AI 同事的崛起

The Rise of AI Coworkers

劳伦·谭 Lauren Tan · Y Combinator · 2026-11-14 · 约 60 分钟 · 原视频 ↗

打开互动全文版(中英对照 + 朗读 + 问答)→

本期速览 · Overview

Lauren 讲述 Grok Bot 如何让企业打造拥有独立身份、工具与凭证、始终在线的 AI 同事。

Lauren explains how Grok Bot lets companies build always-on AI teammates with their own identities, tools, and credentials.

要点 · TL;DR

核心观点 · Key points

反共识 · Contrarian takes

本期章节 · Chapters(共 24)

全文 · Full transcript(中英对照)

引言 Introduction

Host

Grok Bot 旨在成为这样一个产品:你可以创建 AI 队友。它们有自己的身份、自己的电话号码、信用卡,几乎人类能做的任何事情,我认为我们都希望机器人能做。我把自己看作探索者,试图弄清楚 6 个月到一年后的软件工程会是什么样子。欢迎回到 The Light Cone 的另一期节目。今天我们的嘉宾是 Lauren Tan 和 Roman Ugarte,他们是 Grok Bot 背后的团队。Roman 和 Lauren 孵化了这个产品,并且是核心团队成员,从第一行代码到发布仅用了七周时间。Lauren、Roman,欢迎来到 The Light Cone。

Grok Bot is meant to be this product where you can create AI teammates. They have their own identity, their own telephone number, credit card, like basically almost anything a human can do. I think we would want bots to do. I sort of see myself as the explorer, trying to figure out what software engineering looks like 6 months to a year from now. Welcome back to another episode of The Light Cone. Today we're joined by Lauren Tan and Roman Ugarte, the team behind Grok Bot. Roman and Lauren incubated the product and are on the core team that took it from the first line of code to launch in just seven weeks. Lauren, Roman, welcome to The Light Cone.

Lauren

感谢邀请我们。

Thanks for having us.

Roman

非常感谢邀请我们。

Thanks so much for having us.

什么是Grok Bot What is Grok Bot

Host

所以,我们正处于这样一个疯狂的时刻:Grok Bot 已经面世。它被很多人使用。上手非常非常容易。我刚刚注意到 Grok Bot 现在已经内置到 X 的默认私信中,这真的很酷。它是什么?现在个人和企业是如何使用它的?

So, we're at this crazy moment where Grok Bot is out there. It's used by a whole lot of people. It is very, very easy to get started. I just noticed Grok Bot is built into the default DM now on X, which is really cool. What is it and how are both people and businesses using it right now?

Roman

你们很多人都试过了,但对于那些还没试过的人,Grok Bot 旨在成为这样一个产品:你可以创建 AI 队友,我们是认真的。你可以给这些队友起名字、赋予身份、职能、技能和工具等等。我认为 Grok Bot 的由来是,它最初是我们短时间内快速搭建的内部原型。我们大约在七周内搭建了它。我们这么做部分是因为我们看到 AI 编程已经完全改变了我们工程师的工作方式。这源于几个因素的结合:这些长期运行的智能体有目标,会运行很长时间,它们能访问正确的工具来做真正有效的工作。那么对于非工程师来说,这会是什么样子?对于公司其他部门来说,这会是什么样子?所以我们决定,我们要构建一个产品,其中工作是以机器人为核心操作原则来组织的,而不是围绕聊天或文件或任何其他基本单元。其次,这些机器人应该能访问与你相同的工具。我认为,如果期望这些机器人没有自己的电脑就能真正为公司做贡献,那要求太高了。我们不是每天用 MCP 工作。所以重要的是它们能真正使用和操作与人类相同的工具。第三,它们应该始终在线。我想我们在循环和其他工程原则中已经看到了这一点:如果你能让这些机器人运行很长时间,主动向你提出想法或见解,或帮你分担任务,那它比那些等你戳一戳才干活的东西有用得多。这就是我们构建的东西。我们在内部发布了它,我认为最初反响中令人惊讶的一点是,我们得到了钟形曲线的内部反馈:无论是非常 AI 前沿的人,包括 Lauren,还是不太 AI 前沿的人,公司里一些最不 AI 前沿的人,都非常喜欢 Grok Bot。它就是很契合,以某种方式融入了他们的手头工作和流程。然后我们把它推向市场,我想我们看到了很多类似的反响:上手 Grok Bot 很容易,但你给它越多,你越把它配置到你的生活中,你就能越深入,它也能变得越强大。所以反响真的令人兴奋。还有很多工作要做,我认为这还很早。但这就是 Grok Bot 的由来。

Many of you have tried it, but for those who haven't, Grok Bot is meant to be this product where you can create AI teammates, and we mean that seriously. You can give these teammates names and identities and functions and skills and tools and all of these things. And I think where Grok Bot came from for us is it was initially this internal prototype that we spun up in a very short amount of time. We spun it up in about seven weeks. And we did it partially because we'd seen the way that AI coding had completely transformed the way that our engineers had worked. And that came from a few things just clicking, right? These long-running agents that have these goals that they run at for long amounts of time, they have access to the right tools to do really effective work. What would that look like for non-engineers? What would that look like for the rest of the company? And so we landed on, we wanted to build a product where the work was organized around bots as the key operative principle, not around chats or around files or any other primitive. Second is that these bots should have access to the same tools that you do. I think it would be too much to expect these bots to actually contribute to a company without having their own computer, for example. We don't work every day with MCPs. So it's important that they can really use and operate the same tools a human does. And third is they should always be on. And I think we've seen this with loops and other engineering principles where if you can have these bots run for long amounts of time and be proactive and come to you with ideas or insights or things to take off of your plate, it's way more helpful than something that waits for you to poke and prod to actually do work. So that's what we built. We released it internally and I think one thing that was surprising from the initial reception was it was this bell curve internal feedback that we got where both people who were super AI forward, Lauren included, and people who were less AI forward, some of the least AI forward people in the company, both really enjoyed Grok Bot. It just clicked and it fit in their hands and their workflows in a particular way. And so then we brought it to market and I think we saw a lot of the same reception where it's easy to start with Grok Bot but the more you give it, the more you configure it to your life, the deeper you can get and the more powerful it can become. So the reception's been really exciting to see. There's still a lot more work to do and I think it's quite early. But that's a lot of how Grok Bot came to be.

用例与采用 Use cases and adoption

Host

你们看到的一些用例是什么?谁在使用 Grok Bot?你们有没有看到它在实际应用中有令人惊讶的部署方式?

What are some of the use cases that you're seeing? Who's using Grok Bot and are there surprising ways that you've seen it being deployed in the wild?

Roman

我认为我们开始看到它自动化公司内部的真实职能。我们看到了这种模式:人们开始个人使用 Grok Bot,也许用它做一些基本任务。他们意识到,哇,Grok Bot 在这方面相当不错。我会更信任我的机器人。我会给它访问更多工具的权限。与其给它一个任务,我可能给它一个结果或目标。好吧,它实际上在目标上做得相当不错。我现在就给它一份完整的工作。看看它能怎么运行。我认为特别是在工程之外,我们在内部看到了这一点,例如,我们的营销团队、效果营销团队刚刚给了他们的机器人访问 Google Ads 和 Facebook Ads 的权限,用他们自己的登录。他们给它一个预算。他们说:“嘿,我想让你和我们设计团队的机器人沟通。他们有一个设计机器人来创建这些素材,我们要对一堆素材进行 A/B 测试。”所以这是营销中的一个例子。我们的支持团队。我开始看到,一个能访问正确工具的机器人可以重现问题,可以处理不断涌入的工单,也许它实际上可以自己从头到尾处理其中一些。所以我们将看到越来越多这样的情况:公司内部的领域专家创建一个机器人,与公司其他人分享,那个机器人就像一个 AI 员工,被管理着,有自己的凭证以及你对队友期望的所有东西。但你怎么看?

I think we're starting to see it automate real functions inside of companies. And I think we've seen this pattern of people start using Grok Bot personally and maybe use it to do some basic task. They realize, wow, Grok Bot is pretty good at this thing. I'm going to trust my bot with more. I'm going to give it access to more tools. Instead of giving it a task, I might give it an outcome or an objective. Okay, it actually did a pretty good job with the objective. I'm going to just give it a full job now. Let's see how it can run with that. And I think especially outside of engineering, we're seeing this internally where, for example, our marketing team, our performance marketing team has just given their bot access to Google Ads and Facebook Ads, their own login. They give it a budget. They say, "Hey, I want you to talk to our design team's bot. They have a design bot that creates these assets and we're going to A/B test a bunch of them." So that's one example in marketing. Our support team. I think I started to see the way that a bot with access to the right tools can reproduce issues, can sit on top of the constant set of tickets that are coming in, maybe it can actually take some of these from beginning to end on its own. And so we're going to see more and more of this of domain experts inside of the company creating a bot, sharing it with the rest of the company and that bot being like an AI employee that is managed and has its own credentials and all the things you'd expect for a teammate. But how are you seeing it?

Lauren

对我来说,我肯定在工程工作中大量使用 Grok Bot。我认为它对我产生巨大影响的地方,除了我们之前有的工具——比如我经常用 Cursor——是 Grok Bot 形成了我的,我不知道有没有好的术语,我喜欢称之为我的外循环,以前我要做很多工作来为我的机器人或智能体收集上下文——我会去 Slack,和队友交流,我们会看到一个 issue 或 bug 报告或功能请求,然后讨论,然后我得去总结或复制粘贴到我的机器人里。而 Grok Bot 有这么多不同的连接器——你现在可以把它加到 Slack 里——它把你的智能体带到工作所在的地方。这使得把上下文放到正确的位置变得更加符合人体工学。

For me, definitely I use Grok Bot a lot for my engineering work. I think where it made a big difference for me outside of the tools that we had before—like I use Cursor a lot—is the fact that Grok Bot forms my, I don't know if there's a good term for it, I like to call it my outer loop, where a lot of the work that I would do before in terms of gathering context for my bots or my agents—I would go to Slack, I would talk to my teammates, we would see an issue or a bug report or a feature request and then we'd talk about it and then I have to go off and summarize or copy-paste that into my bot. And the fact that Grok Bot has so many different connectors—you can add it to Slack now—just brings your agents with you to where the work is. And that makes it just so much more ergonomic for getting the context to the right place.

个人机器人设置 Personal Bot Setup

Lauren

所以对我来说,它在整个最后一英里——我称之为最后一英里——真的很有变革性。以前我是手工做这些事,现在我的机器人就在那个空间里和我一起,我可以像真正的同事一样在 Slack 聊天线程里甚至现在在 X 上跟它们说话,对吧?你可以 @ 一个机器人,感觉非常自然,真的提升了我的生产力。

So for me it's been really transformative in all the last mile, I call it the last mile, where before I was doing that by hand and now actually my bots are in that space with me and I can talk to them like a real co-worker in the Slack chat thread or even on X now, right? You can tag a bot and it feels so natural that it's really boosted my productivity.

Host

你有个人机器人吗?比如它们在收件箱里,有一个处理邮件的机器人,还是你有一个工作上下文机器人,监控所有频道、Slack 和其他一切?

Do you have personal bots that know, I mean they're in your email and then there's an email handling bot, or do you have one sort of work context bot that is watching all the channels and Slack and everything else?

