代末到两千年代初的时候。

Or, you know, at the end of the the nineties and beginning of the two thousands.

我认为另一个在很多方面与当前市场相似的市场，其实是七八十年代到九十年代外围计算机和小型计算机的发展历程。

I think another market that comes to mind that resembles this market in many ways is actually just like the development of the peripheral computer and mini computers, you know, in the seventies, eighties, nineties.

而且我认为，没错，在这些市场中，发展空间都极其广阔。

And I think that, yes, you know, in each of those markets, the ceiling was incredibly high.

要知道，当时转换赛道是完全可能的。

You know, it was possible to switch.

在相当长的时间里，你都能持续获得价值——无论是来自聪明人每多投入的一小时，还是每多投入的一笔研发资金。

You could keep getting value for like the incremental hour of a smart person's time, the incremental R and D dollar for a really long time.

你永远不会面临无有用之物可建的困境。

You wouldn't run out of useful things to build.

具体到搜索引擎领域（而非计算机领域），拥有分销渠道确实有助于产品优化——你可以调整算法，根据用户数据和反馈来优化学习模型。

And then, you know, in in in search in particular, not in the computer case, having distribution was was helpful for making the product better too, and that you could tune the algorithms, you could tune the learning based off of the the data and the feedback you're getting from users.

我认为，所有这些动态因素在我们的市场中也同样存在。

And I think that, you know, all of those dynamics exist exist in our market too.

所以我认为，也许对我们这样的人来说可悲但可叹的事实是，对世界而言却是个惊人的真相——我认为存在许多可以跨越式发展的机会。

And so I think that, you know, maybe the the sad sad truth for people like us, but then, like, the amazing truth for the world is I think that there are many leapfrogs that exist.

还有很多更有价值的东西等待被创造。

There's many, you know, more useful things to build.

我们距离未来五到十年要竞争的领域还很遥远，而持续推动这个引擎运转是我们的责任。

We're a long way away from where we compete in, you know, five, ten years, and it's incumbent on us to keep that engine going.

那么我理解的是，这听起来更像是一种消费者护城河——通过持续提供最佳体验来留住用户，而不是像Salesforce那样通过绑定全公司的合同来强制使用产品。

So what I'm hearing is it sounds like a lot more like a consumer sort of moat where it's just be the best thing consistently so that people stick with you versus creating lock in and things like that where they're just for like Salesforce where it's just contract to the entire company and you have to use this product.

是的。

Yeah.

我认为关键点在于，如果你所处的领域很快就没有有价值的事情可做，那处境就不太妙。

And I I think the the important thing to note is, you know, if you're in a a space where, like, you kind of run out of useful things to do very quickly, then that's, you know, that's not a great situation to be in.

但如果你处在一个需要重大投资且

But if you're in a place where, you know, big investments and,

你

you

要知道，让越来越多优秀人才沿着正确道路工作能持续创造价值，这样你就能获得研发的规模效益，可以深入地在正确方向上钻研技术，最终达到一个可防御的竞争地位。

know, having more and more great people working on the right path can keep giving you value, then you can get kind of these economies of scale of R and D and you can kind of have you know you know, deeply work on the technology in the right direction and and get to a place where that is defensible.

但没错，确实存在某种消费者倾向性，我真心认为关键在于打造尽可能最好的产品。

But yes, it is it is you know, I think there's there's a consumer like tendency to it, and I really think it's just, you know, about building the best thing possible.

你认为未来这个领域会有一家独大吗？

You think in the future there's one winner in this space?

还是说会出现多个类似产品共存的局面？

Or do you think it's going be a world of a number of products like this?

我认为这个市场非常庞大。

I think the market is just so very big.

这也是你之前问到的关于集成开发环境（IDE）的问题之一。

And this is also one thing that, you know, you asked about the IDE thing early on.

我认为有些人在思考这个领域时犯了一个错误，他们回顾了过去十年的IDE市场，然后问：'谁从中赚到钱了？'

And one thing that I think a trip of some people that were thinking about the space is like, they looked at the IDE market of the past ten years and they said, you know, who's making money off of that?

或者说，这是一个极度分散的市场，每个人都有自己的配置和工具，只有一家公司真正通过打造优秀编辑器盈利，但这家公司的规模也就那样。

Or is like, you know, there's all these, it's this super fragmented space where everyone kind of has their own thing with their own configuration and, you know, there's one company that commercially likes, you know, that actually makes money off of making great, great editors, but like that company is only so big.

于是他们得出的结论是未来也会是这样。

And then the conclusion was it was going to look like that in the future.

我认为人们忽略了一点，就是在2010年代为程序员开发编辑器能做的事情非常有限。

And I think that the thing that people missed was that there was only so much you could do building an editor in the 2010s for coders.

那家靠编辑器赚钱的公司，主要在做的事情是让代码库导航更便捷，以及进行一些错误检查和类型检查。

And the company that made money off of editors was doing things like making it easy to navigate around a code base and doing some error checking and type checking for things.

还有提供优质的调试工具，虽然这些都很有用。

And having good debugging tools, but like which were all very useful.

但我认为这个市场确实非常庞大。

But I think that the the set of things you can build for programmers, think the set of things you can build for knowledge workers in many different areas just goes very far and very deep.

我认为我们所有人面临的问题，实际上是如何自动化处理大量繁琐工作和知识工作，真正改变我们面前所有知识工作领域，使其更加可靠和高效。

And I think that really kind of like the the problem in front of all of us is is like the automation of a lot of busy work and knowledge work and really changing all areas of knowledge work in front of us to be much reliable and more productive.

所以，说了这么多，我想表达的是，我认为我们所在的市场确实非常庞大。

So that that was all, you know, a long winded way to say, I think the market's really, really big that we're in.

我认为它比人们过去意识到的要大得多，比过去为开发者构建工具的市场要大得多。

I think it's much bigger than people have realized than, you know, the other, you know, building tools for developers in the past.

我认为将会出现多种不同的解决方案。

And I think that there will be a bunch of different solutions.

我认为会有一家公司——是否是我们还未可知。

I think that there will be one company and to be determined if it's going to be us.

