'Inside Devin: The AI engineer that''s set to write 50% of its company’s code this year | Scott Wu'

Scott Wu: Our whole team is only like 15 engineers a year.

Scott Wu：我们整个团队每年大约只有 15 名工程师。

We use a ton of Devin when we're building Devin.

当我们构建 Devin 时，我们使用了大量的 Devin。

Most folks on the team are definitely working with up to five Devins at once, and so Devin merges like several hundred pull requests into production in the Devin code bases every month.

团队中的大多数人肯定会同时与最多五个 Devin 一起工作，因此 Devin 每个月都会将数百个拉取请求合并到 Devin 代码库中的生产中。

Lenny Rachitsky: What percentage of your PRs are Devin versus humans right now?

Lenny Rachitsky：现在你们的 PR 中 Devin 与人类的比例是多少？

Scott Wu: It's in the neighborhood of a quarter or so.

Scott Wu：大约四分之一左右。

Lenny Rachitsky: Where do you think this will be at the end of the year?

Lenny Rachitsky：您认为今年年底会是什么情况？

Scott Wu: Honestly, we expect it to be a decent bit more than half.

Scott Wu：老实说，我们预计会超过一半。

Lenny Rachitsky: You guys are so ahead of how companies work with AI engineers.

Lenny Rachitsky：你们对于公司与人工智能工程师的合作方式非常领先。

Scott Wu: AI is going to be the biggest technology shift of our lives, so most of the big tech revolutions that we've had over the last 50 years, like personal computer, and the internet, and the mobile phone, they all had this big hardware component that was a big part of the distribution.

Scott Wu：人工智能将成为我们生活中最大的技术转变，所以过去 50 年我们经历的大多数重大技术革命，比如个人电脑、互联网和移动电话，它们都拥有这种大型硬件组件，这是分布的重要组成部分。

Folks who were building for those industries kind of saw their market grow and grow and grow basically steadily year over year as the number of people with mobile phones increased, right, as the number of people connected to the internet increase.

随着拥有手机的人数增加，随着连接互联网的人数增加，为这些行业建设的人们看到他们的市场不断增长，基本上逐年稳定增长。

One of the things which is already I'd say different in AI, is just how explosive the technology can be.

我想说的人工智能领域的不同之处之一就是该技术的爆炸性。

There's no weight on hardware distribution.

硬件分配没有权重。

It means that the space is just growing so exponentially.

这意味着空间正在呈指数级增长。

Lenny Rachitsky: How is the act of being an engineer and building changing?

Lenny Rachitsky：工程师和建筑的行为有何变化？

Scott Wu: I think there's going to be way more programmers and way more engineers a few years from now.

Scott Wu：我认为几年后将会有更多的程序员和工程师。

Pretty quickly.

很快。

The form factor of what it means to be a programmer obviously is going to change, but at the end of the day, of course the discipline is all about just being able to tell your computer what's do.

程序员的定义显然会发生变化，但归根结底，纪律当然就是能够告诉你的计算机在做什么。

And so in that lens, I really think that programming is only going to become more and more important as AI gets more powerful.

因此，从这个角度来看，我真的认为随着人工智能变得越来越强大，编程只会变得越来越重要。

Lenny Rachitsky: Today my guest is Scott Wu.

Lenny Rachitsky：今天我的嘉宾是 Scott Wu。

Scott is the co-founder and CEO of Cognition, which makes a product called Devin, the world's first autonomous AI software engineer.

Scott 是 Cognition 的联合创始人兼首席执行官，该公司生产的产品名为 Devin，是世界上第一位自主人工智能软件工程师。

Unlike other AI tools that I've highlighted on this podcast, Devin is designed to act like an actual remote engineer that you chat with like you would with any other human engineer through Slack or through its dedicated website.

与我在本播客中强调的其他人工智能工具不同，Devin 的设计就像一个真正的远程工程师一样，您可以通过 Slack 或其专用网站与任何其他人类工程师聊天。

When Devin launched about a year ago, it was very much a junior engineer.

大约一年前，当 Devin 推出时，它只是一名初级工程师。

Over the past year, they've made a lot of progress and Devin is now being used by tons of companies in production.

在过去的一年里，他们取得了很大的进步，Devin 现在被大量公司在生产中使用。

We chatted about how their engineering team of 15 uses Devins to build Devin, including how every engineer uses about five Devins each to help them code and move faster.

我们讨论了他们的 15 人工程团队如何使用 Devins 来构建 Devin，包括每个工程师如何使用大约 5 个 Devins 来帮助他们更快地编码和移动。

How a quarter of their pull requests today are committed by Devins and that they expect this to be over 50% by the end of the year.

今天他们的拉取请求中有四分之一是由 Devins 提交的，他们预计到今年年底这一比例将超过 50%。

Lenny Rachitsky: We also talk about how Scott imagines software engineering is going to look in the future and how the role of an engineer changes from a coder to an architect.

Lenny Rachitsky：我们还讨论了 Scott 对软件工程未来的展望，以及工程师的角色如何从编码员转变为架构师。

We also get into the eight pivots that they went through before landing on this path, why Scott believes AI tools like this will lead to more engineer hiring versus less.

我们还探讨了他们在走上这条道路之前经历的八个关键点，为什么斯科特认为这样的人工智能工具将导致更多而不是更少的工程师招聘。

Also where the name Devin comes from and so much more.

这也是 Devin 这个名字的由来以及更多。

This episode is going to blow your mind.

这一集会让你大吃一惊。

I highly recommend you listen to it if you're at all interested about where engineering, product building, and AI is going.

如果您对工程、产品构建和人工智能的发展方向感兴趣，我强烈建议您听听它。

A huge thank you to Claire Voue for suggesting a bunch of great questions for this conversation.

非常感谢 Claire Voue 为本次对话提出了一系列很好的问题。

Lenny Rachitsky: If you enjoy this podcast, don't forget to subscribe and follow it in your favorite podcasting app or YouTube.

Lenny Rachitsky：如果您喜欢这个播客，请不要忘记订阅并在您最喜欢的播客应用程序或 YouTube 中关注它。

Also, if you become an annual subscriber of my newsletter, you get a year free of Linear, Superhuman, Notion, Perplexity, and Granola.

另外，如果您成为我的时事通讯的年度订阅者，您将获得一年的免费线性、超人、概念、困惑和格兰诺拉麦片。

Check it out at lenny'snewsletter.com and click bundle.

请访问 lenny'snewsletter.com 查看并单击“捆绑”。

With that, I bring you Scott Wu.

接下来，我为您带来斯科特·吴。

This episode is brought to you by Enterpret.

本集由 Enterpret 为您带来。

Enterpret unifies all your customer interactions from Gong calls, to Zendesk tickets, to Twitter threads, to app store reviews and makes it available for analysis.

Enterpret 统一了所有客户交互，从 Kong 通话、Zendesk 票证、Twitter 线程到应用商店评论，并可供分析。

It's trusted by leading product orgs like Canva, Notion, Loom, Linear, monday.com, and Strava to bring the voice of the customer into the product development process, helping you build best in class products faster.

它受到 Canva、Notion、Loom、Linear、monday.com 和 Strava 等领先产品组织的信赖，可将客户的声音纳入产品开发流程，帮助您更快地构建一流的产品。

Lenny Rachitsky: What makes Enterpret special is its ability to build and update customer specific AI models that provide the most granular and accurate insights into your business.

Lenny Rachitsky：Enterpret 的独特之处在于它能够构建和更新客户特定的 AI 模型，从而为您的业务提供最细致、最准确的见解。

Connect customer insights to revenue and operational data in your CRM or data warehouse to map the business impact of each customer need and prioritize confidently and empower your entire team to easily take action on use cases like win-loss analysis, critical bug detection, and identifying drivers of churn with Enterpret's AI assistant Wisdom.

将客户洞察与 CRM 或数据仓库中的收入和运营数据联系起来，以映射每个客户需求的业务影响，并自信地确定优先级，并使您的整个团队能够使用 Enterpret 的 AI 助手 Wisdom 轻松对赢利分析、关键错误检测和识别客户流失驱动因素等用例采取行动。

Looking to automate your feedback loops and prioritize your roadmap with confidence like Notion, Canva, and Linear visit E-N-T-E-R-P-R-E-T.com/lenny to connect with the team and to get two free months when you sign up for an annual plan.

希望自动化反馈循环并自信地确定路线图的优先级，如 Notion、Canva 和 Linear，请访问 E-N-T-E-R-P-R-E-T.com/lenny 与团队联系，并在注册年度计划时获得两个月的免费时间。

This is a limited time offer.

这是限时优惠。

That's enterpret.com/lenny.

那是 enterpret.com/lenny。

Many of you are building AI products, which is why I'm very excited to chat with Brandon Foo, Founder and CEO of Paragon.

你们中的许多人都在构建人工智能产品，这就是为什么我很高兴能与 Paragon 的创始人兼首席执行官 Brandon Foo 聊天。

Hey Brandon.

嘿布兰登。

Brandon Foo: Hey Lenny.

布兰登·符：嘿莱尼。

Thanks for having me.

谢谢你有我。

Lenny Rachitsky: So integrations have become a big deal for AI products.

Lenny Rachitsky：因此集成对于人工智能产品来说已经成为一件大事。

Why is that?

这是为什么？

Brandon Foo: Integrations are mission-critical for AI for two reasons.

Brandon Foo：集成对于人工智能来说至关重要，原因有两个。

First, AI products need contacts from their customer's business data such as Google Drive files, Slack messages or CRM records.

首先，人工智能产品需要来自客户业务数据的联系人，例如 Google Drive 文件、Slack 消息或 CRM 记录。

Second, for AI products to automate work on behalf of users, AI agents need to be able to take action across these different third-party tools.

其次，为了使人工智能产品能够代表用户自动化工作，人工智能代理需要能够跨这些不同的第三方工具采取行动。

Lenny Rachitsky: So where does Paragon fit into all this?

Lenny Rachitsky：那么 Paragon 在这一切中处于什么位置呢？

Brandon Foo: Well, these integrations are a pain to build and that's why Paragon provides an embedded platform that enables engineers to ship these product integrations in just days instead of months across every use case from RAG data ingestion to agentic actions.

Brandon Foo：嗯，构建这些集成是一件痛苦的事情，这就是为什么 Paragon 提供了一个嵌入式平台，使工程师能够在短短几天而不是几个月的时间内交付这些产品集成，涵盖从 RAG 数据摄取到代理操作的每个用例。

Lenny Rachitsky: And I know from firsthand experience that maintenance is even harder than just building it for the first time.

Lenny Rachitsky：我从第一手经验中知道，维护比第一次建造它还要困难。

Brandon Foo: Exactly.

布兰登·符：没错。

We believe product teams should focus engineering efforts and competitive advantages, not integrations.

我们认为产品团队应该专注于工程工作和竞争优势，而不是集成。

That's why companies like You.com, AI21 and hundreds of others use Paragon to accelerate their integration strategy.

这就是 You.com、AI21 等数百家公司使用 Paragon 来加速其集成战略的原因。

Lenny Rachitsky: If you want to avoid wasting months of engineering on integrations that your customers need, check out paragon@useparagon.com/lenny.

Lenny Rachitsky：如果您想避免在客户所需的集成上浪费数月的工程时间，请查看 paragon@useparagon.com/lenny。

Scott, thank you so much for being here and welcome to the podcast.

斯科特，非常感谢您来到这里，欢迎来到播客。

Scott Wu: Thanks so much for having me.

Scott Wu：非常感谢你邀请我。

Excited to be on.

很高兴能上场。

Lenny Rachitsky: I'm really excited to have you here because you are building and you've been building something that is very different from what a lot of other AI companies have been doing for a long time, although they are starting to converge to where you guys are now.

Lenny Rachitsky：我真的很高兴你们能来到这里，因为你们正在构建并且一直在构建一些与许多其他人工智能公司长期以来所做的事情非常不同的东西，尽管它们已经开始向你们现在的水平靠拢。

We're going to talk about that and it's also just such a unique point in the history of AI and just the journey of AI.

我们将讨论这一点，这也是人工智能历史上的一个独特点，也是人工智能之旅的一个独特点。

And so it's really cool to be chatting right now.

所以现在聊天真的很酷。

And I feel like we're going to chat again in a few years and be like, wow, we were so right about so much and so wrong about so much.

我觉得几年后我们会再次聊天，然后我们会说，哇，我们在很多事情上都是对的，但在很多事情上又错了。

Lenny Rachitsky: And so I'm excited to have you here.

Lenny Rachitsky：所以我很高兴你能来到这里。

Let's start with talking about Devin, giving people an understanding of just what the heck Devin is, is the main product that you guys build.

让我们从谈论 Devin 开始，让人们了解 Devin 到底是什么，是你们构建的主要产品。

What is the simplest way to understand what is Devin?

了解 Devin 的最简单方法是什么？

Scott Wu: Absolutely.

吴斯科特：当然。

And so Devin is a fully autonomous software engineer that is going to work on tasks end to end, and so there are a lot of great tools for all parts of the stack of the AI code workflow.

因此 Devin 是一位完全自主的软件工程师，他将端到端地处理任务，因此有很多很棒的工具适用于人工智能代码工作流程堆栈的所有部分。

What Devin does is it is a full asynchronous workflow, and so you can tag Devin on an issue in Slack, you're talking about an issue and you tag Devin, you can tag Devin in Linear, you can have Devin and Devin will make pull requests in your GitHub, and so it's very much built to work with engineering teams as your junior engineer.

Devin 所做的是，它是一个完整的异步工作流程，因此您可以在 Slack 中的某个问题上标记 Devin，您正在谈论一个问题并标记 Devin，您可以在 Linear 中标记 Devin，您可以让 Devin 和 Devin 将在您的 GitHub 中提出拉取请求，因此它非常适合作为您的初级工程师与工程团队合作。

Lenny Rachitsky: Amazing, okay.

莱尼·拉奇茨基：太棒了，好吧。

So I remember when you guys launched this, there was this big pitch of this is your new AI engineer and it was really good at a lot of stuff.

所以我记得当你们推出这个产品时，有一个很大的宣传，这是你们的新人工智能工程师，它真的很擅长很多东西。

It wasn't great at other things.

在其他方面它并不擅长。

It's been a year now about since you guys launched, is that right?

自从你们推出以来已经有一年了，对吗？

Lenny Rachitsky: What's the best way to think about the level of seniority that engineer had back in the day when you guys launched and then the level of seniority of engineer today if that's, I don't know, measure of how to think about Devin?

Lenny Rachitsky：如果我不知道，你们成立时工程师的资历水平是衡量 Devin 的标准，那么最好的方式是什么？

Scott Wu: Yeah, and it's crazy to think about by the way, because a year ago when we did the initial launch, I mean people didn't really believe that an agent was possible.

Scott Wu：是的，顺便说一句，这想想就很疯狂，因为一年前我们首次推出时，我的意思是人们并不真正相信代理是可能的。

Right.

正确的。

And it was, I mean, it was a very different time.

我的意思是，那是一个非常不同的时代。

So like start of 2024, things with model capabilities were definitely quite a bit earlier on, reasoning especially was quite a bit earlier on.

所以就像 2024 年初一样，具有模型功能的东西肯定要早一些，尤其是推理要早一些。

And yeah, I mean, in the time since then it's obviously developed a lot.

是的，我的意思是，从那时起它显然已经发展了很多。

I think in terms of practical skills, there's some comparisons we make.

我觉得在实践技能方面，我们有一些比较。

Sometimes we kind of say, well, when we got started it was kind of like a high school CS student and then as time went on, it became more of a college intern and now it's like a junior engineer.

有时我们会说，好吧，当我们刚开始时，它有点像高中计算机科学学生，然后随着时间的推移，它变得更像是一名大学实习生，现在它就像一名初级工程师。

But I would say though that those are more rough guidelines because I really like the phrase jagged intelligence for example, because obviously there are certain things that it is much better at than a human.

但我想说的是，这些是更粗略的指导方针，因为我真的很喜欢“锯齿状智力”这个词，因为显然在某些事情上它比人类更擅长。

There are certain things that it's much worse at than a human.

在某些事情上它比人类差得多。

Scott Wu: And I think over the last year we've learned a lot especially about not just coding agents but agents in general just really building out how all of us should be working and interacting with agents as part of our flow.

Scott Wu：我认为在过去的一年里，我们不仅学到了很多东西，特别是关于编码代理，而且一般来说，代理都真正构建了我们所有人应该如何与代理工作和交互，作为我们流程的一部分。

And so a lot of the things that we built, I mean, there was no Slack, there was no GitHub integration, there was no Linear, there was no interactive planning phase working back and forth.

所以我们构建的很多东西，我的意思是，没有 Slack，没有 GitHub 集成，没有 Linear，没有来回工作的交互式规划阶段。

There was no way to touch up Devin's code.

没有办法修改德文的代码。

And so a lot of the features that we've built on the product side since then have really been about basically yeah, figuring out how to make working with Devin and handing off tasks to Devin as smooth of an experience as possible.

因此，从那时起，我们在产品方面构建的许多功能实际上基本上都是关于如何与 Devin 合作并将任务交给 Devin 的体验尽可能顺畅。

Lenny Rachitsky: That's so interesting.

莱尼·拉奇茨基：这太有趣了。

So a lot of the work has gone not into how do we just make Devin the best possible engineer, but it's how to work with this new type of entity that we haven't ever worked with.

因此，很多工作并不是我们如何让德文成为最好的工程师，而是如何与我们从未合作过的这种新型实体合作。

Scott Wu: I think it's a 50/50 of both.

Scott Wu：我认为两者各占一半。

I think the capabilities obviously have improved a ton and we've seen these get better and get measurably better.

我认为这些功能显然已经提高了很多，而且我们已经看到这些功能变得更好并且明显更好。

But I think the other side of it is everything to do with yeah, really the product interface and the tools and so on.

但我认为它的另一面与产品界面和工具等有关。

And I think today folks generally know how to use chatbots and to with chatbots, right, and that's an interface that people are familiar with and obviously with agents it's still like a real curve, I think to learn how to use them and how to get the most out of them.

我认为今天人们普遍知道如何使用聊天机器人以及如何使用聊天机器人，对吧，这是人们熟悉的界面，显然对于代理来说它仍然像一条真正的曲线，我认为学习如何使用它们以及如何充分利用它们。

And so it's really exciting to see a lot of others starting to build and do a lot more in the agent space as well.

因此，看到许多其他人开始在代理领域构建并做更多事情，真的很令人兴奋。

But I think this is the kind of thing that we're all really figuring out together as a space.

但我认为这是我们作为一个空间真正共同解决的问题。

Lenny Rachitsky: What can you share about just the scale of Devin at this point, whatever you're comfortable sharing, and then just where do you think the level of Devin's coding abilities will be in a year?

Lenny Rachitsky：关于 D​​evin 目前的规模，您能分享什么吗？无论您愿意分享什么，您认为一年后 Devin 的编码能力会达到什么水平？

Scott Wu: So we work with companies of all stages and sizes.

Scott Wu：所以我们与各个阶段和规模的公司合作。

On the smallest end, it goes to startups of just one or two people who are using Devin to build out a lot of their kind of initial prototype or initial product all the way up to big public companies, Fortune 100 companies or public banks or things like that who are using Devin across their engineering teams.

从最小的角度来看，它适用于只有一两个人使用 Devin 构建大量初始原型或初始产品的初创公司，一直到大型上市公司、财富 100 强公司或公共银行或类似的公司，他们在其工程团队中使用 Devin。

In general, we've seen a huge range of the use cases there.

总的来说，我们已经看到了广泛的用例。

And obviously the kinds of engineering work that you're doing at a one or two person startup, they're very different from the kind of work that you're doing at a public bank.

显然，您在一两个人的初创公司中所做的工程工作与您在公共银行所做的工作非常不同。

Scott Wu: But throughout it's all been basically yeah, being that junior buddy of yours that makes you go faster and really multiplies you, I would say.

Scott Wu：但自始至终，这一切都基本上是的，我想说，作为你的初级伙伴，它会让你走得更快，而且真的会让你倍增。

I think it can multiply you as an engineer obviously by just letting you work with your own team of Devins instead of having to be kind of fully synchronous on a single task.