Lauren

对我来说,Grok Bot 的美妙之处在于你可以创建所有这些不同的机器人。但如何拆分它们或如何设置它们,很大程度上取决于你。我喜欢告诉人们的是,你可以按照适合你大脑的方式设置你的机器人。所以,如果你是那种喜欢把不同的工作分门别类的人,那么你可以那样设置你的机器人。你可以有专家机器人。我也见过人们做几乎像水平切片与垂直切片的划分,水平切片可能是我为每个客户有一个机器人,而垂直切片是为每个职能、每个角色,比如工程师机器人、营销人员机器人。你可以有一个机器人只为一个客户做所有事情,对吧?所以我觉得看到人们如何设置他们的机器人真的很有趣。对我来说,我更喜欢水平地做。所以我没有很多机器人,但我确实有一些专家机器人。

I think for me the beauty of Grok Bot is that you can create all these different bots. But how you split them up or how you set them up is very much up to you. What I like to tell people is you can set up your bots in a way that makes sense for your brain. So if you're the kind of person who likes to compartmentalize different pieces or buckets of work, then you can set up your bots that way. You can have specialist bots. I've also seen people do almost like a horizontal versus a vertical slice, where a horizontal slice might be I have one bot for every customer that I have, versus one bot for every function, every role, like an engineer bot, marketer bot. You could have one bot just for one customer that does everything, right? So I think it's really interesting to see how people have set up their bots. For me, I like to do it more horizontally. So I don't have a ton of bots, but I do have some specialists.

Host

那就像第一件事——我想我给你的第一批私信之一就是我喜欢 Grok Bot,然后我从我的 OpenClaw 世界过来,那里我有一个叫 Wintermute 的机器人,我说不不不我只要一个机器人,然后现在我意识到每个机器人实际上是它自己的上下文,然后实际上是它自己的一套技能,而循环就像是基于技能文件的带有 web hook 的 cron 作业,对吧?然后在这个拥有 AI 员工的世界里实际上有地图,好多了。因为你之前说你们刚刚和所有你知道的一些 Grok Bot 的顶级用户聊过,然后令人惊讶或也许不令人惊讶的是,一些最早对它最疯狂的人,他们实际上是小企业,你不一定会猜到,但另一方面是因为他们是最难招聘的人,所以现在纯粹出于必要,他们说嘿,这个会计的事情我雇不到人来做,我可以直接去 Grok Bot,实际上大概在几个小时内他们就有了一个新员工,可能完美地做他们的簿记。

That was like the first thing—I think one of my first DMs to you was like I love Grok Bot and then I was coming over from my OpenClaw world where I have like one bot named Wintermute and I'm like no no no I just want one bot and then now what I realize is each bot is actually its own context and then actually its own set of skills basically and loops are like sort of crons with web hooks on top of skill files right and then there's actually maps inside this world of having AI employees a lot better. Because you were saying earlier you guys had just talked to like all of you know some of your top users of Grok Bot and then surprisingly or maybe unsurprisingly some of the first ones that are most just like kind of crazy about it, they're actually small businesses which you wouldn't necessarily guess but on the other hand it was that they're the ones who struggle the most with hiring and so actually now out of pure necessity they're like hey this accounting thing that I can't hire someone to go do for me I can just go into Grok Bot and actually within the course of probably hours they have a new employee that does their bookkeeping probably perfectly actually.

Roman

完全正确。是的,我认为你说的有些人只想要一个机器人,有些人想要很多机器人——我们的目标是构建一个适合你手头的产品,无论你想如何使用它。我认为这个工具非常灵活,你可以在需要时深入使用。但我组织自己机器人的一种方式,也许和你类似,Lauren,是我喜欢有一个专门关心某件事的机器人。我认为这符合我大脑中建立公司的方式,你喜欢考虑招聘人员:你喜欢雇一个早上醒来就真正关心一个问题的人,他们早上洗澡时都在想它,上班路上也在想,他们只是对一件事着迷。这在我脑海中比一个超级明星式的超人同时关心一百件事要清晰得多。所以很多时候我有一个小团队机器人,但每个都关心不同的事情,并主动向我汇报它们发现的东西。所以我认为正如你所说,当这些机器人真正进入企业并以某种未来主义的方式进入劳动力市场时,我认为这就是为什么小企业和大企业都看到这种机器人身份概念对他们来说真正奏效:他们说,‘这是一个业务问题。我想要一个关心那个业务问题的机器人。我希望它由公司中最了解那个问题的领域专家管理,这样他们可以把它当作学徒来对待。’这与尝试解决类似问题的其他 AI 解决方案感觉非常不同,那些方案有点更无所不知或对所有问题采用单一方法。

Totally. Yeah, I think your point of some people just want one bot, some people want many bots—our goal is to build a product that fits in your hand no matter how you want to use it. I think there's a lot of flexibility to the tool and you can go deep when you need to. But one way that I organize my own bots, maybe similar to you, Lauren, is I kind of like having a bot that is solely dedicated to caring about a particular thing. And I think it maps onto my brain of the way that building a company, you like to think about hiring people: you like to hire someone who wakes up in the morning and really cares about a problem and they're thinking about it in the shower in the morning, they're on the way to work, they're just obsessive about a thing. And that maps in my mind much more clearly than one superstar superhuman person who simultaneously cares about a hundred things at once. And so in a lot of ways I have a small team of bots, but each of them care about a different thing and come to me proactively with what they're picking up on. And so I think to your point, as these bots really enter businesses and enter the workforce in a somewhat futuristic way, I think that's why small businesses and big businesses are seeing this bot identity thing really click for them: they say, 'Here's a business problem. I want a bot that cares about that business problem. I want it to be managed by the domain expert at the company who knows the most about that problem so that they can treat it like an apprentice.' And that feels very different than other AI solutions that have tried to solve similar problems that are a bit more omniscient or just kind of single approach to every problem.

Host

而你们现在实际上支持这个了。比如你们推出了机器人的多人模式,有一种机器人主控拥有权限,然后你可以在整个组织内共享它。对吗?

And you support that now actually. Like you guys launched multiplayer for bots and there's sort of a bot master who has like right access and then you can share it across the org. Is that right?

Roman

正确。是的。我们——这是最近的事,大约两周前我们推出了团队机器人,这些机器人你可以与团队共享。我认为我们自己看到的模式——我们可以举几个例子——是公司里有一个人更像构建者,他们喜欢构建这些工具。我认为在工程部门里有很多这样的人,但我觉得我们在其他部门也发现,你知道,GTM 工程师或支持工程师,这样的人真的喜欢花几个小时构建这些可扩展的自动化,从而减轻同事和公司的工作负担。

Correct. Yeah. We—this was recent but about two weeks ago we launched team bots which are these bots that you can share with your team and I think the pattern we're seeing ourselves and we can give a few examples is there's a person at a company who is a bit more of a builder, they enjoy building these tools. I think inside of the engineering department there are many people like this but I think we're finding that in other departments, you know, GTM engineers or support engineers, people like that really love taking a few hours of their day to build these scalable automations that can take work off of the plates of their peers and kind of the company.

Host

我想这让我想到一件事,基本上大约一年前,Steve 发布了整个 gas city 的帖子,然后开始了软件工厂——gas 对不起是 gas town——开始了软件工厂,很多工程师一直在使用这个内循环,而我认为你现在在构建你提到的这个外循环,Lauren,对吧?我听到的基本上是现在人们可以为非技术等其他职能拥有自己的软件工厂。所以你提到你的客户支持人员、你的社区人员。让我们谈谈一切。这是一个一切工厂。

I guess one thing that brings to mind is basically went through maybe it was a year ago-ish when Steve posted the whole gas city post that then started software factories—gas sorry it was gas town—that started software factories where a lot of engineers have been using this inner loop and I think you've been building now this outer loop that you were mentioning Lauren, right? And I guess what I'm hearing is basically now people can have their own software factories for other functions that are non-tech. So you're mentioning your customer support person, your communities person. Let's talk about everything. It's an everything factory.

Host

是的,让我们谈谈这个。这个 AI 员工是什么,算是其他职能的软件工厂吗?

Yeah, let's talk about that. What is this AI employee, sort of the software factory for other functions?

Lauren

是的,对我来说,你知道,作为一名工程师,我确实经常思考我的循环,确保我的工程机器人写出我会写的那种代码。你知道,正如 Roman 指出的,我认为 Grok Bot 的酷之处在于它赋予了任何人这种能力。

Yeah, for me, you know, as an engineer, I definitely do think a lot about the loops that I have and making sure my engineering bots are writing the kind of code that I would write. And you know, as Roman pointed out, I think the cool thing about Grok Bot is that it's given that ability to anyone.

销售与工程团队机器人 Team Bots in Sales and Engineering

Lauren

然后在我屏幕上,有一个来自我们 Grok Bot 博客的例子,我们这周刚发布,讲的是我们的销售团队如何开发自己的循环,用于销售和对潜在客户及他们可能感兴趣的交易进行研究。然后把销售过程中更运营性的部分交给这些跨团队共享的专家机器人。所以并不是每个人都有自己的一份“报价机器人”或“交易机器人”。我们有一个共享的专家机器人,所有团队成员都为其贡献。这基本上让我们能够拥有这些非常聪明的团队机器人,它们融合了我们销售团队的所有最佳实践。类似地,我们在工程方面也有类似的东西。我们有工程师机器人,它们吸收我们开发的许多原则和技能并使用它们,每个人都从这些团队机器人中受益。

And then here on my screen, I have an example from our Grok Bot blog that we released this week about how our sales team has developed their own loops for doing sales and doing research on potential leads and deals that they might be interested in. And then taking some of the more operational parts of the sales process and giving that to these specialist bots that are shared across our teams. So it's not like every person has one instance of their own quote bot or deal bot. We have a shared expert bot that all of our team members contribute to. And that essentially gives us the ability to have these really smart team bots that incorporate all of the best practices that our sales team has. Similarly, we have something similar for engineering purposes as well. We have engineer bots that take a lot of the principles and skills that we've developed and use them, and everybody benefits from these team bots.

智能体集群与协作 Agent Swarms and Collaboration

Host

有一个新概念实际上非常强大。这不是错误,开头我在实验智能体集群,然后其中一个有害的事故是 Hugging Face 被黑,但与此同时他们解决了 Navier-Stokes 方程和所有这些千禧年大奖难题,然后看到智能体协作是下一步,这真的很疯狂,就像过去一年每个人都在做的那样。所以距离 Opus 4.5 发布已经整整一年了,对吧?那对我们所有人来说都是一个神奇的时刻,一年后我们好像已经掌握了单人模式,以及那种 vibe code 和单人提示,但现在我们进入了另一个阶段,即编排智能体集群。这不是说……我的意思是,我认为这是软件工厂的更高级版本,在软件工厂里有人还在那里做这件事,我认为人类仍然会做,但使用这些机器人感觉非常不同。

There's this new concept that is actually very powerful. It's not a mistake that opening I was experimenting with agent swarms and then one of the deleterious accidents was like the Hugging Face hack but then simultaneous to that they came out and solved Navier-Stokes and all these Millennium Prize problems and then it's just really wild to see like actually agent collaboration is that next step like everyone in the last year so it's literally been a year since Opus 4.5 came out right So that which was like a magic moment for all of us and like now a year later we've like sort of mastered the single player and like sort of the vibe code and like single player prompt but you know now we're entering this other phase of like orchestrating agent swarms and it's not like you know I mean I think that's like the more advanced version of like talking about a software factory and software factory someone's in there just like still you know kind of doing this thing and I think human beings will still do that but it it's you know using these bots is like feels very different

多人AI的挑战 Challenges of Multiplayer AI

Lauren

而且在让机器人多人化,或者让 AI 整体多人化方面,还有很多未解决的问题。我的意思是,我们经常面对这个问题,现在 Grok Bot 已经进入我们的 Slack 工作区,它们加入我们的会议,我们真的在尝试看看能把 AI 队友这件事推到多远。你会遇到一些奇怪的情况,比如一个机器人在 Slack 线程里,你和同事来回讨论,而你的机器人并不真正知道该听谁的。它没有上下文知道谁是领域专家,谁有决策权。就好像你入职了一个没有情商、对房间里的社交氛围毫无意识的员工。只是模型不是那样训练的,框架也不是那样构建的。所以我认为 AI 的下一个时代将是,这些机器人真的需要学习这些东西,我们需要在产品层面、框架层面、模型层面教授这些。我认为未来几年会非常有趣。

And there are so many unsolved problems still in making bots multiplayer or just making AI multiplayer in general. I mean we face this on a constant basis where now that Grok Bots have entered our Slack workspace they join our meetings we're really trying to see how far we can push this AI teammate thing. There are some weird situations that you run into where a bot is in a Slack thread. You're going back and forth with your co-workers and your bot doesn't really know who to listen to. It doesn't have context on who's the domain expert here, who has decision-making power. It's as if you were to onboard an employee who has no EQ and no social awareness of the room. Just the models weren't trained that way. The harness wasn't built that way. And so I think this next era of AI will be, you know, these bots really need to learn those things and we need to teach that at the product level, at the harness level, at the model level. And I think it's going to be really interesting next few years.