但我确实认为会有一家公司打造出这个通用工具，能够构建世界上几乎所有的软件，这将是一项跨时代的巨大业务。

But I do think that there will be one company that builds the the general tool that builds almost all the world's software and that will be a very, very generationally big business.

但我认为在市场中，你可以占据某些细分领域或软件开发周期中特定环节的利基市场。

But I think that there will be kind of niches you can occupy in doing something for a particular segment of the market or for a very particular part of the software development life cycle.

但通用编程将从编写正式的编程语言转向更高层次的表达方式。

But the general programming shifts from just writing formal programming languages to something way higher level.

这就是你购买并使用来实现这一目标的应用程序。

This is the application you you purchase and use to do that.

我认为那里最终会有一个通吃的赢家，而且这将是一门非常大的生意。

I think that there will be generally one winner there, and and it will be a very big business.

真带劲。

Juicy.

顺着这个思路，有趣的是微软实际上曾处于这个领域的中心位置，拥有出色的产品和强大的分发渠道。

Along those lines, it's interesting that Microsoft was actually, like, right at this set like, at the center of this first with an amazing product, amazing distribution.

你说Copilot是让你突破认知障碍的关键，让你意识到这里可能有大事发生。

Copilot, you said, was, like, the thing that got you over the hump of, wow.

这里可能蕴藏着巨大的机遇。

There could be something really big here.

但感觉他们并没有占据优势。

And it doesn't feel like they're winning.

反而感觉他们正在落后。

It feels like they're falling behind.

你认为这是怎么回事？发生了什么？

What do you think what what do think happened there?

我认为Copilot目前未能达到某些人预期，有其特定的历史原因。

I think that there are, like, specific historical reasons why Copilot might not have lived up or so far have have kind of, lived up to the expectations that some people have for it.

此外还存在一些结构性原因。

And then I think that there are structural reasons.

我认为结构性的原因是——需要说明的是，微软在Copilot案例中显然对我们工作有很大启发。

I think the structural reason is and to be clear, you know, Microsoft, you know, in the Copilot case, obviously, a big inspiration for our work.

总的来说，我认为他们做了很多了不起的事，我们也是许多微软产品的用户。

And in general, I, you know, think they do lots of awesome things and we're users of many Microsoft products.

但我认为这个市场对现有巨头并不太友好。

But I think that this is a market that's not super friendly to incumbents.

在一个对现有巨头友好的市场中，可能只有有限的事情可做，产品很快就会同质化，你可以将其与其他产品捆绑销售。

In that, you know, a market that's friendly to incumbents might be one where there's only so much to do, it kind of gets commoditized fairly quickly and you can bundle that in with other products.

而且不同产品之间的投资回报率差异非常小。

And where the ROI between, you know, different products is, you know, quite quite small.

在这种情况下，购买创新解决方案可能并不划算。

And, you know, in that case, perhaps it doesn't make sense to buy the innovative solution.

更合理的做法是直接购买与其他产品捆绑销售的东西。

It makes sense to just kind of buy that thing bundled in with other stuff.

另一个可能对现有巨头特别有利的市场是，从一开始你就把所有东西集中在一处，切换成本高得令人望而却步。

Another market that might be, you know, particularly helpful for incumbents is one where there's, you know, from from the get go, it's just like you have your stuff in one place and it's like really, really excruciatingly hard to switch.

而且，无论好坏，我认为在我们的情况下，你可以尝试不同的工具，然后决定哪个产品更好。

And, know, for better or for worse, I think in in our case, you can try out different tools and you can decide which product you think is better.

所以这对现有厂商不太友好，而对那些你认为能提供最具创新性产品的厂商更有利。

And so that's not super friendly to own comments and that's more friendly to whoever you think is going to have the most innovative product.

具体的历史原因是，据我了解，最初开发Copilot的那批人基本上都已经去了其他地方做其他事情了。

And then the specific historical reasons, like as I understand them are the group of people that worked on the first version of Copilot have by and large gone on to do other things at other places.

我认为要在所有可能参与开发这类产品的不同部门和团队之间进行协调有点困难。

I think it's been a little hard to kind of coordinate among all the different departments and parties that might be involved in in making something, like this.

我想回到Cursor这个话题。

I wanna come back to Cursor.

我喜欢问每个开发这类工具的人一个问题：如果你能坐在每个首次使用Cursor的新用户旁边，在他们耳边悄悄说几个使用技巧，让他们能更成功、最成功地使用Cursor，你会说哪一两个技巧？

A question I like to ask everyone that's building a tool like this, if you could sit next to every new user that uses Cursor for the first time and just whisper a couple tips in their ear to be more successful, most successful with Cursor, what would be, like, one or two tips?

我认为目前（我们希望在产品层面解决这个问题），要成功使用Cursor很大程度上需要对模型的能力有感觉，既要了解它能处理多复杂的任务，也要知道你需要向模型具体说明多少内容。

I think right now, and we'd wanna fix this at a product level, a lot of being successful with cursor is kind of having a taste for like what the models can do, both what complexity of a task they can handle and like kind of how much you need to specify, you know, things things to that that model.

但就像要对模型的质量有感觉，知道它的短板在哪里，它能做什么不能做什么。

But like having a a taste for the quality of the model and where its gaps exist and what it can do and what it can't.

目前我们在产品中做得不够好，比如没有很好地教育用户理解这一点，也没有提供明确的指导方针和操作范围。

And right now we don't do a good job in the product of like, you know, educating people around that, and maybe giving people some swim lanes, giving people some guidelines.

但要培养这种判断力，我有两条建议。

But so, to develop that taste, would give kind of two two tips.

第一条如前所述，要避免一次性对模型说‘这就是我要你做的完整任务’，然后要么对输出失望，要么全盘接受整个大任务的结果。

So one is, as mentioned before, would bias less toward like, hey, trying to have the model, like, you know, trying in one go to tell the model, hey, here's exactly what I want you to do, then seeing the output, and then either being disappointed or accepting the entire thing for an entire big task.

相反，我会把任务拆分成小块。你可以花同样的总时间进行说明，但要更分散地进行。

Instead, what I would do is I would chop things up into bits, and you can spend basically, you know, the same amount of time specifying things overall, but chopped up more.