我认为，通过让你与自己的 Devins 团队合作，而不是必须在单个任务上完全同步，它显然可以使你作为一名工程师倍增。

And then it's also kind of multiplying your team and multiplying your team's knowledge base because Devin really accumulates a lot of the knowledge from working with every member of your team and is able to bring that into each new session.

然后，这也可以使您的团队和团队的知识库成倍增加，因为 Devin 通过与团队的每个成员合作确实积累了大量知识，并且能够将这些知识带入每个新会议中。

Lenny Rachitsky: Awesome.

莱尼·拉奇茨基：太棒了。

We're going to show people how it actually works later in the podcast.

我们将在稍后的播客中向人们展示它的实际工作原理。

We're going to do a few live demos, but let's actually go to the beginning of the journey.

我们将进行一些现场演示，但让我们真正开始旅程。

What's just the origin story of Devin?

德文的起源故事是什么？

How did this all begin?

这一切是如何开始的？

Scott Wu: The founding team, I mean most of us have known each other for years and years and years actually.

Scott Wu：创始团队，我的意思是我们大多数人实际上已经认识很多年了。

And for almost everyone, this is our first time working together, but we've known each other a long time.

对于几乎所有人来说，这是我们第一次合作，但我们已经认识很久了。

And we all actually had our own kind of journeys in AI for the last decade or so.

在过去十年左右的时间里，我们实际上都在人工智能领域经历过自己的旅程。

And so for myself, I ran a company before this called Lunchclub, which was an AI for a professional networking product and I ran that for about five years.

因此，就我自己而言，我在此之前经营了一家名为 Lunchclub 的公司，这是一个专业网络产品的人工智能，我经营了大约五年。

And one of my co-founders, Steven was one of the first engineers at a company called Scale AI, which has obviously grown a lot and done very well.

我的联合创始人之一史蒂文（Steven）是一家名为 Scale AI 的公司的首批工程师之一，该公司显然已经成长了很多并且做得非常好。

My other co-founder Walden, was an early engineer at a company called Cursor, which has also obviously grown a lot and done really well.

我的另一位联合创始人 Walden 是一家名为 Cursor 的公司的早期工程师，该公司显然也发展了很多并且做得非常好。

And our whole team was kind of like that.

我们整个团队都是这样的。

Many of us knew each other from competitive programming and math competitions, but we had stayed very closely in touch in the decade since then and we've all kind of all had our own journeys.

我们中的许多人都是通过竞争性编程和数学竞赛而认识的，但从那以后的十年里，我们一直保持着非常密切的联系，我们都有自己的旅程。

Scott Wu: And so we had one person who was running teams at Neuro, we had one person who was at Waymo, someone who had their own YC tools startup for machine learning, and we were really excited to build something together.

Scott Wu：我们有一个人在 Neuro 管理团队，我们有一个人在 Waymo，还有一个人拥有自己的 YC 机器学习工具初创公司，我们非常高兴能一起构建一些东西。

And this was around late 2023, so about a year and a half ago at this point.

那是在 2023 年末左右，也就是大约一年半之前。

And yeah, when we got started, I mean I think there were a couple of things that we felt really strongly about and one was that reinforcement learning was really working and was going to be the next big paradigm shift in capabilities.

是的，当我们开始时，我的意思是，我认为有几件事我们非常强烈地感受到，其中之一是强化学习确实有效，并且将成为能力的下一个重大范式转变。

Back then it was the initial ChatGPT launch in 2022, and those models were, to first order were what we would call imitation learning in AI, right, which is basically you have the model read all the texts that you can find on the internet and then train it to talk like somebody on the internet would talk.

当时 ChatGPT 于 2022 年首次推出，这些模型首先是我们所说的人工智能中的模仿学习，对吧，基本上就是你让模型阅读你可以在互联网上找到的所有文本，然后训练它像互联网上的人一样说话。

Right.

正确的。

And there are kind of obviously a lot more details on top of that, but that's kind of the first order pass of what was really done.

除此之外，显然还有更多细节，但这只是真正完成的工作的第一步。

Scott Wu: And it was amazing.

Scott Wu：这太棒了。

Right.

正确的。

I mean, it passed the churn test, it was able to respond and to have encyclopedic knowledge about a lot of things.

我的意思是，它通过了流失测试，它能够做出响应并且对很多事情拥有百科全书式的知识。

And I think this new paradigm which we've gotten into over this last year or year and a half is really high compute RL, which is a very different paradigm, right, which is basically the ability to go and do work on task and put something together and then be evaluated on whether that was correct or incorrect and use that knowledge to decide what to do and to learn from that.

我认为我们在过去一年或一年半中所采用的这种新范式确实是高计算强化学习，这是一种非常不同的范式，对吧，它基本上是执行任务并将某些内容组合在一起的能力，然后评估其正确与否，并使用这些知识来决定要做什么并从中学习。

Right.

正确的。

And so we felt very strongly that that was going to happen.

所以我们强烈地感觉到这将会发生。

I think for us, code was the natural thing to work on for a couple reasons.

我认为对我们来说，出于几个原因，编写代码是很自然的事情。

One, because we're all programmer nerds ourselves, and so teaching AI to code is about as cool as it gets for us, but also because code has this whole automated feedback loop, right, where you can run the code and that is the kind of automated feedback that really feeds into the RL, which makes these models so great at coding.

第一，因为我们自己都是程序员书呆子，所以教人工智能编码对我们来说是一件很酷的事情，而且还因为代码有整个自动反馈循环，对吧，你可以在其中运行代码，这就是真正反馈给强化学习的那种自动反馈，这使得这些模型在编码方面非常出色。

Scott Wu: And then the other thing that we felt very strongly about was that the product experience was going to shift from what I'll call text completion to agents basically.

Scott Wu：我们强烈感受到的另一件事是，产品体验将从我所说的文本完成转向基本上的代理。

Right.

正确的。

And to first order, I would kind of say there's been a lot of great experiences in text completion.

首先，我想说的是，在文本补全方面有很多很棒的经验。

It's been used for marketing, it's been used for customer support, it's been used for education and in Coda obviously as well.

它被用于营销、客户支持、教育，显然在 Coda 中也被使用。

The GitHub copilot was kind of really the dominant product of that initial wave.

GitHub copilot 确实是第一波浪潮中的主导产品。

Right.

正确的。

Scott Wu: But I think that the big shift that we really felt we would see is moving from kind of this text to text model to an actual autonomous system that can make decisions, that can interact with the real world, that can take in feedback, that can iterate and take multiple steps to solve problems.

Scott Wu：但我认为，我们真正认为我们会看到的重大转变是从这种文本到文本模型，再到一个可以做出决策、可以与现实世界交互、可以接受反馈、可以迭代并采取多个步骤来解决问题的实际自治系统。

And now we call that agents, but that was what we were really excited about at the time.

现在我们称之为代理，但这正是我们当时真正兴奋的。

So it was always coding, it was always agents.

所以总是在编码，总是在代理。

And in some ways that kind of feels like it should have been cleared from the start.

从某些方面来说，这感觉像是从一开始就应该被清除。

But even with that, I feel like we've pivoted eight times or something within coding agents over the last year and a half, so.

但即便如此，我觉得在过去的一年半里，我们已经在编码代理内部进行了八次调整，所以。

Lenny Rachitsky: I just noticed recently all the AI, top AI companies sort of, not all, but many of them, the product that is winning is different, has a different name from the company, which is not typical, Cursors, Anysphere, Bolt, StackBlitz, you guys are Cognition Labs, like V0 is Vercel.

Lenny Rachitsky：我最近才注意到所有的人工智能、顶级人工智能公司，虽然不是全部，但其中很多，获胜的产品是不同的，与公司有不同的名称，这并不典型，Cursors、Anysphere、Bolt、StackBlitz，你们是认知实验室，就像 V0 是 Vercel 一样。

And it just tells me these all emerge later in the company's journey and they tried a bunch of stuff and like, oh wow, this thing worked and it's so interesting that it's so common amongst these top AI companies.

它只是告诉我，这些都是在公司发展历程的后期出现的，他们尝试了很多东西，然后，哇哦，这东西成功了，而且很有趣，在这些顶级人工智能公司中很常见。

Scott Wu: Yeah, and there's even, I mean OpenAI, ChatGPT, Anthropic and Claude and Google.

Scott Wu：是的，甚至还有 OpenAI、ChatGPT、Anthropic、Claude 和 Google。

Yeah, it's funny.

是的，这很有趣。

Yeah, yeah, I agree.

是的，是的，我同意。

So when we got started, it wasn't even really a company.

所以当我们开始时，它甚至不是一家真正的公司。

I mean, it was more like a project or a hackathon almost.

我的意思是，这更像是一个项目或黑客马拉松。

We got a bunch, we booked an Airbnb basically for a couple of weeks.

我们有很多，我们在 Airbnb 上预订了几周。

This was around Thanksgiving time and just got a bunch of people together who were just excited to hack on some projects and build something cool.

那是在感恩节前后，一群人聚集在一起，他们很高兴能够参与一些项目并构建一些很酷的东西。

And it's funny, actually the first thing that we were building for actually was more solving these more contest programming problems and using an agentic loop to really do better on that.

有趣的是，实际上我们构建的第一件事实际上是更多地解决这些更具竞赛性的编程问题，并使用代理循环来真正做得更好。

And so obviously if you run your code on the test cases, you can evaluate, there's a lot of agentic work that you can do there to try and do better, and that we spent some time on that initially.

显然，如果您在测试用例上运行代码，您可以进行评估，您可以在那里做很多代理工作来尝试做得更好，我们最初在这方面花了一些时间。

And then we've kind of gone from, I mean the story of the whole company for us in some sense has been going from hacker house to hacker house.

然后我们就从某种意义上说，整个公司的故事从某种意义上说是从一个黑客之家到另一个黑客之家。

Scott Wu: So after that we had another hacker house and that's where kind of some of the initial ideas for Devin came and really building a software engineering agent and not just a coding agent and having to interact with a lot of these tools, but even then there were so many iterations.

Scott Wu：所以在那之后我们有了另一个黑客之家，这就是 Devin 的一些最初想法的来源，真正构建一个软件工程代理，而不仅仅是一个编码代理，并且必须与很多这样的工具进行交互，但即便如此，还是有很多迭代。

And even the idea of talking to Devin for example was like, it was something that we had to come up with.

甚至与德文交谈的想法就像是，这是我们必须想出的东西。

Right.

正确的。

Initially, it was just like you hand off a task and then it works and then it shows you this whole finished code.

最初，这就像您交接了一项任务，然后它开始工作，然后它向您显示整个完成的代码。

Right.

正确的。

And now obviously it's like you can jump in at any time, you can get feedback on the plan, you guys can scope out the task together when you're working with Devin.

现在显然你可以随时加入，你可以得到有关计划的反馈，当你和 Devin 一起工作时，你们可以一起确定任务范围。

And a lot of these things we had to develop obviously, and certainly we've learned a lot about the use cases, the form factor.

显然，我们必须开发很多这些东西，当然我们也学到了很多关于用例、外形因素的知识。

We've made a lot of big improvements and step function improvements on the capabilities and Devin's ability to use tools and to bug and make decisions.

我们对 Devin 使用工具、调试和做出决策的能力和能力进行了很多重大改进和步骤功能改进。

Scott Wu: And so it's yeah, it's been a fun journey.

Scott Wu：是的，这是一次有趣的旅程。

I mean, I think I would say that the grounding question for us really, which is one that we think about all the time, is really just what is the future of software engineering and how should we be working with AI to write code?

我的意思是，我想我想说，对我们来说，我们一直在思考的根本问题实际上就是软件工程的未来是什么，以及我们应该如何使用人工智能来编写代码？

Because I think at the end of the day, of course that's what underlines all of the product decisions that we make, so.

因为我认为归根结底，这当然是我们做出的所有产品决策的重点，所以。

Lenny Rachitsky: I like that you're asking the juicy question I wanted to get to.

Lenny Rachitsky：我喜欢你问了我想问的有趣问题。

Before I ask it just for the history books, when did you guys start kind of hacking around and when did Devin launch?

在我问历史书之前，你们什么时候开始进行黑客活动的？Devin 是什么时候推出的？

How long was that period?

那段时间有多长？

Scott Wu: Yeah, so we started in November of 2023, which was then yeah, just like hackathon mode.

Scott Wu：是的，所以我们从 2023 年 11 月开始，当时是的，就像黑客马拉松模式一样。

We officially made it into a company around the start of 2024, and then our initial launch was in March.

我们在 2024 年初左右正式将其组建为一家公司，然后我们于 3 月份首次推出。

And so it was like nonstop, I mean it's been nonstop for the entire last 17 months, but it's getting to the launch and then obviously working with enterprises and developing the product a lot more, building it and getting it to work for a lot of practical use cases and then making it fully available self-serve in December of last year.

所以它就像是不间断的，我的意思是在过去的 17 个月里一直不间断，但它正在发布，然后显然与企业合作并开发更多产品，构建它并使其适用于许多实际用例，然后在去年 12 月使其完全自助服务。

And now we've rolled out 2.0 obviously just a few weeks ago.

现在我们显然就在几周前推出了 2.0。

And so it's been a very busy time for us.

所以这对我们来说是一个非常忙碌的时期。

Lenny Rachitsky: Understatement of the century.

莱尼·拉奇茨基：本世纪的轻描淡写。

Let me ask this question because you touched on it a bit, this whole idea of Devin as a person and this idea of creating a personality for Devin.

让我问这个问题，因为你稍微触及了德文作为一个人的整个想法以及为德文创造个性的想法。

It's unlike any other, I believe, AI app.

我相信，它不同于任何其他人工智能应用程序。

No one else has a name and you don't think of it as a person.

其他人都没有名字，你也不认为它是一个人。

What made you guys decide to go that approach and just how do you design it to work well that way?

是什么让你们决定采用这种方法？你们如何设计它才能以这种方式良好地工作？

Scott Wu: I would say it's a decision we're pretty proud of, I would say.

Scott Wu：我想说，这是一个我们非常自豪的决定。

I mean, I think there's a lot of different product experiences out there, and I think the thing that really makes Devin unique in what it does is that you can really hand off and more and more what we've seen honestly is that I think a lot of kind of explaining the Devin experience to folks is really just explain it as, yeah, this is your junior buddy.

我的意思是，我认为那里有很多不同的产品体验，我认为真正让 Devin 的独特之处在于你可以真正地交接，而且我们越来越诚实地看到的是，我认为向人们解释 Devin 体验的很多方式实际上只是解释为，是的，这是你的初级伙伴。

And that goes for a lot of the parts of the flow where in the onboarding for example, initially I would say we've definitely had a lot of users come in and just kind of see the blank screen and not really know, or they'd ask, "Hey, I'm going to do this whole big re-architecture, the whole code base."

这适用于流程的很多部分，例如，在入职流程中，最初我会说我们肯定有很多用户进来，只是看到空白屏幕，但并不真正知道，或者他们会问，“嘿，我要进行整个大型重新架构，整个代码库。”

Scott Wu: And basically what we've learned over time is to basically get folks to think more like whoa, whoa, whoa, let's work on getting the repository set up first.

Scott Wu：随着时间的推移，我们学到的基本上是让人们更多地思考，哇，哇，哇，让我们首先建立存储库。

Let's make sure we hand Devin a couple one pointer tasks so it can get familiar with the code base.

让我们确保交给 Devin 几个单指针任务，以便它可以熟悉代码库。

Let's get it the thing.

让我们把事情搞清楚吧。

If Devin needs to be able to test the code or run the linter or CI or things like that.

如果 Devin 需要能够测试代码或运行 linter 或 CI 或类似的东西。

Obviously we want to make sure Devin's got its own virtual machine set up to be able to do that.

显然，我们希望确保 Devin 拥有自己的虚拟机，能够做到这一点。

Scott Wu: And similarly, I think the usage pattern, I think often it wasn't clear and obviously you can sit and just kind of watch Devin do it action by action and work that way, but we found that the best workflow really as a team building a lot of stuff was to work with multiple Devins and to run them asynchronously and to kick them off and to only jump in basically as you needed to provide feedback or steer the plan or anything like that.

Scott Wu：同样，我认为使用模式通常是不清楚的，显然你可以坐下来，只是看着 Devin 一个接一个地做，然后以这种方式工作，但我们发现，作为一个构建很多东西的团队，最好的工作流程实际上是与多个 Devin 一起工作，异步运行它们，启动它们，然后只在需要提供反馈或指导计划或类似的事情时才介入。

And so in many ways, I think Devin as a name really is our attempt to kind of capture the soul of that as a product where it really is treating it like a bit more of an autonomous entity that you can hand off tasks, that you can work with, that you should be teaching and learning with over time.

因此，从很多方面来说，我认为 Devin 作为一个名字确实是我们试图捕捉产品灵魂的尝试，它实际上将其视为一个自治实体，你可以移交任务，你可以与它一起工作，你应该随着时间的推移教学和学习。

Lenny Rachitsky: I want to come back to an area you started us down and then I took us away from which is impact on software engineering and how software engineering is going to change.

Lenny Rachitsky：我想回到你让我们失望的一个领域，然后我把我们带离了这个领域，即对软件工程的影响以及软件工程将如何改变。

So there's kind of two parts to this.

所以这有两个部分。

Just like when people are using Devin today, say this year, how is the act of being an engineer and building changing for those companies?

就像人们今天使用 Devin 时一样，比如今年，作为一名工程师和构建这些公司的行为有何变化？

What does that look like?

那看起来像什么？

Scott Wu: By the way, we're all software engineers ourselves.

Scott Wu：顺便说一下，我们自己都是软件工程师。

It's like I'm a programmer by training and still a programmer at heart certainly.

就好像我是一名经过训练的程序员，但本质上仍然是一名程序员。

And I think the way that we've always thought about it is there's layers of abstraction and there's tools, and one way I would say it at a high level is kind of I think of AI in general as, yeah, I mean computers are obviously getting more and more intelligence and are able to do more and more and it's possible there may come a day where computers truly do everything that we do and humans are not responsible for any of it.

我认为我们一直思考它的方式是有抽象层和工具，我想说的一种高层次的方式是我对人工智能的总体看法，是的，我的意思是计算机显然变得越来越智能并且能够做越来越多的事情，并且可能有一天计算机真正可以做我们所做的所有事情，而人类不需要对其中任何事情负责。

Scott Wu: I don't expect that to come particularly soon, but I guess what I would say is until that point, for as long as we're still part of the equation, one of the most important things to do obviously is for us as humans is to instruct our computers on what we want and what we want to build and what we want to do.

Scott Wu：我预计这不会很快到来，但我想我要说的是，直到那时，只要我们仍然是方程式的一部分，对于我们人类来说，显然要做的最重要的事情之一就是指导我们的计算机我们想要什么、我们想要构建什么以及我们想要做什么。

Right.

正确的。

And software engineering is, we think of it today obviously as Python and C++ and JavaScript and all these things, but at the end of the day, of course the discipline is all about just being to tell your computer what to do.

软件工程，我们今天显然将其视为 Python、C++ 和 JavaScript 以及所有这些东西，但归根结底，当然，这门学科就是告诉你的计算机要做什么。

And so in that lens, I really think that programming is, if anything, is only going to become more and more important as AI gets more powerful.

因此，从这个角度来看，我真的认为，随着人工智能变得越来越强大，编程（如果有的话）只会变得越来越重要。

And I think the thing that's really exciting for us is yeah, it's really seeing that kind of iterative transformation.

我认为对我们来说真正令人兴奋的是，它确实看到了这种迭代转变。

And so you ask what things look like today, and I would say, yeah, it really is like having a junior buddy or really a team of junior buddies that you can work with.

所以你问今天的情况是什么样的，我会说，是的，这真的就像有一个初级伙伴或一个真正的初级伙伴团队，你可以一起工作。

Right.

正确的。

Scott Wu: And so every engineer on our team, we use a ton of Devin when we're building Devin, and so Devin merges several hundred pull requests into production in the Devin code bases every month, which is, I mean, our whole team is only 15 engineers, and so it's a pretty sizable fraction of all the code that we write.

Scott Wu：所以我们团队中的每一位工程师在构建 Devin 时都会使用大量的 Devin，因此 Devin 每个月都会将数百个 Pull 请求合并到 Devin 代码库中的生产中，也就是说，我们整个团队只有 15 名工程师，所以它占我们编写的所有代码的相当大的一部分。

And the way that we use it is basically, yeah, everyone's got their whole team of Devins.