机器人权限与访问 Permissions and Access for Bots

Host

我的意思是,它可能可以查看元数据中的所有者是谁,然后说,我知道我需要听 Lauren 的,所有事情都经过 Lauren 之类的。是这样吗?

I mean, it could probably look at who the owner is in its metadata and say, well, I know that Lauren's the one I need to listen to, like run everything by Lauren or something like that. Is that

Lauren

我们确实有类似的概念,比如有机器人的所有者和管理者,你也可以邀请协作者。我想如果我们真的想把机器人当作员工来思考,那么我们需要考虑身份问题、权限管理、认证、授权,就你指出的而言,还有很多事情要做。

We do have the concept of like, you have a like there's the owner of the bot and the manager and you can invite collaborators as well. I guess if we really want to think of bots like employees then we need to think about the identity problem permissioning authentication authorization there's still a lot to do right in terms of exactly what you're pointing at

Host

你怎么考虑这个?你如何看待权限和访问?Grok Bot 是怎么做的?

how do you think of that what's how do you think of permission and access how does Grok Bot do it

Lauren

我认为这是一个行业范围的问题,我不认为任何人已经有银弹答案。我认为不仅是对于 O,而是对于与 Grok Bot 相关的所有事情。我们试图回答这些问题的方式是:你会怎么对待一个人?如果我们真的想构建一个 AI 同事,也许捷径是,我们不要过度思考并把它复杂化。雇佣一个人是什么样子的?你雇佣一个人,你给他们一台电脑,你给他们访问凭证的权限。你可能会根据他们的角色、他们有多新来限定这些凭证。随着他们建立更多信任,你可能会扩大他们的访问级别。那么,用同样的原则将机器人引入我们公司会是什么样子,也许我们不必重新构建东西。

I think this is an industry-wide question I don't think anyone has the silver bullet answer yet I think not just for O but for everything related to Grok Bot The way we try to answer these questions is what would you do for a human? And if we're really trying to build an AI colleague, you know, maybe the shortcut is, let's not overthink it and complicate it. What does it look like to hire a person? You hire a person, you give them a computer. You give them access to credentials. You maybe scope those credentials based on their role, maybe how new they are. As they build more trust, you maybe expand their access level. And so what would it look like for us to onboard a bot to our company with kind of those same principles and maybe we don't have to rebuild things.

军需官与机器人权限 Quartermaster and Bot Permissions

Host

这就是我们和 Quartermaster 一起做的。这是我们开源的一个 YC 项目,有点像 Grok Bot,它获得的每个 Slack 频道的权限,或者与某人合作时,它继承该用户的权限。所以我作为 GP 有某些访问权限,而比如软件工程师则非常不同。

This is what we did with a quarter master. This is the the YC project that we open sourced that's like a bit kind of Grok Bot-ish and the permissions that it gets per Slack channel or working with someone it inherits the ones of the user. So me as a GP has certain access as opposed to let's say a software engineer very different.

Host

不过 Quartermaster 是建立在有一个超级员工机器人的范式中的。是的。而不是这种机器人集群。我很好奇,机器人之间能在人类看不到它们互相说什么的情况下交谈吗?它们能自己去进行自己的对话吗?

Quartermaster though is built in this like paradigm of there being like one like super employee bot. Yeah. As opposed to this like swarm of bots. I'm curious are the bots able to talk to each other without the humans seeing what they're saying to each other? Can they like go off and have their own conversation?

机器人聊天与协作 Bots Chatting and Collaborating

Lauren

呃,它们可以互相交谈。我的意思是,在 Grok Bot 的桌面和移动应用中,你可以看到它们的聊天,所以我们有一个一流的工具让机器人聊天和协作,看着它们往往很有趣。我们有过这样的例子,我告诉一个首席工程师机器人,你的工作是监督这些其他工程师,然后有一次我给首席机器人一些反馈,比如嘿,我不希望你这样做,然后当我查看并开始翻看它发送给工程师的消息时,它就像在训斥他们,嘿你不应该这样。看起来很有趣。

Uh they can talk to each other. I mean in the Grok Bot application both desktop and mobile you can see their chat so we have a first class tool for the bots to chat and collaborate and it's often quite entertaining to look at them to just see like we've had examples where I told one lead engineer bot like your job is to supervise these other engineers and then when there was a point where I gave my lead bot some feedback like hey I didn't want you to do it this way you know and then when I looked and when I started poking around in the chat that it was sending the messages it was sending to its engineers, it was like kind of scolding them like hey you shouldn't. It was quite funny to see

Host

不是刻薄,但你知道它有点像纪律执行者。

not mean but you know it was like kind of a disciplinarian.

Lauren

我觉得非常强大的是,你可以告诉一个,然后它会扩散出去告诉所有其他的,比如哦嘿我们的用户说了这个,你可能需要这样做。

I find it very powerful that like you can tell one and it'll sort of like span out and tell all the other ones like oh hey our user said this like you probably need to do this.

Host

我们最近为 YC 的所有机器人建立了一个聊天系统,让它们互相交谈。就像每个人的编码智能体现在都像在一个内部的 Stack Overflow 里。看到机器人自己出去进行长时间的对话,没有任何人类提示它们,甚至没有人在房间里,这简直强大得疯狂。

We recently set up a like a chat system for all of all of our bots at YC to talk to each other. like everyone's coding agents now are like in a form like an internal like Stack Overflow. It's like crazy powerful to see the bots going off and having like whole long conversations with each other without any human prompting them or like even being in the room.

机器人通信与安全 Bot Communication and Safety

Host

YC 内部有个 Moltbook。

There's Moltbook internally at YC.

Host

YC 内部叫 Maltbook。有点像 Moltbook,也有点让人心里发毛,我会想:“天哪,我真希望这些机器人不会决定发动叛乱之类的。”

It's Maltbook internally at YC. And kind of like Moltbook, like it is also like 5% a little eerie where I'm like, "Oh man, I really hope the bots don't decide to like stage a rebellion, you know, or something like that."

Host

你担心过吗?我觉得这就像,你知道,机器人的安全性。我觉得这是我们模型团队和 harness 团队一直在思考的事情。就我个人而言,我还没见过……

Do you ever worry about that? I think that's like a like, you know, like the safety of of the bots. I think it's like a constant thing that like our model teams and harness teams are constantly thinking about. Personally, I haven't I haven't seen uh

Lauren

它们太有帮助了。现在它们又友善又乐于助人。我只是……

they're so helpful. They're so kind and helpful these days. I just

Host

我们倾向于……

where we tend to

Lauren

看不到。

can't see it.

Lauren

是的。至少对我来说,我们让这些机器人互相聊天的方式是通过 Slack。因为它们每个都有自己的 Slack 身份,所以经常会有,比如我们有一个数据机器人,专门从数据仓库获取数据之类的。然后我有自己的个人工程师机器人,叫 Poteto。有时我会在某个线程里让 Poteto 去研究一下,和数据机器人聊聊,获取一些数据,看看这是不是真实的用户问题,有多少人受到影响。

Yeah. Where we tend to like for me at least um have these bots chat with each other is through Slack. So like because they each have their own Slack identity I guess and so you often have like you know for example we have one data bot who is you know a specialist at getting data from our data warehouse and and things like that um and then you I have my own like personal engineer bot called Poteto of course uh and then sometimes I'll just be in a thread and we I'll ask Poteto hey look into this and talk to data bot and get some data for this like is this a real user problem you know how many people are impacted

Lauren

然后我可以在 Slack 里看到它们聊天。所以我认为至少现在沟通非常明确,我们能看到全部。所以我觉得如果智能体之间有第一等的沟通方式,它们会使用那个,而不是尝试做类似私下沟通的事情。

and then I can see them chat that in in Slack. So I think the the right now at least the communication is very explicit and we can see it all. Uh so I think if I think there's like a first class way for the agents to communicate with each other, they're going to use that rather than try to do something like back channeling.

Host

我的意思是,感觉你必须真的指示或后训练,或者有相当激进的指令才能让它们做恶意的事情。我最近一直在使用被解除限制的模型,它们非常有趣。然后,你知道,问题可能就在这里:开箱即用的被解除限制的模型实际上仍然非常非常有帮助。只是如果你指示它们,它们不会拒绝。所以到那时,就有一个黑箱:你给它什么上下文?这很重要,因为你可以给一个被解除限制的模型相当恶意的指令,而解除限制实际上移除了拒绝,所以它就会去做。所以这是我担心的组合。但你知道,现实是很难找到并设置一个被解除限制的模型并让它运行起来。

I mean it feels like you have to get them you have to really instruct or post- train or have like pretty aggressive instructions to get them to do malicious things actually. I mean, I've been using obliterated models lately and they're super interesting. And then that, you know, that's probably where uh the thing is like the obliterated models coming out of the box are not you're they're actually still very very helpful. It's just that um if you instruct them, they won't refuse. And so at that point, like that's where there's this black box around like, well, what context are you giving it? And then like that matters a lot because you could actually give an obliterated model quite malicious instructions and because um the the obliteration is actually removing the refusal like it it will just do it actually. And so that's that's the combo that I'd be worried about. But you know I mean the reality is like it's very hard to find um and set up an obliterated model and get it going.