这样你每次说明一小部分，获得一小部分成果，而不是写长篇大论一次性交代所有细节。

So you're specifying a little bit, you're getting a little bit of work, you're specifying a little bit, getting a little bit of work, and you know, not doing as much the like, let's write a giant thing, telling them all exactly what to do.

我认为目前那种方式很容易导致糟糕的结果。

I think that will be a little bit of a recipe for disaster right now.

因此在建议拆分任务的同时，我建议可以先在副业项目而非正式工作中尝试——特别是对于习惯传统软件开发流程的开发者们。

And so by saying toward chopping things up, at the at the same time, and it might make sense to do this on a side project and not on your professional work, You know, I would encourage people to especially, you know, developers who are kinda used to existing workflows for building software.

我鼓励大家主动尝试碰壁，在安全环境（比如副业项目）里大胆探索这些模型的极限，充分释放AI的潜力。

You know, I would encourage people to explicitly try to fall on their face and try to discover the limits of what these models can do by, you know, being ambitious and like kind of a safe environment, like perhaps a side project and trying to kind of go out and to keep you say AI to the fullest.

因为，你知道，有时候我们确实会遇到很多人，他们还没有真正给AI一个公平的机会，某种程度上低估了它的能力。

Because, you know, sometimes we do run or a lot of the time we run into people who haven't given the AI yet a fair shake and are kind of underestimating its abilities.

所以总的来说，建议把事情分解得更小。

So generally, them towards chopping things up and making things smaller.

但要探索你能做到的极限，可以明确地在一个安全环境中放手一搏，亲身体验一下。

But like to discover the limits of what you can do there, like explicitly just kind of try to go for broke in a safe environment and, you know, get get a taste for it.

你可能会惊讶地发现，在某些方面模型的表现出乎意料地好。

You might be surprised in some of the places where the model doesn't break.

我听到的核心意思是，要培养对模型能力的直觉，了解它能将想法推进到什么程度，而不是一味地引导它。

What I'm essentially hearing is kinda build a gut feeling of what the model can do and how far it can take an idea versus just kind of guiding it along.

我猜每次有新模型发布时，你都需要重新建立这种直觉。

And I bet that you need to rebuild this gut every time there's a new model launch.

比如当4.0版本发布时，你就得重新经历这个过程。

Like, when it's on it, I don't know, four point o comes out, you have to kind of do this again.

大体上是这样理解对吗？

Is that generally right?

是的。

Yes.

你知道，过去几年里，变化没有像人们初次接触大型模型时那么显著。

It's not you know, for the past few years, it hasn't been as big as, like, I think the the first kind of experience people have had with some of big models.

不过确实如此。

But, yeah.

这也是我们希望为用户更好解决的问题，减轻他们的负担。

You know, it this is also a problem we would hope to solve much better just for users and take the burden off of them.

但没错，每个模型都有略微不同的特性和个性。

But, yeah, each of these things have slightly different quirks and different personalities.

顺着这个话题，人们一直在争论像Cursor这样的工具，是对初级工程师更有帮助，还是对高级工程师更有帮助？

Kind of along these lines, something that people are always debating, tools like Cursor, are they more helpful to junior engineers, or are they more helpful to senior engineers?

它们能让高级工程师效率提升十倍吗？

Do they make senior engineers 10 x better?

它们能让初级工程师达到高级工程师的水平吗？

Do they make junior engineers more like senior engineers?

或者说，你认为目前谁从Cursor中受益最多？

Or Or do you think most of who do you think it benefits most today from Cursor?

我认为是全面的。

I think across the board.

这两个群体都能从中获得巨大收益。

Both of these cohorts benefit in big ways.

相对排名上有点难以判断。

It's a little hard to say on the relative ranking.

我想说的是，他们会陷入不同的反模式。

I will say they fall into different anti patterns.

我们看到初级工程师有点过于依赖AI处理所有事情。

So I would the junior engineers we see going a little too wholesale relying on AI for everything.

目前我们还没有达到可以在专业工具上端到端这样做的程度，你知道，在一个长期存在的代码库中与数十、数百人协作。

And we're not yet in a place where you can kind of do that end to end on a professional tool, you know, working with tens, hundreds of other people within a long lived code base.

而对于高级工程师来说，对许多人而言并非如此（虽然并非所有人）。

And then the senior engineers, for many folks, is not true for all.

实际上，我们经常看到这些工具被采用的方式之一是通过公司内部的开发者体验团队。

And we actually, often you know, one of the ways these tools are adopted is there's developer experience teams within companies.

通常这些团队由经验非常丰富的高级人员组成，因为他们负责构建工具来提高组织内其他工程师的生产力。

Often those are staffed by incredibly seniors, you know, senior people because often those are people who, you know, are building tools to make the rest of the engineers within an organization more productive.

我们已经看到了一些非常前沿的、突破性的尝试。

And we've seen some very, very, you know, boundary pushing kind of yeah.

就像我们看到有些人站在最前线，尽可能多地尝试采用这项技术。

Like we've seen people who are, you know, on on the the front lines of like really trying to adopt the technology as much as possible there.

但总体而言，我认为平均来看，资深工程师群体低估了AI能为他们带来的帮助，并固守现有工作流程。

But by and large, I would say on average as a group, the senior engineers underrate what AI can do for them and stick to their existing workflows.

所以相对排名有点难确定。

And so the relative ranking is a little hard.

我认为他们都陷入了不同的反模式。

I think they both have they they fall into to different different anti patterns.

但大体上，他们都从这些工具中获得了巨大收益。

But they both, by and large, get get big benefits with these tools.

这完全说得通。

That makes absolute sense.

我喜欢这种光谱两端的比喻，一边期望过高，一边期望不足。

I love that it's like two ends of the spectrum, like, expect too much, don't expect enough.

就像三只小熊的故事。

And it's like the the three bears.

这是个寓言吗？

Is that allegory?

对。

Yeah.

是的。

Yeah.

好的。

Okay.

没错。

Yeah.

那个对。

The yeah.

可能是比较资深，但还不是管理层。

Maybe the sort of senior, but not staff.

