本集简介
双语字幕
仅展示文本字幕，不包含中文音频；想边听边看，请使用 Bayt 播客 App。
你正在研究如何进行复合工程。
You're figuring out how to do compounding engineering.
团队里明明只有两个人，却感觉像有十五个人在工作。
There's two people on the team, but it really feels like there's 15.
用AI编程不仅仅是写代码那么简单。
Coding with AI is more than just a coding part.
无论是用于研究还是工作流程，AI应该被应用到所有领域。
Utilizing it for research, for workflows, it should be used for everything.
过去三周我们完全没碰过WinServe和Cursor。
We haven't touched WinServe or Cursor in the last three weeks.
这两者都具备智能编码能力，但Claude通过简化十倍流程更胜一筹。
Both of those have agentic coding capabilities, but Claude just takes it one step further by simplifying it by factor of 10.
我们最近完全放手让AI干活，我们只负责管理AI。
We've been really leaning into let's AI do the work for us, we're just managing the AI.
你们首先花时间构建了一个能有效生成其他提示词的提示词。
What you did first is spent time building a prompt that effectively builds other prompts.
这真的很酷。
That's really cool.
Kieran、Nitesh，欢迎来到节目。
Kieran, Nitesh, welcome to the show.
谢谢。
Thank you.
非常感谢邀请我们，Tom。
So much for having us, Tom.
能请到你们我太兴奋了。
I'm psyched to have you.
为不了解的观众介绍一下，你们两位都在开发Quora——这是Every公司的AI邮件助手。
So for people who don't know, both of you work on Quora, which is Every's AI email assistant.
Kieran，你是总经理。
Kieran, you're the GM.
Nitesh，你是工程师。
Nitesh, you're an engineer.
除了Cora本身是个非常棒的产品，我很期待将它推荐给所有听众和观众之外，邀请你们两位来做客是因为你们正在探索一种全新的工程开发模式。
And Beyond the fact that Cora is a really cool product, and I'm really excited to bring that to everybody who listens to this show or watches this show, I wanted to do an episode with the two of you because I think that you're figuring out a new way to do engineering.
虽然Quora团队只有两个人，但通过智能代理处理PR请求、开发分支代码、相互审核等机制，实际运作规模相当于15人团队，这完全颠覆了传统软件开发方式。
Because really, Quora has, you know, there's two people on the team, but it really feels like there's 15 because you've got agents who are pulling down PRs and working on branches, then you're pushing them and other agents reviewing, and it's just this kind of crazy thing that it's a new way to build software.
Kieran你之前说过一句让我印象深刻的话：你们正在实践'复利工程'——每完成一项工作都会让后续工作变得更轻松。
Kieran, you said something the other day that really stuck with me, which is you're figuring out how to do compounding engineering, so with each piece of work you do, you're making it easier to do the next piece of work.
我认为你们积累的经验对观众们非常重要，毕竟当我们拥有新工具时，就需要配套的新原则和新工作流程。
And I just think that it's really important to bring what you guys are learning to everybody that watches the show because it's like we have new tools, and so we need new principles and new workflows for using those tools.
所以非常期待能就此展开讨论。
And so I'm really excited to talk to you about that.
是的。
Yeah.
谢谢。
Thanks.
开发Quora很有趣，但更重要的是在Every这个能接触前沿工具、创新思维和工作方式的环境里，我们得以重新思考工程构建的本质。
It's really fun to build Quora, but like being part of every and like being in an environment where you get access to tools, like access to thinking, access to exciting new ways to to work really helps us rethink how we build.
这其实更像是一个实验。
So like, it's really an experiment.
我们正在打造Quora这个产品，但同时也在探索应该如何构建它。
We're building a product, Quora, but at the same time we're figuring out how we should build.
这真的非常有趣。
And that's super interesting.
我们正处于中间阶段，人们会问：你们怎么看这种新模式？
And we're like right in the middle where people say, what do you think of this new model?
比如，我们该如何使用这个研究工具？
Like like, how do we use this research tool?
我们只是在不断尝试各种方法。
And we're just trying things out.
我和娜塔莎最近几周确实感受到了一些变化，可以说事情正在改变，而且不止我们注意到了这点。
And Natasha and I, we've been like really feeling a shift in the last weeks, I would say, where we're like, like things are changing and we're not the only ones.
我们也听到其他人这么说，不过这样的人还不多。
Like we hear other people say that as well, but not a lot of people.
我们学到了很多，也想分享一些我们的发现。
And what we've learned is a lot and we want to share a little bit of what we learned.
而且我们知道，我们才刚刚起步。
And also what we know is like we're just barely starting.
我们只是触及了表面。
We're scratching the surface of this.
现在正发生一场巨大变革，新模型、人们思维方式、MCP等等，变化太多了。
And it's a big shift that's happening right now by new models, by how people think, by MCP, by, like, just it's a lot.
而且，是的，从不同角度讨论这个问题真的很棒。
And, yeah, like, it's great to talk about that from different perspectives.
是啊。
Yeah.
没错。
Yeah.
我同意。
I I agree.
我觉得在Every特别之处在于，每天Discord里都有新人说'我建了个AI代理'，问我们是否愿意在发布前试用。
And and I think it is so special to be at Every because we do like, every day, there's someone new in the Discord who's like, I built an AI agent.
你们想不想在我们正式发布前试用看看？
Do you wanna, like, use it before we launch it?
所以我们能提前接触到OpenAI的模型，有时还有Anthropic的模型。
And so, you know, we get access to OpenAI models before they come out and and sometimes anthropic models.
这样我们就有了先发优势，而你们特别擅长将这些模型整合到实际生产流程中。
And so we have this, like, early edge, and then and then you guys are so good at figuring out how to actually incorporate them into, like, a production process.
Kieran你刚才提到有些变化发生了。
So you said, Kieran, that something changed.
我想了解你认为具体改变了什么，以及你们目前初步形成的工作流程大致是怎样的。
So I guess I I wanna get a sense of what you think changed and what draw the broad strokes of what the workflow is that's starting to emerge for you guys.
对。
Yeah.
对我来说，虽然所有环节都在整合，但最大的转变是我意识到用AI编程远不止写代码本身。
For me, like, obviously, it's it's like everything coming together, but I think the biggest thing is a realization in myself that coding with AI is more than just a coding part.
这实际上关乎如何将其用于研究、工作流程等方方面面。
And it's really about like utilizing it for research, for workflows, for everything.
它应该被用于所有事情。
Like it should be used for everything.
我们现在已经达到一个阶段，智能体足够强大，可以真正处理所有任务。
And we're now at a point where the agents are good enough that they can actually do everything.
所以我们需要重新思考，比如老派的编码方式虽然很棒。
So we need to rethink again like, hey, cursor, windsurf, like the old school way of coding was great.
那种更有氛围感的编码方式，算是前进了一步。
Like more of the vibe coding, like that was one step.
而现在我们意识到，其实只需给出任务，它就能完成。
And then now it's the realization, oh, actually we can just give a task and it will do it.
但工作仍然需要由我们决定具体做什么。
But still the work needs to be done by like, what do we do?
我们该怎么做？
How do we do it?
我们更应该顺应这个趋势并深入探索。
And just a realization that we should lean into that more and really go deep.
就像Cloth Coat那样。
And and it's like cloth coat.
现在已经有优秀的编码智能体，它们能配合Cloth这类擅长遵循指示的新模型工作。
Like, it's just good coding agents or agents available that actually start to work with new models like cloth for like really good at following directions and instructions.
这一切正在融合成型。
And it's all that's coming together.
我意识到，哦，我们已经到达了。
That's that I realized like, oh, we're here.
未来已经来临。
Like the future is here.
我们一直在谈论的这场关于代理进化的变革。
This thing we've been talking about that was going to be the agentic evolution.
突然间它成功了，而且是在现实世界、非实验性的环境中运作。
Suddenly it works and it's working in real world, non experimental playing thing.
就像我们正在开发一个应用，而且它确实能帮我们开发这个应用。
It's just like we're building an app, and it's working building the app.
所以我听到的是，这不仅仅是关于用AI开发，而是将AI应用于开发过程中的所有环节，而你们在这方面最依赖的工具就是Cloud Code。
So what I'm hearing is like, it's not just about developing with AI, it's all the things that go into developing that you're using AI for, and that the the thing that you're using the most for this is Cloud Code.
是这样吗？
Is that right?
如果确实如此，能否为不了解或未使用过Cloud Code的人简单介绍一下它，然后详细说明你们具体是如何使用它的？
And if it's right, like, tell tell me about, like, for people who don't know Cloud Code or haven't used it, give us a little introduction to Cloud Code, then tell us about, like, exactly how you're using it.
是的。
Yeah.
Cloud Code本质上是Anthropic推出的编程代理版本，底层使用了云计算技术。
Cloud Code is basically the coding agent version from Anthropic that uses Cloud under the hood.
它以CLI工具的形式在终端运行，这种设计...
And it runs in your terminal as a CLI tool, which is kind of
你想分享一下屏幕给我们演示一下吗？
Do you wanna share your screen and like show us?
是的。
Yeah.
是的。
Yeah.
ClothCode是一个你在终端中使用的工具。
So ClothCode is a tool that you use in your terminal.
我知道对于非技术人员来说，这听起来有点吓人。
And I know for nontechnical people, this is like, oh, this is scary.
但我已经说服了一些非技术背景的朋友使用ClothCode，他们反馈说这太棒了。
But I've converted friends who were not technical to use Clothcodes, and they were like, oh, this is great.
但其实它非常简单。
But it's really simple.