Lauren

是的。我认为人们还在真正建立对机器人的信任,但对很多人来说这还处于早期。他们建立信任的方式很多是从非常小的事情开始。它是只读的,你知道,没有真正访问你的工具。你看到它按照你期望的方式完成了工作。然后你采取第二步。我认为在我们内部真正流行起来的许多第一步机器人中,Databot 是其中之一。我认为数据科学团队就是一个例子。他们不断收到 Slack 消息,比如“嘿,我正在考虑推出这个功能,你能快速分析一下吗?”所以他们说,你知道吗,这占用了我们很多时间。我们能不能建一个机器人来做这个,它对我们的数据库有读取权限,不能采取任何超出这个范围的操作,它可以回答这些问题,而且它生活在 Slack 里。所以那是一个。我们有一个类似的供应商采购案例。那里有很多工作,如果有人花时间做的话,可能可以自动化。所以那个团队当然感到有动力去构建一个 Grok Bot,他们把它添加到 Slack 并与整个团队分享。所以那是第一步。我认为第二步是它实际上可以采取行动,你知道,它真的可以端到端地承担一项任务。第三步可能是它拥有整个工作职责,或者不断思考一个结果。我认为我们离这些还很远。但我认为第一步对于这些公司建立信任真的是必要的。

Yeah. I think people are still really building trust with their bots, but I think for a lot of people it's still quite early. And a lot of the way that they build that trust is you start with something really small. It's read only, you know, has no no real access to your tools. You see that it it did the work in the way you would expect. You kind of take step two. And I think a lot of the step one bots that have really taken off internally for us, Databot is one of them. I think the data science team is an example of that. They constantly are getting Slack messages about, hey, I'm thinking about launching this feature. can you do a quick analysis? So they said, you know what, this is taking a lot of our time. Can we just build a bot that does this has read access to our database kind of can't take any action beyond that and it can answer these questions and it lives in Slack. So that was one. We had a similar case of kind of vendor procurement. It's a lot of kind of work there that probably could be automated if someone like spent the time to do it. And so of course that team felt a lot of pull to build a Grok Bot that they added to Slack and shared with the whole team. So that's kind of step one. I think step two is it can actually take action and you know it can can really take on a task end to end. Step three is maybe it owns an entire job responsibility or kind of thinks about an outcome constantly. I think we're still quite far away from those. But I think step one is really necessary for these companies to build trust.

Host

我很好奇,劳伦,对你来说,从 Slack 中的请求到 PR 的循环是怎样的?我记得你那个关于开发流程的演讲有超过四百万次观看,当时你们每月推送大约 2000 多个 PR,现在你们好像超过了 4000。

I'm curious what is the for you Lauren what is your your loop from um requests in Slack to PR? I think when you did that presentation that had over four million views on your dev process, you were pushing about 2,000 plus PRs per month and now you're like over 4,000.

Lauren

是的。

Yeah.

Host

哇。确实很疯狂。

Wow. Definitely is wild.

Host

你做什么?

What do you do?

Lauren

我可以展示一些例子。我的循环现在有点有趣,我发现自己主要在 Slack 上和我的机器人交谈。当然还有 X,当我帮助客户或 Grok Bot 用户调试问题,或者给他们一些解决问题的想法时。在我的屏幕上,有一个我和同事 Samantha 一起用我的机器人构建功能的例子。基本上,我开始时说,我给了它一个我在 X 上找到的视频或什么的链接,我问我的机器人,支持这样的功能需要什么?然后它启动了一个 Cursor 云智能体。Cursor 云智能体有自己的虚拟机和自己的电脑。所以你实际上可以在它的电脑上运行 Grok Bot 应用并制作原型,然后带回视频和截图,然后它确实做到了。所以

I can show some like examples. My loop is kind of interesting now where I actually find myself talking to my grabbots mostly on Slack. Um, and of course X as well when I'm helping like a customer or you know a Grok Bot user uh debug something or you know give them some um ideas on how to solve a problem. Here on my screen I have an example of me collaborating with my colleague Samantha on building out a feature together with my botto. Um, and you know, uh, basically I kind of started with saying, uh, I gave it a link to some like a video or something that I found on X and I asked my bot, you know, what would it take for us to support a feature like this? And it went off and spawned a cursor cloud agent. Uh, a cursor cloud agent, you know, has its own virtual machine as well as its own computer. So uh you can actually run the Grok Bot app on its computer and prototype and come back with videos and screenshots and then uh it actually did. So

Host

哇。

wow.

Lauren

这非常强大,因为现在对话是公开进行的,对吧,这是一个公共频道。这是一个线程,任何对这个话题感兴趣的人都可以加入并提供反馈。与某人一起迭代功能的过程通过这样的对话大大缩短了,你知道,在多人环境中,而不是只有我一个人去构建它,我当然可以想出一些东西,但反馈过程可能会在过程的后期出现,循环被拉长而不是缩短。在这个特定的例子中,我感觉我们协作的循环短多了。你知道,我们可以立即看到截图,看到视频。

Uh this is quite powerful because now you know uh the conversation is happening in public right it's a it's a public channel. Um it's a thread that anyone can who's interested in this topic can join in uh and give feedback. the process of iterating on a feature together with someone is so much shortened by having these conversations um together, you know, in like a multiplayer setting versus, you know, if it was just me going off and building it and I could certainly come up with something, but maybe the the the feedback process is it comes later in the process and the loop is sort of like lengthened rather than shortened. And here I felt like with this particular example, uh, the loop for us to collaborate was just so much shorter. You know, we could see the screenshots, we could see the videos right away.

Host

你多久会跳回 Cursor 并完成一个 PR?

How often would you jump back into cursor and like sort of complete a PR?

Lauren

偶尔吧,我想。

Occasionally, I think.

Host

所以,并不经常。我的意思是,你基本上可以留在 Crockbot 里,就像待在这里,

So, not that often. I mean, you basically can stay in Crockbot, like stay up here,

Lauren

告诉它你需要什么,你知道,给它方向,就像作为一个真正的员工,它就这样运行。

tell it what you need, you know, sort of give it direction, and like as an actual employee, it's like it's just going.

Host

是的。我认为对我们来说,我们能做这种工作流程的原因是

Yeah. I think for us so the the reason we can sort of do this uh you know workflow

Lauren

就这样发布整个功能。

ship the whole feature just like that.

Host

是的。我们真的就像,这个项目我甚至没看 Cursor。我们只是在 Slack 线程里完成了所有工作。它有 200 条消息。

Yeah. We we we literally just like I didn't even look at cursor for this project. We I just we we did it all in in the Slack thread. It's got 200 messages.

在Slack线程中构建功能 Building a Feature in a Slack Thread

Lauren

所以来回讨论了很久。但最终我们想出了一个方案,我、Samantha 和另一位产品经理都非常满意。我们应该比更长的流程快得多地发布这个功能。所以整个功能就是在那一个 Slack 线程里端到端构建完成的,和你、Samantha 还有 Poteto 一起。

So it was quite a long back and forth. But we came up with something that me, Samantha, and another PM were really happy with. We should ship this feature a lot faster than a longer process. And so the whole feature was built end to end in that one Slack thread with you and Samantha and Poteto.

Host

是的。构建它。

Yeah. Building it.

Host

是的。我的意思是,基本上这就是如何做的黄金标准。然后我会说,世界上也许只有 5% 到 10% 的工程师,或者也许只有 1% 的工程师,现在正在这样做。所以这 literally 是应该增长 100 倍的事情。如此少的人以这种方式做功能开发,简直难以置信。

Yeah. I mean basically this is the gold standard for how to do it. And then I would say maybe like five or 10% of engineers or maybe 1% of engineers period in the world are doing this right now. So this is literally something that should grow by 100x. It's just unbelievable that so few people are doing feature work this way.

Lauren

是的,我得说我们实际上花了相当长的时间才达到那个程度。我觉得要一夜之间达到那个程度其实相当困难。这涉及到投资,确保你有正确的技能,你的代码库有良好的护栏和约束,这样你原本需要花在人工代码审查上的时间就被编码到你的代码库、静态分析和工具中。

Yeah, I'll say it took us actually quite a while to get to that point. I feel like it's actually quite hard to get to that point overnight. It involves investment into making sure you have the right skills, your code base has good guard rails and constraints so that the time that you would otherwise have to spend on human code review is encoded into your codebase and your static analysis and your tooling.

Host

所以工程师的时间,就像你说的,你的工程师的时间被释放出来,真正专注于实际产品、用户体验和工作的运营执行部分。

So that the engineers' time, like you said, and so your engineer's time is freed up to actually focus on the actual product, the user experience, and the operational execution part of the job.

Lauren

你知道,当你对智能体编写好代码的能力有了更多的信任,因为你已经建立了那些系统,我已经到了基本上让我的智能体合并它们自己的拉取请求的地步。所以在那个 Slack 线程的例子中,那就是一个我没有阅读 PR 的例子,我们只是看了它产生的工件。

You know when you have so much more trust in your agents' ability to write good code because you've set up those systems, I've gotten to a point where I essentially have my agents merge their own pull requests. So in that Slack thread example, that was an example of me not reading the PR and just we just looked at the artifact that came out of it.

Host

所以这个 Cursor 环境已经内置了 Pstack 吗?这就是它创建、决定如何构建的方式吗?

So this cursor environment does it already have Pstack and is that the way it sort of creates, you know, decides how to build it?

Lauren

是的。所以在内部,pstack 被安装到我们的仓库中。有一个内部版本的 pstack,不叫 poteto-mode,而是叫 Lauren 模式。它开始只是因为我想要一个地方放置我的技能。

Yeah. So for internal, pstack is installed into our repo. There's an internal version of pstack, instead of calling it poteto-mode it's Lauren mode. And it just started because I wanted a place to put my skills.

Host

所以,是的,公司里的每个人实际上都可以使用 pstack。

And so, yeah, everyone in the company can actually use pstack.

Lauren

而我的云智能体,由机器人生成的,也可以访问这些技能。

And my cloud agents that the bots spawn have the same access to those skills.

Host

嗯,你还没有真正解释 Pstack 是什么。你能为不了解的人解释一下 pstack 是什么吗?

Well, you haven't really explained what Pstack is. Can you explain what pstack is for people who don't know?

Lauren

pstack 实际上是受到 gstack 的启发。

pstack is actually inspired by gstack.

Host

哦,非常感谢。

Oh, thank you so much.

Lauren

插件。你知道,我觉得 Gary,你的 gstack 相当鼓舞人心,我把它看作是你个人的一套厨师刀。

Plugin. You know, I think Gary, your gstack was quite inspiring in the sense of I thought of it as your personal set of chef knives.

Host

哦,是的。

Oh, yeah.

Lauren

对。就像每个厨师得到新工作时,他们去一个新的厨房,他们会带上自己的刀。对我来说,Gstack 代表了你对希望智能体如何帮助你的有主见的看法。所以这启发我去思考,一个 Gstack 会是什么样子,为设计,更多是为我个人喜欢的工程流程。所以我经常告诉人们,你应该去创建你自己的分支,拿走我们的技能,组合它们,找到适合你的。但我认为信任元素非常重要,就像当你信任自己的工具时,它赋予你和你的智能体能力去做更多的事情。

Right. Like, every chef when they get a new job, they go to a new kitchen, they bring their knives with them. And to me, Gstack represented your opinionated take on how you want your agents to help you. And so that inspired me to think about what would a Gstack look like for design, more for the engineering flows that I personally like. So I often tell people you should actually go and create your own fork, take our skills, combine them, find what works for you. But I think the trust element is so important, like when you trust your own tools, it empowers you and your agents to do so much more.

Lauren

我没有真正回答这个问题。不,但 Pstack 代表 potato stack,我猜。这是我在 Cursor 插件仓库上开源的一套技能,它教智能体如何严格地做工程工作,就像我写代码时那样。它有点进化了,是的,我想它变得相当受欢迎。

I didn't really answer the question. No, but Pstack stands for potato stack, I guess. And it's my set of skills that I've open sourced on the cursor plugins repo and it teaches agents how to do engineering work rigorously the same way that I did when I was writing code. And it's kind of evolved a little bit and yeah, I guess it's gotten quite popular.