对。

Yeah.

你知道，

You know,

正好在中间。

it's right right in the middle.

嗯。

Mhmm.

有意思。

Interesting.

好吧。

Okay.

还有几个问题。

Just a couple more questions.

在担任这个角色之前，你希望自己知道些什么？

What's something that you wish you knew before you got into this role?

如果能回到Cursor刚开始时的迈克尔，虽然时间并不久远，你会给他什么建议？

If you could go back to Michael at the beginning of Cursor, which was not that long ago, and you could give him some advice.

你会告诉他些什么？

What's something that you would tell him?

困难之处在于，很多关键知识都是隐性的，很难用言语传达。

The tough thing with this is feels like so much of the, the hard one knowledge is tacit and a bit hard to communicate verbally.

生活的悲哀在于，在某些人类活动领域，你似乎必须经历失败才能真正学会正确的方式，或者需要身边有个杰出榜样。

And the sad fact of life feels like for, you know, for some areas of human endeavor, like you kind of do need to fall on your face to either either need to fall on your face to to learn the correct thing or you need to be kind of around someone who's a great example of kind of excellence in the thing.

我们在招聘方面就深刻体会到了这一点。

And one area where we have felt this is is is hiring.

我认为我们实际上在招聘方面尝试了极其高效的方式。

I think that we actually were so we we tried to be incredibly efficient on the hiring front.

对我们来说，无论是出于个人原因还是公司战略考虑，拥有一支世界一流的工程师和研究团队与我们共同开发Kirser都至关重要。

It was really important to us that, you know, for personal reasons and also for, I think actually for the company's strategy, having a world class group of engineers and researchers to work on on Kirser with us was going to be incredibly important.

同时还需要找到那些兼具求知欲和实验精神的人才，因为我们需要构建太多新事物。

Also getting people who fit, you know, a a sort of mix of, you know, intellectual curiosity and experimentation because there can be so many new things we need to build.

此外还需要保持学术诚信，甚至需要些许微观层面的悲观和直率——随着各种噪音增多，特别是公司业务规模扩大时，保持清醒头脑同样极为重要。

And then also kind of an intellectual honesty and maybe micro pessimism and bluntness because, you know, with all the noise and, you know, especially as the company's grown and, the business has grown, you know, keeping a level head, I think is an incredibly important too.

但引进合适的人才团队这件事，可以说除了产品开发之外，是我们最为重视的。

But getting the right group of people into the company, you know, was, you know, the the thing that maybe more than anything else apart from apart from building the product, we really really, you know, fussed over.

正因如此，我们实际上等待了很长时间才扩大团队规模。

And, you know, I we actually waited a long time to grow the team because of that.

我认为大多数人都听说过招聘太快的案例，但我们最初其实是招聘得太慢了。

And I think that most you know, many people you hear hire too fast, think we actually hired too slow to begin with.

这个问题本可以得到改善。

I think it could have been remedied.

我们本可以做得更好。

I think we could have been better

在这方面。

at it.

我们最终采用的招聘方法——虽然并不新颖，就是寻找我们认为真正世界一流的人才，并在某些情况下花费多年时间招募他们——最终对我们很有效，但一开始我们并不擅长。

And, the the method of recruiting that we ended up eventually falling into and working really well for us, which which isn't that novel of like going after people that we think are really world class and like recruiting them over the course of, in some cases, many years, ended up working for us in the end, but I I don't think we were very good at it to begin with.

因此我认为，我们在确定合适人选的标准（即哪些人真正适合团队、卓越的表现是什么样子）以及如何与潜在候选人沟通机会（特别是当他们本无求职意向时激发其兴趣）方面，都经历了艰难的学习过程。

And so I think that there were hard won lessons around both who was the right profile, like who actually made sense on that team, like what did what did greatness look like, And then how to, you know, talk with someone about about the opportunity and, you know, get them excited if they really weren't looking for anything.

在这方面我们积累了大量关于如何有效开展工作的经验，这确实花费了我们不少时间。

There there were lots of kind of learnings there about how to do that well, and that took us a bit of time.

对于当前正在招聘的人，有哪些值得分享的经验教训？

What are some of those learnings for folks that are, you know, hiring right now?

你们曾忽略或学到了哪些关键点？

What's something you missed or or learned?

我认为最初我们可能过于偏向寻找符合名校出身、非常年轻、在名校环境中取得高含金量成就的模板型人才。

I think, you know, to start with, we maybe we actually biased a little bit too much towards looking for people who fit the archetype of well known school very young, had done the things that were like, you know, high credential in those well known school environments.

实际上我们很幸运地在早期就遇到了许多优秀的资深从业者，他们愿意与我们共同奋斗。

And and actually, like, you know, I think Shand, we're lucky early on to find a lot of, you know, to find fantastic people who are willing to, you know, to do this with us, who were later career.

所以，是的，我认为我们一开始可能在人才定位上花费了过多时间。

And so, yeah, I think we should kind of spend a bunch of time on maybe a little bit of the wrong profile to begin with.

部分原因在于资历问题。

And part of that was a seniority thing.

还有部分原因在于兴趣和经验方面的考量。

Part of that was like, you know, kind of an interest and experience thing too.

我们也雇佣过非常年轻但极其优秀的人才，尽管在某些情况下他们与标准模板略有不同。

We have hired people who are excellent, excellent, excellent, and very young, but they maybe look, in some cases slightly different from, you know, being straight out of central casting.

另一个经验教训是我们的面试流程经历了很大演变。

You know, another lesson is just like we very much evolved our interview loop.

现在我们有一套精心设计的面试问题，核心环节是让候选人在现场进行两天的工作测试项目。

And so now we, you know, we have like a hand rolled set of interview questions and then, you know, kind of core to our core to how we interview too is is actually we have people on-site for two days and do do a project with us, a work test project.

这种方式效果很好，但我们也逐渐发现...

And that has worked really well, but increasingly you're finding that.

还有就是如何了解人们的兴趣所在，在对方完全没有求职意向时展现我们的优势，并展开这类对话。

And then, yeah, I think how to to learn about what people are interested in and, you know, put our best foot forward and and letting them know about the opportunity when they're really not looking for anything and have those conversations.