你只需启动终端，输入Cloth，界面就会弹出。
You just hit you start your terminal, you say Cloth, and an interface will pop up.
简单来说，对于只听不看演示的听众们，他现在正在终端里操作。
And basically, for people who are who are listening instead of watching, he's in his terminal.
就是那种经典的黑屏界面，感觉像在用DOS系统，他刚刚输入了Claude。
It's, you know, the the classic black screen that you're you know, it feels like you're using DOS or something, and he just typed Claude.
然后屏幕上就出现了'欢迎使用Claude code'的提示，还有个可以输入命令的小文本框。
And then we just got a a thing that says welcome to Claude code, and there's a little text box for him to type in any command.
没错。
Yeah.
那它有什么特别之处？或者说区别在哪里？
And why is this different or what makes this different?
它可以访问目录或计算机。
This has access to the directory or the computer.
所以它已经可以查看我电脑上的文件了。
So it can look through files on my computer already.
它能在我的电脑上运行程序。
It can run things on my computer.
它可以对网页进行截图。
It can take screenshots of websites.
它能够搜索网络。
It can search the web.
就像它拥有工具，但比普通布艺版本可用的工具要多得多。
Like it has tools, but way more tools than available in a normal cloth version.
这很重要，因为像构建东西这样的工程工作，你确实需要比基础工具更多的工具。
And that's important because engineering work like building stuff, you do need more tools than just like the basics.
你需要GitHub来查看需要构建什么、当前状态如何，或者CI/CD流水线做了什么，比如测试是否失败，所有这些功能集中在一个编程代理中，实际上让我能够拥有一个工作流程，或者说让代理来完成我实际要做的事情。
You need GitHub to see what you need to build or what the status is or what the pie, the CI the pipeline does, like do the test fail, like having all these things available in one coding agent actually makes it possible for me to have a workflow or like a thing I do actually be done by an agent.
这才是重点所在。
And that's that's the important thing.
实际上，复合词的出现是因为工作远不止编码——如果你和工程师交流，会发现大部分工作可能是编码，但也许只占20%，80%的工作可能是确定下一步该做什么，或者理解人们的反馈意见及其解读方式。
Like, really the compound word comes in by doing more than just coding because lots of like if you talk to an engineer, like most of the work is maybe coding, but maybe it's actually 20%, maybe 80% of the work is like figuring out what to do next or understanding what people like, what their feedback is and how to interpret it.
在这里你可以做的事情是，比如一个有趣的方式是使用它来查询'我们上周发布了什么内容？'
And what you can do here is you can, for example, like a fun way is like to use it to say, let's say, what did we ship in the last week?
所以它确实知道很多事情。
So like it knows stuff.
所以我在询问它我们发布了什么，它很可能会查看Git日志，因为那是我们追踪实际发布内容的方式。
So I'm asking it what we shipped and it will most likely look at the Git log because that's how we track what we did ship.
是的，它会浏览Git日志，查看我们合并到主分支的内容。
And yeah, so it looks through the Git log, it looks at what we merged to main.
没错，这是一种很有趣的使用方式。
And yeah, that's a fun way to use it.
例如，我们可以将其用于产品营销，它会显示：哦，这些是修复的bug、简略的跳过功能、聊天面板状态、邮件摘要、XML标签、主要功能、简略的健康监测、时区自动检测。
And for example, we can use this for product marketing and it says, oh, these are the bug fixes, a brief skip functionality, chat panel state, email summary, XML tags, major features, brief health monitoring, time zone, auto detection.
这些都是我们发布的内容。
These are all things we released.
现在我可以说了
And now I can say
而且它用漂亮的文笔写道
And it's written in nice write up that
任何人都能读懂。
anyone can read it.
是啊。
Yeah.
实际上对两个人来说工作量很大。
And it's actually a lot for for two people.
这里有什么来着？
There's there's what?
大概有六个主要功能、五个重要bug修复和三个基础设施更新？
Like, six major features and, like, five important bug fixes and three infrastructure updates?
比如说，这工作量可不小。
Like, that's a lot.
是的。
Yes.
确实，工作量很大。
It it's a lot.
而且这周我们一直在尝试让AI替我们完成工作，我们只需要管理AI就行。
And and like this week we've been really leaning into like, let's AI, do the work for us and we're just managing the AI.
再举个例子，如果有人来问你‘这个项目的进展如何？’
One other thing, for example, is if if you have someone come to you oh, what is the status on this?
或者‘你们下周要交付什么？’
Or what are you going to ship next week?
让我们看看它会怎么做。
Let's see what it will do.
你能看看流水线上有什么，以及即将产出什么吗？
Can you see what is in the pipeline and what will come out soon?
这真是太棒了。
So this is awesome.
对了Nitesh，在讨论进行中，你随时可以加入发言。
And, Nitesh, while this is going, if you if you wanna jump in at any point, feel free to.
等会儿我会把话题抛给你，不过你也可以随时主动加入讨论。
At some point, I'll I'll lob it to you, but also, like, just I'll you know, feel free to jump in.
好的。
Yeah.
当然。
Sure.
是的。
Yeah.
我们拭目以待吧。
So we'll see.
比如，我不确定它是否有项目权限，但你应该明白我的意思。
Like, I don't know if it has project access, but, like, you you get the gist.
如果你把信息连接到代理上，使用起来会非常方便。
Like, if you have the information connected to the agent, it's very easy to use it.
使用熟悉的工具非常重要。
And it's very important to use a tool you're familiar with.
目前我认为Clothecode对我来说最合适。
And at this point, I think Clothecode works the best for me.
它最灵活的地方在于不仅能解决编码问题。
It is the most flexible because it doesn't only solve coding issues.
这很关键。
And that's important.
很多编码代理只能编程，但我想做的远不止编码。
Lots of these coding agents are made to code, but I want to do more than coding.
我希望它能成为整个工程领域的支持。
I want it to be like a support in engineering in general.
我觉得Klot团队确实考虑到了这一点。
And I think the Klot team really thought about that.
他们没有设计得过于具体，而是保持通用性，同时在实际解决问题、审视行为、思考错误并自我修正方面表现出色。
They made it not too specific and they kept it general while actually being really good at solving things and looking at what it did, thinking about the mistakes it made and self correcting.
正是这些难以实现的特质汇聚在一起，才使得它现在能够投入使用。
So that is stuff coming together that's very hard that makes it possible to use now.
是的。
Yeah.
在Cursor中进行编码与代理式编码有什么区别？
What's the difference between coding in cursor and agentic coding?
Claude代码与我们习惯的Cursor和Vinsurf有着本质上的简单区别。
Claude code is such a, simple departure from the cursor and Vinsurf that we're used to.
这两者虽然都具备代理式编码能力，但Claude通过简化流程将其提升了一个层级——我认为简化程度大约有十倍之多。
Like, both of those have agentic coding capabilities, but Claude just takes it one step further by simplifying it by, I I think, like, a factor of 10.
所以，Kieran之前提到的，Claude代码可能因为是个终端而让人感到害怕。
So, what Kieran was, telling earlier about how, Claude code may feel intimidating because it is a terminal.
但实际上，它比WinSurf和光标简单多了，因为这里除了一个文本按钮和文本框什么都没有。
But in reality, it is, like, so much simpler than the WinSurf and cursor because there is nothing except a text button, text box here.
就是说，没有命令键，没有快捷键，没有接受、删除、拒绝、移除这些选项。
Like, there's no no, like, command key, no shortcuts, no accept, delete, reject, remove.
真的什么都没有。
Like, there's nothing.
就只是一个文本框，它能工作是因为底层Claude模型现在能力强大多了。
It's just a a text box, and it works because the model, the underlying Claude model is so much more capable now.
所以它能运行更长时间并执行工具调用。
So it's able to work for longer and do tool calls.
所以它就像一个更简洁的用户界面，同时却更加强大。
So it it's it's like a simpler UI, which makes it at the same time more powerful.
尽管Cursor和Plotco背后的底层模型是相同的。
Even though even though, like, the underlying model behind like Cursor and Plotco is the same.
是的。
Yeah.
举个例子，今天早上我正在汇总一些指标。
And an example of this is this morning, I was pooling some metrics.
我当时想，为什么这个表单没有收到任何回复？
I was like, why didn't we get any responses to this form?
然后然后
And then And
为了提供背景信息，我们基本上有一个表单，询问人们如果不能再使用Chorus会有多失望，这样我们就能知道我们的表现如何。
for for for context, like, basically, we have a form that we ask people how disappointed they would be if they could no longer use Chorus so we can tell how how well we're doing.
你注意到我们有一个每周例会，会上我们会过一遍所有指标，我们知道你发现没人填写那个表单。
And you noticed we have a we have a weekly meeting where we go through all the metrics, and we know that you noticed that no one had filled out that form.
所以你进入Cloud Code，询问，嘿，为什么没人填写这个表单？
So you're going into Cloud Code, and you're asking, hey, like, why is no one filling out this form?
是的。
Yeah.
我当时想，肯定有什么问题，比如这个表单没发出去。
I was like, there has to be something like this form was not sent.
然后我问，嘿，十四天前是不是出了什么差错。
And I asked like, hey, fourteen days ago, something went wrong.
你能看出问题出在哪里吗？
Can you see what went?
它所做的就是列出一个待办事项清单，比如获取控制器最近的日志变更、搜索代码库。
And what it did, it made a checklist to dos like fetching recent log changes to the controller, searching the code base.
于是它查看了那个日期前后的变更，发现我们移除了一个添加人员的代码片段，就在这里，它提示说：嘿，其实你只需要加上这个。
So it looked through what changed around that date and it found we removed a piece of codes that adds people there, which is here, it says, hey, actually you just need to add this.