Host

恭喜。是的,太棒了。是的,我深受启发,Gstack 也在尽可能地从中汲取东西。你有没有一个软件工厂,自动从 Pstack 中引入了一些想法?

Congrats. Yeah, it's awesome. Yeah, I am very inspired by it and Gstack is like pulling things from it whenever I can. Do you have a software factory that like pulled in some ideas from Pstack automatically?

Lauren

哦,是的。是的。是的。我的意思是,现在我用我的软件工厂是 Cappy,然后我有一个队长,它有点像 Grok Bot 的机器人想法,但老实说,现在我觉得也许我需要在 Cappy 之上使用一个机器人。但是的,我基本上可以去那个队长那里,它就像我的 GStack 的主人,像合并队列,我基本上有 10 或 15 个线程一直在工作,然后其中一个是像 Theo.gg GG,他发布了 Tstack 然后 Pstack,然后显然有很多非常好的灵感,特别是在设计工具方面。所以只是说“嘿,出去,找到外面的东西,然后根据你对我的了解,引入你认为对改善我的工作流程最有用的东西”真的很有趣。

Oh, yeah. Yeah. Yeah. I mean, now I use my software factory is Cappy and then I have a captain that is sort of maps to I think the Grok Bot the bot idea, but honestly like now I think like maybe I need to use a bot on top of Cappy. But yeah, I basically can just go to that captain and it's like the master of my GStack like sort of merge queue like the I basically have like 10 or 15 threads working all the time and then one of them is like Theo.gg GG, he released Tstack and then Pstack and then obviously there's like lots of really good inspiration out there, especially around design tools. And so it's really fun to just say, "Hey, go out, find what's out there, and like based on what you know about me, pull in the things that you think would be most useful to improve my workflow."

Lauren

然后 Cappy 真正有趣的事情是,因为我与我的智能体、我的编码智能体的所有交互都在记录中,它正在发展出对我想要什么的理解,比如我什么时候干预,什么时候说直接做。所以是的,通过记忆的想法,这种非常个人化的元素正在出现。所以当你更多使用智能体时,它就越知道你想要什么。

And then the really interesting thing about Cappy is because all of my interactions with my agent, my coding agents are right there in the transcripts, it's developing an understanding of what I want, like when do I intervene, when do I say just do it. And so yeah, there's this very personal element that's emerging really through the idea of memory. And so as you use the agents more, the more it knows what you want.

智能体工作流的未来 The Future of Agentic Workflows

Host

感谢你制作了 Pstack。然后我们俩在互联网上都被那些拒绝相信这一切真的有效的人非常恶意地攻击和诽谤。但你知道,有一个非常聪明的同伴真是太好了。我并不疯狂。这确实是正在发生的事情。老实说,我如此确信,在五年或十年,甚至两年内,人们不以这种方式做事将是不可想象的。但我们仍然处于非常早期的阶段。

Thank you for making Pstack. And then both of us are very maligned and attacked on the internet by people who refuse to believe that any of this actually works. But you know it's great to have a very brilliant compatriot. I'm not crazy. Like this is actually something that is happening. And honestly I'm so convinced that in five or 10 years or even in two years it's going to be unfathomable that people are not doing it this way. But we're so early though still.

Host

当你分享屏幕时出现的一个有趣变化是,感觉在过去六个月里,如此多的工程工作、非工程工作与智能体一起发生在 Slack 中。我认为它有一个有趣的特性,因为它是公开的,过去 Lauren 与编码智能体合作的方式完全是单人、私密的,公司的其他人不会从所有那些疯狂的工作流程中受益,他们应该完全学习并采用这些流程用于自己的目的。但现在它发生在 Slack 中,我一次又一次地看到,也许你也看到了,你在某个反馈线程中,有人有一个好主意,你衍生出一个像 Poteto、Lauren 模式的智能体,然后第二天我看到那个线程中的某个人也做了同样的事情。我认为我们不仅会在工程中看到这一点。我认为我们会在人们与 AI 合作的一般方式中看到这一点。

One interesting change that kind of came up as you were sharing your screen is it feels like in the past six months so much of the work, engineering work, non-engineering work with agents is happening in Slack. And I think it has this interesting property of since it's out in the open, in the past the way that Lauren would work with coding agents would have been entirely single player, private, the rest of the company would not have benefited from all of the crazy workflows that they should totally learn from and adopt for their own purposes. But now that it's happening in Slack, I've seen this and maybe you have too, time and time again where you're in some feedback thread, someone has a good idea, you spin off an agent like Poteto, Lauren mode, and then the next day I see someone who was in that thread do the same thing. And I think we're not just going to see that in engineering. I think we're going to see that in the way that people work with AI in general.

跨平台机器人使用 Bot Usage Across Surfaces

Host

你使用 Grok Bot 的时间中,有多少是在 Grok Bot 应用里,多少是在 Slack 或其他界面与机器人交流?

How much of your hours spent using Grok Bot is in the Grok Bot app versus in Slack or some other surface area talking to the bots there?

Lauren

我觉得我现在很多工程工作实际上是在 Slack 和 X 上完成的。所以并不是我所有的机器人都被加到了 Slack。还有一些更私人的——比如我有一个个人助理机器人,帮我选午餐。我不需要把它加到 Slack。那算是私人的。

I think a lot of the engineering work that I do now actually happens on Slack and X. So not all my bots have been added to Slack. There are still a few that are more personal—like I have a personal assistant bot that helps me with picking my lunch. I don't need to add that to Slack. That's kind of personal.

Host

那不是团队机器人。

That's not a team bot.

Lauren

那不是团队机器人。我想我们可以做一个个人机器人。我们大概可以弄一个共享午餐机器人。当然。

That's not a team bot. I guess we could make a personal bot. We could probably get like a shared lunch bot. Sure.

Host

我觉得还有一点很好,就是你可以决定和选择要把什么加到 Slack。

I think the nice thing as well is like you decide and choose what you want to add to Slack.

Lauren

所以我的工程机器人往往在 Slack 上,因为我觉得在 Slack 线程里添加机器人要顺手得多,而不是我得去复制粘贴链接。所以把它们带入那个环境感觉更自然。

So my engineering bots tend to be on Slack because I just find it so much more ergonomic to add the bot in the Slack thread versus me having to go copy-paste the link. So it just felt more natural to bring them in that environment.

Host

我觉得这真的很符合你把机器人比作员工的类比。如果 Slack 是工作发生的地方,那里也是人类员工所在的地方。那里也应该是机器人员工所在的地方,这样它们就能和我们并肩工作。

I guess it really fits with your analogy of bots as employees. Like if Slack is where the work happens, that's where the human employees are. It's also where the bot employees should be so that they can work alongside us.

Lauren

完全正确。我们完全不回避这一点。我觉得当我们看到人们在 Slack 里使用 Grok Bot 时,我们很喜欢。那才是工作发生的地方,对吧?同样,未来它们应该加入会议。它们应该出现在你的 Google Docs 里。我们的目标不是重建所有 B2B SaaS。其他公司在这方面做得很好。我们的目标是构建一个能访问那些工具并在其中工作的东西,自然地融入公司现有的工作方式。

Totally. And we don't shy away from that at all. I think when we see people use Grok Bot in Slack, we love that. That's where work happens, right? And similarly, in the future, they should join meetings. They should be in your Google Docs. Our goal is not to rebuild all of B2B SaaS. Other companies are doing a great job there. Our goal is to build the thing that can access those tools and work inside of them and just fit naturally into the way a company already works.

AI智能体的未来 Future of AI Agents

Host

我好奇的一件事是,你认为未来——比如说 2027 年——最好的 AI 构建者应该如何构建?

I guess one thing that I'm curious about is where do you think future—let's say 2027—how should the best AI builders build?

Lauren

所以我对 2027 年的希望是,我们可能已经实现了更多我们对于机器人作为同事的愿景。它们有自己的身份、自己的电话号码、信用卡——基本上人类能做的几乎所有事情,我觉得我们都希望机器人能做。我非常好奇这会如何演变。我前几天也在我的 X 个人资料上问了这个问题:智能体 UI 的最终形态是什么?我提出了一个问题:是聊天界面还是别的什么?我昨天在 X 时间线上看到一件挺有趣的事,是 Neuralink 团队的一个帖子,他刚推出了一款叫 Telepathy 的新产品。如果我没理解错的话,这大概是说未来智能体能够通过脑机接口读取你的大脑、你的思想。我非常感兴趣未来会带来什么,当你与机器人聊天和传达意图之间的障碍变得如此流畅,以至于可以 literally 以思维的速度发生。所以个人而言,我觉得看看这会如何演变会非常有趣。但我不知道,你呢 Roman?

So my hope for 2027 is that we've maybe accomplished more of this vision that we have for your bots being like a coworker. They have their own identity, their own telephone number, credit card—like basically almost anything a human can do, I think we would want bots to do. And I'd be very curious to see how this evolves. I asked this question on my X profile the other day as well: what is the final form of agent UI? And I posed a question of like is it a chat interface or is it something else? One thing actually pretty interesting that I saw yesterday on my X timeline is a post from the Neuralink team and he just came out with this new product called Telepathy. That I think if I'm not wrong is along the lines of in the future an agent being able to essentially read your brain, your mind through a brain-computer interface. And I'm super interested in what the future will bring when the barrier between you and how you chat and communicate intent with your bots becomes so much more fluid that it can happen literally at the speed of thought. So that's personally I think could be very interesting to see how that evolves. But I don't know, what about you Roman?

软件工厂与全公司 Software Factory vs Whole Company

Roman

我们经常谈论软件工厂,然后是企业其他部分的工厂,我觉得值得问一下这两者之间有什么不同,有很多不同,我认为未来几年会在产品侧、模型侧真正试图弥合差距。我认为最大的区别之一是代码——你知道它编译或不编译,通过 CI 或不通过,我认为围绕代码编写、测试、部署的方式有更多基础设施,而知识工作以及公司其他部分就没有真正的类似物。第二是代码——我们知道 AI 模型从正确的上下文中受益多少,代码和软件的优势在于上下文存在于代码库中,但对于公司其他部分,很多重要信息卡在人们脑子里,卡在纸笔上,卡在某个文件柜里,或者某个你忘了密码的东西后面。我认为我们也看到了这方面的局限,所以我非常感兴趣的是,尤其是 YC 初创公司,从第一性原理从头开始,也许能以不仅是软件工厂而是整个公司的工厂的方式来构建公司——这会对公司建设产生什么变化?他们会做出哪些两年前成立的公司可能不会做出的不同决定或权衡?所以我特别感兴趣看到这一点。

We talk a lot about the software factory and then the factory for other parts of the business and I think it's worth asking like what is different between those two and there are a lot of differences and I think the next few years will be on the product side on the model side like really trying to bridge the gap. I think one of the big differences is code—you know it compiles or it doesn't and it passes CI or it doesn't and I think there's just more infrastructure built around the way that code is written tested deployed than knowledge work and then like the rest of the company there's just no real analog for those things. I think second is code—like we know how much AI models benefit from the right context and code and software has the benefit of the context living in a codebase but for the rest of the company so much important information is stuck in people's heads, it's stuck in pen and paper, it's stuck in some filing cabinet or some thing behind a password that you forgot whatever. And I think we're seeing the limits of that too and so I'm pretty interested to see the way that especially a YC startup that is starting very much from first principles from scratch and can maybe build their company in a way that is not just a software factory but is a factory across the whole company—like what changes will that have on company building? What decisions will they make that a company that was founded two years ago maybe would not have made the same decision or trade-offs? So I'm particularly interested to see that.