我们使用它的方式基本上是，是的，每个人都有他们的整个德文团队。

If you're going to be looking through various issues, if you're going through feature requests, if you're going through bugs, if you're going through new paradigms that you want to build, then it is naturally the case that there's a lot of handoff points where you just say, "Hey, at Devin, here's what's going on.

如果你要研究各种问题，如果你要处理功能请求，如果你要处理错误，如果你要处理你想要构建的新范例，那么很自然地会有很多交接点，你只需说，“嘿，在 Devin，这就是发生的事情。

Can you please take a pass at this?" Right.

能请你通过一下吗？”正确的。

Scott Wu: And sometimes Devin will be able to do the task 100% autonomously and just makes the PR and then you merge the PR and that's great.

Scott Wu：有时 Devin 能够 100% 自主地完成任务，只需制作 PR，然后合并 PR，这很棒。

Sometimes you want to be able to jump in for the 10 or 20% that really needs your help.

有时您希望能够帮助真正需要您帮助的 10% 或 20%。

Maybe there's a few details with how exactly you want to scope it or how you're architecting this feature, or maybe you want to go and test the front end at the end yourself to make sure it looks exactly the way that you want and give your one or two lines of feedback after that.

也许有一些细节涉及您想要如何精确地确定范围或如何构建此功能，或者您可能想自己在最后测试前端，以确保它看起来完全符合您想要的方式，然后提供一两行反馈。

Right.

正确的。

But a lot of it is really yeah, is kind of like yeah, learning to work with Devin to be able to just do more in parallel and build more.

但很多事情确实是的，有点像是的，学习与 Devin 合作，能够并行地做更多的事情并构建更多的东西。

Lenny Rachitsky: What percentage of your PRs are Devin versus humans right now?

Lenny Rachitsky：现在你们的 PR 中 Devin 与人类的比例是多少？

Scott Wu: Yeah, I'd have to look, but it's in the neighborhood of a quarter or so of all of our-

Scott Wu：是的，我得看看，但它就在我们四分之一左右的附近——

Lenny Rachitsky: And then what was it like six months ago?

Lenny Rachitsky：那么六个月前是什么样子呢？

Scott Wu: Oh yeah, it's grown a ton for, I mean, we've seen it grown exponentially internally ourselves as well.

Scott Wu：哦，是的，它已经增长了很多，我的意思是，我们自己也看到它在内部呈指数级增长。

And so it's kind of an interesting one where again, it's always both the capabilities and the product interface.

因此，这是一件有趣的事情，同样，它始终是功能和产品界面。

And so I think the intelligence has increased a lot.

所以我认为智力提高了很多。

But the other thing of course is that we've spent a lot of time in figuring out how to build and to really kind of build for an interface where you can get Devin's value on tasks where Devin is able to do the 80 or 90%.

但另一件事当然是，我们花了很多时间来弄清楚如何构建以及真正构建一个界面，在该界面中，您可以在 Devin 能够完成 80% 或 90% 的任务上获得 Devin 的价值。

And so Devin is obviously not perfect and it'll make mistakes and so on.

所以 Devin 显然并不完美，它会犯错误等等。

And a lot of the question is basically, yeah, how do you scope out your initial task with Devin and then just kind of set Devin off and have it go and do the things that you want do?

很多问题基本上是，是的，你如何与德文一起确定最初的任务，然后让德文开始做你想做的事情？

How do you come in at the end and review and give feedback?

最后你如何参与审查并提供反馈？

How do you make sure Devin learns over time?

你如何确保 Devin 随着时间的推移不断学习？

How are you able to kind of just check in as needed and course correct if you want to?

如果您愿意的话，您如何能够根据需要进行登记并纠正路线？

Lenny Rachitsky: Okay, so today about quarter of your PRs are Devins.

Lenny Rachitsky：好的，今天大约四分之一的 PR 是 Devins。

Where do you think this will be at the end of the year?

您认为今年年底会是什么情况？

What would you guess?

你猜怎么着？

Scott Wu: I think by the end of this year, we expect it to be more than half.

Scott Wu：我认为到今年年底，我们预计会超过一半。

And I mean, as time goes on, one of the things that we've seen is just you're able to do more and more and more work asynchronously.

我的意思是，随着时间的推移，我们看到的一件事就是您能够异步完成越来越多的工作。

Right.

正确的。

And you're able to hand off more and more.

而且你能够交出越来越多的东西。

I think the soul of programming, the soul of software engineering has really been about through all the areas, not just now, but even when it was assembly, right, and even when it was Pascal and even when it was punch cards or whatever, I think the soul of it has really been basically just about defining the problem that you're facing and really thinking through exactly...

我认为编程的灵魂，软件工程的灵魂确实贯穿了所有领域，不仅仅是现在，甚至在汇编时，对，甚至在 Pascal 时，甚至在打孔卡或其他什么时，我认为它的灵魂基本上就是定义你所面临的问题并真正思考清楚......

Scott Wu: ...

吴斯科特：...

about defining the problem that you're facing and really thinking through exactly what is the solution that you want to build.

定义您面临的问题并真正思考您想要构建的解决方案是什么。

Thinking through the architecture, thinking through the details, and really mapping out in your mind exactly what you want to build basically and what you want to have your computer do.

思考架构，思考细节，并在你的脑海中真正准确地规划出你想要构建的基本内容以及你想让你的计算机做什么。

I think that's what makes software engineering really great and I think that's the funnest part of software engineering.

我认为这就是软件工程真正伟大的原因，也是软件工程中最有趣的部分。

Scott Wu: I think, at the same time, that's probably in the neighborhood of 10% of the average software engineer's time, right?

Scott Wu：我认为，与此同时，这可能相当于软件工程师平均时间的 10% 左右，对吗？

Because 90% of the time is you've got this Kubernetes error, then you've got to debug, and you have to see what went wrong and the system crash, or you left some port open and this is messing up, or there's a bug report that you have to take care of, or you've got to migrate your code, or you got it upgraded to a new version or things like that.

因为 90% 的情况是你遇到了这个 Kubernetes 错误，然后你必须进行调试，你必须查看出了什么问题以及系统崩溃，或者你打开了某些端口，这导致了混乱，或者有一个错误报告你必须处理，或者你必须迁移你的代码，或者你将其升级到新版本或类似的事情。

A lot more implementation.

更多实施。

Scott Wu: One of the ways that we've kind of thought about Devin in building Devin is really allowing engineers to go from bricklayer to architect, so to speak.

Scott Wu：我们在构建 Devin 时考虑的方式之一是真正允许工程师从砖瓦匠转变为建筑师，可以这么说。

A lot of it is just getting to the point where you can do the high-level directing and you can basically specify things exactly how you want.

很多事情只是达到了你可以进行高水平指导的程度，你基本上可以按照你想要的方式指定事情。

I think it's very much about still having the human in control and having the human able to do the full specification, but just multiplying the magnitude of what you can do and what you can build in one day or one hour or however long.

我认为这在很大程度上仍然是让人类保持控制并让人类能够完成完整的规范，但只是将您可以做的事情以及您在一天或一小时或无论多长时间内可以构建的内容相乘。

Lenny Rachitsky: In the future, say someone is trying to get into software engineering, thinking about becoming an engineer, first of all, do you think people should...

Lenny Rachitsky：在未来，假设有人试图进入软件工程领域，考虑成为一名工程师，首先，你认为人们应该……

Classic question everyone's getting these days.

这些天每个人都会遇到的经典问题。

Should you still learn to code?

你还应该学习编码吗？

Lenny Rachitsky: I'd love your perspective there.

Lenny Rachitsky：我很喜欢你的观点。

And then two, for people that are engineers today, what skills do you think will be more and more important and then less important in this discussion of moving from bricklayer to architect?

第二个问题，对于今天的工程师来说，您认为在讨论从砖瓦匠转向建筑师的过程中，哪些技能会变得越来越重要，然后又变得越来越不重要？

Scott Wu: Yeah, for sure.

吴斯科特：是的，当然。

I love this question.

我喜欢这个问题。

First of all, the question of whether you should still learn to code, my answer would be absolutely yes.

首先，对于你是否还应该学习编码的问题，我的答案是肯定的。

I think, to a large extent, when you take computer science classes and when you learn these fundamentals, sure you're learning a little bit about how a particular language is, syntax works or something like that.

我认为，在很大程度上，当你参加计算机科学课程并学习这些基础知识时，你肯定会了解一些特定语言的原理、语法的工作原理或类似的东西。

But honestly, most of what you're learning really is about the ability to logically break down problems for number one.

但老实说，你学到的大部分内容实际上都是关于逻辑地分解第一个问题的能力。

And two, I would say is just the model of a computer and a lot of these decisions and a lot of the abstractions that we've built over time.

第二，我想说的是计算机模型以及许多这样的决策以及我们随着时间的推移构建的许多抽象。

Scott Wu: What is a database and how should you think about a database?

Scott Wu：什么是数据库？您应该如何看待数据库？

What is a garbage collection system and how do those work and all of these different pieces?

什么是垃圾收集系统？它们如何工作以及所有这些不同的部分？

The reason I think that's important is because it's the same with a lot of these other...

我认为这很重要的原因是因为它与许多其他......

Arguably we've already gone through these phases in programming and I think this next one is going to be somewhat faster and somewhat bigger, but in many ways a similar flavor, which is when you work with Python today, obviously, a lot of things are already abstracted away from you.

可以说，我们已经经历了编程的这些阶段，我认为下一个阶段会更快、更大，但在很多方面都有相似的风格，这就是当你今天使用 Python 时，显然，很多东西已经从你身上抽象出来了。

Scott Wu: In some sense, someone from 50 years ago might already call Python.

Scott Wu：从某种意义上说，50 年前的人可能已经称之为 Python。

You just get to explain in English what you want and now the computer does it for you.

您只需用英语解释您想要什么，现在计算机就会为您完成。

That's great and I think it's really powerful.

这太棒了，我认为它真的很强大。

It's opened it up.

它已经打开了。

I mean, we have far more programmers, obviously, than we ever have before because of that, but I would say certainly as you're building your skills as an engineer, it really helps a lot to understand the abstractions and to be able to peel the layers beneath.

我的意思是，显然，我们的程序员比以前多得多，但我肯定会说，当你正在培养工程师的技能时，理解抽象并能够剥离下面的层确实有很大帮助。

Folks will use assembly for example if they're really performance optimizing a piece of code, but also in order to build good systems and to understand these things, you certainly want to understand these abstractions of how does networking work.

例如，如果人们确实在优化一段代码的性能，那么他们会使用汇编，但为了构建良好的系统并理解这些东西，您当然想了解网络如何工作的这些抽象。

Scott Wu: What is TCP/IP like exactly or what happens with this Python code when it gets interpreted or all of these details.

Scott Wu：TCP/IP 到底是什么样的，或者当这个 Python 代码被解释时会发生什么，或者所有这些细节。

Similarly, I think we will get to a state where, with no experience at all, you're going to be able to build some pretty cool stuff and to do some pretty amazing work just by explaining what it is that you want.

同样，我认为我们将达到这样一种状态，在没有任何经验的情况下，你将能够仅仅通过解释你想要什么来构建一些非常酷的东西并做一些非常令人惊奇的工作。

But I think that, for quite some time, you really want to be able to think precisely about the details, to peel back the abstractions, to be very precise about what it is that you want to build and how.

但我认为，在相当长的一段时间里，你确实希望能够精确地思考细节，剥离抽象，非常精确地了解你想要构建什么以及如何构建。

Lenny Rachitsky: And then for skills that you think are more and more valuable for engineers, where should engineers today be leaning more and more into versus like, "Forget this.

Lenny Rachitsky：对于你认为对工程师来说越来越有价值的技能，今天的工程师应该越来越倾向于什么，而不是像“忘了这个。

I don't need to think about this anymore"?

我不需要再考虑这个了？”

Scott Wu: For sure.

吴斯科特：当然。

I think architect...

我认为建筑师...

I mean, we already have a term for architect and engineering and I think it is directionally the right term.

我的意思是，我们已经有了一个关于建筑师和工程的术语，我认为这在方向上是正确的术语。

It's, I think, one thing to just do a routine implementation and write boilerplate code and things like that.

我认为，仅仅进行例行实现并编写样板代码之类的东西是一回事。

I would say that, in many ways, AI coding has already made us much faster at that, but I think a lot of the core questions of understanding very complex systems and working in the context of the whole company and thinking about the product that you're building or the work that you're doing and understanding, "Okay, what are the problems that we want to solve?

我想说，在很多方面，人工智能编码已经让我们更快了，但我认为很多核心问题是理解非常复杂的系统，在整个公司的背景下工作，思考你正在构建的产品或你正在做的工作并理解，“好吧，我们想要解决的问题是什么？

How do we want to solve those problems?

我们想如何解决这些问题？

What is exactly the solution that we want to build?

我们想要构建的解决方案到底是什么？

What are all of these key decisions and trade-offs that we're going to be making?"

我们将要做的所有这些关键决策和权衡是什么？”

Scott Wu: Basically, I think folks who are able to do that really, really well are just going to be able to leverage themselves more and more.

Scott Wu：基本上，我认为那些能够真正非常出色地做到这一点的人将能够越来越多地发挥自己的作用。

If anything, I think there's going to be way more programmers and way more engineers a few years from now than there are today.

如果有的话，我认为几年后将会有比现在更多的程序员和工程师。

I think pretty quickly the form factor of what it means to be a programmer, obviously, is going to change.

我认为，作为一名程序员的意义显然很快就会发生变化。

And in some sense, it already has, but I think there's just going to be so much more for us to build.

从某种意义上说，它已经做到了，但我认为我们还会做更多的事情。

Folks talk about Jevons Paradox all the time.

人们一直在谈论杰文斯悖论。

I mean, software is truly the shining example of Jevons Paradox, where we have always managed as a society to find more and more things that we want to build software for and build more code for.

我的意思是，软件确实是杰文斯悖论的光辉例子，我们作为一个社会总是设法找到越来越多的东西，我们想要为其构建软件并为其构建更多代码。

I really think there's a lot more out there to do.

我真的认为还有很多事情要做。

Lenny Rachitsky: For people that don't know Jevons Paradox, can you briefly explain it?

Lenny Rachitsky：对于那些不了解杰文斯悖论的人，你能简单解释一下吗？

Scott Wu: Absolutely.

吴斯科特：当然。

Yeah.

是的。

Jevons Paradox just says that as the price of something goes down, it can still be the case that the total spend on it actually goes up.

杰文斯悖论只是说，随着某种东西的价格下降，其总支出实际上仍然可能上升。

You can think about this with money, you can think about this with time or resources, but the direct version here is, I think, as it becomes easier and easier to program and as programming becomes more and more effective, I think we're going to have a lot more programmers.

你可以用金钱来思考这个问题，你可以用时间或资源来思考这个问题，但我认为，这里的直接版本是，随着编程变得越来越容易，随着编程变得越来越有效，我认为我们将会有更多的程序员。

I think in a kind of zero-sum view, you might say, "Well, we're going to be 10 times faster at software engineering and it means that we're going to need 10 times fewer software engineers." But I think in practice, what really is going to happen is actually we're going to build even more than 10 times as much code.

我认为，从某种零和观点来看，你可能会说，“嗯，我们的软件工程速度将提高 10 倍，这意味着我们需要的软件工程师将减少 10 倍。”但我认为在实践中，真正发生的情况是我们将构建 10 倍以上的代码。

And because all of the work that we do is so capped, obviously, on our ability to actually build and execute and iterate, we're going to have so many great ideas out there, we're going to have so many great products out there.

显然，因为我们所做的所有工作都受到我们实际构建、执行和迭代的能力的限制，所以我们将有很多伟大的想法，我们将有很多伟大的产品。

People are going to build a lot more personalized experiences, for example, and there's going to be a lot to do.

例如，人们将打造更多个性化的体验，而且还有很多事情要做。

Lenny Rachitsky: Going back to the way you guys use Devin, you said that every engineer has this fleet of Devins.

Lenny Rachitsky：回到你们使用 Devin 的方式，你们说每个工程师都有这支 Devin 车队。

How many Devins per engineer do you find most people are working with these days at your company?

您发现现在你们公司的大多数人都与多少位 Devins 一起工作？

Scott Wu: Yeah.

吴斯科特：是的。

It's very asynchronous.

这是非常异步的。

Obviously, you can kick them up and start them up and shut them down basically as you see fit, but most folks on the team are often working with up to five Devins at once, I would say.

显然，你可以根据自己的需要启动和关闭它们，但我想说，团队中的大多数人经常同时与最多五个德文一起工作。

It's a nice flow where you think through, "All right, what are the five things that we want to get done today?" You have Devin one do number one, you have Devin two do number two, Devin three...

这是一个很好的流程，你可以思考：“好吧，我们今天想要完成的五件事是什么？”你让德文一做第一，你让德文二做第二，德文三……

For what it's worth, I think it's taken us some time to adjust to it and get to the point where it's really intuitive for us, but I think it's definitely a different experience where you're handing off most things asynchronously.

就其价值而言，我认为我们花了一些时间来适应它并达到对我们来说真正直观的地步，但我认为这绝对是一种不同的体验，你异步地交接了大多数事情。

Scott Wu: The goal for each of your tasks is to be there for the parts that really need your expertise, either you really, really need to define exactly what it is that you're solving for and what you're building or maybe some of the more complex parts where you want to steer Devin towards, particularly what kinds of changes you want to make.

Scott Wu：你每项任务的目标是完成真正需要你专业知识的部分，要么你真的真的需要准确定义你正在解决的问题和你正在构建的内容，要么是你想要引导 Devin 走向的一些更复杂的部分，特别是你想要做出什么样的改变。

"I want the class to be set up this way and we should go and change all the downstream references to this as well or whatever." But basically having Devin do the bulk of the work asynchronously with you.

“我希望以这种方式设置类，我们应该去更改所有对此的下游引用或其他内容。”但基本上让 Devin 与您异步完成大部分工作。

Lenny Rachitsky: And then how many engineers do you guys have roughly?

Lenny Rachitsky：那么你们大约有多少工程师？

Scott Wu: Yeah, our engineering team today is about 15 people.

Scott Wu：是的，我们今天的工程团队大约有 15 人。

Lenny Rachitsky: Holy moly.

莱尼·拉奇茨基：天啊。

Okay.

好的。

And then each one has five-ish Devins.

然后每个人都有五个左右的德文。

Lenny Rachitsky: So there's five times the number of Devins as engineers.

Lenny Rachitsky：所以工程师数量是 Devins 的五倍。

What I love about this is just a glimpse into where the future is going.

我喜欢这只是对未来发展方向的一瞥。

You guys are so ahead of how companies work with AI engineers.

你们对于公司与人工智能工程师的合作方式非常领先。

Seeing how you operate is going to be, as a sense, essentially how most companies will end up operating.

从某种意义上说，了解你的运营方式基本上将成为大多数公司最终的运营方式。

Scott Wu: Yeah.

吴斯科特：是的。

For what it's worth, we've already seen this shift, I would say, ourselves where...

无论如何，我们已经看到了这种转变，我想说，我们自己......

In terms of the team, obviously, folks don't spend that much of their time just writing out boilerplate or just doing pure implementation of features.

就团队而言，显然，人们不会花费太多时间来编写样板或纯粹执行功能。

People get to spend much more of their time focused on really just thinking about the core questions of, "How do we make Devin better?

人们会花更多的时间专注于思考核心问题，“我们如何让 Devin 变得更好？

What is the right interface for Devin?

Devin 的正确界面是什么？

What is the right flow or the right set of features that's really going to make this as great of an experience as possible?" That's how we like things.

什么才是真正能让这种体验尽可能出色的正确流程或正确功能集？”这就是我们喜欢事物的方式。

Lenny Rachitsky: When is the point you reach, where there's takeoff of this being the by...

Lenny Rachitsky：你什么时候到达这个点，哪里才是这一切的起飞......

Your Devin starts moving so much further ahead of everyone else.

你的德文开始比其他人走得更远。

Once you have enough Devins doing all these things, they're just like wherever and you're 10 years, 20 years, 30 years, 100 years ahead.

一旦你有足够多的 Devins 来做所有这些事情，他们就像在任何地方一样，你就会领先 10 年、20 年、30 年、100 年。

Scott Wu: Honestly, as a community, I think all of us as engineers around the world, I think, are going to have to think about this and build for this and adapt to these new technologies.