随着时间的推移，我们在这方面确实变得越来越熟练了。

There's definitely been, you know, gotten gotten better at that over time.

你有什么特别喜欢问的面试问题吗？

Do you have a favorite interview question that you like to ask?

我认为这个为期两天的工作测试——我们原以为它无法适用于更多人——却展现出了惊人的持久效果。

I think this two day work test, which we thought would not scale past a few people has been has had surprising staying power.

它的优点在于能让应聘者像对待真实项目一样从头到尾参与其中。

And the great thing about it is it lets someone go end to end on it like a a real project.

这并不是那种我们用作‘样板工程’的工作。

It's it's not, you know, work that we use as kind of a can candelissa project.

但它能让你通过两天时间观察一个真实的工作成果。

But it gives you two days of seeing like a real work product.

而且这对团队的时间投入要求并不高。

And it doesn't have to be incredibly time intensive on the team's time.

你可以把原本用于半天或一天现场面试的时间用在这上面。

You know, you can take the time you would spend in like a half day or one day on-site.

你可以把这两天时间分散开来，给候选人充足的时间来完成他们的项目工作。

You kind of spread it out over those two days and give someone a lot of time to do to do work on their projects.

这样实际上有助于扩大规模。

And so that can actually help it help it scale.

这能真正帮助你验证——你是否愿意和这个人共事，因为你会和他们相处几天，共进多餐。

And then it really helps you it helps you enforce, you know, do you wanna be around this person type test, because you are around this person, you know, for a few days and so the, you know, bunch of meals with them.

我们原本没预料到这个方式会持续使用，但它对我们的选拔流程价值非常大。

And so that one, we didn't expect that one to stick around, but that has been really, really important to our value to process.

这对激励员工也特别重要，尤其是在公司初创阶段。

And then also important to getting people excited at the especially the very early stages of the company.

因为在产品尚未被使用、知名度低，且产品相对不够完善时，你唯一能依靠的就是这个让某些人觉得特别并愿意加入的团队。

Because before people are using the product and know about it, and you know, when the product is comparatively like not very good, really the only thing you have going for you is, you know, a team of people that, you know, some some people find special and and want to be around.

这两天的相处让我们有机会让候选人了解我们，有时候甚至能说服他们决定加入我们。

And, know, the two days, it would would give us a chance to just like, you know, have this person, meet us and, in some cases hopefully get get convinced that they they wanna throw in with us.

所以这个方式的效果出乎意料。

And so, that one that one was unexpected.

严格来说不算面试问题，更像是某种预面试。

Not exactly an interview question, but kind of like a, you know, a forward interview.

终极面试问题。

The ultimate interview question.

为了明确你描述的内容，就是给他们一个任务，比如在我们的实际代码库中构建某个功能，与团队协作完成编码并交付。

So just to be very clear about what you're describing, it's a you give them an assignment, like build this feature in our actual code base, work with the team to, code it and ship it.

大致是这样吗？

Is that roughly right?

是的。

Yes.

不是那种...我们不会使用知识产权，不是完整交付，但类似一个模拟项目，

Not not like, so we don't use the IP, not shipped end to end, but yeah, it's like a mock, like,

你

you

知道，通常在我们的代码库中，这是个真实的小型两天项目，你需要独立完成从开始到结束的全流程。

know, very often in our code base, here's a real mini two day project, you're going to do it end to end, largely being left alone.

你知道，这其中也包含协作。

You know, there's there's collaboration too.

而且，我们是一家相当封闭的公司，在几乎所有情况下，候选人实际上就是和我们一起坐在办公室里。

And then, you know, we're we're a pretty imprisoned company, so and almost all cases, yeah, it's actually just sitting in office with us too.

你一直在说这种方法甚至沿用到了今天。

And you've been saying that this has scaled to even today's.

你们现在规模有多大？

How how big are you guys at this point?

我们目前接近60人。

So we are going on 60 people.

以你们的影响力和规模来说，这个团队很小。

So small for the scale and impact.

我原以为会比这个数字大得多。

That's I was I was thinking it'd be a lot larger than that.

是啊。

Yeah.

我猜其中占比最大的是工程师。

And I imagine the largest percentage is engineers.

是的。

Yeah.

最重要的是，明确地说，我们未来工作的很大一部分是组建一个更大、更优秀的团队，能够持续改进产品和我们为客户提供的服务。

The thing that's more than anything, and to be clear, you know, a big part of the the work ahead of us is is is building a group of people that is is bigger and awesome and can continue to make the the product better and the service we give to customers better.

所以你们并不打算长期保持这么小的规模。

And so you don't plan to stay that small, for longer.

我们希望不会。

We wouldn't hope so.

但，是的，人数较少的部分原因是，公司内部工程、研究和设计的占比非常高。

But, yeah, part part of the reason that, that number is is small is, the percentage of of engineering and and research and design is very high within the company.

因此，许多软件公司在拥有大约40名工程师时，总人数会超过100人，因为有很多运营工作，而且通常从一开始就非常以销售为导向。

And so, many software companies when they have, you know, roughly 40 engineers would be over a 100 people because there's lots of operational work and often they're very very sales led from the get go.

这确实相当耗费人力。

And that's just quite labor intensive.

要知道，我们最初是以极其精简和产品导向的方式起步的，如今我们已经服务了许多优质客户并逐步发展壮大，但在这方面仍有大量工作要做。

And, you know, we started from a place of being like incredibly lean and product led and like, you know, we now serve lots of our good customers and it built that out, but, you know, there's much more to do there.

我想问你一个问题。

A question I wanted to ask you.

AI领域正在发生翻天覆地的变化。

There's so much happening in AI.

每天都有新产品发布——现在甚至出现了专门报道AI领域每日动态的新闻简报，而且数量不少。

There's things launching every there's like newsletters, like, many newsletters whose entire function is to tell you what is happening in AI every single day.

作为处于这个领域核心地带（可以说是最炙手可热的中心）的公司，你们如何保持专注？又是如何让团队保持专注、埋头苦干而不被这些光鲜亮眼的新事物分散注意力的？

Running a company that's at the center, kind of the white hot center of the space, how do you stay focused, and how do you help your team stay focused and heads down and just build and not get distracted by all these shiny things?