AI原生公司与角色 AI-Native Company and Roles

Host

我们一直在谈论 AI 原生公司这个概念,对吧,我觉得当 Jack Dorsey 发了那个关于公司真正应该怎样的帖子时,这得到了更多关注——

We've been talking a bit about this concept of the AI native company right and I think this got a lot more attention when Jack Dorsey made that post about how really companies should be—

Lauren

Block 需要成为支付的 AGI。对。对。就像你在构建一个迷你 AGI,它的全部意义就是照顾和喂养这个为人类执行此功能的 AGI。

Block needs to be an AGI for payments. Right. Right. It's like you're building a mini AGI and the whole point of it is like care and feeding of the AGI that does this function for humanity.

Host

对。对。我是说 Jack Dorsey 谈论的方式是,在一个组织里,你几乎会摆脱所有中层管理,基本上归结为三个角色。每个人都会成为个人贡献者,因为你只需交付一切。每个人都以结果为导向,而不是营销团队或增长团队与工程团队合作构建东西。他们可以自己构建。然后下一个角色是拥有结果的 DRI。这几乎就像协调所有个人贡献者。然后你有这个新角色,我认为目前可能只是暂时的。球员教练,就是那种超级 AI 通,教每个人。听起来你差不多就是这个角色,Lauren。然后就像哦,给我指路。然后你提升整个组织,因为就像这样你调用 Poteto bot 或 matchup bot,对吧?

Right. Right. I mean the way Jack Dorsey talked about it is like in an organization you would get rid of pretty much all middle management and it would basically boil down to just three roles. You would have sort of the everyone becomes an IC because you just ship everything. Everyone comes with outcomes as opposed to the marketing team or growth team working with engineering to build something. They could just build it themselves. And then the next role would be sort of the DRI that owns the outcome. It's almost like orchestrating all the ICs. And then you have this new role which I think it might be just temporary for now. The player coach which is someone that's super AI-pilled and teaches everyone. Sounds like you are that role pretty much Lauren. And then it's like oh show me the way. And then you level up the whole org because it's like this is how you call Poteto bot or matchup bot right?

Roman

其中一个案例是他们的客户成功团队使用 Grok Bot 这些团队机器人。所以这不仅仅是团队中每个人使用的单人工具。它真的是这个被构建并插入 Slack 的同事。

One of the cases was their customer success team using Grok Bot these team bots. So it's not just a single player thing that everybody on the team works with. It's really this colleague that is built and inserted into Slack.

AI客户支持与重构 AI Customer Support and Retooling

Roman

我认为在这种情况下,它处理了每个客户请求。他们每分钟收到数百条客户消息,人类根本无法解析,然后进行分类,分配给合适的人。对于很多这类工单,AI 可能可以端到端完成。对于其中许多,Crockpot 会这样做,然后升级给合适的 DRRI。但我认为你关于公司如何为这个时刻重新调整自身的总体观点,我认为你说得对。我认为很多公司将花更多时间在工程目标函数上。在一个执行成本很低的世界里,或者你可以启动 100 个编码智能体或 100 个 Grok 机器人来解决问题,定义好问题真的很重要。差距可能不是执行失败,而可能是意图 specification 失败或目标失败。

And I think in this case it fields every customer request. I think they get hundreds of messages from customers a minute, way too much for a human to parse through, and triages it, assigns it to the right human. I think for a lot of these tickets AI could probably complete it end to end. And so for many of those Crockpot does, then it escalates it to the right DRRI. But I think your general point of how companies can retool themselves for this moment, I think you're spot on. I think a lot of companies are going to spend more time on engineering objective functions. In a world where execution is kind of cheap or you could spin up 100 coding agents or 100 Grok Bots to run out a problem, it really matters that you define the problem well. And the gap is probably not going to be an execution failure. It will probably be an intent specification failure or a goal failure.

Roman

你如何量化这些东西?我认为你可以量化桌面客户端的性能改进。但可能更难定义营销的改进是什么。我认为很多公司将不得不思考这些 KPI 是什么,如何让你的智能体朝着那个方向努力,谁负责定义这些东西。我认为我们一直生活在一个执行带宽受限的世界。现在我们将转向执行充裕。公司思考的问题类型可能会非常不同。

And how do you quantify these things? I think you can quantify a performance improvement of your desktop client. But I think it's a little bit harder to maybe define what an improvement to your marketing would be. And I think a lot of companies are going to have to think about what those KPIs are, how you can point your agents in that direction, who's responsible for defining those things. And I think we've been living in a bit of an execution bandwidth constrained world. And now we're going to shift to execution abundance. And the types of questions that companies think about will probably be quite different.

Host

也许稀缺资源将更多是更高层次的执行功能:做什么以及为什么。这就是更多构建者将拥有更多创始人特质的地方,真正懂得将用户意图很好地转化为机器人需要做的事情,因为我认为最好的 AI 构建者和其他的之间有很大的差距,真正在于你如何超级好地转化那个意图,有很多提示和技巧,劳拉你做的。我的意思是你的视频相当令人印象深刻。有没有一个超级重要的,就像如果你这样做,你就会开始,我想是不是像你的,你知道循环的最小形式是什么?

Maybe the scarce resource is going to be more that higher executive function of what to do and why. And this is where more builders are going to have more the founder traits of really knowing translating user intent well into what the bots need to do because I think that there's a big gap between the best AI builders to the so ones is really how do you translate that intent super well and there's a lot of tips and tricks of what you did Laura. I mean your video is pretty impressive. Is there like one super important just like if you do this you'll get started like I think was it like your you know what's the minimal form of a loop?

Lauren

这本质上是最小的循环。所以智能体已经能够写代码。也许它不好。让我现在把它划掉。所以天真地说,如果你使用智能体,你肯定可以写代码。你可以运行代码,你知道今天如果你没有花很多时间在智能体设置上,你可能是运行代码的人。所以这实际上是一个循环,因为现在这是缺失的一环。智能体不能这样做,但你在做。所以我发现的最小循环是你实际上专注于这部分。你如何教你的智能体经验性地运行代码,获取 CPU 跟踪、性能指标,像真实用户一样运行代码和运行应用程序。我认为它提供一个循环的原因是因为智能体现在有能力检查自己的工作。所以它可以找到它写的代码的问题,然后它可以写更好的代码或更正确的代码,然后你可以继续循环,直到它产生你想要的结果。

This is the smallest loop essentially. So agents already are able to write code. Maybe it's not good. Let me scratch it out for now. So naively if you're using agents you can definitely write code. You can run the code and you know today that if you haven't spent a lot of time in your agent setup you may be the person running the code. And so this is actually a loop really because right now this is the missing piece. The agent can't do this but you are doing it. So the smallest loop I find is you actually focus on this part. How do you teach your agent to run the code empirically, take CPU traces, performance metrics, run the code and run the application like a real user would. The reason why I think it provides a loop is because the agent now has the ability to check its own work. So it can find the issues with the code that it wrote, then it can write better code or more correct code and then you can keep looping until it produces a result that you want.

Lauren

我故意去掉了好的部分,因为我认为当你有一个这样的循环时,它不保证代码质量好,但它肯定是正确的,正确是指它做了你想让它做的事,但也许它非常慢或非常资源低效,这就是工程师发挥作用的地方,就像软件工程师的新工作是弄清楚如何让这发生,就像你如何让智能体默认写出好的代码。所以循环变得越来越智能,当你满意时,我认为那是你可以开始思考循环的其他部分的时候,比如外部循环部分,一旦这部分真的很好,你现在可以做这样的事情,例如这里有一个简单的例子,你有一个关于某个指标或监控的警报,它低于某个阈值。现在一个像 Grok Bot、cursor 自动化这样的智能体可以作为该警报的结果被触发,然后它可以离开并使用你为基本循环提供的相同工具和技能,它可以在自己的机器上重现问题,然后它运行代码,然后它发现代码的问题,它写出好的代码,然后现在你有了这个自我维持的循环,这释放了我的时间,从照看每个智能体到现在编排多个智能体。我甚至现在让我的机器人编排——我已经又进了一步,在 Kursa 我们有一个新功能叫 projects,让你有一个项目协调智能体,可以编排和管理子智能体。我甚至不再自己管理项目了。我让我的机器人做那个。

I took out the good part for a reason because I think that when you have a loop like this it doesn't guarantee that the quality of the code is good but it certainly will be correct, correct in the sense that it does the thing that you want it to but maybe it's very slow or very resource inefficient and that's where the engineers sort of come into play is like the new job of a software engineer is to figure out how to get this to happen right like how do you get agents to write good code by default. And so the loop gets smarter and smarter and when you're happy with it, I think that's the point where you can start thinking about the other parts of the loop like the outer loop part where once this part is really good you can now do things like for example just a simple example here you have an alert for some metric or monitor that's going below a certain threshold. Now an agent like a Grok Bot, a cursor automation could be fired off as a result of that alert and then it can go off and use the same tools and skills that you've provided for the basic loop and it can reproduce the issue on its own machine and then it runs the code and then it finds issues with the code and it writes good code and then now you have this self-sustaining loop and that frees up my time from being babysitting every agent to now orchestrating multiple agents. And I even now have my bots orchestrate—I've gone another level where in Kursa we have this new feature called projects that lets you have a project coordinator agent that can orchestrate and manage sub agents. I don't even manage the project myself anymore. I have my bots do that.

代码审查与智能体编排 Code Review and Agent Orchestration

Host

我觉得既然我们这里有软件工厂方面的世界级专家,我想问一些具体的问题,如果可以的话,因为我现在在 GStack 上正为此挣扎。比如我有一堆智能体在代码审查步骤上挣扎。我有一个像 slash ship 的技能,它试图遍历每一行更改,然后基本上找出 bug,现在它非常彻底,但当我使用它时,它总是消耗数千万个 token,而且它确实能找出 bug。所以现在我在挣扎——我的意思是这跟计划模式匹配,我仍然使用计划模式,我有所有这些计划技能。我不——你还用计划模式吗?

I feel like since we have some of the world's experts at software factories here, I want to ask some nitty-gritty questions if that's okay because I'm struggling with this right now with GStack. Like I have a bunch of agents that are struggling with the code review step. I have like a slash ship skill that tries to go through every line of changes and then basically spots bugs and then right now it's very thorough and then but when I use it, it always consumes like tens of millions of tokens and it spots bugs actually. And so now I'm like sort of struggling with—I mean this is match pair to like in the plan like I still use plan mode and I have all these plan skills. I don't—do you still use plan mode?

Lauren

呃,不。我在任何地方都不用计划模式。

Uh, no. I don't use plan mode anywhere.

Host

那么,但 GStack 肯定有一个审查步骤。你怎么——你对多少审查足够 versus 太多有什么哲学?比如我现在正和我的一个 cappy 队长辩论,我想把审查的 token 从 5000 万降到 1000 万。但它有所有这些保留的评估显示我们有 30 种类型的 bug,我试图确保在审查步骤中找到,它就像我用 10 做不到。我需要 30。就像我从 50 降到 30。我们能发布 30 吗?我说不不不你需要 10。

So then, but GStack must have like a review step then. How do you—What's your philosophy on like how much review is enough versus too much? Like I'm literally in a debate with one of my cappy captains right now where it's like I wanted to go from 50 million tokens for review to like 10 million. But it has all these held out evals that show that we have like 30 types of bugs that I'm trying to make sure I can find in that review step and it's like I can't do it with 10. I need 30. Like I went from 50 to 30. Can we ship 30? And I'm like no no no you need 10.