Scott Wu：老实说，作为一个社区，我认为我们作为世界各地的工程师，都必须考虑这一点，并为此进行构建并适应这些新技术。

But what I would say is, I think more and more...

但我要说的是，我越来越觉得……

Especially as capabilities get better, but certainly even in studies say today, I think more and more things are going to shift towards this asynchronous flow.

特别是随着能力变得更好，但即使在今天的研究中，我认为越来越多的事情将转向这种异步流程。

And one of the reasons I would say for that is in the real world, you're just capped by real world constraints.

我想说的原因之一是在现实世界中，你只是受到现实世界的限制。

I think that one way to put it is kind of like...

我认为一种表达方式有点像......

Don't take these numbers exactly, but it's kind of like the first order math of it is, of course, being able to write files or to complete this function or complete this line or things like that.

不要完全接受这些数字，但这有点像它的一阶数学，当然，能够写入文件或完成此功能或完成这一行或类似的事情。

Scott Wu: It helps a ton.

Scott Wu：这有很大帮助。

It's a really great experience.

这真是一次很棒的经历。

There's a lot of parts of building software that obviously are almost not that at all, right?

构建软件的很多部分显然几乎根本不是那么回事，对吧？

If you have a bug that you're trying to fix and so you spin up the local server, you click around on your own product on the front end and try to reproduce the bug yourself.

如果您有一个试图修复的错误，因此您启动了本地服务器，则可以在前端单击您自己的产品并尝试自己重现该错误。

Once you have the error, you take a look at Datadog and you see what happened and you try to find other errors in the logs.

一旦出现错误，您可以查看 Datadog，了解发生了什么，并尝试在日志中查找其他错误。

You look at those files and you see what went wrong, you make some edits, maybe you go and rerun the whole process again now that you make sure your change looks right.

您查看这些文件，发现出了什么问题，您进行了一些编辑，也许您现在可以确保您的更改看起来正确，然后再次重新运行整个过程。

That's a lot of what it means to be a software engineer.

这就是成为一名软件工程师的意义所在。

These are processes that take real time.

这些过程需要实时进行。

I think we're going to shift more and more towards this agentic workflow, because that's in some ways the way to really get to that 200%, 500%, 1000% gains that we'll be getting to with software engineering over the next few years.

我认为我们将越来越多地转向这种代理工作流程，因为从某种程度上来说，这是我们在未来几年软件工程中真正实现 200%、500%、1000% 收益的方式。

Lenny Rachitsky: Enough talk.

莱尼·拉奇茨基：够了。

Let's show people what the heck this actually looks like.

让我们向人们展示这到底是什么样子的。

You've got a couple demos prepped that show a few use cases that you found helpful.

您已经准备了几个演示，其中展示了一些您认为有用的用例。

You're going to pull up your screen and then we'll kick it off and then we'll talk as it's happening.

你要拉起你的屏幕，然后我们将开始它，然后我们将在它发生时进行交谈。

Scott Wu: Yeah.

吴斯科特：是的。

The whole process obviously of working with Devin is working asynchronously.

显然，与 Devin 合作的整个过程都是异步进行的。

I thought it'd be cool for us to actually just watch Devin a little bit in action and then we can go through some other examples of work that Devin's done or things that Devin does for us, even on our team for example.

我认为对我们来说，真正观看一下 Devin 的实际行动会很酷，然后我们可以浏览 Devin 所做的其他一些工作示例或 Devin 为我们所做的事情，甚至是在我们的团队中。

But then we can check back in asynchronously with our Devin after.

但之后我们可以与 Devin 异步检查。

Scott Wu: I'll share this real quick.

Scott Wu：我会很快分享这一点。

The key thing that I would just emphasize here is a lot of it obviously is really just about thinking about as a software engineer or as engineers ourselves or engineering teams, PMs, and so on.

我在这里要强调的关键是，很明显，其中很多实际上只是作为软件工程师或我们自己的工程师或工程团队、项目经理等的思考。

What are the things that we would want to build that we would want to hand off?

哪些是我们想要构建但想要交付的东西？

We have Devin set up with our own Devin code base, for example, and so I'll go ahead and kick off with Devin for that.

例如，我们让 Devin 设置了我们自己的 Devin 代码库，所以我将继续与 Devin 一起开始。

I'll just say, "Devin, I'm on with my friend, Lenny."

我只会说，“德文，我和我的朋友莱尼在一起。”

Scott Wu: "Can you modify Devin web app two?" Let's feature your newsletter as part of the Devin website.

Scott Wu：“你能修改 Devin Web 应用程序二号吗？”让我们将您的时事通讯作为 Devin 网站的一部分。

Lenny Rachitsky: Let's do it.

莱尼·拉奇茨基：让我们开始吧。

On the real Devin website.

在真正的德文网站上。

Scott Wu: We're going to kick this off.

Scott Wu：我们要开始了。

As you can see, Devin gets started instantly and goes ahead and responds.

正如您所看到的，Devin 立即开始并做出回应。

Again, you can work with this asynchronously, you can work with it synchronously as well.

同样，您可以异步使用它，也可以同步使用它。

For this, we'll just kind of go in a little bit and see exactly what's going on.

为此，我们将稍微深入一下，看看究竟发生了什么。

But as you can see here, Devin's going through files and taking a look through a lot of stuff.

但正如您在这里看到的，德文正在浏览文件并查看很多内容。

We can follow here basically as we need to and see what makes sense.

我们基本上可以根据需要遵循这里，看看什么是有意义的。

You can see Devin's already called out a few particular pieces where there's the sidebar, which we have implemented on the front end, and there's pieces there.

你可以看到 Devin 已经调出了一些特定的部分，其中有侧边栏，我们已经在前端实现了，并且那里有一些部分。

We're going to have a new component and that component's going to link to Lenny's website.

我们将有一个新组件，该组件将链接到 Lenny 的网站。

That all sounds good.

这一切听起来都不错。

Devin's asking us any questions if there's anything that we have here.

如果我们这里有什么东西，德文会问我们任何问题。

Same story here where you can let Devin make its own decisions and hand off, or you can go ahead and give some more thoughts.

同样的故事在这里，你可以让 Devin 自己做出决定并移交，或者你可以继续给出更多想法。

Should the button open in a new tab or within an application?

该按钮应该在新选项卡中还是在应用程序中打开？

I'll say let's open it in a new tab.

我会说让我们在新选项卡中打开它。

Lenny Rachitsky: And you could answer these at any point.

Lenny Rachitsky：你可以随时回答这些问题。

Is it waiting for the answer?

是在等待答案吗？

Scott Wu: You answer these at any point, you can hand off, hand back off.

Scott Wu：你可以随时回答这些问题，你可以把手递过去，也可以把手收回来。

Lenny Rachitsky: It's not going to be like, "Just god damn it.

莱尼·拉奇茨基（Lenny Rachitsky）：不会是这样，“天哪。

I just wrote it this way.

我就是这样写的。

Why didn't you tell me earlier?"

为什么不早点告诉我？”

Scott Wu: That's right.

吴斯科特：是的。

One of the big pieces with Devin is Devin will always be enthusiastic, will always be ready to put in the hours.

德文最重要的事情之一是德文将永远充满热情，永远准备好投入时间。

Lenny Rachitsky: Thanks for changing scope.

Lenny Rachitsky：感谢您更改范围。

Thanks, Scott.

谢谢，斯科特。

Scott Wu: We'll give Devin a chance to work and it's going to go through these files and it'll make a pull request for us and we'll see and go from there.

Scott Wu：我们会给 Devin 一个工作的机会，它会检查这些文件，它会为我们提出一个 Pull 请求，我们会看到并从那里开始。

But I thought it'd be fun to show some other examples of Devin in action as well.

但我认为展示德文的其他一些实际例子也会很有趣。

One of the examples actually this morning, which I just used Devin for, is I asked Devin to help me brush my own facts up for this podcast.

实际上，今天早上我刚刚使用 Devin 的例子之一是，我要求 Devin 帮助我为这个播客整理我自己的事实。

Obviously, a huge fan of the podcast and the newsletter.

显然，他是播客和时事通讯的忠实粉丝。

I asked Devin, "Hey, Devin.

我问德文：“嘿，德文。

I'm to be on the podcast.

我要参加播客。

Could you please research everything you can about him and make a nice website quiz for me so that I can make sure I know my facts?" This was just this morning, I asked Devin to do this and I'll just show what Devin did, it looks like.

您能否研究有关他的一切信息，并为我做一个很好的网站测验，以便我可以确保我了解我的事实？”就在今天早上，我要求德文这样做，我将展示德文所做的事情，看起来像这样。

Went to Wikipedia first.

首先去了维基百科。

Unfortunately, it's not a page in Wikipedia, which is...

不幸的是，这不是维基百科中的页面，这是......

Lenny, we will work on that.

莱尼，我们会努力解决这个问题。

Lenny Rachitsky: I'm not a big enough deal yet.

Lenny Rachitsky：我的地位还不够高。

Scott Wu: They did you dirty.

Scott Wu：他们把你弄脏了。

I mean, we need a page list.

我的意思是，我们需要一个页面列表。

And so then it went and found it on Spotify.

然后它就在 Spotify 上找到了它。

Lenny Rachitsky: So you're watching what it's researching live?

Lenny Rachitsky：所以你正在现场观看它正在研究什么？

Scott Wu: Yeah, yeah, yeah.

吴斯科特：是啊，是啊，是啊。

This was this morning, obviously.

显然，这是今天早上。

Lenny Rachitsky: This is a playback of what Devin did.

Lenny Rachitsky：这是 Devin 所做事情的回放。

This is part of Devin you could just watch what it did.

这是 Devin 的一部分，你可以看看它做了什么。

Scott Wu: Yeah, yeah.

吴斯科特：是的，是的。

Especially when you're building engineering products or something like that, you can see each of the steps that Devin was doing, or if Devin tested the code locally, obviously you want to be able to go and look and see what Devin was cooking around with and testing or things like that.

特别是当你正在构建工程产品或类似的东西时，你可以看到 Devin 正在执行的每个步骤，或者如果 Devin 在本地测试了代码，显然你希望能够去看看 Devin 正在做什么和测试或类似的事情。

It found the newsletter.

它找到了新闻通讯。

It's going and looking at this and it's going and reading all of this.

它会去查看这个，它会去阅读所有这些。

And then it says, "Okay, let's get started with putting the code together." It says, "Hey, I've researched." It's going through and writing all of this, putting the app together.

然后它说：“好吧，让我们开始将代码放在一起。”它说：“嘿，我已经研究过了。”它正在经历并编写所有这些，将应用程序放在一起。

It plays its own quiz itself.

它自己进行测验。

Actually, we should just play this quiz actually.

事实上，我们应该直接玩这个测验。

Let's see how much I know.

让我们看看我知道多少。

Lenny Rachitsky: What is the name of the podcast?

Lenny Rachitsky：播客的名称是什么？

Lenny's Podcast.

莱尼的播客。

For people not watching, so to say, approximately how many subscribers?

对于不观看的人来说，大约有多少订阅者？

A million.

一百万。

Very good.

非常好。

Lenny Rachitsky: What are three main topics Lenny is focused on?

Lenny Rachitsky：Lenny 关注的三个主要主题是什么？

Oh, product, growth, and career.

哦，产品、成长和职业。

Very good.

非常好。

It's a good quiz of people.

这是对人们的一个很好的测验。

Lenny Rachitsky: What does Lenny do besides podcasting?

Lenny Rachitsky：Lenny 除了播客之外还做什么？

Scott Wu: I'd say writing, investing, and advising.

Scott Wu：我想说的是写作、投资和咨询。

How often does Lenny-

莱尼多久-

Scott Wu: Yeah.

吴斯科特：是的。

We can go through all these and do all these.

我们可以经历所有这些并做到所有这些。

I took this quiz, by the way, obviously, to make sure that I was well-prepped, but this is kind of one of the more fun examples, obviously, of

顺便说一句，显然，我参加了这个测验，以确保我做好了充分的准备，但这显然是更有趣的例子之一

Lenny Rachitsky: Scott, how many subscribers do I have at my newsletter?

Lenny Rachitsky：斯科特，我的时事通讯有多少订阅者？

Scott Wu: And then one last one I'll show and then maybe we can come back to our initial run after.

Scott Wu：然后我会展示最后一个，然后也许我们可以回到我们最初的运行。

Like I was saying, a lot of this is really built to work with all of the existing code workflows out there.

就像我说的，其中很多内容实际上是为了与所有现有的代码工作流程一起使用而构建的。

For example, we were doing some exploration with the DeepSeek repository on GitHub and we imported it into Devin and we got our own fork of it set up in Devin.

例如，我们正在 GitHub 上的 DeepSeek 存储库中进行一些探索，并将其导入到 Devin 中，并在 Devin 中设置了我们自己的分支。

A couple of things I just wanted to show here.

我只是想在这里展示一些事情。

One is Devin sets up its whole wiki with all of its internal understanding.

一是 Devin 建立了整个 wiki，并包含其所有内部理解。

When Devin indexes the code base, obviously, building a representation of the code base and learning it and improving it over time is one of the big things that Devin does.

显然，当 Devin 对代码库进行索引时，构建代码库的表示并学习它并随着时间的推移对其进行改进是 Devin 所做的重要事情之一。

Funnily enough, we found that naturally humans really are interested to understand this code base representation as well.

有趣的是，我们发现人类自然也确实有兴趣理解这种代码库表示形式。

Scott Wu: Devin Wiki is something that we built here and you can take a look at all these different pieces and see each of these different things.

Scott Wu：Devin Wiki 是我们在这里建立的，你可以查看所有这些不同的部分并看到每一个不同的东西。

Here are the FP8 operations, here's an SG integration.

这是 FP8 操作，这是 SG 集成。

There's diagrams of how the different layers are built and put together.

有一些图表显示了不同层是如何构建和组合在一起的。

There's deployment operations.

有部署操作。

There's a lot of details about the architecture as well.

还有很多关于架构的细节。

You can ask questions about it as well.

您也可以提出相关问题。

For example, you can say, "How does DeepSeek handle multi-token prediction designed for spec deck?" It'll go through and it's able to search through the entire code base and give you an informed answer based off of that.

例如，您可以说，“DeepSeek 如何处理为规范组设计的多令牌预测？”它会进行检查，并且能够搜索整个代码库，并据此为您提供明智的答案。

Scott Wu: We use this a lot.

Scott Wu：我们经常使用这个。

It helps when you're scoping out a task for Devin and doing an initial prompt.

当您为 Devin 确定任务范围并进行初始提示时，它会很有帮助。

It also helps, obviously just in a vacuum, you often have questions about your code base that are really nice.

它也有帮助，显然只是在真空中，您经常对代码库有非常好的疑问。

Lenny Rachitsky: This episode is brought to you by Attio, the AI-native CRM.

Lenny Rachitsky：本集由人工智能原生 CRM Attio 为您带来。

Attio is built to scale with your business from day one.

Attio 从第一天起就根据您的业务进行扩展。

Connect your email and calendar and Attio instantly builds a CRM that matches your business model, with all of your company's Contacts and interactions enriched with actionable insights.

连接您的电子邮件和日历，Attio 即可立即构建与您的业务模式相匹配的 CRM，并通过可操作的见解丰富您公司的所有联系人和交互。

Sync in your product's usage, billing info, or any other data sources and Attio's flexible data model will handle it all without any rigid templates or workarounds.

同步产品的使用情况、账单信息或任何其他数据源，Attio 灵活的数据模型将处理所有这些，无需任何严格的模板或解决方法。

Lenny Rachitsky: With Attio, AI isn't just a feature.

Lenny Rachitsky：对于 Attio，人工智能不仅仅是一个功能。

It's the foundation.

这是基础。

You can do things like instantly prospect and route leads with research agents, get real-time insights from AI using customer conversations, and build powerful AI automations for your most complex workflows.

您可以执行诸如立即与研究代理寻找潜在客户和路由线索、使用客户对话从 AI 中获取实时见解以及为最复杂的工作流程构建强大的 AI 自动化等操作。

Industry leaders like Flatfile, Replicate, and Modal are already experiencing what's next for CRM.

Flatfile、Replicate 和 Modal 等行业领导者已经在体验 CRM 的下一步发展。

Go to A-T-T-I-O.com/lenny to get 15% off your first year.

访问 A-T-T-I-O.com/lenny 即可享受第一年 15% 的折扣。

That's A-T-T-I-O.com/lenny.

那是 A-T-T-I-O.com/lenny。

Lenny Rachitsky: Something that I've learned as I've been talking to more and more AI building companies and apps is there's a big difference in how large of a code base they could integrate into.

Lenny Rachitsky：当我与越来越多的人工智能构建公司和应用程序交谈时，我了解到，它们可以集成到多大的代码库中存在很大差异。

That's a big deal for companies that are existing versus startups, people that have large existing code base.

对于现有公司和初创公司以及拥有大量现有代码库的公司来说，这是一件大事。

How should people think about what kind of code base Devin can plug into?

人们应该如何考虑 Devin 可以插入什么样的代码库？

Scott Wu: Yeah.

吴斯科特：是的。

We go all the way to the biggest code base as possible.

我们一直致力于尽可能最大的代码库。

One way I'd kind of put it is how...

我想说的一种方式是......

The way that we as engineers would think about a large code base is certainly when you're making changes or when you're thinking about a particular task.

作为工程师，我们考虑大型代码库的方式肯定是在您进行更改或考虑特定任务时。

You're not bringing in every single line of the code base at once.

您不会立即引入代码库的每一行。

You have a high level of traction that you're able to think about and look into and then you're obviously able to zoom in and get to higher resolution on each of these different things.

你有很高的吸引力，你能够思考和研究，然后你显然能够放大并在每个不同的事情上获得更高分辨率。

Devin works in much the same way, where the first thing it'll do is it's going to figure out the high level architecture of what's going on here and what this is built for and so on.

Devin 的工作方式大致相同，它要做的第一件事就是弄清楚这里发生的事情的高级架构以及它的构建目的等等。

But within each of the components, it's obviously also going to be able to zoom in and give some more detail about each of these.

但在每个组件中，显然还能够放大并提供有关每个组件的更多细节。

Here's FP8 to be float 16 and how exactly a lot of that is set up.

下面是 FP8 为 float 16 以及其中的大部分是如何设置的。

Here's each of the different parts of the code base.

这是代码库的每个不同部分。

Similarly, we built this to be scalable.

同样，我们将其构建为可扩展的。

Lenny Rachitsky: It's essentially coming back to the engineer as architect.

Lenny Rachitsky：本质上，这又回到了工程师作为建筑师的问题。

Now, it's helping you understand the architecture, kind of circling back to that.

现在，它正在帮助您理解架构，有点回到这一点。

Scott Wu: Yeah.

吴斯科特：是的。

Yeah, exactly.

是的，完全正确。

One of the fun use cases that we've seen actually with folks is they'll often actually get Devin's help to onboard new engineers on the team.

我们在人们身上看到的一个有趣的用例是，他们经常会得到 Devin 的帮助来加入团队中的新工程师。

When you're new and you're joining, there's obviously a lot of questions that you have about the code base or about how things are set up.

当您是新加入时，显然您对代码库或如何设置有很多疑问。

It also sometimes can be a little bit awkward to ask your mentor or your manager the questions and if you're worried that they're going to be really dumb questions.

有时，如果您担心这些问题会是非常愚蠢的问题，那么向您的导师或经理提出问题可能会有点尴尬。

It's nice to just be able to ask Devin and to go through Devin's wiki and to understand these internal representations.

很高兴能够询问 Devin 并浏览 Devin 的 wiki 并理解这些内部表示。

Lenny Rachitsky: I think that's really interesting, because it comes back to your point that Devin is not just a junior engineer.

Lenny Rachitsky：我认为这真的很有趣，因为这又回到了你的观点，Devin 不仅仅是一名初级工程师。

It's what you call a jagged engineer.

这就是所谓的锯齿状工程师。

Lenny Rachitsky: A jagged intelligence, where it's almost like a staff engineer at understanding the code base.

Lenny Rachitsky：一种参差不齐的智能，它几乎就像一个理解代码库的工程师。

Usually, you have to ask an engineer that's been there a long time, " What does this do?