然后我说：好的，帮我处理一下。
And I said, okay, do it for me.
创建一个拉取请求。
Create a pull request.
它照做了。
And it did that.
我又说：哦对了，我还会写个脚本把之前遗漏的人员都加进去，完成迁移。
And I said, oh yeah, by the way, I'm also going to create a script that will then add everyone that we missed to it, migrate it.
就这样搞定了。
And that was it.
最妙的是，这根本没耗费我任何精力。
And the fun part was like, didn't it didn't cost me any energy.
就像在GitHub上随手记个待办事项那么简单。
Like, it was as easy as me writing it down in GitHub to look at later.
我根本不需要动手。
I don't need to.
我只要开口，它立刻就办妥了，这体验太棒了。
I just ask it and it does it immediately, which is really nice.
这就像收件箱清零一样，用不了五分钟就能搞定吧？
It's like the inbox zero, does it take less than five minutes?
就是做这类事情。
Do it kind of thing.
对。
Yeah.
我认为人们可能没有完全意识到的是，在没有AI的情况下，这项任务可能需要花费三十分钟到几个小时不等的时间。
I think the thing that people may not fully realize is that that's a thing that task could take anywhere from, like, thirty minutes to a couple hours without without AI.
不仅如此。
And it's not just that.
它还需要你集中精力，专门腾出时间坐下来完成。
It would require you to, like, focus on it and, like, put aside time to, like, sit down and do it.
而现在你只需要像这样发出请求，然后可以一个接一个地继续发送。
And now you just sort of, like, send off requests like that, and then you can send off another one and another one.
你可以让很多这样的任务并行处理。
You have a bunch of these sort of working in parallel.
所以具体描述一下这个场景吧，你实际的工作流程是怎样的，你具体在做什么，开了多少个标签页，你自己有没有手动编码，是不是同时处理五个任务，还是只用Cloud Code，给我们讲讲这些细节。
So give me a snapshot of what that looks like concretely, what your actual workflow is, what are you actually doing, how many tabs do you have open, like, are you actually doing any hand coding yourself, do you have like five in parallel, are you just using Cloud Code, like, give give give us a give me a sense of that.
好的。
Yeah.
我也会给你看看我的屏幕。
I'll show you my screen as well.
也许Nitesh你可以说说我们之前做了什么，比如当我们早期获得Cloth权限时，我们当时兴奋地做了哪些尝试。
Maybe, Nitesh, you can you can tell what we did before, like when we got early access to Cloth, like we were excited what we did.
我来分享我的屏幕。
I'll show share my screen.
对。
Yeah.
对。
Yeah.
所以，这大概是在直播前一天的情况。
So, this is like, like one day before the live stream was scheduled.
我们当时就想，好吧，明天一切都会改变。
We were like, okay, tomorrow, is gonna change.
我们将拥有一个更强大的模型，可以满足我们所有的需求。
We'll have a much more capable model, which we'll be able to watch for everything that we want.
基本上我们就像获得了一个编程精灵。
And we're basically gonna get like a coding genie for us.
所以今天我们最有效率的事，不是按常规流程工作，而是应该开个两小时会议，列出一大堆我们希望明天那个更强大的模型能解决的问题。
So the best, most productive thing for us to do today, instead of doing our regular studio programming, we should just jam for a two hour call where we make a massive list of issues that we want the future, like tomorrow's superior model to solve.
我们确实这么做了。
And we did that.
我们创建了大约20个任务项，包括需要修复的问题、计划开展的工作，并为新云端模型做好了系统准备。
We we created like 20 issues in terms of, you know, like, what we want to fix, what were the things that we were planning to work on, and prepared the system for the new cloud model.
对。
Yeah.
有趣的是，Nitesh让ChatGPT说：嘿，明天我们就要实现通用人工智能了。
And it was funny because Nitesh had like he prompted ChatGPT to say, hey, tomorrow we have a we we reached AGI.
你能帮我们列出所有需要做的事，并准备好AGI来解决我们完成的所有问题吗？
Can you can you but yeah, can you help us come up with everything we need to do and like prepare the AGI to solve everything we did?
然后我们将其输入Anthropic的提示优化器，接着将其作为提示使用，最终我们创建了
And then we fed that into the prompt improver of Anthropic, and then we used that as a prompt, and we created
等等。
Wait.
在你继续之前，先别急着往下说。
Before you move on before you move on.
对于正在收听的听众来说，基本上你们在GitHub看板里有个类似Trello的板块，每项你们确定要完成的任务，看起来都有一份详细文档，无论是功能开发还是bug修复，都详细说明了具体内容和实施方法。
So so for people for people who are listening, so basically, you have this sort of Trello board type thing inside of GitHub Kanban board, and for each thing that you've identified as what you wanna do, it looks like you have a document that lays out in detail, okay, if it's a feature or it's like a bug fix or whatever, lays out in detail what it is and how to actually do it.
你能打开其中一份文档看看吗？
Can you open up one one of them?
好的。
Okay.
比如有个功能是希望AI生成合成数据，这份文档从问题陈述到解决方案构想，再到所有需求和技术要求，包含了一大堆内容。
So like a feature is you wanna generate you wanna have AI generate synthetic data, and it has this document has everything from a problem statement to, like, a solution vision to all the requirements and all the technical requirements and, like, a bunch of a bunch of stuff.
而且看起来甚至包含了带天数估算的实施步骤之类的细节。
But it's and even it has seems like it has implementation steps with day counts and stuff like that.
这还挺有意思的。
So Which is funny.
所以这是个
So this is a
很多变化发生在好吧。
lot changes in the Okay.
是的。
Yeah.
是的。
Yeah.
是的。
Yeah.
所以我们使用封闭代码，并定制了这个生成的提示词来创建这些内容，因为手动创建确实工作量很大。
So this we use closed code and we have this custom prompt that we generated to create these because, like, it's a lot of work to create these.
即便用JetGPT，也需要查看大量代码步骤，必须
And even with JetGPT, like, there's a lot of steps you need to look at all the codes, need
你想到一个方案
to come You up think a have
仔细考虑。
about it.
没错，需要大量思考，所以要做好确实很难。
Yeah, there's a lot of thinking, so it's really hard to do well.
我们的做法是在Clothecode里创建了一个命令，这种命令就像你常用的自定义提示词，而我们的命令类似这样：嘿，这里有个
So what we did, we created an command in Clothecode, a command is kind of a custom prompt that you use a lot, and ours is like, hey, there's a
功能 抱歉，这个命令是在ClothCode里还是Cursor里？
feature Sorry, is a command in ClothCode or a command in Cursor?
因为你开着Cursor对吧，
Because you have Cursor Yeah,
我确实开着Cursor因为这是我编辑文件的方式，但命令是在ClothCode里的。
have Cursor open because that's how I edit files, but it's ClothCode.
如你所见，我们正在使用Cloth，我可以通过输入CCY命令来操作，这是布料代码。
So you can see we're in Cloth, and and I can use this command by hitting CCY, which is cloth codes.
然后我会描述遇到的问题，比如一个bug、一个问题，任何情况都可以。
And then I say something like a problem I have, like a bug, a problem, anything.
所以操作起来非常顺畅。
So it's very low friction.
我有这些CCI命令，当时我和娜塔莎正在即兴讨论。
So I have this CCI commands and Natasha and I were just jamming.
我们就在想，如果这样做会怎样？
We're like, oh, what if we do this?
哦，听起来很酷。
Oh, that sounds cool.
然后语音转文字功能就启动了。
And then a voice to text and it starts.
那么让我们看看这个功能如何运作。
So so let's let's see how this works.
在程序运行的同时，我们可以检查其他内容。
And then while it's running, we can go over the thing.
我想要Quora实现无限滚动功能，当我读完一个简报时，自动加载下一个未读简报，直到所有未读内容都加载完毕。
So I want infinite scroll in Quora where if I am at the end of a brief, it should load the next brief and it should go until every brief that's unread is read.
就像这样
So like
是啊。
So yeah.
我只是希望大家明白，基兰几乎从不打字，全靠语音转文字。
I I just want people to understand, like, Kieran almost never types anything and and does all voice to text.
所以他当时正通过终端使用语音转文字功能，我相信是在Cloud Code里用了一个尚未发布的内部孵化项目Monologue，她是该项目的第四大用户，但目前仍处于保密阶段，不过在这里可以小小预告一下，即将上线。
So he was just doing voice to text into his into his terminal, into Cloud Code with, I believe, an internal, as of yet unreleased internal every incubation called Monologue, which she is the number four biggest user of, but still still under wraps, but but, you know, a little preview in here coming soon.
基本上，看起来它的功能是——它是在把语音转换成我们之前看到的那份文档，还是说它实际上会去执行那些指令？
And, basically, what what it seems like it's doing is it's it's taking that is it turning is it turning that into the that document that we were looking at earlier or is it actually going and executing it?
对。
Yeah.
它的工作机制是：会把我说的内容插入到未来的描述中，然后按这些步骤执行。
So what it does is it will insert whatever I said here in the future description, and then it will follow all these steps.
这些步骤包括：研究，研究最佳实践。
And these steps are research, research best practices.
第一步是立足于代码库本身。
So one is grounding itself in the code base.
先研究现有内容，再研究最佳实践。
So researching what exists, then it's researching best practices.
也就是搜索网络，寻找开源模式。
So it's searching the web, finding open source patterns.
相当于从整体上遵循最佳实践来奠定基础。
So it's like grounding it in like best practices in general.
接着它会呈现一个计划。
Then it's will present a plan.
当我说‘好的，听起来不错’时。
And when I say, yep, sounds good.