我认为招聘是非常关键的一环。

You know, I think hiring is a big part of it.

如果能找到态度正确的人才——当然这一切都要打个星号备注，我觉得我们在这方面做得还不错。

And if you get people with the the right attitude and, you know, all of this should be asterisked in, like, you know, I think we're doing well there.

我认为我们在这方面其实还可以做得更好。

I think that like, you know, we'd probably be doing better there too.

而且，我认为作为一家公司，我们或许应该更多地讨论这个话题。

And, you know, it's something that we should probably talk even more about as a company.

但我认为，雇佣那些具备正确心态的人——那些不太关注外部认可，更专注于打造卓越产品、追求高质量工作成果的人，以及那些总体而言头脑冷静、情绪波动不大的人。

But I think that, you know, hiring people with the right disposition, you know, people who are less focused on external validation, more focused on building something really great, more focused on doing really high quality work, and people who are just generally kind of level level headed and maybe like the highs aren't very high and lows aren't very low.

我认为招聘能帮你克服很多困难，实际上这是全公司都明白的道理：你需要流程、需要层级、需要很多东西。

I think hiring can can get you through a lot here and I think that's that's actually like, you know, a learning throughout the company is that, you know, for any you you need process, you need hierarchy, you need lots of things.

但对于任何引入公司的组织工具而言，你最终希望从这个工具中获得的结果才是关键。

But for for any kind of organizational tool that you're intro introducing into a company, you know, the the result you're looking to get from that tool.

此外，通过雇佣那些具备你期望行为特质的人，你就能在很大程度上实现组织工具想要达成的效果。

Also, you know, you can go pretty far on like hiring people with the right behaviors that you want like, you know, to result from that for organizational thing.

具体来说，比如在工程方面，我们目前还能用较少的流程来运作。

You know, the specific example that comes to mind is we've been able to get away with not a ton of process yet on the engineering front.

虽然我认为我们需要增加一些流程，但就公司规模而言，通过雇佣真正优秀的人才，我们并不需要太多流程。

And I think we need a little bit more process, but for our size, not a ton of process by hiring people who I think are really excellent.

其中一点就是雇佣那些头脑冷静的人。

You know, one is, you know, hiring people who are level headed.

我认为第二点是经常讨论它。

I think two is just talking about it a lot.

我认为第三点是以身作则。

I think three is hopefully leading by example.

对我们个人而言，自2021、2022年以来，我们一直专业从事这项工作并研究人工智能。

And, yeah, for us personally, you know, we've, you know, since 2021, 2022, been professionally working on this and working on AI.

我们见证了各种技术和理念的潮起潮落——若回溯到2021、2022年，那时只有GPT-3。

And we've just seen a sea change of the comings and goings of various technologies and ideas of if you have to transport yourself back to 2021, 2022, this is GPT-three.

InstructGPT尚未问世。

InstructGPT doesn't exist.

没有DALL·E。

There's no DALL E.

没有Stable Diffusion。

There's no stable diffusion.

随后我们经历了所有这些图像技术的诞生，以及ChatGPT的崛起。

And then, you know, we've gone through all of those image technologies existing, TrakGPT in that rise.

还有GQ四这些新模型，各种不同的模态，所有视频相关的东西。

And, you know, GQ four, all of these new models, all these different modalities, all the video stuff.

实际上，只有极少数事物真正会对业务产生影响。

And only, you know, a very small number of these things really kind of affects affect the business.

所以我觉得我们已经建立起某种免疫系统，能够辨别哪些事件真正会对我们产生重要影响。

So I think we've kind of just built up a little bit of an immune system and kind of know know when when an event comes around that actually is really gonna matter for us.

这种动态也体现在：虽然讨论声浪很多很多，但真正重要的事情可能只有几件。我认为过去十年AI领域也反映了这种现象，学术界有太多关于深度学习的论文。

And this is, you know, this dynamic too of there being lots and lots and lots of chatter, but then maybe only a few things that really matter, think I has been mirrored in AI over the last decade where, there have been so many papers on deep learning in academia.

学术界有太多关于AI的论文。

So many papers on AI in academia.

但令人惊讶的是，AI的很多进步其实可以归功于一些持续存在的简单而优雅的想法。

Then the amazing thing is there are really a lot of I mean, a lot of the progress of AI can be attributed to some very simple, elegant ideas that have stayed around.

而绝大多数被提出的想法都没有持久力，也没有产生太大影响。

And the vast majority of of ideas that have been put out there haven't had staying power and haven't mattered a ton.

所以这种动态某种程度上反映了深度学习作为整个领域的发展轨迹。

And so the dynamic is a little bit mirrored in kind of the evolution of deep learning as a field overall.

最后一个问题

Last question.

你认为人们对AI发展趋势及其将如何改变世界建造方式的最大误解或尚未完全理解之处是什么？

What do you think people still most misunderstand or maybe don't fully grasp about where about where things are heading with AI in building and the way the world will change?

人们仍然有点过于偏向两个极端，要么认为一切会发展得非常快，要么觉得这些都是虚张声势。

People are still a little bit, you know, occupied too much either end of a spectrum of, you know, it's all going to happen very fast and, you know, this is all, you know, bluster and type and snake well.

我认为我们正处在一场将产生深远影响的技术变革之中。

And, you know, I think we're in the middle of a technology shift that's gonna be incredibly consequential.

我认为它的影响将超过互联网。

I think it's gonna be more consequential than the Internet.

我认为它的影响将超过自计算机诞生以来我们见过的任何技术变革。

Think it's gonna be more consequential than, you know, any shift in tech that we've seen since since the advent of computers.

而且我认为这需要时间。

And I think it's going to take a while.

我认为这将是一个持续数十年的过程。

And I think it's going be a multi decade thing.

我认为许多不同的群体将在推动这一进程中发挥重要作用。

And I think many different groups will be consequential in pushing it forward.