通常，你必须问一位在那里工作很长时间的工程师，“这个是做什么的？

Where is this thing?

这东西在哪里？

How does this work?" It feels like Devin's very good at that.

这是如何运作的？”感觉德文在这方面非常擅长。

Scott Wu: Yeah.

吴斯科特：是的。

Yeah.

是的。

Obviously, the retrieval and processing a lot of code and a lot of tokens at once is something that language models are really great at.

显然，同时检索和处理大量代码和大量标记是语言模型真正擅长的事情。

Basically being able to get those gains in the places that you need them is really great.

基本上能够在您需要的地方获得这些收益真的很棒。

Yeah.

是的。

Sweet, cool.

甜甜的，凉快的。

Lenny Rachitsky: All right.

莱尼·拉奇茨基：好的。

You got a couple more use?

你还有几个用吗？

Scott Wu: Yeah.

吴斯科特：是的。

One last thought I'll show is just...

我要展示的最后一个想法是......

We just rolled this out last week actually, but it's a full Devin automation setup with Linear.

事实上，我们上周刚刚推出了这个功能，但它是一个完整的 Devin 自动化设置与 Linear。

If you have tasks that you're doing on the DeepSeek repository, for example, and it's all set up, all you have to do is you just add the Devin label and Devin will come through and it'll give you this.

例如，如果您正在 DeepSeek 存储库上执行任务，并且所有任务都已设置完毕，您所要做的就是添加 Devin 标签，Devin 就会通过并为您提供此信息。

It's going to give you its thoughts on what the tasks looks like and you can take a look at each of the particular files that you see, or it'll point out snippets that it thinks are important.

它将向您提供有关任务的想法，您可以查看您看到的每个特定文件，或者它会指出它认为重要的片段。

From there, if you feel good about what was built or the conclusions that we came to, then you can just start off the Devin session that will go and actually do that work.

从那里开始，如果您对所构建的内容或我们得出的结论感到满意，那么您可以开始 Devin 会议，该会议将进行并实际完成这项工作。

Lenny Rachitsky: That is insane.

莱尼·拉奇茨基：这太疯狂了。

That sounds like such a simple idea, but essentially what you're saying is there are tasks in Linear that are fixes and features and now Devin just goes off and can just do them for you.

这听起来像是一个简单的想法，但本质上你所说的是 Linear 中有一些任务是修复和功能，现在 Devin 就可以为你完成这些任务。

Scott Wu: Yeah.

吴斯科特：是的。

Definitely it's a hands-on process.

毫无疑问，这是一个亲自动手的过程。

You certainly want to be involved when Devin is scoping out the task or giving you its thoughts.

当 Devin 确定任务范围或向您提供想法时，您当然希望参与其中。

The nice thing, too, by the way, is Devin will give you its confidence level.

顺便说一句，好的事情是 Devin 会给你信心水平。

Here's how likely I am to really understand this piece or that piece or whatever, but it helps make things a lot faster.

这是我真正理解这篇文章或那篇文章或其他内容的可能性，但它有助于使事情变得更快。

To your point, a lot of product managers, for example, obviously love to be able to use Devin in Linear to understand the code base better or things like that.

就您的观点而言，例如，许多产品经理显然喜欢能够使用 Linear 中的 Devin 来更好地理解代码库或类似的东西。

Claire Vo, for example, from LaunchDarkly is a big Devin user and she loves basically going and scoping out tasks or asking data questions or asking, "Hey, what's going on?

例如，LaunchDarkly 的 Claire Vo 是 Devin 的大用户，她基本上喜欢探索任务范围或提出数据问题或询问，“嘿，发生了什么事？

Or is this merged into production yet?

或者这已经合并到生产中了吗？

Or is this a feature flag right now?

或者这是现在的一个功能标志？

Or what percent of people are getting this or that feature?" It's a clean way basically to make that intelligence much more accessible.

或者有多少人获得这个或那个功能？”基本上，这是一种让情报更容易获取的干净方法。

Lenny Rachitsky: I love, just with the integration with Linear, that you can still keep it really simple.

Lenny Rachitsky：我喜欢与 Linear 的集成，你仍然可以保持它非常简单。

You add a little ticket like, "Hey, this link to this home page, do this," and Devin will be really good at understanding what you mean and then show you, "Here's what I'm thinking." Is this right?

您添加一个小票，例如“嘿，这个主页的链接，执行此操作”，德文将非常擅长理解您的意思，然后向您展示“这就是我的想法”。这是对的吗？

Scott Wu: Yeah.

吴斯科特：是的。

Yeah, cool.

是的，很酷。

Okay.

好的。

Yeah, Devin did finish working.

是的，德文确实完成了工作。

It seems like there's something going on with the CI and it's debugging that right now, but it went ahead and put up the initial first pass pull request and we can take a look.

CI 似乎发生了一些事情，现在正在调试，但它继续并提出了最初的第一遍拉取请求，我们可以看一下。

Scott Wu: This is the Devin website, obviously, in this custom deploy and we have Lenny's newsletter right here.

Scott Wu：显然，这是 Devin 网站，在这个自定义部署中，我们在这里有 Lenny 的时事通讯。

Lenny Rachitsky: Let's ship this to production.

Lenny Rachitsky：让我们将其投入生产。

We won't be so confused.

我们就不会这么迷茫了。

Lenny Rachitsky: That's amazing.

莱尼·拉奇茨基：太棒了。

Okay.

好的。

Show it again real quick.

很快再次显示它。

Just added it to the home page of Devin.

刚刚将其添加到 Devin 的主页。

Scott Wu: Yeah.

吴斯科特：是的。

Devin, obviously, has access to our Devin code base.

显然，Devin 可以访问我们的 Devin 代码库。

It does a lot here and so it's super familiar with all the pieces here.

它在这里做了很多事情，所以它对这里的所有部分都非常熟悉。

Scott Wu: Yeah, I like how that looks.

Scott Wu：是的，我喜欢它的样子。

We've got Devin search, we got Devin , and we've got Lenny's newsletter.

我们有德文搜索，我们有德文，我们有莱尼的时事通讯。

Lenny Rachitsky: .

莱尼·拉奇茨基：。

You link to my site.

你链接到我的网站。

We'll get some PageRank going.

我们将获得一些 PageRank。

Lenny Rachitsky: Okay.

莱尼·拉奇茨基：好的。

Is that a good example?

这是一个很好的例子吗？

Oh, there it is.

哦，就在那里。

What a beautiful website for my newsletter.

我的时事通讯网站真是一个漂亮的网站。

Is that just a good example of the kind of thing Devin is very good at like, "Here's a very specific thing to change on the website"?

这只是德文非常擅长的事情的一个很好的例子，比如“网站上有一个非常具体的事情需要改变”？

How does your people think about what Devin is very good at and maybe where it starts to fall apart?

你们的员工如何看待 Devin 擅长的领域以及它开始失败的领域？

Scott Wu: The way that we often describe it is, I think, Devin is best when it is working on tasks that are well-defined.

Scott Wu：我认为，我们经常描述它的方式是，当 Devin 处理明确定义的任务时，它是最好的。

One way to put it, you might...

一种说法是，你可能...

Scott Wu: On tasks that are well-defined.

Scott Wu：关于明确定义的任务。

One way to put it is, you want to be giving Devin tasks, not problems.

一种说法是，你想给德文分配任务，而不是问题。

And a lot of these things like what you just saw, which was kind of like a quick front-end feature request or a bug fix or adding testing and documentation or things like that.

很多事情就像你刚刚看到的那样，有点像快速前端功能请求或错误修复或添加测试和文档或类似的东西。

Scott Wu: One of the things that makes a loop really nice obviously is a quick way to iterate and test.

Scott Wu：使循环变得非常好的原因之一显然是一种快速的迭代和测试方法。

And so with something like this, obviously super easy for us, for example, to just go pull up the preview and see that the link worked.

因此，对于这样的事情，显然对我们来说非常容易，例如，只需打开预览并查看链接是否有效。

Obviously it would be easy for Devin to do as well.

显然，德文也很容易做到。

Devin will often go and log in to Devin and start a Devin session and make sure when it's working on our code base, which is kind of hilarious.

Devin 经常会登录 Devin 并启动 Devin 会话，并确保它何时在我们的代码库上运行，这有点搞笑。

But yeah, you generally want something that is kind of easy to verify and easy to test is the main thing.

但是，是的，您通常想要一些易于验证和易于测试的东西，这是主要的事情。

And you can work on bigger projects or bigger asks as well, obviously.

显然，您也可以从事更大的项目或更大的要求。

But in that case you should certainly expect to need to steer Devin more to make sure it's going the right direction.

但在这种情况下，你当然应该期望需要更多地引导 Devin 以确保它朝着正确的方向发展。

Lenny Rachitsky: It's interesting because that's very similar to the way people talk about synthetic data and reinforcement learning, creating data that's very easy.

Lenny Rachitsky：这很有趣，因为这与人们谈论合成数据和强化学习的方式非常相似，创建数据非常容易。

There's a very definitive answer, yes and no.

有一个非常明确的答案：是和否。

Lenny Rachitsky: Okay.

莱尼·拉奇茨基：好的。

Let me ask you this question.

让我问你这个问题。

What's something that you guys debated a lot as you were designing and building Devin?

在设计和构建 Devin 时，你们经常争论的是什么？

Scott Wu: I'll give a couple that comes to mind.

Scott Wu：我会举几个我想到的。

One I would say is a question of, I'll call it how opinionated we should be.

我想说的一个问题是，我们应该有多么固执己见。

We had the workflows that we used to Devin for, which was very much as you can see for basically integrating to our Slack and GitHub making pull requests for us in our repos, responding to issue reports or things like that.

我们拥有 Devin 所使用的工作流程，正如您所看到的，基本上集成到我们的 Slack 和 GitHub，在我们的存储库中为我们发出拉取请求，响应问题报告或类似的事情。

And naturally we've had certainly a lot of other different things that have come up that folks have tried.

当然，我们确实已经尝试过很多其他不同的事情。

I mean, we have folks who order their DoorDash with Devin, for example.

我的意思是，例如，我们有人向 Devin 订购 DoorDash。

Even we have folks, certainly a lot of people who are building cool websites from scratch or working on things like that.

即使我们也有人，当然有很多人正在从头开始构建很酷的网站或从事类似的事情。

Scott Wu: Yeah, I mean it is been an interesting trade off for us where I think the way that I would describe it is in our product, certainly the large bulk of the features that we build are for this kind of making pull requests and engineering teams use case.

Scott Wu：是的，我的意思是，这对我们来说是一个有趣的权衡，我认为我在我们的产品中描述它的方式，当然，我们构建的大部分功能都是针对这种提出拉取请求和工程团队用例的。

But I think basically our kind of general stance with all the others is obviously if folks want to use Devin for that, that's great and we want to just make sure that they're fully aware about the limitations and about where things can get caught up.

但我认为基本上我们对所有其他人的总体立场显然是，如果人们想使用 Devin 来实现这一点，那就太好了，我们只想确保他们充分意识到局限性以及事情可能会陷入困境的地方。

Scott Wu: It is funny with AI and especially because I would say one of, I would say the most common pieces of advice out there I would say is focus on a really niche cohort.

Scott Wu：人工智能很有趣，特别是因为我想说，最常见的建议之一是关注真正的利基群体。

Do things that don't scale, make one use case that's really great and then you grow from there.

做一些无法扩展的事情，创建一个非常好的用例，然后从那里开始成长。

And I think that's great advice across the board.

我认为这是一个全面的好建议。

But yeah, it's kind of interesting because I think with generative AI, you naturally see this where a lot of product experiences can turn out to be more general.

但是，是的，这很有趣，因为我认为通过生成人工智能，你自然会看到很多产品体验可以变得更加普遍。

And so it's an interesting trade-off for us.

所以这对我们来说是一个有趣的权衡。

This is something that we still always go back and forth on and how much do we want to do more to support all the other kind of use cases out there to handle other things that folks might want to do with Devin.

这是我们仍然经常反复讨论的事情，我们想要做更多的事情来支持所有其他类型的用例，以处理人们可能想要与 Devin 一起做的其他事情。

Scott Wu: I think another one that comes to mind is how much Devin should be, let's say a single comprehensive project experience versus a suite of tools.

Scott Wu：我认为另一个想到的问题是 Devin 应该是多少，比如说单一的综合项目经验与一套工具。

And as you can see here, we have Devin search, we have Devin Wiki, we have the linear ticket scooping, and certainly these tools interact with each other, but I think as time has gone on, we've seen it more and more as really building this suite of tools.

正如您在这里所看到的，我们有 Devin 搜索，我们有 Devin Wiki，我们有线性抢票，当然这些工具彼此交互，但我认为随着时间的推移，我们越来越多地看到它真正构建了这套工具。

And I think the core agent experience and the core kind of agent that will go off and build each of these build things for you, for example, is always of course going to be that's Devin, and that is the core piece.

我认为核心代理经验和核心代理类型将为您构建这些构建内容，例如，当然总是德文，这是核心部分。

I think that will always be what's really special about our work.

我认为这永远是我们工作的真正特别之处。

But I think that all of the other features out there, there is a complex suite of work that's required for real-world software sharing and engineering is just messy at the end of the day.

但我认为所有其他功能，现实世界的软件共享和工程都需要一套复杂的工作，最终会变得混乱。

Scott Wu: And so I think there are a lot of different flows and a lot of different use cases that makes sense.

Scott Wu：所以我认为有很多不同的流程和很多不同的用例都是有意义的。

And an obvious thing to point out is you could ask the same questions to Devin search as you could to Devin and Devin will go through and it'll do the same thing.

需要指出的一个明显的事情是，您可以向 Devin 搜索提出相同的问题，就像向 Devin 提出相同的问题一样，Devin 也会经历同样的事情。

It'll go through and look through the files and give you an answer and stuff.

它会仔细检查文件并给你答案之类的东西。

Scott Wu: But with that said, on the one hand, I think on the capability side, there's certainly a lot you can do to really optimize things for very specifically question answer about this repository and that made sense to really build into a specific kind feature.

Scott Wu：但话虽如此，一方面，我认为在功能方面，您当然可以做很多事情来真正优化有关此存储库的非常具体的问题答案，并且真正构建特定类型的功能是有意义的。

Scott Wu: And then on the other side, I would say we found that users actually really, really like having this access of control.

Scott Wu：另一方面，我想说我们发现用户实际上非常非常喜欢拥有这种控制权。

Sometimes you have a task that you're thinking about, but you actually don't want Devin to get started on the task just yet.

有时您正在考虑一项任务，但实际上您不希望 Devin 立即开始执行该任务。

You want to ask Devin and understand what parts of the code base might be relevant.

您想询问 Devin 并了解代码库的哪些部分可能相关。

And so you want to be very direct about saying, "This is just an ask and I just want to see the snippets of the code base that relate," or, "I just want to look at the Wiki and understand the existing representation."

因此，您需要非常直接地说，“这只是一个询问，我只想查看相关的代码库片段”，或者“我只想查看 Wiki 并了解现有的表示形式。”

Scott Wu: And so it's on both the capabilities and on the UX side, we've found that that's kind of what's naturally made sense over time.

Scott Wu：因此，无论是功能还是用户体验方面，我们发现随着时间的推移，这自然是有意义的。

Lenny Rachitsky: Well, let's talk about the landscape then of just other companies in the space, which is something a lot of people are always thinking about.

Lenny Rachitsky：好吧，让我们谈谈该领域其他公司的情况，这是很多人一直在思考的事情。

There's all these different approaches.

有所有这些不同的方法。

You guys are going full on AI engineer, there's obviously ID companies.

你们正在全力招募人工智能工程师，显然还有 ID 公司。

There's also just models being built that are really good at engineering.

还有一些正在构建的模型非常擅长工程。

Everyone's kind of starting to build agents now.

现在每个人都开始建立代理。

You guys are ahead on this in a lot of ways.

你们在很多方面都处于领先地位。

OpenAI just recently said they're going to build a software engineering agent.

OpenAI 最近表示他们将构建一个软件工程代理。

Anthropics got something there.

人类学在那里得到了一些东西。

Cursor and Windsurf have their only agents and Replit.

Cursor 和 Windsurf 有他们唯一的代理和 Replit。

Thoughts on just where you guys fit in the landscape and then how do you think you win long term.

思考一下你们在这个领域的定位，以及你们认为如何才能赢得长期胜利。

How do you think about that?

您对此有何看法？

Scott Wu: Yeah, and for what it's worth, I think all of these are incredible teams.

Scott Wu：是的，就其价值而言，我认为所有这些都是令人难以置信的团队。

I think really smart and really forward-thinking folks who are building a lot of great products out there.

我认为那些非常聪明、非常有远见的人正在构建许多伟大的产品。

And I think there's a lot to do honestly over the next few years with the advent of AGI or whatever you want to call it.

我认为，随着通用人工智能（AGI）或任何你想称之为的事物的出现，未来几年有很多事情要做。

I think one of the quotes that I love is in 2017, if you asked if we had AGI, the answer is no. And in 2025, if you ask if we have AGI, the answer is, "Well, you have to define AGI.

我认为我最喜欢的一句话是在 2017 年，如果你问我们是否有 AGI，答案是否定的。到 2025 年，如果你问我们是否有 AGI，答案是：“嗯，你必须定义 AGI。

And it depends on your subject." And I think it does get to the point of, I mean there is a lot of really amazing stuff happening.

这取决于你的主题。”我认为这确实达到了这样的目的，我的意思是有很多非常令人惊奇的事情正在发生。

I think that it's easy to underrate, I would say just how big of a shift it is that we're seeing where I think there are a lot of great products out there, for example, over the last 10 years, 20 years, 30 years, that have made each of these different niches of the life cycle of building a product a little bit easier, for example, right?

我认为这很容易被低估，我想说的是，这是一个多么大的转变，我们看到了我认为有很多伟大的产品，例如，在过去的 10 年、20 年、30 年里，这些产品使构建产品生命周期的每个不同领域变得更加容易，例如，对吗？

Scott Wu: There's great products out there for instant response, there's great products out there for logging, there's great products out there for billing, there's all of these different tools.

Scott Wu：有很棒的产品可以即时响应，有很棒的产品可以用于日志记录，有很棒的产品可以用于计费，还有所有这些不同的工具。

And the obvious thing is what we're seeing with AI is all of these spaces are going to be moving multiple times faster and it's going to be like an order of magnitude shift, if anything.

显而易见的是，我们在人工智能中看到的是，所有这些空间的移动速度都将加快数倍，如果有的话，这将是一个数量级的转变。

Scott Wu: And so I think from our perspective, we've obviously had a very specific lens that we've bet on this whole time, and that is autonomous coding agents.

Scott Wu：所以我认为从我们的角度来看，我们显然一直在押注一个非常具体的镜头，那就是自主编码代理。

And there's a lot of problems to solve there, to be honest, right?

老实说，那里有很多问题需要解决，对吧？

There's still a ton to do on the core capabilities, certainly, and we see cases all the time where it's like, wow, why did Devin make that decision?

当然，在核心功能上还有很多工作要做，我们总是看到这样的情况，哇，为什么德文做出了这个决定？

That seems, no human engineer would've ever done that.

看来，没有人类工程师会这么做。

There's all sorts of spots where, with the product interface, there's obviously a lot to think about.

在产品界面的各个方面，显然有很多需要考虑的地方。

Scott Wu: And I think it's by the way, not just a single thing that we're working towards, but something that will change with every edition of capabilities.

Scott Wu：顺便说一句，我认为这不仅仅是我们正在努力实现的单一目标，而且会随着每个版本的功能而改变。

I kind of think of it as there's 20 generations of agent product, agent coding experiences to come.

我认为，未来将有 20 代代理产品和代理编码体验。

I think the one that we'll get to over the course of several years is probably something where you don't even look at the code at all, right?

我认为我们将在几年内实现的目标可能是您根本不看代码的，对吧？

And you're actually just looking at your own product and you're just able to look and specify and say, "Hey, this button should be a little bit rounder.

实际上，您只是在查看自己的产品，您只需查看并指定并说：“嘿，这个按钮应该更圆一点。

Let's do that.

让我们这样做吧。

And by the way, let's add a new tab here and maybe we should save this information.

顺便说一句，让我们在此处添加一个新选项卡，也许我们应该保存此信息。

Let's start up a database table and let's index it on X, Y, and Z columns." And you're just basically working with your products in real time and having your agent build out those things for you.