比如，我喜欢这个计划中的人工审核环节，因为它有时会出错，但大多数时候是正确的。
Like, I like that review human in the loop for the plan because sometimes it does it wrong, but most of the time it's right.
然后我说，嗯，听起来不错。
And I say, yep, sounds good.
接着它会在GitHub上创建问题，并将其分配到正确的任务队列等等。
And then it creates the GitHub issue and it will put it in the right lane and all that.
哦，有意思。
Oh, interesting.
所以就像我们之前在GitHub上看到的看板系统，你创造了一种方式让你可以通过语音将功能需求输入Cloud Code，然后它会完成所有调研生成详细文档，最后直接添加到GitHub issues里。
So it's like that whole Kanban we were looking at in GitHub earlier, you've created a way for you you speak your feature into Cloud Code, and then it does all the research to create that long document, and then just adds it into into GitHub issues.
这真的很酷。
That's really cool.
是啊。
Yeah.
这是个重要步骤，因为这与光标编程不同——在光标编程时通常会跳过这步，因为工具本身不是为此设计的，那些工具主要是用来写代码的。
It's it's an important step because it's like, this is different from cursor coding because in cursor, normally you skip this step because the tool is not really made for it, like the tool's made to code.
当然你也可以创建Markdown文件之类的。
Yes, you can do create you can create markdown files and all of that.
但我们应该充分利用现有的问题追踪系统，它运作良好、用户广泛，并且已经与现有工作模式无缝衔接。
But let's lean into like an issue tracker that exists and it works well and people use it and it already hooks into existing like, patterns.
就像，我们可以把这个交给开发人员，他们就能直接实现它。
Like, it like, we we can give this to a developer and they can implement it.
没错。
Yeah.
需要特别指出的一点是，当你运行这个系统时——我们第一次看到Opus四代时的反应简直是‘天哪’——它能在完全无人干预的情况下持续运行，并给出相当不错的结果。虽然我们之前也接触过一些代理类系统，但这次在自主性和质量上完全是前所未有的突破。
And one of the things that just to point out is, like, you're running this, and I I think one of the special things that when we saw Opus four for the first time, were like, holy shit, is that it just runs forever with without any intervention and then gives you a pretty good result, which we've had sort of agentic type things for a little while, but it it's just a way different level of autonomy and quality than than we've ever had before.
而且它就像是在逐项核销待办事项清单，我认为其他代理循环系统在这方面会显得远不够彻底。
And and it's like just checking things off of this to do list in a way that I think other agent loops are just gonna be a lot less thorough.
是的。
Yes.
完全同意。
Absolutely.
我和Kieran在那儿搞了个有趣的比赛。
Me and Kieran have like a fun thing going on there.
我们正在比拼谁能让ClorCode运行最长时间。
We're trying to see who can have ClorCode running for the maximum amount of time.
Kieran现在暂时领先。
Kieran is stopping the list right now.
他跑了25分钟。
He Twenty five minutes.
持续运行了25分钟。
Ran it for twenty five minutes.
我现在才跑到8分钟。
I'm only at eight minutes right now.
哇塞。
Oh, man.
Kieran，你是怎么让它跑这么久的？
How, how did you get it to go so long, Kieran?
一个非常、非常长的计划，包括是的。
A very, very long plan, includes yeah.
这只是一个非常、非常复杂的漫长计划，还包含大量测试，要确保它能运行并通过所有测试。
It's just very, very complicated long plan and also include a lot of tests and just make sure that it runs all the tests and fixes all the tests.
而且很有趣。
And Interesting.
它持续了相当长时间。
It goes pretty long.
是啊。
Yeah.
嗯哼。
Mhmm.
带我过一遍，等等。
Take me through wait.
我想弄明白。
I wanna understand.
你是怎么制作那个能生成提示的提示的？
How did you make that prompt that that creates the prompt?
或者说，那个能生成研究文档的提示？
Or, like, the prompt that creates the research document?
所以，你是怎么知道该放哪些元素的？
So, like, how did you know which elements to put in?
你是不是用了同样的方法，就是用了那个Claude提示优化器，Anthropic的提示优化器来制作的？
Did you just use did you just do the same thing where you you use, like, a the Claude prompt improver, the anthropic prompt improver to make that?
或者，是的。
Or, yeah.
你是怎么想到要把这些组合在一起的？
Why how did you think about putting that together?
是的。
Yeah.
这就像是复合效应的一部分。
This is part of, like, the compounding effect.
就像是一个能产生很多很多结果的想法。
It's like having an idea that has, like, a lot of a lot of outcomes.
这就是尼泰什发给我的内容。
So this was what Nitesh sent me.
他说，我们刚刚获得了通用人工智能。
He said, we just got AGI.
它已经交付，我们可以编写软件了。
It got delivered and we can write software.
这...这就是你最初的提示，还挺有趣的。
That that that this was your initial prompt, which is kind of fun.
就像，非常戏剧性。
Like, like it's very dramatic.
然后ChatGPT说，我准备好了。
And then ChatGPT said, I'm ready.
好的。
Okay.
那就这么做吧。
So now do this.
我当时就想，好吧。
And I was like, okay.
是啊。
Yeah.
那没关系。
That's fine.
那没关系。
That's fine.
但你知道Anthropic控制台的提示优化器吗？
But like, do you know the Anthropic Console prompt improver?
你可能会问，那是什么？
You're like, what is that?
好吧，不知道的人注意了，这是...哦，他们改版了。
Well, anyone that doesn't know, this is the Oh, they changed it.
这是控制台吗？
Is this console?
对，是的。
Yeah, it is.
哦，没错，确实是。
Oh, yeah, it is.
好的。
Okay.
他们稍微调整了一下，但这很棒，因为你基本上可以粘贴一个提示词之类的东西，然后说‘对，我们正在思考’，点击生成，它就会自动优化提示词。
They changed it a little bit, but this is great because basically you paste in a prompt or something like that and you can say, yeah, we're thinking and you click generate and it will improve the prompt automatically.
你会想，这能有多好呢？
And you think like, how good can it be?
效果相当不错，因为操作起来也很顺畅。
It's pretty good because it's also very low friction.
所以就像，花一分钟看看能不能出结果、是否有效，非常简单。
So like, it's very easy to just take a minute to see if something comes out, if it works.
如果不行，删掉就好。
If it doesn't work, delete it.
无所谓。
Doesn't matter.
我们当时正在即兴讨论，想着要列出30个研究任务，所以觉得最好准备个提示词。
We were just jamming, and we were like, well, we're going to come up with 30 research tasks, so, like, we better have a prompt.
我就直接复制了这个提示词。
So I just copied this prompt.
然后这就成了正式文档。
And that became the document.
是啊。
Yeah.
好的。
Okay.
粘贴到这里并修改参数，然后通过输入斜杠来触发这些子句，我们这里还有两个自定义提示词。
Into here and change the arguments, and then you can trigger those in clause by doing slash, and we have these two custom prompts here.
然后我觉得这让我更清楚地理解了什么是复合工程，因为你最初花时间构建了一个能有效生成其他提示的提示，那些研究文档本质上就是为云代码准备的提示。
And then I think that actually gives me a much better idea of, like, what you mean by compounding engineering, because what it says to me is what you did first is spent time building a prompt that effectively builds other prompts, because those research documents are effectively prompts for Cloud Code.
所以现在你有了一个能生成提示的提示，每次想开发新功能时，需要指定的内容就减少了。
And so, now that you have a prompt that builds prompts, every time you want to make a new feature, you have to specify less.
比如你只需简单描述小功能，它就会自动研究并扩展成完整文档。而以前每次开发功能时，你都得先要求它做调研，再考虑各种边界情况或我偏好的实现方式等等，这实在太酷了。
Like, you just say the little feature, and then it'll go do the research to build it out into a big document, versus before, every single time you have to do a feature, you have to say, at first, want you to research it, and then I want you to think through all these different corner cases or the ways that I like things built or whatever, I think that's so cool.
还有一点特别有趣的是，它在我们交谈时就在持续工作，这完全是另一种编程方式。
And what's also really interesting to point out is it's working while we've been talking, and that's just a different way to code.
记得我们上周还是上上周通电话时，正在一起测试这个功能。结果就在通话期间，我推送了一个直接上线的功能——而我根本没碰代码库，这简直不可思议。
We were, you know, we were on the phone together, like, last week or the week before, and we were testing this out together, and I shipped a feature that went to prod while we were talking, which I'm not in the code base at all, so it's like kind of crazy that that actually happened.
这像是更社交化的编程方式，就像我们现在边聊边构建，这在以前是不可能的。
And it's it's like it's a kind of more social way to code, like we're coding right now building stuff, which was not possible before.
想要一套真正能帮你拿下大单的制胜演示稿吗？
Do you want a winning deck that'll actually help you land big deals?
Pitch将设计、协作与交付整合在单一平台。
Pitch combines design, collaboration, and delivery all in one platform.
展开剩余字幕（还有 408 条）
使用Pitch再没有版本混乱或风格不符的幻灯片。
With Pitch, there's no version chaos and no off brand slides.
它提供创建和展示专业精美演示稿所需的一切。
It's everything that you need to create and present professional, beautiful decks.
Pitch的新AI功能可即时为不同受众重写幻灯片文本。
Pitch's new AI actions let you instantly rewrite slide text for any audience.
它能保持品牌调性，自动生成幻灯片摘要和演讲备注，甚至能优化视觉设计而无需推倒重来。
They let you maintain brand tone and automatically generate slide summaries and speaker notes, and they even let you enhance visuals without starting over.