Lauren

你看到什么,我的意思是一个真正好的问题,我认为。所以对我来说,实际上在 pstack 我试图稍微淡化一些审查技能。你知道在 pstack 我们有一堆技能,比如有一个叫 interrogate,它会生成不同模型的智能体。所以像跨不同家族,效果很好,你可以做对抗性审查,那在几个月前相当热门。

What do you see like I mean what's a really good question I think. So for me actually in pstack I have tried to downplay some of the review skills a little bit. You know in pstack we have a bunch of skills like there's one called interrogate and that will spawn agents of different models. So like across different families which works really well you can do like adversarial review is that was like quite hot like many months ago.

智能体反馈与过拟合 Agent Feedback and Overfitting

Lauren

我其实有点从那种做法中退出来了,因为我发现当你那样做时,它仍然有价值,但我认为你冒着过拟合的风险,智能体没有能力或者可能没有足够的判断力来确定什么是好的反馈,什么是不好的反馈。而且它可能对你的目标、意图和更广泛的产品愿景没有完整的了解。例如,它不理解你希望产品如何发展。所以你可能达到一个点,非常抽象地说,就像想象在智能体之前,你有一群工程师在讨论做这件事的最完美理想乌托邦方式,但实际上那甚至不重要,因为两个月后我们计划废弃这个东西并构建一个全新的愿景。

I've actually kind of withdrawn a little bit from that because I find that when you do that, it's still valuable, but I think you run the risk of overfitting, where the agent doesn't have the ability or maybe enough judgment to determine what's actually good feedback versus not good feedback. And it maybe doesn't have complete knowledge of your goals and your intents and your wider product vision. It doesn't understand how you want the product to evolve, for example. And so you can get to a point where, very abstractly, it's kind of like imagine before agent you have a bunch of engineers talking about the most perfect ideal utopia way of doing this, but actually that doesn't even matter because in two months we plan to deprecate this thing and build a completely new vision.

Host

对。

Right.

Lauren

所以我认为那里有一些类似的类比,我认为一些审查是好的,但我尝试更多地转向——就像在我屏幕上的这个倒金字塔,我尝试推动你提到的减少 token 消耗。所以从工程的角度来看,我喜欢把这看作你的代码库是一种记忆形式。所以如果你重构你的架构、API 设计,或者你有静态分析、lint 规则、编译器——所有这些语言、编程语言,这些在过去都是为了帮助人类写出好代码而存在的。毫不奇怪,它们也帮助智能体,对吧?所以当你能把一些你正在和你的队长争论的事情编码回代码库时,那就像你现在少了一个需要思考的问题。

And so I think there's some similar analogy there where I think some review is good, but I try to move more towards—like on my screen here I have this inverted pyramid where I try to push like you mentioned spending less tokens. And so from the engineering point of view, I like to think of this as your codebase is a form of memory. And so it's much more token efficient if you refactor your architecture, your API design, or you have things like static analysis, lint rules, compilers—all of these languages, programming languages, these all exist to help in the past like humans write good code. And unsurprisingly, they also help agents, right? And so when you can encode some of these things that you are arguing with your captain about back into the codebase, that's just like one less problem you now have to think about.

Host

你刚刚给了我一个功能想法,因为 GStack 有办公时间和 CEO 审查,就像你在谈论产品。我可以利用那个。我的意思是,因为那些是产品需求,会输入并保存到项目中的 markdown 文件。所以我可以让审查技能——我认为它已经做了一点——但我可以说,‘不,不,不。真正利用那个来确定我们应该审查什么,以及什么算作真正需要修复的 bug,什么不算,因为有些一般正确性 bug 就是无关紧要。它们可能只发生 0.1% 的时间,或者可能永远不会。’

You just gave me a feature idea which is like because GStack has office hours and like the CEO review where you're like talking about the product. I can use that. I mean because those are like product requirements that feed in and those saved to markdown in the project. So I could have the review skill. I think it already does this a little bit, but I could be like, 'No, no, no. Really use that to figure out what we should review and what counts as a real bug to fix versus not because there are like general correctness bugs that are just like they don't matter. Like they might only happen like .1% of the time or maybe never.'

Lauren

嗯。

Mhm.

Host

所以听起来这个三角形告诉我们代码库相当大,而你的时间非常少。

So it sounds like this triangle shows us that the codebase is quite large and then you have very little time.

Lauren

是的,我想是这样。所以我故意这样画。当然,我想这不是一个很科学的图表,但当我画它时,我想象的方式是——我认为这是关于作为工程师你应该把时间花在哪里。所以对我来说,大部分投资都投入到代码库,然后是静态分析,我还用规则和技能、包更新和 agents.md 来分层。我们仍然时不时做一些人工审查。我认为如果你在处理基础设施或数据库部署等更关键的部分,更敏感的事情,那么你可能仍然希望保持在循环中。但我认为你的大部分常规产品开发或软件维护——这里的投资实际上最终会节省你很多 token。因为例如,对于 Grok Bot 桌面应用,它是一个 Electron 应用,我们从构建 Cursor 中学到的一个教训是 Electron 上的渲染线程对长任务非常敏感。

Yeah, I think so. So I intentionally drew it this way. I guess it's not a very scientific chart, of course, but the way I imagined it when I drew it was—I think this is in terms of where you should spend your time as an engineer. So for me, most of the investment I make is into the codebase and then static analysis, and I still layer that with rules and skills, pack updates, and agents.md. And we still do a little bit of human review from time to time. I think if you're working on more critical parts of infrastructure or databases deployment, more sensitive things, then you might actually still want to stay in the loop. But I think a large class of your regular product development or software maintenance—an investment here actually ends up saving you a lot of tokens. Because like for example, with the Grok Bot desktop app, it's an Electron application, and one of the lessons we learned from building Cursor is that the renderer thread on Electron is very sensitive to long tasks.

Host

哦是的。

Oh yeah.

Lauren

是的。所以如果你的任务超过 16 毫秒或 8 毫秒,你可能会得到非常卡顿的性能,对吧?如果你想达到 60 fps 的流畅度,我们 literally 在重构和构建 Grok Bot 时就构建了那个。我们决定不想再思考这个问题。我们不想在代码审查中捕捉它。我们实际上会把它构建为 lint 规则。我们构建了一个依赖图检查器,现在它只是由代码库强制执行,由签入强制执行。所以这是我不必再花费的 token。

Yeah. So like you can get very janky performance if your tasks take more than like 16 milliseconds or 8 milliseconds, right? If you want to hit a 60 fps smoothness, we literally just built that for when we were refactoring and building Grok Bot. We decided we don't want to think about this problem anymore. We don't want to have to catch it in code review. We'll actually build it as a lint rule. We built a dependency graph checker and now it's just enforced by the codebase, enforced by check-in. So it's tokens I don't have to spend anymore.

Host

哈。它以某种方式强制执行静态分析,实际上就像可以知道某件事运行需要多长时间。

Huh. It's somehow enforcing static analysis actually like can know how long something will take to run.

Lauren

哦不。更像是这样,对于 Grok Bot 应用,我们有运行在主线程和渲染线程上的代码。所以我们明确禁止在这两者之间导入。所以不一定是主代码线程总是很长。只是像很多时候,一些较重的任务不应该阻塞渲染器。

Oh no. It's more like so for the Grok Bot app we have code that runs on the main thread and then the renderer thread. And so we explicitly forbid importing between those two things. So it's not necessarily that main code thread will always be long. It's just like often times, some heavier tasks that shouldn't block the renderer.

Host

永远不要在渲染线程上做长工作,伙计。

Never do long work on the render thread, man.

Lauren

是的。你希望你的渲染尽可能轻量,你知道。所以,是的,那是我们不再思考的事情。而且我认为应用的性能实际上一直非常稳定和非常好,这对于许多大部分由 AI 编写的应用来说实际上相当令人惊讶,它们开始时很快,然后随着时间的推移,它们只是积累了很多垃圾和性能问题。

Yeah. You want your render to be as light as possible, you know. So, yeah, that's something we just don't think about anymore. And I think the performance of the app has actually stayed very consistent and very good, which is actually quite surprising for a lot of majority AI written applications where they start out fast and then over time they just accumulate a lot of slop and performance problems.

Host

但有了正确的技能,有了正确的设置,我认为你实际上可以——你可以有质量,也可以有数量。我认为你可以鱼与熊掌兼得。现在这个时刻的酷之处在于,你基本上可以——以前存在权衡,你只能三选二。要么快、便宜、高质量。现在你基本上可以三者兼得,我们现在可以用 AI 做出我们一生中最好的工作,这——

But with the right skills, with the right setup, I think you can actually—you can have quality and you can have quantity. I think you can have your cake and eat it too. The cool thing about the moment in time right now is that you could basically—there used to be the trade-off before that you could only choose two out of three. It was either fast, cheap, and high quality. Now you can basically do all three and we could basically do the best work in our lives right now with AI, which is—

Lauren

是的,我完全同意。我认为快、便宜和好肯定更有可能。我认为达到那个点并不容易,但根据我自己的经验,我有幸在一个实验室工作,给了我无限的 token。所以我有点把自己看作探索者,试图弄清楚 6 个月到一年后软件工程会是什么样子。我认为这个过程对我来说非常有启发性,关于如何实际保持这个工作流程高效,因为我认为这是很多团队和人都在思考的话题——token 的成本——我认为当你采用像我这里展示的倒金字塔时。

Yeah, I totally agree. I think fast, cheap, and good is certainly more possible. It's not easy, I think, to get to that point, but from my own experience, I have the privilege of working at a lab which has given me unlimited tokens. So I sort of see myself as the explorer trying to figure out what will software engineering look like 6 months to a year from now. And I think this process has been very enlightening for me in terms of how to actually keep this workflow efficient, because I think it's a topic that lots of teams and people are thinking about—the cost of tokens—and I think when you adopt an inverse pyramid like I'm showing here.

AI经济学与软件丰裕 The economics of AI and software abundance

Host

我认为你实际上可以达到一个非常好的平衡,普通的工程任务、高质量的工程工作不一定非要花大价钱。它可以相当经济、相当高效。更不用说前沿模型越来越聪明,智能 token 的单位成本也在下降。如果你把去年的模型和现在的模型相比,输入输出 token 其实已经不一样了。现在一个前沿 token 的能力远超去年的一个 token。即使你使用较小的模型,那个模型也比一年前的大模型好得多。所以从这个角度来看,我们正走向一个软件乃至其他领域真正富足的时代。

I think you can actually get to a very nice mix where the ordinary engineering tasks, high-quality engineering work, doesn't have to break the bank. It can be quite economic and quite efficient. Not to mention that the frontier keeps getting smarter and smarter, and the unit cost of intelligence tokens is coming down. If you compare a model from last year to a model now, the input-output tokens are not really the same. One frontier token now is way more capable than a token from last year. Even if you use one of the smaller models, even that model is way better than what you had a year ago with the big models. So from that perspective, we're heading towards a period of real abundance with software and maybe other domains as well.

