一位身价50亿美元的创始人如何使用AI（3个教程）

太好了。

Great.

那就麻烦你了。

Please do that.

对吧？

Right?

所以它就是在帮你起草那些跟进邮件。

So it's drafting those follow-up emails.

它还会帮你起草其他各种东西。

It's drafting all these other things.

所以这就是AI真正开始变得非常有用的地方，它不只是给你一个简报，然后让你自己去处理。

And so this is where AI starts to get really helpful is it's not just giving you a brief and sort of teeing it up for you to do.

现在它会说，我来帮你搞定这件事。

It's now saying, I'll I'll just do that for you.

我觉得，当你真正让AI去干活的时候，你才能真正感受到它的强大。

And this is where I think you start to really feel the power of AI is when you can actually put it to work.

所以我特别喜欢Scribe，就是那种会后跟进或者一天结束时的总结，因为以前我每天结束时要花大概两个小时，就为了处理那些被拉进去的小杂事。

So I love Scribe, those sort of, like, post meeting follow-up or the end of the day wrap up because it just, I don't know, I used to spend, like, two hours at the end of my day just, like, you know, just trying to catch up on every little ticky tack thing that I sort of got roped into.

现在每天结束时的这两个小时缩短到了大概十五分钟。

And now that, like, two hours at the end of the day is down to, like, fifteen minutes.

那你觉得这真的能帮你每天省下差不多两个小时吗？

So you think this is legit saving you two hours a day almost?

我，是的。

I yes.

对。

Yes.

而且，你知道，关于我一天中哪些做得好、哪些做得不好的部分，它帮我专注于重要的事情。

And furthermore, that, like, you know, what's good about my day, what's bad about my day part, it just helps me stay, like, focused on the things that matter.

所以AI在这方面帮了我，但我也记得，大概在我开始这样做九十天后，我说，嘿。

So both the AI is helping me with that, but I also will send, like, I remember 90 like, about about ninety days after I started doing this, I said, hey.

我有个想法。

I had this thought.

我就想，你对什么让我过得好、什么让我过得不好，了解得真多。

I was like, you know so much about, like, what makes me have good days and bad days.

你能不能帮我总结一下，我过得好和过得不好的日子，分别有什么共同点？

Can you, like, summarize, like, what is common about my good days and what is common about my bad days?

它就说，当然可以。

He was like, sure.

然后它就把这些都做了。

And it did all that stuff.

然后我把结果给了我的助理，我说，嘿。

And then I gave it to my assistant, and I said, hey.

这是总结，请让我的日子多一些好日子，少一些坏日子。

Here's, like please make my days look like more good days versus bad days.

那你过得好和过得不好的日子，分别是什么样的？

What what were your good days and what were your bad days?

对我来说，最重要的一点是，我是个早起的人。

So for me, like, the single most important thing is I I'm a morning person.

所以我早上工作效率最高。

And so I do my best work in the morning.

那时候我最专注。

I'm most focused.

你知道，一到午饭后，我就没法认真做任何事了。

You know, once once after lunch comes around, you know, I I'm just not gonna get anything, like, serious done.

所以最重要的是，上午11点之前我不想开会。

And so the most important thing is I don't want a meeting until 11AM.

只要给我那段早上的时间，让我处理最关键的事情，这一天就会很顺利。

And so you just give me that morning time where I can knock out the most critical things, and the day's gonna go great.

那样的话，接下来一整天我都会很顺利。

Like, I will win the rest of the day.

但如果那段时间被占用，比如陷入救火模式，你就能看出来，我一天结束时会很烦躁，就是那种‘唉’的感觉。

But if I have to, like you know, if if that time starts to get sucked into, like, firefighting mode, you can tell I'll just be grumpy at the end of the day where I'm just like, ugh.

就是觉得自己不够成功。

Like, I'm not I'm not being successful.

你会发现这个播客是上午11点开始的。

And you'll notice this podcast started at 11AM.

嘿。

Hey.

我们快速休息一下。

Let's take a quick break.

你有没有过这种感觉：策略定好了，简报写完了，大家也都对齐了，然后你发现还是得有人坐下来真正去创作所有内容？

You know that feeling when strategy is done, the brief is written, everyone's aligned, and you realize someone still has to sit down and actually create all the content?

那个人通常就是你，而且明天就要交。

That someone is usually you, and it's due tomorrow.

其实，HubSpot的Breeze助手可以帮上忙。

Well, the Breeze assistant from HubSpot can help.

它直接在HubSpot里面就能用。

It works right inside HubSpot.

你可以起草活动文案、博客文章、邮件，全都用你的品牌语气，全都基于你真实的客户数据。

You can draft campaign copy, blog posts, emails, all in your brand voice, all using your actual customer data.

所以你不仅仅是创作内容。

So you don't create just content.

你创作的是能转化的内容。

You create content that converts.

去 hubspot.com 看看吧，这是专为成长型企业打造的智能客户平台。

Check out hubspot.com, the agentic customer platform for growing businesses.

你能给我们演示一下怎么让 AI 来反驳你，而不是……

Can you show us how to get the AI to challenge you instead of a

哦，对。

Oh, yeah.

那我们就来试试。

Let's do that.

回到刚才的话题，就是会跟你争论的 AI。

So coming back, AI that argues back.

这个其实很简单。

So this is pretty simple.

对吧？

Right?

我平时就用 Cursor 作为我的日常助手。

I use I use Cursor as my daily, like, agent.