Pitch全面管理从设计到交付的演示文稿工作流程，且全程不牺牲设计品质。
Pitch manages your entire presentation workflow from design to delivery, and it does all of this without compromising on design quality.
它具备云端协作、品牌资源库、交互式嵌入和智能缩放工具等功能。
It features cloud based collaboration, brand libraries, interactive embeds, and smart scaling tools.
借助Pitch，您创建的每张幻灯片都能保持品牌一致性，这得益于自定义模板和品牌色调。
With Pitch, every slide you create stays on brand, thanks to custom templates and brand tones.
您分享的每份演示文稿都会追踪参与度，让您清楚潜在客户是否已打开查看。
And every presentation you share tracks engagement, so you know whether prospects have opened it.
若面对新受众，AI会即时重写您的宣传信息。
If you have a new audience, AI rewrites your messaging instantly.
想了解哪些幻灯片最受欢迎吗？
You wanna track which slides resonated?
参与度分析数据会为您揭示答案。
Engagement analytics show you.
如果客户需要专属空间来推进交易，可以创建定制化的提案室。
If your client needs a dedicated space for their deal, create a custom pitch room.
Pitch平台能适应您的实际工作方式。
The pitch platform adapts to how you actually work.
准备演示时，可通过实时链接或定制房间进行分享。
When you're ready to present, share via live links or custom rooms.
Pitch负责托管服务、数据追踪和专业润色。
Pitch handles the hosting, the tracking, and the professional polish.
您是否准备好在一个平台上完成设计、协作和成交全流程？
Are you ready to design, collaborate, and close deals all on one platform?
立即免费在pitch.com上开始创作，几分钟内就能让你的团队准备就绪。
Start creating for free at pitch.com and get your team set up in minutes.
就是pitch.com。
That's pitch.com.
今天就开启你的免费试用吧。
Start your free trial today.
现在回到节目内容。
And now back to the episode.
是的。
Yeah.
完全正确。
Absolutely.
哦对了，就在我们讨论的时候，我们做了研究并创建了这个议题，这很酷。
And oh, so it's while we were talking, we did the research and we created this issue, which is cool.
我们当时大概同时运行了六七个项目，因为我们就是想到新点子就立刻行动。
And we had, I think, six or seven running at the same time because we were just like, new idea, let's go.
新点子来了就立刻执行。
New idea, let's go.
我们还做了这些：梳理用户反馈、阅读邮件，尽一切可能收集信息，然后进行头脑风暴。
And what we also did, we went through user feedback, we read emails, we just everything we could, we gathered and we were just like brainstorming.
这过程非常有趣，因为在头脑风暴阶段，你可以随时启动各种方案，看看会有什么结果，观察它们能产生什么创意，之后再找时间复盘。
And it's really fun because if you're in this brainstorming place, you can just kick off agents and see what comes up, what they come up with and take another time to then review.
我们的另一个做法是，正如你所说，通过电话会议共同完成这个过程特别有意思，因为奇迹往往就是这样发生的。
So what we do also is, like to agree with you on like, it's really fun to do this together on a call because that's where magic happens.
这里仍然保留人工审核环节，因为我们发现需要人工检查内容是否合理、是否有遗漏，就像需要品味、经验和直觉那样。
And there is still a human review step here because we found that we want to look at it, see if it makes sense, if anything is missing like this is having taste, experience, intuition like this.
这就是我之前解决的邮件发不出去的bug。
This the bug I solved earlier with the email not going out.
娜塔莎对他的服装代码做了同样处理，但没得到正确答案。
Natasha did the same with his clothes code, but it didn't give the right answer.
是啊。
Yeah.
所以这里面还是需要人类直觉的参与。
So so there is like there is still like a human touch of intuition.
就像我暗示说查看历史记录，这确实让它朝着正确方向思考了。
Like I hinted at look at the history and that actually made it think into the right direction.
然后泰什没有添加查看历史记录，结果系统就说一切正常。
And then Tesh didn't add look at the history and then it said, no, everything works fine.
所以这里面还是存在直觉因素
So there is still like intuition and like
这仍然是项技能。
It's still a skill.
这仍然是项技能。
It's still a skill.
这确实是项技能。
It is a skill for sure.
没错。
Yeah.
不是这样的。
It's not yeah.
这绝对是一项技能。
It's absolutely a skill.
没有能解决一切问题的神奇提示词。
There's no magic prompts that does everything.
关键在于正确使用并发挥其优势，这是肯定的。
Like, it is about using it the right way and using it to its strengths for sure.
是啊。
Yeah.
是啊。
Yeah.
Nitesh，你对这一切感觉如何？
Nitesh, how have you found this all?
因为我知道，Kieran是长期从事Rails开发的专家级程序员，非常厉害，而你在编程道路上还处于相对早期阶段。
Because I know, you know, Kieran is, like, a long time, like, Rails, like, expert person who's just, an incredible programmer, and I think you're a little bit earlier in your programming journey.
那么以这种方式开始参与Quora的工作，对你来说是什么体验？
So what has that been like to come to every start working on Quora and start working on it in this way?
是啊。
Yeah.
不。
No.
说实话，这真的让我大开眼界。因为我的编程经验是，两年前ChatGPT问世时，我才想着，好吧。
This has been, like, incredibly eye opening, I would say, because honestly, like, my experience with programming is, two years ago when ChatGPT came out, I thought, okay.
现在对我来说，自学编程并构建那个我一直想要的SaaS应用再合适不过了。
Now it's perfect for me to teach myself programming and build that SaaS application, and that I always wanted to.
所以从第一天起，我就用ChatGPT自学编程。
So I taught myself programming using ChatGPT from the very first day.
因此我经历了所有的转变过程。
So I have gone through all the transitions.
我从ChatGPT开始，当Cursor问世后，就把工作流程转移到了Cursor上。
I went from ChatGPT, and then when Cursor came out, I shifted the workflow to Cursor.
后来Windsurf变得更好了。
Then the Windsurf got better.
我们就转用Windsurf了。
We shifted it to Windsurf.
而且，你知道，我总觉得自己处于技术前沿。
And, you know, I was always thinking like, okay, I am at the forefront.
我认识的朋友里没人像我这样深度使用AI并走在最前沿。
I don't know any of my friends who are doing so much with AI and are at the forefront.
后来我加入了Avery，开始和Kieran共事。
And then I joined Avery and start working with Kieran.
而Kieran完全是另一个级别的存在。
And Kieran is at a whole other level.
在我们的会议中，他现在已经开始写代码了。
In our meetings, he's like, now he's writing code.
他现在正在打字。
He's now typing.
他总是对着电脑说话。
He's always speaking into the computer.
于是我就想，好吧，我得把时间记录到工作流程里。
And so I was like, okay, I need to log time to the workflow.
后来即使Claude代码发布了，其实是Kieran硬推着我去用的。
And then even when Claude code came out, Kieran actually pushed me into using it.
现在很明显，这就是编程的正确方式了。
And clearly, it is now the way to program.
比如我和Kieran，我们俩最近大概三周都没碰过WinServe或Cursor了。
Like me and Kieran, both of us, like, we haven't even touched WinServe or Cursor in the last, like, three weeks or so.
就算偶尔用一下，通常也只是为了查阅些资料。
Or even if we do touch it, like, it's it's usually just because we wanna read something.
基本上就像是因为电脑上没装VS Code才将就用它。
It's it's basically like we're using it because we don't have Versus Code on our computer.
其实用VS Code老版本还是Clerc什么的都无所谓。
Like, it wouldn't matter if it was Versus Code, like the older Versus Code or Clerc or something.
因为现在所有AI功能都在ClotCode上实现，身处这种编程生态每三个月就彻底变革的时代真的很有趣。
Because all the AI stuff is happening with ClotCode now, and it's really fun to, you know, have and be in this position where the entire coding landscape just changes completely every three months.
你会意识到，其实根本没有人站在技术最前沿。
And you realize, like, nobody's at the forefront.
不得不说，我嫉妒你能在ChatGPT问世时学编程，我二十年前可是靠看书学的。
I gotta say, I'm jealous of you learning to code right when ChatGPT came out because I learned to code from books, like, twenty years ago.
《B h b四步入门》
B h b four for dummies.
是啊。
Yeah.
就像那种《24天学会基础》之类的教材，比如《山姆自学基础》什么的。
Like, basic for learn basic in twenty four days, like Sam's Teach Yourself Basic or whatever.
对。
Yeah.
德尔菲五号。
Delphi five.
没错。
Yeah.
而且而且而且，你说这个真的很好笑，我本来以为自己在AI编程领域算是前沿了，直到开始和基兰共事——这让我想起《星球大战》前传第一部里那个场景，他们在水下被海怪袭击眼看要完蛋了。
And and and also, it's so funny for you to to say, like, I thought I was I thought I was sort of at the at the forefront of AI coding, and then I joined every and started working with Kieran because it just reminds me of I don't know if, like, there's a scene in Star Wars, the prequel episode one where, like, they're they're under the water and and they're, like, being attacked by a sea monster, and it looks like they're gonna die.
然后突然出现个更大的海怪，直接把袭击他们的那只给吞了。
And then another bigger sea monster comes out and just, like, eats the eats the the one that's killing them.
奎刚当时就说：总有更大的鱼。
And Qui Gon is like, there's always a bigger fish.
总有更大的鱼。
There's always a bigger fish.
确实，基兰就是那条更大的鱼。
And, yeah, Kieran Kieran is the bigger fish.
但我也有同感。
But I feel the same.
虽然你这么说我，但其实我自己也完全摸不着头脑。
Like you say that about me, but I'm like, I have no idea what I'm doing.
就像我必须做，就像我落后了。
Like I need to, like I'm running behind.