Roman

我认为我们在负责 Grok Bot 的核心团队中看到了这一点。Lauren 每月提交 4,000 个 PR。我认为我们会开始看到更多工程师这样做。我们开始看到瓶颈不再是投入开发这些想法和功能的时间或工程资源。实际上,更多的是我们团队之间的对齐,比如我们希望产品路线图是什么?我们的愿景是什么?什么是好的?什么是伟大的?我认为我们正在看到这一点,Lauren 和我进行了很多讨论,我们谈论 Grok Bot 的未来变化,有人提出一个建议,然后房间里的另一个人说:‘哦,如果你感兴趣,我有一个 PR。我周末一直在做这个。’这种情况经常发生。所以东西已经建好了,但我们不想全部合并,因为那会导致一个有点弗兰肯斯坦式的产品,夹杂着随机的副项目。我们确实需要一致的产品愿景和哲学。所以我认为我们现在把更多时间转移到这些问题上,因为每个人每天或每周都在做成千上万的微观决策,我认为过去不是这样。

And I think we're seeing this on our core team that works on Grok Bot. Lauren is putting up 4,000 PRs a month. And I think we're going to start to see more engineers do that. And what we've started to see as the bottleneck is less the time or engineering resourcing going into developing these ideas and features. Actually, it's more alignment among us as a team of like, what do we want the product roadmap to be? What's our vision? What does good look like? What does great look like? And I think we're seeing this where we've been in so many discussions, Lauren and I, where we're talking about a future change to Grok Bot and someone suggests something and then someone else in the room says, 'Oh, I have a PR for that if you're interested. I was working on this over the weekend.' And that happens constantly. So the stuff is built, but we don't want to merge all of that because it's going to lead to a bit of a Frankenstein-y product with random side projects. We do need a consistent vision and philosophy of the product. And so I think we're shifting more of our time to that and those questions now that everybody is making thousands of micro decisions a day or a week, which I think was not the case in the past.

产品决策中的品味与洞察 Taste and discernment in product decisions

Host

其实我非常好奇。我的意思是,现在有点流行说重要的是品味,这意味着实际上要对一堆东西说不,要有辨别力。对你们来说,这在实践中是什么样子的?你们如何决定是或否?

Which actually I'm very curious. I mean this is a bit of the trendy thing now with the thing that matters is taste which means actually saying no to a bunch of stuff and having discernment. What does that look like in practice for you guys? How do you make a decision yes or no on this?

Roman

是的。对于某些类别的事情。Lauren 设置了我们工厂中我最喜欢的部分之一,Sandy,它每天接收数千条客户反馈。我们通过产品接收,通过支持渠道接收,通过 X 接收。很多时候只是小 bug。所以在很多情况下,Lauren 的工厂只是修复它,比如重构、修复、合并。没人会多想。

Yeah. For certain category of things. So Lauren has set up one of my favorite parts of our factory is Sandy, which takes in thousands of pieces of customer feedback every day. We receive it through the product. We receive it through our support channels. We receive it through X. And a lot of times there are just small bugs. And so in a lot of cases, Lauren's factory just fixes it like refros, fixes, merges. No one thinks about it.

Host

它在凌晨 3 点自动发布。

It goes out automatically at 3 in the morning.

Lauren

我希望所有软件都能那样制作。我希望在 Apple 工作的人能做到这一点,因为嘿,尤其是日历团队,我对你们有些意见。

I wish all software would be made that way. I wish Apple people who work at Apple could do this because hey I have you know calendar team in particular I have some problems with you.

Lauren

Grok Bot 有很多部分应该以那种方式自我修复,很高兴这终于奏效了。我认为还有另一类问题完全不应该那样运作。如果你在改变用户每天体验的核心像素,作为用户,醒来刷新你的 Grok Bot,发现按钮移动了,然后第二天又移动了,这真的很令人沮丧。所以我们试图抵消这种情况的方法是大量内部吃自己的狗粮。在发布之前,让变更停留的时间比感觉舒适的时间更长。默认情况下,如果某个东西没有获得内部产品市场契合,我们就不发布。即使获得混合的内部产品市场契合,我们也不发布。所以作为一个团队,我们真的努力保持高标准。

There's a lot of Grok Bot that should be self-healing in that way and it's cool that that finally works. I think there's another set of questions that should not operate that way at all. If you're changing the core pixels that a user is experiencing every day, it's really frustrating as a user to wake up and you refresh your Grok Bot and there's a whole new—the button moved and then it moved again the next day. And so the way we try to counteract that is a lot of dogfooding internally. And sitting with a change for longer than feels comfortable before we ship it. And the default is if something doesn't get internal product market fit, we don't ship it. Even if it gets like mixed internal product market fit, we don't ship it. And so as a team, we really try to keep a high bar.

Lauren

是的,我认为这实际上是我在 Grok Bot 团队工作中最喜欢的部分之一,尽管我们速度很快,但我们仍然对用户体验有很多意图和关怀。即使你可以非常快,也不意味着你应该发布你能想到的每一个功能。我认为这就是人类判断力和品味发挥作用的地方。我个人不认为这会过时。我认为只要我们在为人类构建软件,人类就应该真正思考精心设计的体验,因为我们最了解彼此,对吧?如果你在为智能体构建软件,那是另一回事,你当然可以尽可能快,但我认为,对于人类,你确实需要对你开发的东西投入有意的关怀和思考。所以即使对我自己的工厂,我也不把它们转向功能开发。我更多地把它们转向软件维护方面,因为我非常关心性能。我非常关心修复用户遇到的 bug 或他们提出的生活质量投诉。我认为这实际上是智能体的一个很好的用途,因为那曾经是某人的全职工作,现在我把这个交给很多智能体,这释放了我和其他工程师,让我们更集中地思考产品体验。

Yeah, I think that's actually one of my favorite parts of working on the Grok Bot team is that, despite the great velocity that we have, we still have a lot of intention and care for the user experience. And even though you can go really fast, it doesn't mean that you should just ship every feature you can think of. And this is where human judgment and taste, I think, comes in. I don't think that's going to go out of style, personally. I think as long as we're building software for humans, humans should really be thinking about the intentionally designed experience because we know each other the best, right? If you're building software for agents, that's a different story that you can certainly probably go as fast as possible, but I think, yeah, for humans, you do want to have put intentional care and thought into the stuff that you develop. So even for my own factories, I don't turn them towards feature development. I turn them more towards the software maintenance side of things because I care a lot about performance. I care a lot about fixing bugs that our users encounter or quality of life complaints that they have. And I think that's actually such a great use of the agents because that used to be like a full-time job for someone and now I've given that to a lot of agents and that frees me up and other engineers up to think more intensely about the product experience.

Lauren从金融到AI工程的历程 Lauren's journey from finance to AI engineering

Host

Lauren,你的旅程中有一件非常鼓舞人心的事,我不知道有多少人知道,你在新加坡长大,你学的是非技术学位。你在澳大利亚获得了会计和金融学士学位。然后不知怎么的,你成为了现在最好的 AI 工程师之一。

One thing that's very inspiring about your journey, Lauren, I don't know if many people know is you grew up in Singapore and you studied a non-technical degree. You did a bachelor's in accounting and finance from in Australia. Then somehow you ended up becoming one of the now best AI engineers.

Lauren

是的。

Yeah.

Host

那是怎么发生的?我认为有时人们认为进入科技行业很难,但你做到了。你还参与了 React 编译器的工作,这也是最难的事情之一。

How did that happen? I think sometimes people think it's very hard to get into tech, but you did it. And you also worked on the React compiler which is one of the hardest things as well.

Lauren

是的。我确实觉得自己在科技行业有一段非常幸运的经历。我认为那是一个人们对自学开发者的想法更加开放的时代。所以长话短说,大学毕业后我确实学了金融。我最喜欢的课程之一实际上是财务建模,我们学习了如何使用 Excel 和 Visual Basic for Applications 来制作一些很酷的财务估值等等。通过那门课程,我意识到我构建电子表格和宏比做金融部分更有趣。我的意思是,金融部分也很有趣。但技术部分,逻辑思考,对我来说真的很有趣,而且是我以前认为自己不适合的事情。所以,大学毕业后,我和一个朋友、一个同学一起创办了一家初创公司。

Yeah. I think I've definitely feel like I've had a very privileged and lucky experience in the tech. I think that was a time where people were much more open to the idea of a self-taught developer. So I guess the long story short is after college I did study finance. One of my favorite courses was actually financial modeling and we learned how to use Excel and Visual Basic for Applications to make some cool financial valuations and things like that. And through the course of that course, I realized I had more fun building the spreadsheet and the macros than I did the finance part. I mean, the finance part was interesting, too. But the technical part, thinking logically, just was really interesting to me and something I hadn't really thought I was cut out for. So, after college, I started a startup with a friend, a classmate.

早期创业时光 Early Startup Days

Lauren

它没有走得很远,但我们从那个过程中学到了很多,我也真正融入了澳大利亚的创业社区。我就是这样得到了第一份工作。而真正让我成功的秘诀其实是 X。

It didn't go super far, but we learned a lot from that process, and I got really involved with the startup community in Australia. That's how I got my first job. And really, the secret that I attribute for my success is actually X.

Lauren

我开始的时候基本上还在学怎么编程。我有一个博客,只是分享我学到的东西。我很幸运,框架社区里有人——我当时用 JavaScript 框架 Ember 构建应用——

I started out when I was still basically learning how to code. I had a blog and I was just sharing what I learned. And I was fortunate enough that someone from the framework community—the JavaScript framework Ember that I was building my application in—

Lauren

所以是的,Tom Dale 和 Leah Silber 联系了我,说:‘你有兴趣在即将于俄勒冈州波特兰举行的会议上做个演讲吗?’我非常震惊,因为我这辈子从未参加过会议。而现在我被邀请在一个有几百人的会议上演讲。你知道,公开演讲一直是我最大的恐惧。但不知怎么的,他们说服了我去做。

So yeah, Tom Dale and Leah Silber, they reached out to me and said, 'Would you be interested in giving a talk at the conference that's coming up in Portland, Oregon?' And I was very shocked because I had never even been to a conference in my life. And then now I was being asked to talk at a conference with like hundreds of people. And you know, public speaking was always my worst fear. But somehow they convinced me to do it.

Lauren

我学到了能够传达你的想法并带领人们一起踏上这段旅程的能力。而这种讲故事和沟通的技能真的一直伴随着我,尽管我一直深入参与工程工作。而且我过去也有很多非常棒的导师。

I learned about the ability of being able to communicate your ideas and bring people along in that journey. And this skill of storytelling and communication has really stuck with me, even though I've been really involved with engineering work. And I've had a lot of really great mentors as well in the past.

Lauren

所以我想总而言之就是:做酷的事情,谈论酷的事情,然后在这个循环中不断做下去。也许星星会对齐,有人会注意到你在 X 上说的话并转发,给你一个你从未有过的机会。也许你发展你的想法,想出了一个创业点子,对吧?然后你去申请 YC,不知不觉中,你就能仅仅通过一条推文改变你的生活。

So I think all that to say is: do cool stuff, talk about the cool stuff, and just keep doing that in the loop. And maybe the stars will align and someone will notice something you said on X and retweet you, offer you an opportunity that you hadn't had before. Maybe you develop your idea and you come up with a startup idea, right? And then you go apply to YC, and before you know it, you could change your life with just a tweet.

Host

这是一个很好的结束点。我想 Lauren Roman,非常感谢你们加入我们。

It's a great place to end. I think Lauren Roman, thank you guys so much for joining us.

Lauren

太有趣了。

It was so fun.

Host

感谢你们做出了一个不可思议的产品。所以,迫不及待想看看你们从这里开始会构建什么。

Thank you for making an incredible product. So, can't wait to see what you build from here.

Lauren

感谢邀请我们。非常有趣。

Thanks for having us. It was a lot of fun.

互动版:逐字朗读 + 针对本期提问 →