要知道，要实现计算机能为我们承担越来越多工作的世界，需要攻克所有这些独立问题并取得进展。

And, you know, to get to a world where computers can increasingly do more and more and more for us, there's all of these independent problems that need to be knocked down and progress needs to be made on them.

其中部分属于科学范畴——让这些模型理解不同类型数据、更快速、更经济、更智能，符合我们关注的模式，并在现实世界中采取行动。

Some of those are on the the science side of things of getting these models to understand different types of data, be faster, cheaper, smarter, you know, conform to modalities that we care about, take actions in the real world.

而另一部分则关乎我们将如何与它们协作。

And then some of it's on like how we're going to work with them.

人类在计算机上实际应该看到什么、控制什么，以及如何与这些事物协同工作。

And what's the experience that human should actually be seeing and controlling on a computer and working with these things.

但我认为这将需要数十年时间。

But I think it's gonna, you know, it's gonna take decades.

我认为未来会有大量令人惊叹的工作要做。

I think that there's gonna be lots of amazing work to do.

同时我认为，这里特别重要的一类群体模式是——不是自卖自夸——那些专注于自动化与增强特定领域知识工作的公司，他们既构建底层技术，整合供应商的最佳方案（有时自主研发），又打造相应的产品体验。

I think that also, you know, one of the most like, a pattern of a group that I think will be especially important here, you know, not not to talk our own book, but I think it's like, you know, the company that works on automating and augmenting a particular area of knowledge work builds the both the technology under, you know, under the surface for that, integrating the best parts from providers, sometimes doing it in house, and then also builds product experience for that.

我认为从事这项工作的人——我们正尝试在软件领域实现这一点，其他领域的人们也在做类似的事——这些人将会产生极其深远的影响。

I think people who do that and we're doing it in trying to do it in software, people do that in other areas, I think those folks will be really, really, really consequential.

不仅仅是为了用户看到的最终价值，我认为随着规模扩大，他们将对推动技术进步起到至关重要的作用。

Not just for like, you know, the end value that users see, but then I think as they get to scale, they'll be really important for pushing forward the, you know, the technology.

因为他们将能够建立——最具可验证性的那些企业——将能够打造非常庞大的商业体系。

Because I think they'll be able to build, you know, the most testable of them will be able to build very, very big businesses.

所以非常期待看到其他领域涌现更多这样的公司。

And, so excited to see the rise of, you know, other companies like that in other areas.

我知道你们正在招聘对这类工作感兴趣的人。

I know you guys are hiring for folks that are interested in, hey, I wanna go work here and build this sort of stuff.

目前你们主要在寻找哪些类型的职位？

What kind of roles are you looking for right now?

有没有特别急需填补的岗位？或者你最期待尽快招到哪些人才？

Anyone specifically you're trying to any roles you're most excited about filling ASAP?

如果感兴趣的话，求职者应该了解哪些信息？

What should people know if they're curious?

这群人需要做的事情太多了，我们目前还无法胜任。

There are so many things that this group of people need to do that, like, we are not yet equipped to do.

所以，你知道的，首先整体上来说是相当通用的。

And so, you know, kind of generic across the board, first of all.

所以如果你认为我们没有某个职位，也许你主动联系后会发现事实并非如此。

And so if you don't think we have a role for something, maybe if you reach out that that won't actually be the case.

也许我们真的可以从你那里学到东西，并意识到我们需要一些我们之前没意识到的东西。

And maybe we can actually learn from you and kind of decide that we we need something that we weren't yet aware of.

但总的来说，我认为今年对我们最重要的两件事是打造该领域最好的产品并实现增长。

But, you know, by and large, I think that, you know, two two of the most important things for us to do this year are have the best product in the space and then grow it.

我们正处于这种抢占市场的模式，世界上几乎所有人要么没用过我们这样的工具，要么在用开发速度较慢的产品。

And we're kind of in this land grab mode where almost everyone in the world is either using no tool like ours or they're using one that's maybe developing less quickly.

因此扩大Kershaw二是我们的重要目标。

And so growing Kershaw two is a big goal.

我想说，是的，我们一直在寻找优秀工程师、设计师、研究员，当然也包括业务端的各类人才。

I would say, yeah, especially always on the hunt for folks who excellent engineers, designers, researchers, but then folks in all across the business side too.

我不禁要问这个问题，既然你谈到了工程师。

I can't help but ask this question now that you talk about engineers.

这有点像在问，代码会自己写完我们所有的代码。

There's kind of this question of just like, you know, code's gonna write up all all our code.

AI会编写我们所有的代码，但大家仍在疯狂招聘工程师。

AI's gonna write all our code, but everyone's still hiring engineers like crazy.

所有的基础模型，太多了

All the foundational models, so many

我们并不是在外面，你知道，像对待So公司的人那样。

We're not out there treating the, you know, the horn of people from the So yeah.

你认为工程类职位会迎来一个放缓的转折点吗？

Do you think there's gonna be an inflection point of, like, engineering roles start to kinda slow down?

我知道这是个很大的问题，但你觉得工程师在这些公司中会越来越被需要，还是认为最终会有这些自动代理程序来为我们构建一切？

I know this is, a big question, but just it's do you see engineers being more and more needed across all these companies, or do you think at some point, there's all these cursor agents running, building for us.

我们再次持有这样的观点，你知道，这中间会有一个漫长而混乱的过程，不会突然就变成你只需退后一步，提出所有需求就能让工程部门完成一切。

Again, we we kind of have the view that like there's this, you know, both long messy middle of, you know, it it not jumping to a just like you step back and you ask for all your stuff to be done and you have your engineering department.

要知道，我们非常希望从现有的编程模式中进化出来。

And you know, very much like you want to evolve from programming as it exists today.

我们希望人类能始终掌握主导权。

We want humans to be in the driver's seat.

我们认为即使在最终阶段，保持人们对一切的控制至关重要，仍需要专业人士来决定软件的形态。

And you know, we think even in the end state like that's you know, folks control over everything is is really important You will need professionals to do that and kind of decide what the software looks like.

因此我两方面都认同：工程师绝对是必需的。

So both I think that yes, engineers definitely needed.

我认为工程师们将能够承担更多工作。

I think that engineers will be able to do much more.