我们还有上百万件事要做。
We need to do like a million more things.
这就是现实的状况。
So that's just the reality of the landscape.
就像总有更多事情，但这真的关乎练习。
Like there is always more, but it's really about practice.
就像你应该练习使用AI。
Like you should practice using AI.
你应该每天鞭策自己。
You should push yourself every day.
如果不这样做，你就会错过很酷的东西。
If you don't, like you'll miss very cool stuff.
是啊。
Yeah.
嗯，我个人很好奇，也为观众们问问，这有什么问题？
Well, are I I guess I'm curious, personally, and also for people in the audience, what are the problems with this?
对吧？
Right?
所以基本上，听起来你正在转向一种不接触代码的编程形式。
So, basically, it sounds like you're moving to a form of coding where you don't touch the code.
你高了一个层级。
You're one level above.
那么，由此产生的问题有哪些，你们又是如何解决的？
And so what are the problems that come up with that, and how are you solving them?
比如，为了确保一切顺利，你们需要引入哪些新的工程实践？
Like, what are the new engineering practices that you need to incorporate in order to make sure that things go well?
对我来说，最重要的领悟是，我总是会不断回顾这一点，尤其是在处理短代码时。
For me, like, the the most important realization for me has been, like, this thing that I always keep going back to, especially with shortcode.
我在那本管理书籍《高产出管理》中读到过，这本书是ITAL CEO大约五十年前写的。
I read this in that management book, that high output management, which the ITAL CEO wrote, like, fifty years ago.
在第一章中，他提到在任何生产流程中，都应在价值最低的阶段解决问题。
And the first chapter, he mentioned something like how in in any production process, you should fix any problem at the lowest value stage.
我一直在思考这个观点，因为AI和绘图代码现在能为我们做很多事情，所以聚焦在事情的最早期阶段变得尤为重要。
And I just can't stop thinking about that statement because because AI and plot code can now do so many things for us, it has become really important to focus on the earliest part of things.
具体来说，当我们使用Kiran刚才展示的工作流创建GitHub问题时（比如一个非常详细的GitHub工单），很容易就会想再启动Cloud Code让它去解决这个工单。
So what I mean by that is when we see that, you know, when we are using the workflow that Kiran just showed to create a GitHub, like a very detailed GitHub issue, then it's very tempting to like, start another Cloud Code to ask it to just, hey, go now, work on this GitHub issue and fix it.
但这实际上会带来问题，因为Claude在该工单中给出的方案可能并非你期望的方向。
But that's actually gonna be a problem because there are chances that, you know, the plan that Claude was able to give in that issue, it wasn't the direction that you wanted to go.
你需要在让Claude去实施解决方案之前就发现这一点，并在那个阶段进行修正。
And you want to catch that before you ask Claude to go and implement the solution, and then you wanna fix it over there.
这完全说得通。
That makes perfect sense.
我特别特别喜欢这个想法。
I really, really like that idea.
这让我联想到，所有这些就像是一个杠杆。
The thing it reminds me of is just it's like all this stuff is like it's like a lever.
就像，你在杠杆上越往外移动，拥有的力量就越大，但偏离正确方向的力量也同样越大。
And, like, the further out you get on the lever, like, the more power you have, but also the more power you have to go in the wrong direction.
每一寸微小的移动最终都会造成巨大差异，所以我认为尽早调整才能确保不会偏离轨道。
Every little inch makes a big difference at the end, and so trying to catch it earlier, I think, is the thing that makes sure that you're not shooting off into space.
莉兹的杠杆比喻完全失效了，但你应该明白我的意思吧？
Liz Lever metaphor is totally breaking, but, like, you know what I mean?
就像如果你对准月球发射火箭，一英寸的偏差就意味着数千英里的误差。
Like, if you if you point a rocket at the moon, like, one inch means thousands of miles of difference.
所以我想AI领域也是同样的道理，这对我来说是个很好的教训，因为我总是急于跳过规划阶段。
And so I guess the same thing is true with AI stuff, and I think that's actually a good lesson for me because I I tend to wanna, like, rush through the planning stuff.
对我来说，要专注阅读像写作规范那样的文档真的很困难。
It's like it's just hard for me to, like, look at a document like that, like the thing that's that caught us writing, and concentrate on it.
你们是怎么解决这个问题的？
So how how have you guys found that?
是啊。
Yeah.
大部分时候读这些确实挺无聊的。
It's it's kind of boring to read most of the time.
但你可以让它变得有趣些。
But you can make it more fun.
比如你可以说'极简、极简'。
Like, you can say just like minimal, minimal.
这个太过了。
This is too much.
可是然后问题又来了，它又开始遗漏东西。
Just but but then the thing is then it misses things again.
所以这实际上很重要。
So it's actually important.
所以对于代码，我喜欢它专注于用户故事，或者像是提问和回答问题那样。
So for code, I like it to focus on user stories or like asking questions and answering them.
比如说，嘿，一个好的产品经理会针对这个提出哪些问题，我们应该考虑并给出两个选项？
So let's say like, hey, what are some questions a good PM would ask about this, like that we should consider and give like two options?
这样读起来比‘第一周做这个，第二周做那个’要有趣得多。
Like that's it's more fun to read that than like week one we'll do this, week two we'll do that.
就像是产品需求文档很无聊，但你可以让它们稍微有趣些，或者提供更多例子，或者你可以塑造那种研究。
Like that's it's like PRDs are boring and you can make them a little bit more fun or give more examples or like you can shape that research.
这通常就是我们在人工审核阶段要做的工作。
And that's normally what we do in the in the human review step.
比如，我们是否发现了任何危险信号？
It's like, we see any red flags?
是否需要补充更多材料？
Do we need more stuff to be added?
因为这将节省大量时间。
Because it will save so much time.
是的。
Yeah.
这倒让我想起了我们在业务另一个板块的发现。
That actually reminds me of something that we're finding in another part of the business.
丹尼上过这个节目，他是Spiral的总经理，在Spiral内部，我们正在开发一个写作助手。
So Danny, who's been on this show, is the GM of Spiral, and inside of Spiral, we're building a writing agent.
你可以把它想象成某种热代码，但专门用于写作任务。
So you can think of it sort of like hot code, but specifically for writing tasks.
我觉得这其中有相似之处，有时候你会希望这个写作助手切换到类似采访模式，试图更深入地了解你是谁、你想要什么，而不是简单地输出一堆需要你阅读的内容。
And I think there's something similar about that, where sometimes you want that writing agent to shift into, like, an interview mode where it, like, tries to understand more about what who you are and what you want, rather than just, like, spitting out a bunch of stuff that you then you have to read through.
听起来ClaudeCode或这类编码工作流程中可能缺少这个功能，如果能通过提问的方式，而不是让你阅读冗长的文档，就能使输出内容更准确，而不必全部重做，那会非常好。
And it sounds like there's maybe something missing here in ClaudeCode or these sort of coding work coding workflows where it would be really nice, instead of having to read that, like, long document, it's finding ways to, like, ask you questions so that the thing it outputs is more likely to be right without you having to redo the whole thing.
是的。
Yeah.
完全同意。
Absolutely.
这个自定义命令的想法很有意思。
That's an interesting idea for a custom command click here.
我们绝对应该试试。
We should totally try that.
没错。
Yeah.
当然。
For sure.
这确实是我们应该自动化并改进的事情，毫无疑问。
Like, this is something we should automate and make better, like, for sure.
同时，它知道很多，因为它能访问你的代码库和风格规范。
And at the same time, it it knows a lot because it has access to your code base and your style.
就像，那非常强大。
And like, that's that's very powerful.
就像你有了代码库，实际上做得相当不错。
So like you have the code base and it's actually pretty good doing it.
我认为，除了从一开始就做好，那些看似无聊的传统测试和防范措施也同样重要。
Like, I think in addition to like making it very good at the beginning, I think just boring traditional tests and evils are very important as well.
否则你怎么知道你做的确实有效呢？
Because how do you know what you did is actually working?
你可以打开控制台手动点击测试，但何必呢？
Well, you can open a console and click through it, but like, why?
直接给它写个测试，哪怕是最基础的也好。
Just have it tested, write a test for it, like just just the bare minimum.
冒烟测试就很棒，你只需要看它是否能基本运行，否则工作量就太大了。
Smoke tests are great where you just see does does it kind of work because otherwise it does way too much.
但这是让它自行迭代和修复问题的好方法。
But it's a very good way to have it iterate and fix things by itself.
我们还没怎么尝试，但我们用了Figma MCP功能，直接说'按这个Figma设计实现'。
And we haven't tried it as much yet, but we use the Figma MCP where we say, hey, implement this from Figma.
现在你可以用Puppeteer给移动版截图，然后进行对比。
And then now there is like you can have Puppeteer take a screenshot for a mobile version and then say, compare the two.
我们还没真正试过，但我们想多尝试这类方法。
Like, we haven't really tried it out, but like we want to try more of that out.
所以这些检查机制和测试流程，通常都是手动完成的。
So there are these checks in place, tests in place that you normally do manually.
同样适用于像提示词评估这样的提示词。
And the same for prompts like evals for prompts.
所以我某种程度上认为评估就像是给代码写测试。
So I kind of think of an eval as like writing a test for code.
评估就是针对提示词的测试。
An eval is a test for a prompt.
我上周还看到，我让Clothes Code运行了一个评估，结果发现实际上十次里有四次失败。
And what I've seen last week as well, I had Clothes Code run an eval and then say, actually, it fails four out of 10 times.
我说，运行十次看看。
I said, run it 10 times.
它每次都能通过吗？
Does it always pass?