让我们启动一个数据库表，并在 X、Y 和 Z 列上对其建立索引。”您基本上只是实时处理您的产品，并让您的代理为您构建这些东西。

Obviously there's going to be a lot of generations in between here and there, but I think the product experience itself is going to change every single time.

显然，这之间会有很多代，但我认为产品体验本身每次都会发生变化。

Scott Wu: And then obviously there, there's all of the practicality of just getting it out there in the world.

Scott Wu：很明显，将其发布到世界各地具有所有的实用性。

And so folks obviously need to learn how to use the new technology.

因此，人们显然需要学习如何使用新技术。

There's a lot to do to deploy into all of the messiness of real world software.

要部署到现实世界中所有混乱的软件中，需要做很多工作。

There's a lot of COBOL out there still.

仍然有很多 COBOL。

There's a lot of FORTRAN out there still.

仍然有很多 FORTRAN 语言。

There's lots of kind of abstractions and details that folks have done.

人们已经做了很多抽象和细节。

Scott Wu: And so I think from our perspective, we have since the beginning have been laser-focused on agentic coding, and that is the one thing that we've really believed in.

Scott Wu：所以我认为从我们的角度来看，我们从一开始就专注于代理编码，这是我们真正相信的一件事。

It's the one thing that we've designed for and that goes all the way to even the revenue model with ACUs and having the usage-based setup.

这是我们设计的目的之一，它一直延伸到带有 ACU 的收入模型并具有基于使用情况的设置。

It does into obviously all the product experiences of thinking how, okay, where do you want to talk to Devin?

显然，它涉及到思考如何，好吧，你想在哪里与 Devin 交谈的所有产品体验？

You want to be able to talk to Devin in Slack, you want to be able to spin this up from your issue.

您希望能够在 Slack 中与 Devin 交谈，您希望能够从您的问题中解决这个问题。

You want to be able to all of these things and then of course the capabilities.

您希望能够实现所有这些事情，当然还有功能。

Scott Wu: And so I don't think there's any one easy answer.

Scott Wu：所以我认为没有一个简单的答案。

I think it's obviously a combination of things, but this has been the space that we've lived in and spent all of our time in for the last year and a half, and it's going to be that way for the next five or 10 years too.

我认为这显然是多种因素的结合，但这是我们在过去一年半中居住并度过所有时间的空间，并且在未来五年或十年内也将如此。

Lenny Rachitsky: Along these lines, a big question everyone always has in AI's moats and defensibility, it's a question I've been asking every founder that comes on.

Lenny Rachitsky：沿着这个思路，每个人都对人工智能的护城河和防御能力抱有一个大问题，这是我一直在问每一位创始人的问题。

How do you just think about how to build a moat in this space when it's so much easier to build and so much is built on these models that are themselves advancing so quickly?

当这个领域的建造变得如此容易，而且这么多的东西是建立在这些本身发展如此迅速的模型之上时，你如何思考如何在这个空间中建造一条护城河？

Scott Wu: I'd give one slight tweak on that, which is I think it's often less about moats and more about stickiness.

Scott Wu：我想稍微调整一下，我认为这通常不是关于护城河，而是更多关于粘性。

And what I mean by that is moats are in some sense, typically what folks mean by moats is something that means that a competitor couldn't even enter the market.

我所说的护城河在某种意义上是指，通常人们所说的护城河意味着竞争对手甚至无法进入市场。

And I agree that at a high level, a lot of different folks at different layers of the AI spectrum, the foundation labs or the application layer or so on, I don't think there's any kind of hard barrier that would prevent others from entering.

我同意，在高水平上，人工智能领域的不同层、基础实验室或应用层等有很多不同的人，我认为不存在任何类型的硬障碍可以阻止其他人进入。

I think what does exist is stickiness, which I would kind of define as once you have a product experience that you really like, are you excited to keep using that experience or is there an effect where it is just as easy from now on to just switch on to a new one and learn a new one and so on.

我认为确实存在的是粘性，我将其定义为一旦您拥有了真正喜欢的产品体验，您是否会兴奋地继续使用该体验，或者是否有一种效果，从现在开始就可以轻松地切换到新产品并学习新产品等等。

Scott Wu: And I think from that perspective, I think there's a few things that are really great about coding agents in particular.

Scott Wu：我认为从这个角度来看，我认为编码代理有一些真正伟大的事情。

One I would say is there is a lot of just inherent stickiness and learning and buildup over time, which is that as you use Devin and as your whole team uses Devin, it's the same thing with an engineer.

我想说的是，随着时间的推移，有很多固有的粘性、学习和积累，也就是说，当你使用 Devin 和你的整个团队使用 Devin 时，这对于工程师来说是一样的。

If you're joining on day one versus you've been at the company for five years, you wrote half the code yourself, you've touched every file you've built every single piece, you know all the engineers.

如果你是第一天加入公司，而你已经在公司工作了五年，那么你自己编写了一半的代码，你接触过你构建的每一个部分的每个文件，你认识所有的工程师。

And so similarly, it's like Devin will really learn and build its representation of your code base and of your stack and of your process over time and will be able to do a lot more with that.

同样，随着时间的推移，Devin 将真正学习并构建代码库、堆栈和流程的表示，并且能够用它做更多的事情。

Scott Wu: And then the other piece of it, which I think is really exciting I'd say is there really is a lot to do of what I would call a multiplayer aspect of code, which if you think about it, is how a lot of things get done in the real world, certainly.

Scott Wu：然后，我认为真正令人兴奋的另一部分是，我所说的代码的多人游戏方面确实有很多工作要做，如果你仔细想想，这就是现实世界中很多事情的完成方式，当然。

And so it's one thing to have your own experience, which you use yourself as just an engineer, but for example, ourselves, we see this all the time where some engineers are working with Devin and teaching Devin things and as I mentioned, folks will have Devin on board their new engineers and kind of convey that knowledge to them.

因此，拥有自己的经验是一回事，你只是将自己作为一名工程师使用，但例如我们自己，我们一直看到一些工程师与 Devin 合作并向 Devin 教授东西，正如我所提到的，人们会让 Devin 加入他们的新工程师并向他们传达这些知识。

Scott Wu: Or similarly, it's like I'll start a session with Devin in Slack and I'll say, "Hey, it'd be cool if we could do this thing." And some other engineer will chime in and say, "Oh, by the way, the reason we did it initially was X and Y." And so Devin, just make sure when you do this change that you still support that workflow.

Scott Wu：或者类似地，就像我会在 Slack 中与 Devin 开始一个会议，我会说，“嘿，如果我们能做这件事，那就太酷了。”其他一些工程师会插话说：“哦，顺便说一句，我们最初这样做的原因是 X 和 Y。”因此，Devin，只需确保在进行此更改时您仍然支持该工作流程即可。

And Devin will say, "Okay, sounds great." Or Devin will make a PR, I'll be working with Devin, we'll make pull requests and GitHub and somebody else will be reviewing that PR or give some comments and Devin will work on that too.

德文会说：“好吧，听起来很棒。”或者 Devin 将制作 PR，我将与 Devin 合作，我们将提出拉取请求，GitHub 和其他人将审查该 PR 或给出一些评论，Devin 也会对此进行处理。

You'll be in linear.

你将处于线性状态。

Scott Wu: So all these kind of spaces, it really does just kind of set up for an experience basically where Devin can just grow in the value that it can provide for your whole work over time.

Scott Wu：所以所有这些类型的空间，它确实只是为一种体验而设置，基本上，德文可以随着时间的推移为你的整个工作提供价值增长。

And so I think from that perspective, if anything, we want there to be a lot of innovation and a lot of new products and so on.

所以我认为从这个角度来看，如果有的话，我们希望有很多创新和很多新产品等等。

I don't think that the goal is to try to lock other people out of building.

我不认为我们的目标是试图将其他人拒之门外。

There's a lot of stuff to build, and I think there's going to be a lot of different experiences.

有很多东西需要构建，我认为会有很多不同的体验。

I think from our perspective, what we think about is more like how can we make Devin more and more and more useful as you're using it more?

我认为从我们的角度来看，我们所考虑的更像是如何让 Devin 随着你使用它的次数越来越多而变得越来越有用？

Lenny Rachitsky: It's very similar.

Lenny Rachitsky：非常相似。

We had Michael from Cursor, the CF Cursor on the podcast, and he had a similar point of just he thinks moats are just kind of like consumer, like Google is the way, he thinks it's like Google where people can easily switch.

我们有来自 Cursor 的 Michael，播客上的 CF Cursor，他也有类似的观点，他认为护城河有点像消费者，就像 Google 一样，他认为这就像人们可以轻松切换的 Google。

You just have to stay the best, and that's the answer.

你只需要保持最好的状态，这就是答案。

And it feels like you're adding to that of just like, but also if you can create some stickiness where it is very hard to leave because it's so good at what it's doing and it's built knowledge and integrated to your workflows and builds on that on that stickiness.

感觉就像你在添加类似的东西，但如果你能创造一些很难离开的粘性，因为它非常擅长它正在做的事情，它建立了知识并集成到你的工作流程中，并建立在这种粘性的基础上。

Scott Wu: Yeah, and I think one of the things that's nice about our space too is, software engineering for better for worse has a very clear tie to value.

Scott Wu：是的，我认为我们这个领域的优点之一是，软件工程无论好坏都与价值有着非常明显的联系。

And what it means is, I guess one way to put it is there is always kind of a clear next level, at least for the next while.

这意味着，我想一种说法是总会有一个明确的下一个水平，至少在接下来的一段时间内是这样。

I think there could be some point where you're just like, "All right, just build the entirety of YouTube for me." And Devin just does the whole, it's like there's probably been a hundred million hours of human engineering time building YouTube, building the algorithm, building all the infrastructure, all of the, everything, every little detail.

我认为在某些时候您可能会想，“好吧，为我构建整个 YouTube 吧。” Devin 只负责整体工作，就好像可能花费了一亿小时的人类工程时间来构建 YouTube、构建算法、构建所有基础设施，所有的一切、每一个小细节。

And maybe there's some time where Devin just does that out of the box.

也许有时候德文会开箱即用。

That's obviously going to be a long time from now.

显然这将是很长一段时间之后的事了。

Scott Wu: I think on the interim, on every level in between, obviously it makes a difference the quality of software engineering.

Scott Wu：我认为在过渡期间，在其间的每个层面上，显然它都会对软件工程的质量产生影响。

And I think one of the cool things with developers obviously is developers are really willing to learn new experiences and to put in effort if it means that they're able to have a higher and higher quality experience.

我认为对于开发人员来说，一件很酷的事情显然是开发人员真的愿意学习新的体验并付出努力，如果这意味着他们能够拥有越来越高质量的体验。

Lenny Rachitsky: Awesome.

莱尼·拉奇茨基：太棒了。

I'm going to spend a little time on the tech that enables Devin.

我将花一些时间讨论支持 Devin 的技术。

Without divulging trade secrets, just what allowed you to make Devin so good?

在不泄露商业秘密的情况下，到底是什么让你们让 Devin 如此出色？

Was there an unlock with a certain model?

是否有特定型号的解锁？

Some folks have shared three points on a 3.5 was a huge unlock for a lot of their products.

有些人分享了三点，认为 3.5 对他们的许多产品来说是一个巨大的解锁。

Just what's kind of the key to the way you've architected or built Devin that makes it work so well?

您设计或构建 Devin 的方式的关键是什么，使其运行如此良好？

Scott Wu: We obviously, we've been betting on agents for a long time.

Scott Wu：显然，我们长期以来一直在经纪人上下注。

I think that agents were doable and workable a lot earlier than most folks might've thought.

我认为代理的可行性和可操作性比大多数人想象的要早得多。

But certainly I think as the community has really rallied around it, I mean you see the impacts of that in the pre-training, you see the impacts of that in a lot of the work that's done with these models.

但我当然认为，随着社区真正团结起来，我的意思是你会在预训练中看到它的影响，你会在使用这些模型完成的许多工作中看到它的影响。

I actually don't think there's been any, from our perspective, I don't think there's been any single step function based model shift or anything that has been kind of like a night and day difference in Devin, but I certainly think that the curve of every point on the chart, I mean there's a new model that comes out every week now has obviously made a big difference in terms of what we've been able to do.

实际上，从我们的角度来看，我不认为有任何基于单步函数的模型转换或任何类似于德文的昼夜差异的东西，但我当然认为图表上每个点的曲线，我的意思是现在每周都会出现一个新模型，显然对我们能够做的事情产生了很大的影响。

And then obviously on top of that, we work with the research teams at all these foundation labs to do a lot of our work on top.

显然最重要的是，我们与所有这些基础实验室的研究团队合作，在上面做了很多工作。

Scott Wu: And so I think that my hot take here at which I would give is, I think in terms of base intelligence, we're honestly basically already there.

Scott Wu：所以我认为我在这里要给出的最热门观点是，我认为就基础情报而言，老实说我们基本上已经在那里了。

And I think a lot of what we see actually and what we spend our time on is less so, obviously, we don't our own models or things like that.

我认为我们实际看到的很多东西以及我们花时间做的事情并不是那么明显，显然我们没有自己的模型或类似的东西。

It's less so increasing the base IQ of a model, for example, and more about teaching it all of the idiosyncrasies of real-world engineering and thinking about here's how you use Datadog and do this, and here's how you might diagnose this error and here are the different things that you could run into and here's how you handle each of those.

例如，提高模型的基本智商并不是那么重要，而是更多地教授它现实世界工程的所有特性，并思考如何使用 Datadog 并执行此操作，如何诊断此错误，以下是您可能遇到的不同情况，以及如何处理每个问题。

And when you're ready, here's how you make GitHub PR.

当您准备好后，请按照以下步骤进行 GitHub PR。

Scott Wu: And this is true in engineering.

Scott Wu：这在工程领域也是如此。

It's true in every other space as well.

在其他所有空间中也是如此。

I mean, there's so much detail and idiosyncrasy to the work that we all do obviously day to day.

我的意思是，我们每天所做的工作显然有很多细节和特质。

And a lot of it is kind of like teaching the model to mirror the complexity of the real world, I would say, rather than getting it to some higher fundamental level of problem solving, which I think the foundation labs are doing a really great job.

我想说，其中很多都有点像教授模型来反映现实世界的复杂性，而不是让它达到解决问题的更高的基本水平，我认为基础实验室在这方面做得非常出色。

Lenny Rachitsky: There's something you shared when we were chatting before we started recording around the growth of previous transformative technologies were very hardware oriented and there was a limiting factor to their growth and AI is not that.

Lenny Rachitsky：在我们开始记录以前的变革性技术的发展之前，您在聊天时分享了一些内容，这些技术的发展非常面向硬件，它们的发展存在限制因素，而人工智能则不然。

Can you just share that insight?

您能分享一下您的见解吗？

Scott Wu: For a number of reasons?

Scott Wu：有很多原因吗？

I think AI is going to be the biggest technology shift of our lives, but I think one thing, which is what we were just talking about before this, which is most of the big tech revolutions that we've had over the last 50 years, I mean, I'm thinking about personal computer and the internet and the mobile phone and stuff, they all had this big hardware component that was a big part of the distribution.

我认为人工智能将成为我们生活中最大的技术转变，但我认为有一件事，这就是我们之前谈论的，这是我们在过去 50 年中经历的大部分重大技术革命，我的意思是，我正在考虑个人电脑、互联网和移动电话等东西，它们都有这个大型硬件组件，这是发行版的重要组成部分。

And so you had the internet, and initially it was just these universities that were talking with one another, but obviously over time we got the whole world plugged into the internet and it took years and years and years.

所以你有了互联网，最初只有这些大学在相互交谈，但显然随着时间的推移，我们让整个世界都接入了互联网，这花了很多年。

Same thing was true with mobile phones.

手机也是如此。

Same thing was true with PC.

PC 也是如此。

Scott Wu: And the thing that's interesting about that in particular, which is I would say we're already seeing the effects of that, is in these hardware distribution machines, obviously there's a lot that depends on real time.

Scott Wu：特别有趣的是，我想说我们已经看到了它的影响，在这些硬件分发机器中，显然有很多东西依赖于实时。

And so folks who were building for those industries saw their market grow and grow and grow basically steadily year over year as the number of people with mobile phones increased, as the number of people connected to the internet increased.

因此，随着拥有移动电话的人数的增加以及连接到互联网的人数的增加，为这些行业建设的人们看到他们的市场不断增长，并且基本上逐年稳定增长。

And many of those businesses, it's still crazy to think, but many of those businesses got started right in the beginning.

许多这样的业务，现在想想仍然很疯狂，但其中许多业务是从一开始就开始的。

I mean, Apple and Microsoft were started right around the same time.

我的意思是，苹果和微软大约在同一时间成立。

And same is true for a lot of the great internet businesses or wherever, but certainly it was something that touched whole world with time or a large fraction of the whole world.

对于许多伟大的互联网企业或任何地方来说都是如此，但毫无疑问，它随着时间的推移影响了整个世界或整个世界的很大一部分。

And it had a really massive impact, but it took place over several years because of the time that it took.

它确实产生了巨大的影响，但由于所花费的时间，它花了几年的时间才发生。

And I think one of the things which is already I'd say different in AI, is just how explosive the technology can be.

我认为在人工智能领域我已经说过的不同之处之一就是这项技术的爆炸性。

Once AI could, and I think we're firmly past the inflection point in AI code where it's, as an engineer, if you're not using AI at all to write code, I mean you're falling behind honestly.

一旦人工智能能够实现，而且我认为我们已经坚定地越过了人工智能代码的拐点，作为一名工程师，如果你根本不使用人工智能来编写代码，我的意思是你老实说已经落后了。

And it is a technology that everyone should have and should be using, and there's no kind of weight on hardware distribution that is causing that.

这是一项每个人都应该拥有并且应该使用的技术，并且硬件分配上没有任何压力造成这种情况。

It means that the space is just growing so exponentially, basically.

这意味着空间基本上呈指数级增长。

Lenny Rachitsky: Michael Ballin has this interesting point that cliches are cliches because they're so true and that's why they're cliches.

莱尼·拉奇茨基：迈克尔·巴林有一个有趣的观点，即陈词滥调之所以是陈词滥调，是因为它们是如此真实，这就是为什么它们是陈词滥调。

I heard that a million times and I think it's like people hear this like, "Yeah, yeah, I know," but it's actually insane what is happening.

我听过一百万次，我想人们听到这句话就像是，“是的，是的，我知道”，但实际上正在发生的事情是疯狂的。

That's why you're here to help us through this transition.

这就是为什么您来这里帮助我们完成这一转变。

Scott Wu: Yeah, no, I mean, it is a fun time and I think there will be real investment and real work that it takes.

Scott Wu：是的，不，我的意思是，这是一段有趣的时光，我认为这需要真正的投资和真正的工作。

But I think from the perspective of us as engineers, for example, I think it just means it's so important to stay in the loop with everything that's happening.

但我认为，从我们作为工程师的角度来看，我认为这只是意味着了解正在发生的一切是非常重要的。

And as we're seeing it's not only because of your learning and your ability to work these technologies, but it's also about basically teaching the AI what there is to know about your code base in order to make it really effective at building with you and doing more of the things that you would want it to do.

正如我们所看到的，这不仅是因为你的学习和使用这些技术的能力，而且还在于从根本上教会人工智能关于你的代码库的知识，以便让它真正有效地与你一起构建并做更多你希望它做的事情。

Lenny Rachitsky: So along those lines, for people listening that they're like, "Hey, we should be using Devin at our company," what are things you've found to be helpful in helping an engineer at a company get adoption and be able to use Devin either culturally or logistically?

Lenny Rachitsky：因此，对于那些听他们说“嘿，我们应该在我们公司使用 Devin”的人来说，您发现哪些东西有助于帮助公司工程师获得采用并能够在文化或后勤方面使用 Devin？

Scott Wu: So a pattern which we often see with folks is there will be a few folks at the team who are really excited and want to try out the new thing, and they want to put in the investment and are really excited to get it going.

Scott Wu：所以我们经常在人们身上看到的一种模式是，团队中会有一些人非常兴奋，想要尝试新事物，他们想要投入投资，并且非常兴奋能够让它继续下去。

And they'll go through all the setup.

他们将完成所有设置。

They'll give Devin the repos, they'll teach Devin how to run the lint and the CI and all of those details.