软件需求具有持久性——这虽非最新观点，但想想现在构建那些看似简单、易于定义的项目所需的高昂成本和人力投入，实在令人难以置信。

I think the demand for software is very lasting, which is not the most novel thing, but I think it's it's kind of crazy to think about how expensive and labor intensive it is to build things that are pretty simple and easy to specify or would look like it to the outside observer.

你也知道，目前要实现这些目标有多困难。

And you know, just how hard those things are to do right now.

如果能将现有受成本与需求制约的各类开发工作，其成本降低几个数量级，我们就能在电脑前完成海量新事物，开发出无数新工具。

And so if you can you know, all of the stuff that exists right now that's, you know, justified by the cost and demand that we have now, if you could bring that down by orders of magnitude, I think you'd have tons and tons and tons of more stuff that we could do at our computers, tons more tools.

要知道，我深有体会——我早期的一份工作其实是在一家生物技术公司，为他们构建内部工具。

And, you know, I've I've felt this where, you know, one of my early jobs actually was working for a biotechnology company and it was building internal tools for them.

当时市面上现成的工具非常糟糕，完全不符合他们的使用场景。

And the off the shelf tools that existed were horrible and did not fit their use case at all.

而我正在构建的那些内部工具，显然存在大量可开发功能的需求。

And then the internal tools I was building, there was definitely a ton of demand there, for things that could be built.

这些需求远超出我在职期间能完成的工作量。

And you know, that far outstrip just the things that I could I could build in the time that I was with them.

但确实，我认为计算机工作的物理特性依然如此美妙——

But yes, I think that it's still so, you know, the physics of working on computers are so great.

它本应

It should be

能够让你应该

able to you should be

基本上可以随意调整一切，实现所有你想做的事。

able to kind of basically just move everything around, do everything that you want to do.

目前仍然存在很多摩擦。

There's still so much friction.

我认为对软件的需求远超过我们当前能开发的水平，现在开发简单的生产力软件成本堪比制作一部大片。

I think that there's much more demand for the software than for software than what we can build today with, you know, things costing like a blockbuster movie to make kind of simple productivity software.

因此我认为在很长远的未来，确实会对工程师有更多需求。

And so I think long into the future, yes, there will actually be more demand for engineers.

还有什么我们没谈到的话题你想补充吗？

Is there anything that we didn't cover that you wanted to mention?

最后还有什么负面智慧想留给听众吗？

Any last negative wisdom you wanted to leave listeners with?

你也可以说没有，因为我们已经谈了很多。

You could also say no because we've done a lot.

我们经常思考如何组建团队，使其既能持续改进现有产品，又能开发新东西。

We think a lot about how how you set up a a team to be able to make new stuff in addition to, like, continuing to improve the stuff that you have right now.

我认为如果我们要成功，集成开发环境（IDE）必须经历巨大变革。

And I think if we're to be successful, like, yeah, IDE is gonna have to change a ton.

未来的形态将不得不发生巨大改变。

What it looks like is gonna have to change a ton going into the future.

环顾四周，看看我们尊敬的公司，确实有一些企业能持续乘势而上，不断推动前沿发展。

And, you know, if you look around, the the companies we respect, there are definitely examples of companies that have continued to really, you know, ride the wave of many leapfrogs and continue to kind of actually push the frontier.

但这样的公司其实也相当罕见。

But, you know, they're they're kind of rare too.

因为这是一件非常困难的事情。

Like, it's a hard thing to do.

所以部分关键在于持续思考这个问题，并在状态好的时候尝试从第一性原理出发进行反思。

And, so, you know, part of that is is just kind of thinking about the thing and trying to reflect on it, you know, in our our good days and, you know, the first principle side of things.

另一部分则是深入研究历史上那些卓越的案例。

Part of it is also, you know, trying to get in and and study past examples of of greatness here.

这也是我们经常思考的一个重点。

And, you know, that that's that's something that we think about a lot too.

确实。

Yeah.

你刚才说的那些话，就是我们开始录音前，你身后那些书是怎么回事？我还好奇那是什么来着。

That what you just told is we were before we started recording, you had all these books behind you and I was like, what's that over there?

那好像是某个老牌计算机公司的历史，这家公司在很多方面都很有影响力，但我却从未听说过。

It's like the history of some old computer company that was influential in a lot of ways that I've never heard of.

我觉得这很能说明你的创新源泉——研究过去和历史，研究哪些方法奏效、哪些没有。

And I think that says a lot about you, of where a lot of this innovation comes from, is studying the past and studying history and what's worked and what hasn't.

好的。

Okay.

如果听众想联系你或者申请职位，该去哪里找你呢？

Where can folks find you online if they wanna reach out and maybe apply?

你提到有些职位可能连他们自己都不知道存在。

You said that there may be roles you they may not even be aware of.

他们该去哪里找这些信息？

Where do they go find that?

另外听众怎样才能帮到你呢？

And then how can listeners be useful to you?

好的。

Yeah.

如果大家对这些工作感兴趣，我们非常乐意交流。他们可以访问cursor.com，既能了解产品也能找到联系我们方式。

I you know, if if folks are, you know, interested in working on this stuff, would would love to speak, and they can find, if they go to cursor.com, they can kind of both find the product and find out how to reach us.

所以

So

很简单。

Easy.

Michael，非常感谢你参与这次访谈。

Michael, thank you so much for being here.

这里是Engrave。

This was Engrave.

非常愉快。

It was wonderful.

谢谢。

Thank you.

大家再见。

Bye, everyone.

非常感谢大家的收听。

Thank you so much for listening.

如果你觉得本期节目有价值，可以在苹果播客、Spotify或你喜欢的播客应用上订阅我们的节目。

If you found this valuable, you can subscribe to the show on Apple Podcasts, Spotify, or your favorite podcast app.

也请考虑给我们评分或留言，这能帮助其他听众发现这个播客。

Also, please consider giving us a rating or leaving a review as that really helps other listeners find the podcast.

你可以在Lenny的播客官网上找到所有往期节目或了解更多关于本节目的信息。

You can find all past episodes or learn more about the show at Lenny's podcast dot com.

下次

See you in

节目见。

the next episode.