不，有四次没通过。
No, four times it doesn't.
我说，哦，看看输出结果。
I said, oh, look at the output.
比如，为什么它没有调用那个工具？
Well, like, why didn't it call that tool?
那是个很酷的工具，名叫test。
It was a cool tool tool called test.
结果显示，哦，是因为提示词不够具体。
And it says, oh yeah, it wasn't specific enough.
我说，好吧，继续调整提示词直到它能每次都稳定通过。
And I said, okay, just keep going and change the prompt until it's passing consistently all the time.
它做到了。
And it did it.
就像我只是下楼买了杯咖啡，上楼就完成了。
It's like I just walked downstairs, got a coffee, walked up and that was it.
评估同样非常强大，因为它们能告诉你一个提示是否有效，就像编写代码时测试能验证代码是否工作一样。
So evals are also very powerful because they will tell you if a prompt works and similar to writing code, tests says your code works.
因此，采用那些更传统、看似枯燥的方法也同样非常有效。
So leaning into those more boring traditional ways is also very powerful.
这很合理。
That makes sense.
我有
I have
一个想法。
a thought.
因为我认为其中一点非常特别，而且Nitesh，我想你也这么认为，如果我错了请告诉我。
And because one of the things I think is really special, and I think, Nitesh, you're in this boat too, so tell me if I'm wrong.
但Karen，我认为你真正特别的一点是你会测试所有东西。
But one of the things I think is really special about you, Karen, is that you just test everything.
比如，你测试过每一个代理。
So, like, you've tested every single agent.
Nitesh，你也用过很多代理吗？
Nitesh, have you have you used, like, a lot of the agents as well?
没有。
No.
那是凯尔。
That's Care.
好的。
Okay.
嗯，我觉得我们还是可以做到的。
Well, I I think we could I think we could still do this.
我觉得会挺有趣的。
I think it'd be kinda fun.
我想花五分钟和基兰一起给特工们做个从S级到F级的排名。
I wanna spend five minutes with Kieran doing a s tier through f tier ranking of agents.
接下来我会共享屏幕，我会喊出一个特工名字，然后你告诉我它的排名。
And so what I'm gonna do is I'm gonna share my screen, and I'm gonna I'll call out an agent, and then you tell me where it ranks.
你愿意玩吗？
Are are you game?
好啊。
Yeah.
我们开始吧。
Let's do it.
好的。
Okay.
酷。
Cool.
让我们看看。
Let's see.
我猜是光标？让我们用光标吧。
Does I guess cur let's do let's do cursor.
对。
Yeah.
所以这很有趣，因为光标，什么光标？
So it it's fun because cursor like, what cursor?
是布料四号吗？
Is it cloth four?
是最大值吗？
Is it max?
是
Is it
最佳设置下的光标。
Cursor on the best possible settings.
是后台代理吗，还是那个，好吧，传统最佳设置关闭的光标。
And is it the background agent, or is it the the okay, cursor traditional best possible setting close.
这就是光标冲浪让人困惑的地方。
Like, that's the confusing part about cursor windsurf.
好像有无数个版本。
Like, there are like a million versions of it.
为什么不用最好的版本呢？
And like, why don't you just have the best version?
这就是我喜欢某些代理的原因。
And that's what I love about certain agents.
他们只是说，看，这是最好的特工。
They just say, look, this is the best agent.
所以这就是为什么它不是最好的。
So that's why it wouldn't be the best.
我会说
I would say
a。
a.
A？
A?
好的。
Okay.
光标是
Cursor is
光标与Cloak四号配合得非常好。
Cursor is very good with Cloak four.
好吧。
Alright.
风帆冲浪。
Windsurf.
C，因为他们没有clot四号。
C, because they don't have clot four.
这很荒谬，因为三周前他们还是a，现在却不是了。
It's ridiculous because three three weeks ago, they would be a, and now now they're not.
哇。
Wow.
哇。
Wow.
好吧。
Okay.
因为我几个月前从风帆冲浪切换到光标，或者说从光标切换到风帆冲浪，但后来又切换回来了。
Because I I switched from windsurf to or from cursor to windsurf, like, a few months back, but I switched back.
好的。
Okay.
所以我们有风帆冲浪是C，光标是A。
So we've got windsurf is a windsurf is a c, cursor is an a.
我们来看看。
Let's see.
德文。
Devin.
是个B。
It's a B.
B。
B.
为什么？
Why?
感觉它不够一体化。
It's like it's not as integrated.
配置起来有点困难，代码质量方面，它不如Cursor或ClothCode那么全面。
It's a little bit hard to set up and the code quality is like, it's not as well rounded as cursor or ClothCode.
我不确定他们是否用Cloth做背景，但它的可用性确实不如其他产品。
I don't know if they use Cloth for the background, but, like, it's not as usable as the the others.
查理。
Charlie.
查理主要是用于代码审查的。
Charlie is like for code reviews.
所以我们主要用查理来做代码审查。
So we use Charlie for code reviews mostly.
我其实没怎么把它当作智能代理来用。
I haven't really used it as an agent as much.
我觉得查理作为代理只能算B级，但作为代码审查工具绝对是A级。
I think Charlie as an agent is B, but it's A as a code reviewer.
我特别喜欢它做的代码审查。
Like, I really like the code reviews it does.
这挺有意思的。
So that's interesting.
它在某些方面确实很出色。
Like, it's really good at something.
那星期五呢？
And then what about Friday?
我把星期五排在Cursor之上，大概在S级和A级之间。
I put Friday higher than Cursor, maybe between S and A.
这很有趣，因为他们甚至还没用到Cloth四代。
And it's funny because they don't even use Cloth four yet.
他们仍在研究如何真正让它良好运作。
They're still they're still working on how they like really make it work well.
现在是3.7版本，我喜欢它的设计理念与Clothcote不同，但Friday有着非常明确的工作方式。
It's 3.7, but like why I like it there, it's different than Clothcote, but Friday has a very opinionated way of working.
我欣赏他们的理念，实际效果很好，完全能达到预期。
And I love their opinions and it really works well and it just does it.
比如你提出一个问题，他们制定计划，你批准后就能自动执行。
Like you give an issue, they make a plan, you approve and it does it.
它会自动创建拉取请求。
It creates a pull request.
我见过它完成一些我无法用传统代码实现的功能。
And I've seen it do this stuff that I couldn't do with cold codes.
举个例子，它能直接实现Figma的设计稿。
Like, for example, implement this Figma design.
它曾展示过一个助手的Figma设计方案。
It just one showed a Figma design for the assistant.
我多次见证这样的时刻，它完成的工作让我惊叹：哇，这就是未来的味道，非常独特。
And and I've seen moments where more multiple moments like that, where it did things are like, wow, okay, this I taste the future, which is really unique.
而且团队规模也很小。
And it's a small team as well.
真的很酷。
So really cool.
Codex。
Codex.
这对我来说是b。
This is b for me.
好吧。
Alright.
Codex是个b。
Codex is a b.
Copilot。
Copilot.
我还没用过Copilot。
I haven't used Copilot.
你从没用过GitHub Copilot？
You never used GitHub Copilot?
没有。
No.
我是说，我三年前用过，但我，呃，没用。
I mean, I used it three years ago, but I, like, no.
我还行。
I okay.
咱们公平点说。
Let's be fair.
我大概半年前试过，用了一秒就再没用过了。
I tried it maybe a half a year ago, and after one second, stopped using it.
那你把它排在什么位置？
So where where do you rank it?
D级。
D.
就是说，它不具备代理性，不过，我确实应该试试新版本。
Like, it was not agentic, but but, I mean, I should try the new version for sure.
我们还没试过呢。
We have not tried yeah.
我们还没试过那个代理副驾驶功能，所以这个评价不完全公平，不过好吧。
We've we haven't tried the co the the agentic copilot, so that's that's not totally fair, but okay.
我们还漏了什么吗？
Are we missing anything?
我觉得我们是。
I feel like we are.
克劳德。
Claude.
克劳德的代码。
Claude's code.
很明显，克劳德代码。
Obviously, Claude code.
我猜是S级，宝贝。
I assume s tier, baby.
我们也有工厂。
We have factory as well.
哦，是的。
Oh, yeah.
工厂在你心目中排名如何？
What's how where do you rank factory?
这很有趣。
It's interesting.
工厂在某些方面确实比其他选择更好，但不是我的风格。
Factory, with certain things, is like better than any others, but it's not my style.
它更像是为那些非常书呆子气、追求极致代码的企业级用户设计的。
It's like factories for more enterprise y people that are very nerdy and want like absolute bangers of code.
实际上它很不错，比如多仓库管理这类功能。
And it's actually good, like multi repo stuff like that.
使用起来有点困难，因为它既在网页端又在本地运行。
It's a little bit hard to use because it's on the web, but also local.
所以我给它打B级，可能略低于Codex和dev in。
So I rate it B, maybe a little bit below Codex and dev in.
是的。
Yeah.
但它确实有它的用武之地。
But it's like, there is a use for it for sure.
有些东西正在发展。
There's something going.
那里有些好东西，但可能不适合我们。
There's something good there, but it's maybe not not for us.
这不是我的菜。
It's not my thing.
是啊。
Yeah.
对啊。
Yeah.
AMP也是。
AMP also.
AMP。
AMP.
那是什么？
What's that?
AMP。
AMP.
A m p。
A m p.
AMP。
AMP.
哦，AMP。
Oh, AMP.
嗯。
Yeah.
抱歉。
Sorry.
那是什么？
What's that?
是的。
Yeah.
我会把它放在布料外套下的S级，介于哇哦之间。
I would put it s tier under cloth coat between Woah.