他们会给 Devin 存储库，他们会教 Devin 如何运行 lint 和 CI 以及所有这些细节。

And they'll start off by giving it those initial tasks and help Devin build a foothold basically.

他们将首先给它那些初始任务，并帮助德文基本上建立一个立足点。

And as time goes on, eventually folks will see, "Wow, Devin's writing all these PRs, Devin's doing this .

随着时间的推移，最终人们会看到，“哇，德文正在写所有这些 PR，德文正在做这个……。

Lenny Rachitsky: Who's this Devin person that just joined the company's just knocking out PRs?

Lenny Rachitsky：这个刚加入公司、刚刚搞公关的 Devin 人是谁？

Scott Wu: Yeah.

吴斯科特：是的。

And they'll see that, and then naturally they'll get on and they'll get an account.

他们会看到这一点，然后很自然地他们就会继续下去并获得一个帐户。

And one of the cool things of course is by the time they join, Devin already knows a good amount of detail about the repositories that they're already working in and they're working with that.

当然，最酷的事情之一是，当他们加入时，Devin 已经了解了有关他们已经在其中工作的存储库的大量详细信息，并且他们正在使用这些存储库。

And so one of the really cool things which we often see is that the early adopters themselves can really pave the way I think, for everyone else on the team.

因此，我们经常看到的一件非常酷的事情是，早期采用者本身可以为团队中的其他人铺平我认为的道路。

Scott Wu: But yeah, I think the main thing I would just kind of call out is it really does take, it's a very different product experience, and I think for what it's worth, I think there's still a lot more that we can do to make it as intuitive and as clear as possible to folks like how to use Devin and what the right steps are and how to really maximize value out of Devin.

Scott Wu：但是，是的，我想我要强调的是，它确实需要，这是一种非常不同的产品体验，而且我认为，就其价值而言，我认为我们还可以做更多的事情，让它对人们来说尽可能直观和清晰，比如如何使用 Devin，正确的步骤是什么，以及如何真正最大化 Devin 的价值。

But I think that, yeah, it's the kind of thing where if you put in the investment and understand exactly what it takes to get Devin to be successful, we've found ourselves that as time has gone on, we just use Devin more and more and more with every next update.

但我认为，是的，如果你投入投资并准确了解如何让德文取得成功，我们就会发现，随着时间的推移，我们在每次下一次更新中都会越来越多地使用德文。

Lenny Rachitsky: So let me follow that thread.

Lenny Rachitsky：所以让我继续关注这个话题。

There is a question I ask every AI app building founder, which is, if you could sit next to every new user of Devin and whisper something in their ear to help them be successful with Devin, one or two tips, what would those tips be?

我问每个人工智能应用程序开发创始人一个问题，那就是，如果你可以坐在 Devin 的每个新用户旁边，在他们耳边低语一些，帮助他们通过 Devin 取得成功，一两个技巧，这些技巧会是什么？

Scott Wu: I think the biggest thing I would say is it really is just treat Devin like your new junior engineer.

Scott Wu：我认为我想说的最重要的事情是，就像对待新的初级工程师一样对待 Devin。

And I think that's the biggest thing.

我认为这是最重要的事情。

I think folks come in and they see the blank page and they think of all sorts of various things that they want to try out.

我认为人们进来后会看到空白页，他们会想到他们想要尝试的各种不同的东西。

They think of lots, where I think typically the flow that we see that works best is obviously you can try demos and you can do things, but a lot of it is just like, "Yeah, let's figure out what tickets we want to get done today or this week and let's have Devin get started on those and let's start with the easier ones and then work with Devin and understand what things Devin needs to get set up to be able to test its own code and do this well.

他们想到了很多，我认为通常我们看到的最有效的流程显然是你可以尝试演示，你可以做一些事情，但很多都是这样的，“是的，让我们弄清楚今天或本周我们想要完成哪些任务，让 Devin 开始处理这些，让我们从更简单的开始，然后与 Devin 一起工作，了解 Devin 需要设置哪些东西才能测试自己的代码并做好这项工作。

And then let's scale up over time." And then obviously as you work with your engineer, you understand better how to communicate with them or what are the right tasks or projects to bring them in on.

然后随着时间的推移，让我们扩大规模。”显然，当您与工程师一起工作时，您会更好地了解如何与他们沟通，或者让他们参与哪些正确的任务或项目。

But I think that really is the one-liner for us.

但我认为这对我们来说确实是一句俏话。

Lenny Rachitsky: Okay.

莱尼·拉奇茨基：好的。

There's a question I'd be meeting task.

有一个问题我要完成任务。

I just want to get back to this because it's something I think a lot about with Devins.

我只想回到这个话题，因为这是我和德文斯经常思考的事情。

Everyone's going to have five Devins, let's say 10 Devins.

每个人都会有 5 个德文，假设有 10 个德文。

Everyone's kind of turning into basically an engine manager with a bunch of junior engineers, which isn't necessarily the best job in the world because it's just a bunch of, at least you don't have to do performance reviews and one-on-ones, but it's sitting around, checking a lot of PRs all day.

每个人基本上都变成了一个引擎经理，带着一群初级工程师，这不一定是世界上最好的工作，因为它只是一堆，至少你不需要做绩效评估和一对一，但它整天坐着，检查很多 PR。

There's a sense of you become an architect, which is kind of what every engineer wants to become eventually, right?

有一种你成为一名建筑师的感觉，这是每个工程师最终都想成为的，对吗？

They're all, "I just want to think about the architecture.

他们都是，“我只想考虑一下架构。

I don't want to code all these stupid, fix bugs."

我不想编写所有这些愚蠢的代码，修复错误。”

Lenny Rachitsky: So I get that there's a good side to that, but just I imagine you're thinking a lot about this, just like how do you make life pleasant and fun and enjoyable as basically an engine manager of say, 500 Devins in the future?

Lenny Rachitsky：所以我知道这有好的一面，但我想你对此思考了很多，就像你如何让生活变得愉快、有趣和愉快，就像未来的 500 Devins 引擎经理一样？

Scott Wu: Yeah, I can just imagine the performance, "Devin, you've done a really great job on your task, but I really would like you to be more proactive in the team meetings." So what I'd say-

Scott Wu：是的，我可以想象这样的表现，“Devin，你的任务做得非常出色，但我真的希望你在团队会议上更加积极主动。”所以我想说的是——

Scott Wu: It's funny actually because something that in terms of the wording that we thought a lot about as well is just, we've used the term manager of Devins in the past, which of course I think is a big part of it.

Scott Wu：这实际上很有趣，因为我们在措辞方面也考虑了很多，我们过去使用过德文斯经理这个词，当然我认为这是其中很重要的一部分。

But I think that the only thing I would point out here is I think that the bricklayer versus architect is closer to the experience than being a manager.

但我认为我在这里唯一要指出的是，我认为瓦工与建筑师比成为经理更接近体验。

Because I think a lot of the difficulty of management or the reason that people shy away from it is more because of a lot of the various.

因为我认为很多管理上的困难或者说人们回避它的原因更多的是因为很多方面。

Scott Wu: ...

吴斯科特：...

because of a lot of the various, let's say...

因为有很多不同的情况，比如说......

There's all of the context, and the ownership, and the responsibility and stuff, and then there's also all of the emotional aspects of it.

有所有的背景、所有权、责任等等，还有所有的情感方面。

Where, I think working with Devin is a little more like just being, more as having an interface to hand off tasks and build tasks.

我认为与 Devin 的合作更像是存在，更像是拥有一个交接任务和构建任务的界面。

And so, the parallel that I would draw is when we invented Python, obviously...

因此，我想与我们发明 Python 时进行比较，显然......

It's like, in many ways the description and the outlining of tasks, obviously it was a different paradigm, but I think certainly it was nowhere near what folks typically think of as management bureaucracy today.

从很多方面来说，任务的描述和概述显然是一种不同的范式，但我认为它肯定与当今人们通常认为的管理官僚机构相去甚远。

Scott Wu: And I think that with Devin, a lot of it is just like, it's more like finding the right levels of abstraction that you could work with Devin on, and just finding the workflows that work really well, and the obvious thing to say here is, it's like you can always have Devin take a first pass at things.

Scott Wu：我认为对于 Devin，很多事情就像是找到你可以与 Devin 一起工作的正确的抽象级别，并且只是找到运行良好的工作流程，这里明显要说的是，就像你总是可以让 Devin 在事情上先行一步。

And so you have Devin take the first pass, if it's great, you merge it right away, if it needs some touch up, you can obviously give that feedback, for example, but a lot of it is like, it's more about, basically making Devin part of your flow than it losing control, which I think is the main thing that folks are scared of with management.

所以你让 Devin 进行第一遍，如果它很棒，你会立即合并它，如果它需要一些修改，你显然可以提供反馈，例如，但很多都是这样，它更多的是，基本上让 Devin 成为你流程的一部分，而不是失去控制，我认为这是人们对管理感到害怕的主要事情。

Lenny Rachitsky: Are you thinking about a manager Devin?

Lenny Rachitsky：你在考虑经理德文吗？

Like a Devin that manages other Devins?

就像一个德文管理其他德文一样？

Scott Wu: Yeah, yeah.

吴斯科特：是的，是的。

So, for what it's worth, Devin can start other Devins through the API, right?

因此，无论如何，Devin 可以通过 API 启动其他 Devin，对吗？

And so, we've seen this happen quite a bit of times, where naturally if you have some big tasks that you want to do, Devin will do this all the time, it'll chunk up and then into smaller Devins.

因此，我们已经看到这种情况发生了很多次，当然，如果你有一些想要完成的大任务，Devin 会一直这样做，它会分成大块，然后分成更小的 Devins。

And so, it's the kind of thing that you need to give Devin the credentials to be able to do that, it's not currently something that is default enabled, but I can certainly imagine as time goes on that there's more and more of that

因此，您需要向 Devin 提供凭证才能做到这一点，目前这不是默认启用的东西，但我当然可以想象随着时间的推移，这种情况会越来越多

Scott Wu: Yeah, yeah.

吴斯科特：是的，是的。

I think the thing that's kind interesting too is, with humans, the way I almost say it, in technical terms, it's like there's this coupling of a context and a thread, and what I mean by that is basically, each human can only operate a single threaded on the work that they do, and they have their set of context, and then there's other humans obviously who can do other stuff at the same time, but they have their own context.

我认为，对于人类来说，同样有趣的事情是，用技术术语来说，就像我所说的那样，上下文和线程之间存在这种耦合，我的意思基本上是，每个人只能在他们所做的工作上操作单个线程，并且他们有自己的一组上下文，然后其他人显然可以同时做其他事情，但他们有自己的上下文。

With agents, one of the cool things is you can have an agent that's doing multiple lines of exploration at once, but is sharing all of the context of everything that they find, and so I think that this is very early, and I think we'll see this, but folks obviously love to talk about basically systems of agents communicating with one another, and I think that there will be a lot of new paradigms to build for, once we get there.

对于代理来说，最酷的事情之一是你可以让一个代理同时进行多条探索，但会共享他们发现的所有内容的所有上下文，所以我认为这还为时过早，我想我们会看到这一点，但人们显然喜欢谈论代理之间相互通信的基本系统，我认为一旦我们到达那里，将会有很多新的范例可供构建。

Lenny Rachitsky: And it's so interesting what you said about the decision between having one Devin and only one Devin do all the things, and you just tell them things and they fire off jobs versus you have five Devins, and they're each doing individual things, it's such an interesting decision to make.

Lenny Rachitsky：你所说的关于让一个德文和只有一个德文做所有事情之间的决定非常有趣，你只要告诉他们事情，他们就会解雇工作，而你有五个德文，他们每个人都做自己的事情，这是一个非常有趣的决定。

Lenny Rachitsky: Okay, two more questions.

Lenny Rachitsky：好的，还有两个问题。

Maybe the most counterintuitive thing you've learned so far building Devin, that maybe goes against startup wisdom, common startup wisdom?

也许是迄今为止你在构建 Devin 时学到的最违反直觉的事情，这可能违背了创业智慧，常见的创业智慧？

Scott Wu: Something I've thought about a lot lately as we've built this is, this is not my first company, actually for a lot of us, it's not our first company, I think of our 26 or 27 people total on the team, I think 18 of us have started our own company before this.

Scott Wu：在我们建立这个公司的过程中，我最近思考了很多，这不是我的第一家公司，实际上对于我们很多人来说，这不是我们的第一家公司，我认为我们团队总共有 26 或 27 人，我想我们中有 18 个人在此之前已经创办了自己的公司。

And yeah, one of the things I think about is, there's actually your point about cliches I think really spoke to me as well, which is, there's the really common things which you hear all the time in startups, where you're like, you got to move fast, or you got to hire great people, it's like, okay, well, obviously you do, I wasn't planning on not hiring great people, I wasn't planning on going slow.

是的，我想到的一件事是，实际上你关于陈词滥调的观点也对我很有意义，那就是，你在初创公司中一直听到一些非常常见的事情，你就像，你必须快速行动，或者你必须雇用优秀的人才，就像，好吧，好吧，显然你这样做，我不打算不雇用优秀的人才，我不打算放慢脚步。

And similarly it's like, yeah, you really got to build something that people want.

同样，是的，你真的必须构建人们想要的东西。

And there's these three to five things which are always repeated, and they're always the common wisdom in startups.

这三到五件事总是被重复，而且它们始终是初创公司的共同智慧。

Scott Wu: And I definitely had this idea as a founder, when I was starting initially, that, all right, so those are the three to five basic things, but as you get really deep into it, you spend a lot of years into it, you learn all of the thousands of other things that you have to learn to build a company.

Scott Wu：作为创始人，我确实有这个想法，当我刚开始的时候，好吧，这些是三到五个基本的东西，但是当你真正深入它时，你花了很多年的时间，你学到了建立一家公司必须学习的所有其他东西。

And I think to some extent that's, of course, true, and there's lots of little details that you'll get into with all these different things, including team building, and product, and strategy, and engineering decision-making, and fundraising, and sales, and every other component.

我认为在某种程度上，这当然是正确的，而且你会涉及到所有这些不同的事情的很多小细节，包括团队建设、产品、战略、工程决策、筹款、销售以及所有其他组成部分。

But I also realized that as time has gone on, more and more, I felt like building companies well sometimes just comes down to doing those three to five things just even more than you could possibly expect.

但我也意识到，随着时间的推移，我越来越觉得要建立好公司，有时只需要做三到五件事，甚至比你想象的还要多。

Scott Wu: And so, with us, it's like...

Scott Wu：所以，对于我们来说，就像......

And everyone says we go fast, but it's like, yeah, we had a hackathon in November, we had another hackathon in December, we started the company officially in January, we got the prototypes out to initial users in February, we did a launch in March, we got our first customers in April...

每个人都说我们进展很快，但就像是的，我们在 11 月举办了一场黑客马拉松，在 12 月举办了另一场黑客马拉松，我们在一月份正式成立了公司，我们在二月份向初始用户提供了原型，我们在三月份进行了发布，我们在四月份获得了第一批客户......

It's just like, basically truly pushing the pace in every spot where we possibly could has really made a difference for us.

就像，基本上真正在我们可能做到的每个地方加快步伐，确实对我们产生了影响。

And similarly, it's like, yeah, everyone always says you should hire great people, but I think that the truth within that truth is basically you should fight to all ends basically, to get the folks that you really want to bring in.

同样，是的，每个人总是说你应该雇用优秀的人才，但我认为这个真理中的真相基本上是你应该全力以赴，以获得你真正想引进的人。

And one of my favorite stories to share is we had a candidate who came and interviewed, he was a junior at MIT, so he was very, very young, and we gave him an interview, and he did way better than almost any of the full-time candidates that we had ever talked to.

我最喜欢分享的故事之一是，我们有一位候选人来面试，他是麻省理工学院的大三学生，所以他非常非常年轻，我们给了他面试机会，他的表现比我们曾经交谈过的几乎所有全职候选人都要好。

And so, we said, hey, what do you think about taking some time off of school and working with us and building out Devin?

所以，我们说，嘿，你觉得休学一段时间和我们一起工作并打造德文怎么样？

Scott Wu: We really think you're just going to be able to come in and just have a ton of impact already from day one.

Scott Wu：我们真的认为你从第一天起就能够进来并产生巨大的影响。

And he thought about it for a while, and he came back and he said, you know what?

他想了一会儿，然后回来说，你知道吗？

I'm down, I want to do it, but my parents really want me to graduate from school, and I'm just not sure there's a way to make it work.

我很沮丧，我想做，但我的父母真的希望我从学校毕业，而我只是不确定是否有办法让它发挥作用。

And so, we talked him more and just understood the situation, and then we flew to North Carolina, went straight from the airport to his parents' house, had dinner with him and his parents, we talked a lot, a really nice Gujarati family, we gave them some gifts and just talked to them about it, and try to understand what would it take, and what would we need to make work.

因此，我们与他进行了更多交谈，了解了情况，然后我们飞往北卡罗来纳州，从机场直接前往他父母家，与他和他的父母共进晚餐，我们聊了很多，这是一个非常好的古吉拉特家庭，我们给了他们一些礼物，只是和他们谈论了这件事，并尝试了解这需要什么，以及我们需要什么才能开展工作。

And they just said, it sounds like a great opportunity, but we really want our son to be able to graduate.

他们只是说，这听起来是一个很好的机会，但我们真的希望我们的儿子能够毕业。

Scott Wu: And we talked that through and we figured out a setup basically, where he could work for us essentially full time, but then come in for his required classes, and do what he needed to do to get the diploma basically, but no more than that.

Scott Wu：我们讨论过这个问题，我们基本上找到了一个设置，他可以基本上全职为我们工作，然后参加他所需的课程，并做他需要做的事情来获得文凭，但仅此而已。

And we talked that through, and then we got to a point where everyone was happy with that, and then went straight back to the airport and flew right back, basically.

我们讨论了这个问题，然后我们达到了每个人都对此感到满意的地步，然后直接返回机场，然后基本上就飞回来了。

And that was the first and only time that I've ever been in North Carolina, and it was just a great trip.

那是我第一次也是唯一一次来到北卡罗来纳州，这是一次很棒的旅行。

And it is the kind of thing where it's like, hiring great people is one thing, but truly just never giving up, and really giving it everything that you can to make it work for people who really makes sense to be on the team.

就像这样的事情，雇用优秀的人才是一回事，但真正要永不放弃，并真正尽你所能，让它为真正有意义的团队成员服务。

And he's been with us on the team for over a year now, and he's been an incredible, incredible engineer, and we wouldn't be here without him.

他已经在我们团队中工作一年多了，他是一位令人难以置信的工程师，没有他我们就不会在这里。

Scott Wu: And similarly, we had someone else who was again, really, really talented candidate, did amazingly well, very young, and had a lot of great offers at a lot of other companies, and we were talking to them about, he wanted to start his own company someday as well, and we were talking to him about certainly a lot of the obvious things, which are having him meet our investors, or get to do work with customers, or see a lot of these other components so that when the time came that he would have all the experience he needed to start his own company.

斯科特·吴：同样，我们还有一个非常非常有才华的候选人，他表现得非常好，非常年轻，并且在很多其他公司收到了很多很棒的邀请，我们正在与他们谈论，他也想有一天创办自己的公司，我们正在与他谈论很多显而易见的事情，这些事情让他会见我们的投资者，或者与客户合作，或者看到很多其他组件，以便当时机成熟时，他将拥有他所需的所有经验。创办自己的公司。

But one of the other things that was big is he was talking with a lot of great companies already, he didn't want to burn any bridges, and so we actually worked with him and basically hand wrote all of his rejection responses to each of the other companies, and worked with him on it to say, here's how you should say it in a way that's going to come off as that you really did appreciate the time with them, and that obviously you want to remain close with them and stay in touch.

但另一件重要的事情是，他已经在与很多伟大的公司进行谈判，他不想毁掉任何桥梁，所以我们实际上与他合作，基本上手写了他对其他每家公司的所有拒绝回复，并与他一起工作，并与他一起说，这就是你应该如何表达的方式，因为你真的很感激与他们共度时光，显然你想与他们保持密切联系并保持联系。

And it was the kind of thing obviously where it's like, look, obviously our job is to make sure that he's happy enough that he doesn't want to leave at any time in the near future, but I think it's the kind of thing where the way that you put together a really, really great team is by fighting for what's right for them too.