又一个S级。
Another s tier.
是的。
Yes.
好吧。
Alright.
为什么？
Why?
它非常擅长完成工作。
It's it's very good at just getting work done.
经济性已经很不错了，工具也很好。
The economics are pretty good good tools already.
就像，人们使用那个工具来构建它。
Like like, people people use that tool that build it.
他们在内部试用。
They're dogfooding.
比如，你能从Clothcodes和Antler的开发者那里感受到他们对代理的热爱，他们正在构建最好的东西，并尝试新事物。
Like, you can feel from Clothcodes and Antler developers that love agents and they're just building the best thing and they're trying new things out.
嗯，就是这样。
So, yeah, that's it.
我们看看。
Let's see.
这正是基兰如此重要的原因。
This is exactly why Kieran is the big fish.
我是说，就是'你是对的'。
I mean, it's You are yeah.
你把它们串联起来了。
You're stringing them together.
就像你同时在用Cloud Code、Friday和其他工具，这确实很酷。
Like, you're using Cloud Code and Friday and and other stuff all at the same time, which is, yeah, the thing that is really cool.
是的。
Yeah.
就像，确实存在这样的情况。
Like, and like, there are yeah.
按我的理解，就像面试一个职位时，你需要找一个开发者来解决特定问题。
Like, how I think about it, like, I'm I'm thinking about it more, it's like you're interviewing for a role and you find a developer to, like, solve a certain problem.
我认为编程代理也是类似的。
I think it's similar with coding agents.
比如Friday现在就很擅长做UI。
Like, Friday is good at like doing UI now.
所以如果我需要UI相关的工作，就会找Friday。
So if I need UI work, will go to Friday.
如果需要进行研究，我会去Clothes Code。
If I need to do research, I go to Clothes Code.
是的，如果我想要代码审查，就用Charlie。
And yeah, there is a if I want a code review, I use Charlie.
这很有趣，而且代理们会协同工作。
Like it's it's fun, and agents work together.
你不需要只有一个代理。
You don't need to have one agent.
我们有所谓的代码。
We have called codes.
因为Charlie在GitHub工作，你可以直接CC Charlie，他就会对PR进行代码审查。
And that's because Charlie, like, works in GitHub, so you can just, like, CC Charlie, and and Charlie will do the the code review on the PR.
没错。
Yeah.
所以我们使用GitHub和拉取请求，遵循常规的开发流程。
So we use GitHub and pull requests and normal developer flows.
这样人类开发者就能介入。
So humans can hook in.
我们可以聘请擅长特定领域的人来审查代码，然后关闭的代码会自动完成工作。
So we can hire someone that's very good at specific thing and review code and then close code will just do the work.
但这非常强大，因为这是我们花了大约二十年完善的生态系统，而且行之有效。
But it's very powerful because it is just an ecosystem that we refined over like twenty years or whatever, like, and it works.
所以让我们充分利用这一点。
So let's lean into that.
这大概就是为什么Copilot应该没问题，因为它已经内置在里面了。
And that's probably why Copilot will probably be fine since it's in there already.
是啊。
Yeah.
等等。
Wait.
你最近确实这么做了。
You actually did that recently.
比如，我们之前遇到一些基础设施问题，要知道Quora处理过海量的邮件，所以我们有些基础设施问题需要解决。你似乎找来了一位真正的专家，然后用一种特定的代理方式与他们合作，既得到了你需要的东西，又让他们工作量更少。
Like, we had some infrastructure things where, you know, we've handled tons and tons and tons of emails at Quora, so we had some infrastructure issues to work out, and you it seems think I you brought in someone who's like a real expert and then worked with them in a specific agentic way that you got what you needed from them, but it was less work for them.
对。
Yeah.
没错。
Yeah.
所以，虽然当时还没出现问题，但我们希望对最重要事项的交付有更高可见度。
So, like like, there was no issue yet, but we wanted more visibility in delivery of like the most important things.
而且我不太擅长这个，虽然懂一些，但觉得还是该找专业人士。
And like, I'm not very good at it or like, I know stuff, but like, let's bring in someone.
我们的做法就是进行了两小时通话并全程录音。
And what we did, we just had a conversation like a two hour call and I recorded everything.
结束后我把录音喂给系统，让它整理出两个资源问题。
And at the end, I just fed that into cloth and say, okay, can you make two issues, resource issues from this?
大概十分钟后，系统就反馈说问题整理好了。
And like ten minutes later, said, okay, here are the issues.
你能审查一下吗？
Can you review them?
然后他就说，天啊，什么？
And he was like, holy, what?
这家伙，虽然他不是AI怀疑论者，但他确实非常擅长他的工作。
This guy, like he's not an AI skeptic, but he's like, he's he's very good at what he does.
通常他做的事情，AI目前还不太擅长，因为有些领域AI还达不到那个水平。
And normally what he does, AI is not good at yet because like there are things AI is not as good at yet.
但他对此印象深刻，并提出了很好的改进意见。
But he was very impressed with it and he had like very good comments on it to iterate over it.
我们基本上就是能更快地迭代想法，因为有了讨论的基础。
And like what we basically did, we just iterated more quickly through ideas because we had something to talk about.
第二天他说要做人工审查时，我就说，那咱们开始吧。
And then I said the next day when he was like, did the human review, like, let's go.
我直接用Clothecode实现了。
I just use Clothecode to implement it.
我们坐下来进行了代码审查。
And we sat down and did the code review.
这就像把原本可能需要两周的工作，加速到几小时完成，真的很棒。
So it's like it's just accelerated what would have taken two weeks maybe, is now in like a few hours, which is really cool.
我很喜欢这样。
I love it.
好吧，就是这样。
Well, there you have it.
你已经有了一份智能体分级列表。
You've got your tier list of agents.
Claude代码拔得头筹。
Claude code takes the cake.
AMP紧随其后排名第二，而GitHub Copilot不幸垫底，不过一旦我们测试它们的代理能力，仍有改进空间。
We've got AMP coming up coming up in second, and GitHub Copilot, unfortunately, bringing up the rear, but with room for improvement once we try out their their agentic capabilities.
在结束今天之前，大家还有什么想说的或讨论的吗？
Anything else you guys wanna wanna say or talk about before we end today?
每个人都应该使用冷代码或尝试一下，即使你不懂技术。
Everyone should use cold code or try it out, even if you're not technical.
订阅他们的高级或专业版计划。
Subscribe for their max or pro plan.
每月只需100美元。
It's only a $100 per month.
你可以无限制地使用。
You have unlimited access.
如果你对技术方面持怀疑态度，其实它非常简单。
If you're skeptical about being technical, that it's very easy.
我见过一些人，我有个朋友用了Cursor，我就说直接用Claude Coat更好。
And I've seen people, a friend of mine, he used cursor and I said, just use cloth coat, it's better.
他当时问：能好多少？
He was like, how much better can it be?
然后他说：确实更好。
And he said, yes, it's better.
他将所有用光标感觉编码的内容都重构进了紧密的代码中。
And he rebuilt everything he did with cursor vibe coded into close code.
然后他说，没错，这太棒了。
And he's like, yeah, this is great.
他感受到了下一步，认为每个人都应该尝试，真正地推动他们的工具。
He felt that next step and you should everyone should try it and really push push their tools.
是的。
Yeah.
尼泰什，还有什么智慧之言吗？
Nitesh, any other words of wisdom?
一定要在最初阶段就检查AI的工作成果。
Just be sure to check the AI's work at the lowest starting stage.
你希望尽早发现问题。
You wanna catch those problems early.
没错。
Yeah.
这真是个好建议。
That's that's a great one.
还有，记得使用Quora。
And, also, use Quora.
Quora。
Quora.
电脑。
Computer.
快来看看。
Check it out.
简直太棒了。
It's pretty awesome.
我们随时都在给你寄送好东西。
We're shipping you things all the time.
非常感谢两位的到来。
Thank you both for coming on.
这真是莫大的荣幸。
This is a true pleasure.
我已经迫不及待想看看你们接下来几个月还会搞出什么新花样，我们很快会再聊的。
I cannot wait to see what else you cook up over the next couple months, and we'll talk soon.
谢谢。
Thank you.
太感谢你了。
Thank you so much.
天哪，各位。
Oh my gosh, folks.
你们必须立刻马上点赞并订阅《AI与我》频道。
You absolutely positively have to smash that like button and subscribe to AI and I.
为什么？
Why?
因为这档节目堪称精彩绝伦。
Because this show is the epitome of awesomeness.
这就像在后院发现一个宝箱，但里面装的不是黄金，而是关于聊天GPT纯粹无修饰的知识炸弹。
It's like finding a treasure chest in your backyard, but instead of gold, it's filled with pure unadulterated knowledge bombs about chat GPT.
每一集都是一场情感、洞见与欢笑的过山车，让你欲罢不能，渴望更多。
Every episode is a roller coaster of emotions, insights, and laughter that will leave you on the edge of your seat craving for more.
这不仅仅是一个节目。
It's not just a show.
这是一场由Dan Shipper担任飞船船长的未来之旅。
It's a journey into the future with Dan Shipper as the captain of the spaceship.
所以帮自己一个忙。
So do yourself a favor.
点赞、猛戳订阅，系好安全带准备迎接人生中最刺激的旅程。
Hit like, smash subscribe, and strap in for the ride of your life.
现在无需多言，让我直说吧，Dan，我已经无可救药地爱上了
And now without any further ado, let me just say, Dan, I'm absolutely hopelessly in love with
你。
you.
关于 Bayt 播客
Bayt 提供中文+原文双语音频和字幕，帮助你打破语言障碍，轻松听懂全球优质播客。