显然，这就是这样的事情，看，显然我们的工作是确保他足够开心，以至于他在不久的将来任何时候都不想离开，但我认为，组建一支真正非常伟大的团队的方式就是为适合他们的事情而奋斗。

Lenny Rachitsky: Wow, those are incredible stories.

Lenny Rachitsky：哇，这些都是令人难以置信的故事。

And it makes so real these, as you say, cliches, hire the best people, this is what it sounds like to hire the best people.

正如你所说，这让这些陈词滥调变得如此真实，雇用最好的人才，这就是雇用最好的人才的听起来。

This is what it takes.

这就是它所需要的。

Scott Wu: Yeah, no, and I was just saying, a lot of things we've fought very hard to just reimagine things from the ground up because a lot of it really is just thinking about where do we think the technology is going over the next five, 10 years, and what is the place that we want to have in that future?

Scott Wu：是的，不，我只是说，我们非常努力地从头开始重新构想很多事情，因为很多事情实际上只是在思考我们认为技术在未来五年、十年内会走向何方，以及我们希望在未来拥有什么？

Lenny Rachitsky: I wonder if people are going to be fighting for the best Devin someday.

Lenny Rachitsky：我想知道有一天人们是否会为最好的德文而战。

They're just going to Devins.

他们要去德文斯。

Scott Wu: I'll give you overtime pay, benefits, free healthcare, and everything, and then the Devin's like...

Scott Wu：我会给你加班费、福利、免费医疗等等一切，然后德文就像......

Lenny Rachitsky: Three Devins like Magic: The Gathering cards.

Lenny Rachitsky：三个德文喜欢万智牌。

Lenny Rachitsky: And then just going back to your three to five things.

Lenny Rachitsky：然后回到你的三到五件事。

So, essentially this is incredible advice, essentially, it's like you always hear hire the best people, move fast, build things people want.

所以，本质上这是一个令人难以置信的建议，本质上，就像你总是听到雇佣最好的人，快速行动，构建人们想要的东西一样。

Scott Wu: Yeah, build something people want, stay as close as possible to your customers, and then I think the other thing is just always think about where things are going, not where they are today.

Scott Wu：是的，打造人们想要的东西，尽可能靠近你的客户，然后我认为另一件事就是始终考虑事情的发展方向，而不是它们今天的情况。

I feel like those are the five things, which is...

我觉得这就是五件事，那就是……

Especially in AI, with things moving so fast, and there's so much great talent, I feel like a lot of these are even more true, where it's like it's not just thinking about where things are going to be in 10 years, it's like thinking about what's going to happen next week.

尤其是在人工智能领域，事情发展得如此之快，而且有如此多的优秀人才，我觉得其中很多都是更真实的，这不仅仅是考虑 10 年后事情会发生什么，就像考虑下周会发生什么。

And obviously, things are moving very quickly, and it is very hard to predict, but you really have to be very rigorous with yourself, I'd say, about thinking through those things and evaluating all of the decisions that you make in that lens.

显然，事情发展得非常快，而且很难预测，但我想说，你真的必须对自己非常严格，思考这些事情并评估你在这个镜头中做出的所有决定。

Lenny Rachitsky: And staying focused is the big takeaway to me here, is it ends up feeling like there's 1000 things you should do, but it's always these five things.

Lenny Rachitsky：保持专注对我来说最大的收获是，最终感觉你应该做 1000 件事，但总是这五件事。

Lenny Rachitsky: Scott, we covered a lot of ground.

Lenny Rachitsky：斯科特，我们涵盖了很多内容。

We went through every question I had, which is great, is there anything else that you want to share?

我们解决了我遇到的每一个问题，这很好，您还有什么想分享的吗？

Anything else you want to leave your listeners with, maybe a final nugget or something really you want to double down on that we said before we let you go?

您还想留给听众什么其他内容，也许是最后的金块，或者您确实想在我们让您离开之前所说的基础上加倍努力的东西？

Scott Wu: The biggest thing that comes to mind for me is there's a lot of different perceptions about AI, right?

Scott Wu：我想到的最重要的事情是关于人工智能有很多不同的看法，对吧？

I think basically every emotion under the sun right now.

我基本上想到了现在阳光下的所有情绪。

There's a lot of fear, for example, there's also a lot of skepticism, and we're very skeptical types as well, and we always wanted to try it ourselves to really see it and believe it.

例如，有很多恐惧，也有很多怀疑，我们也是非常怀疑的类型，我们总是想亲自尝试一下，真正看到它并相信它。

And I think the main thing that comes to mind for me, is I'm honestly really optimistic about what we're building here with AI, not just with code and with Devin, but the whole space, and everything that's getting done.

我认为我想到的最重要的事情是，老实说，我对我们在这里用人工智能构建的东西非常乐观，不仅仅是代码和 Devin，而是整个空间，以及正在完成的一切。

And I think one of the cool things that is really actively happening is just the ability for everyone to multiply themselves, and that's how we've always thought about it, it's how we've thought about what we're building.

我认为真正积极发生的一件很酷的事情就是每个人都有能力自我繁殖，这就是我们一直以来的思考方式，这就是我们对我们正在构建的东西的思考方式。

And I think there is a lot more to do out there in the world, I'm not too worried about us running out of things to do, and from that lens, I think that the thing that we've always been most excited about is, how can we all do more?

我认为世界上还有很多事情要做，我不太担心我们会无事可做，从这个角度来看，我认为我们一直最兴奋的事情是，我们如何才能做得更多？

Lenny Rachitsky: I hear you Scott.

莱尼·拉奇茨基：斯科特，我听到了。

Well, with that optimism, we've reached our very exciting lightning round.

好吧，带着这种乐观态度，我们已经进入了非常激动人心的闪电回合。

Are you ready?

你准备好了吗？

Lenny Rachitsky: Okay, here we go.

莱尼·拉奇茨基：好的，我们开始吧。

First question, what are two or three books that you find yourself recommending most to other people?

第一个问题，您发现自己最向其他人推荐的两三本书是什么？

Scott Wu: In terms of nonfiction, I think for folks in startups, I think one of the things that I've really enjoyed is just learning and understanding the history of Silicon Valley.

Scott Wu：就非小说类作品而言，我认为对于初创公司的人们来说，我真正喜欢的事情之一就是学习和了解硅谷的历史。

And it's all of these things that we think about, somebody invented them.

我们所思考的所有这些事情都是有人发明的。

It's one of the great realizations I feel like, is somebody invented the idea of a seed route, somebody invented the idea of venture capital, somebody invented the idea of product-market fit, and all of these different principles that we talk about.

我觉得这是一个伟大的认识，有人发明了种子路线的想法，有人发明了风险投资的想法，有人发明了产品市场契合的想法，以及我们谈论的所有这些不同的原则。

And so, for that, there's a book called The Power Law by Sebastian Mallaby, which I really like.

为此，塞巴斯蒂安·马拉比（Sebastian Mallaby）写了一本名为《幂律》（The Power Law）的书，我非常喜欢。

And basically, it's just a tour of many of the great businesses and the great products that have been built over the last 60, 70 years in Silicon Valley, which I really love.

基本上，这只是对过去 60、70 年来在硅谷建立的许多伟大企业和伟大产品的参观，我真的很喜欢。

I think in terms of fiction, I actually have always really liked The Great Gatsby by F.

我觉得就小说而言，我其实一直很喜欢F的《了不起的盖茨比》。

Scott Fitzgerald, it's one of my personal favorites as a fiction book.

斯科特·菲茨杰拉德，这是我个人最喜欢的小说之一。

Lenny Rachitsky: Do you have a favorite recent movie or TV show that you've really enjoyed?

Lenny Rachitsky：您最近有最喜欢的电影或电视节目吗？

Scott Wu: I have to admit, I have not watched...

Scott Wu：我必须承认，我没有看过……

I can't think of a single movie or TV show that I have watched in the last while, so I'm sure, I'm looking forward to watching a lot of great ones post-AGI.

我想不起最近一段时间看过的任何电影或电视节目，所以我确信，我期待着在 AGI 之后观看很多精彩的电影或电视节目。

Lenny Rachitsky: That's got to be in the trailer, that's great, I like that.

Lenny Rachitsky：那必须出现在预告片中，太棒了，我喜欢。

And that just shows how hard you're working, just how much shit is going on, and how fast everything's moving.

这只是表明你工作有多努力，发生了多少糟糕的事情，以及一切进展得有多快。

Do you have a favorite product you've recently discovered that you really love?

您最近发现了您真正喜欢的最喜欢的产品吗？

Could be an app, could be something physical, could be a toothbrush.

可以是一个应用程序，可以是一个物理的东西，可以是一把牙刷。

Scott Wu: One I would say is I got an Aura frame recently, it's just like a frame that shows photos, and you can show a new photo every day, or every hour, or every 15 minutes, or whatever you like.

Scott Wu：我想说的是我最近得到了一个 Aura 相框，它就像一个显示照片的相框，你可以每天、每小时、每 15 分钟或任何你喜欢的方式显示一张新照片。

I've actually, I've really enjoyed it a lot, I think it's a nice way to basically just have a picture frame memories that come up.

事实上，我真的非常喜欢它，我认为这是一种很好的方式，基本上只是让相框记忆浮现出来。

And then, the other thing I would say as a general purpose thing, it's not particularly new, but I would say, I think AirPods are extremely well-built and well-designed.

然后，我想说的另一件事是，作为一个通用的东西，它并不是特别新，但我想说，我认为 AirPods 的制作和设计都非常精良。

And I realize now that it's like, I basically use them for all...

我现在意识到，我基本上都用它们......

I'm taking calls on a walk and I'm using AirPods, obviously I'm doing work at my computer, at my desk, I'm plugged into AirPods, and it works quite well, honestly, for a lot of different situations, and they're very comfortable, they're very consistent.

我在散步时接听电话，我使用 AirPods，显然我是在电脑前工作，在办公桌前，我插入了 AirPods，说实话，它在很多不同的情况下都工作得很好，而且它们非常舒适，非常一致。

Lenny Rachitsky: I'm going to double down on the Aura frame, I also, I got one of these for my mom and my mother-in-law, and they're so great for just sharing photos of your kids with your family.

Lenny Rachitsky：我会在 Aura 相框上加倍努力，我也为我妈妈和我岳母买了一个，它们非常适合与家人分享孩子的照片。

And people have, they've heard of digital picture frames, but the Aura just does it really well, and it's really easy to add photos, and they're just really nice looking.

人们听说过数码相框，但 Aura 确实做得很好，而且添加照片真的很容易，而且它们看起来真的很漂亮。

Scott Wu: You can imagine not that long from now, we'll have the Aura Frame except Studio Ghiblifies every photo that you have in it, and then it's...

Scott Wu：你可以想象，不久之后，我们将拥有 Aura Frame，除了 Studio Ghiblizes 你在其中的每张照片，然后它......

Yeah.

是的。

Lenny Rachitsky: Or just imagines things you've done that are really cool.

Lenny Rachitsky：或者只是想象你所做的事情非常酷。

Look at my sweet life.

看看我甜蜜的生活。

Yeah.

是的。

Cool.

凉爽的。

And it's Aura, it's A-U-R-A, I believe is how you spell it.

它是 Aura，它是 A-U-R-A，我相信这就是你的拼写方式。

Folks who want to check it out, we'll link to it.

想要查看的人，我们将链接到它。

Not affiliated.

不隶属。

Okay, two more questions.

好吧，还有两个问题。

Do you have a favorite life motto that you often come back to and find useful in work or in life?

您是否有一条最喜欢的人生格言，并且经常回顾并发现它在工作或生活中很有用？

Scott Wu: Yeah, something I've thought about a lot is a lot of the proverbs out there are actually contradictions, right?

Scott Wu：是的，我想了很多，很多谚语实际上是矛盾的，对吧？

It's like birds of a feather, and then you also have opposites attract, and you have all...

就像物以类聚，异性相吸，你就拥有了一切……

And it's kind of funny, because you feel that both of them are true, and often they both are true, and a lot of it is about understanding why.

这有点有趣，因为你觉得它们都是真的，而且通常它们都是真的，而且很多都是为了理解为什么。

And one of those that I feel like, especially in the world of startups that I think about all the time is, I think it is very important to be focused and driven, and to really maximize your potential, and then at the same time, it's also very important to not let your own personal emotion get tied up in your success or failure, and I think especially with startups, because there's always ups and downs, honestly, even in the most successful companies ever, it's just like, it's a rocky road, there's a lot that happens, and a lot that goes down.

我的感觉之一，特别是在我一直思考的初创公司的世界里，我认为专注和驱动，真正发挥你的潜力非常重要，同时，不要让你自己的个人情感与你的成功或失败联系在一起也非常重要，我认为特别是对于初创公司，因为总是有起起落落，老实说，即使在有史以来最成功的公司中，它就像是一条崎岖的道路，发生了很多事情，也发生了很多事情。

Scott Wu: And I think one of the things which I've thought a lot about is that somehow you really want to do your best, and put everything you can into it, and do everything you can to...

Scott Wu：我想我想了很多的一件事是，你真的想尽力而为，尽你所能，尽你所能......

Basically, you want to put it all out on the field.

基本上，你想在球场上全力以赴。

But at the same time, you want to be okay with both wins and losses, and you want to be able to move on, and go into the next one each time.

但与此同时，你希望能够接受胜利和失败，并且希望能够继续前进，每次都进入下一场。

And something, yeah, it's funny, but what I've found personally is that obviously it's really important for your own emotional state and mental state to be able to do that, and we've had lots of mistakes, and I've had a lot.

是的，这很有趣，但我个人发现，显然，能够做到这一点对于你自己的情绪状态和精神状态来说非常重要，我们犯了很多错误，我也犯了很多错误。

Scott Wu: I had my first company, which obviously, which was cool, but there are a lot of tricky spots there, and then over the course of Cognition, it feels like it's been already eight years compressed into one year, and it's still going at that pace.

Scott Wu：我有我的第一家公司，显然这很酷，但是那里有很多棘手的地方，然后在 Cognition 的过程中，感觉已经把八年压缩成一年了，而且仍然以这个速度进行。

But somehow it also actually makes you more successful, I think, too.

但我认为，不知何故，它实际上也让你更加成功。

It's like, you are just more able to give it your best, and to do the things that will lead to success if you're not tying it up in your own personal worth.

就好像，如果你不把自己的个人价值捆绑在一起，你就更有能力尽自己最大的努力，做一些能带来成功的事情。

Lenny Rachitsky: That is so interesting, I just had a podcast recording recently where with an executive coach, Jerry Kelowna, that I think will come out before this, might be after this, that's one of his big pieces of advice, and it's a very Buddhist approach of just and attaching to a certain outcome.

Lenny Rachitsky：这太有趣了，我最近刚刚与执行教练 Jerry Kelowna 录制了一段播客，我认为这会在此之前发布，也可能在此之后发布，这是他的重要建议之一，这是一种非常佛教的方法，公正并重视特定的结果。

Lenny Rachitsky: Okay.

莱尼·拉奇茨基：好的。

Final question, I'm curious if there's a story here, but we could keep it short, is there a story behind Devin as the name, or is there another contender for Devin being the ?

最后一个问题，我很好奇这里是否有一个故事，但我们可以简短地说，Devin 这个名字背后是否有故事，或者 Devin 是否还有另一个竞争者？

Scott Wu: Devin was the name from pretty early on, we were interested, we were working on coding agents from the beginning, and my co-founders are Steven and Walden, for example, and we had this idea, all right, let's get started, and let's try to expand the box as much as we can.

Scott Wu：Devin 是很早就有了的名字，我们很感兴趣，我们从一开始就致力于编码代理，例如我的联合创始人是 Steven 和 Walden，我们有这个想法，好吧，让我们开始吧，让我们尝试尽可能地扩展盒子。

So, have everyone think out of the box and do their own thing, and let's have everyone do their own thing first for a bit, and then we'll consolidate and take everything that we've learned.

所以，让每个人跳出框框思考，做自己的事情，让每个人先做自己的事情一段时间，然后我们将巩固并利用我们所学到的一切。

And so, Walden made a virtual developer version of him, which was called DevWalden, and then Steven made one of him, which is called DevSteven, and we had all these...

因此，Walden 为他制作了一个虚拟开发者版本，称为 DevWalden，然后 Steven 制作了他的一个，称为 DevSteven，我们拥有了所有这些......

And then we were kind of combining it all into one thing, and we're like, okay, you know?

然后我们将所有这些结合成一件事，我们就像，好吧，你知道吗？

It's Devin.

是德文。

And that was the thing.

就是这样。

And so Devin was, yeah, Devin stuck for us quite early on, I would say.

所以德文，是的，我想说，德文很早就为我们坚持了。

Scott Wu: One thing which we did have a big decision on though actually is what the image of Devin would be.

Scott Wu：我们确实做出了重大决定的一件事是 Devin 的形象。

And so, as folks know, there's the hexagons and then people have seen this more recently, but there's actually also an otter, a little otter with a laptop in its lap, and that is Devin as well.

所以，正如人们所知，有六边形，然后人们最近才看到这个，但实际上还有一只水獭，一只膝盖上放着一台笔记本电脑的小水獭，这也是德文。

And we had this debate over what to go with, and what not to go with and stuff, and it's been a while now, but somehow we still have both the hexagons and the otter.

我们就该搭配什么、不搭配什么等等问题进行了争论，已经有一段时间了，但不知何故，我们仍然拥有六边形和水獭。

Lenny Rachitsky: You skipped over where the Devin, did you have just have, it just came to you?

Lenny Rachitsky：你跳过了 Devin 的地方，你刚才有没有，它就来找你了？

Scott Wu: Oh, so Devin is, it's a dev.

Scott Wu：噢，Devin 就是这样，它是一个开发者。

Yeah.

是的。

Scott Wu: And so, it was kind of like, when we were consolidating all the names, it just seemed clear then that this would be the universal dev that we all liked to work with.

Scott Wu：所以，有点像，当我们整合所有名称时，很明显这将是我们都喜欢合作的通用开发者。

Lenny Rachitsky: Incredible.

莱尼·拉奇茨基：难以置信。

Scott, this was so much fun.

斯科特，这太有趣了。

Oh my God, I learned a ton, which is always a really good sign.

天哪，我学到了很多东西，这总是一个好兆头。

Two final questions, where can folks find you/Devin/anything else you want to point them to, and how can listeners be useful to you?

最后两个问题，人们在哪里可以找到你/德文/你想要向他们指出的任何其他东西，以及听众如何对你有用？

Scott Wu: Awesome.

吴斯科特：太棒了。

Yeah, no, we're at App.devin.ai, and you can find us as well on Twitter or a lot of other social media.

是的，不，我们在 App.devin.ai，您也可以在 Twitter 或许多其他社交媒体上找到我们。

We'd obviously love to hear any feedback you have about the Devin product, there's so much to figure out, and I think the, like I said, I think we're all still 20 steps away from really the future of software engineering, and so it really means a lot to hear what folks think about the product as they're trying it out.

显然，我们很乐意听到您对 Devin 产品的任何反馈，有很多东西需要弄清楚，而且我认为，就像我说的，我认为我们距离软件工程的真正未来还有 20 步之遥，所以听到人们在尝试该产品时对此产品的看法确实意义重大。

And so, please let us know anytime if there's things that we can do to make it better.

因此，如果我们可以采取措施使其变得更好，请随时告诉我们。

Lenny Rachitsky: Scott, thank you so much for being here.

Lenny Rachitsky：斯科特，非常感谢你来到这里。

Scott Wu: Thank you so much for having me.

Scott Wu：非常感谢您邀请我。

I had a great time.

我玩得很开心。

Me too.

我也是。

Bye everyone.: Thank you so much for listening, if you found this valuable, you can subscribe to the show on Apple Podcasts, Spotify, or your favorite podcast app.

大家再见。：非常感谢您的收听，如果您觉得本文有价值，您可以在 Apple Podcasts、Spotify 或您最喜欢的播客应用程序上订阅该节目。

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

另外，请考虑给我们评分或留下评论，因为这确实有助于其他听众找到播客。

You can find all past episodes or learn more about the show at LennysPodcast.com.

您可以在 LennysPodcast.com 上找到所有过去的剧集或了解有关该节目的更多信息。

See you in the next episode.

下一集见。