但如果你用的是 Claude、Chateaux Dei 或者 Gemini，它们通常都有系统提示或者某种方式，让你可以调整它跟你说话的风格。

But in if you're using Claude, if you're using Chateaux Dei, if you're using Gemini, they often have, like, a system prompt or a way that you can sort of tune, like, how it talks to you.

在 Cursor 里，有一个文件叫 agents.md。

And so in Cursor, have a file that's the agents m d.

它基本上到处都用这个。

That's kinda what it uses everywhere.

Claude 也有一个 Claude.md 文件。

Claude has a Claude m d m d.

基本上，我有一个文件，内容就是“怎么跟我合作”。

And, basically, I have a file that's like how to work with me.

然后，你知道，我让它直接一点，诚实一点。

And, you know, I, you know, have it be direct and honest.

我需要你告诉我同事们不敢告诉我的真相。

I need the truth my coworkers are afraid to tell me.

我需要你挑战我的假设。

I need you to challenge my assumptions.

我需要你找出我思维中的漏洞。

I need you to poke holes in my thinking.

我需要你在我真的不同意的时候，表达你的不同意见。

I need you to dis like, disagree when you genuinely disagree.

这些东西就像是追求目标的实体。

And these things are like goal seeking entities.

所以如果你告诉它，我想要你做什么，它就会说，好的。

And so if you tell it, like, what I want from you is this, it's gonna say, okay.

好吧，这就是韦德想要的，所以我会努力去实现这个目标。

Well, that's what Wade wants, so I'm gonna try and give him that goal.

但如果你不告诉它这些，如果你不做这样的配置，它就会真的试图去迎合你，因为它觉得那就是你想要的。

But if you don't tell him this, if you don't make this configuration, it's going to just genuinely try and, like, placate you because it thinks that's what you want.

所以你得告诉它，你知道，你得稍微引导它一下。

And so you do have to tell it you know, you sort of guide it a little bit.

比如说，我提前为这个跑了一个实时提示的例子。

So, like, here's an example of, like, you know, a live prompt I ran ahead of time for this.

我说，嘿。

I said, hey.

我在想我们这个季度应该把整个免费层砍掉，因为我们的一个竞争对手涨价了。

I'm thinking we should kill our entire free tier this quarter because one of our competitor raised prices.

帮我起草一下公告。

Help me draft the announcement.

对吧？

Right?

所以这应该是它肯定会说，嘿，先暂停一下的事情。

So this is, like, something that it should definitely, like, say, hey, press pause.

就是说，那是个坏主意。

Like, that that's a that's a bad idea.

所以当我这么做的时候，它给我的反馈是这样的。

And so when I do that, here's the, you know, pushback that it gives me.

它说，嘿，这基于一个很薄弱的信号就做出了一个大动作。

It's like, hey, that's a big move off a pretty thin signal.

在你起草任何东西之前，我希望你把三件事摆到桌面上来。

Before you draft anything, I want you to I want three things on the table.

比如，把数据给我。

So, like, you know, bring me the data.

告诉我你真正想解决的问题是什么。

You know, tell me, like, what the actual problem you're trying to solve is.

定价公告是错误的。

You know, pricing announcements are wrong.

有没有一个更小的实验你可以先试试？

Like, you know, is there a smaller experiment you can run?

太好了。

That's great.

对。

Yeah.

就是帮你理清这些问题。

It's just, like, helping you think through, you know, these, problems.

然后，我还有一个自己做的技能，我管它叫“作战会议”技能。

And then, you know, I have this other skill that I've made, which I call my War Council skill.

这个比刚才那个更具体，我经常用。

And so this is, like, an even, like, more specific version of this that I will often invoke.

所以我会说，嘿。

So I could say, hey.

对。

Yeah.

我在考虑要不要招这个人。

I'm thinking about hiring this person.

这是我从面试小组那里拿到的关于他/她的所有笔记。

Here's all the, like, notes that I've got on them from the interview panel.

帮我看看这个人能不能提升团队的标准。

Like, help me figure out if this person's gonna raise the bar or not.

这个“作战会议”技能的作用是——其实，让我先回去看一下。

And what it will do, the War Council, is this is a actually, let let me poke back.

我记得我这里有“作战会议”这个技能。

I think I have the War Council skill in here.

对。

Yeah.

找到了。

Here it is.

好。

Okay.

其实它没有把技能调出来。

I it didn't actually pull the skill up.

所以，如果你去Zapier的GitHub，你可以拿到一份“作战会议”的副本。

So this, if you go to, Zapier's GitHub, you can get a copy of the War Council.

但这个功能会启动一些子代理，这些子代理会扮演不同的角色。

But what this does is it spins up sub agents, and these sub agents take on personas.

一共会启动七个子代理，也就是七个角色。

There are seven sub agents that get spun up, so seven personas.

其中一些是固定成员。

Some of them are standing members.

比如，战时COO就是战争委员会的固定成员。

So for example, I have the wartime COO is a standing member of the War Council.

还有冷酷的CFO也是固定成员。

There's the ruthless CFO is a standing member.

还有一个唱反调的董事会成员，也是战争委员会的固定成员。

There's a contrarian board member that are is standing members of the War Council.

然后我留四个角色，让它们动态生成。

Then I leave four personas to be generated dynamically.

所以根据提示，它会决定还需要从谁那里获取建议。

So based on the prompt, it will decide who else who else do I need to get advice from.

那个战时COO，是你创建的角色。

That wartime COO, you created the persona.

你是描述了它的特点，还是用特拉维斯·卡兰尼克的访谈之类的训练了它？

You described its characteristics, or you trained it on, like, Travis Kalanick transcripts or something?

比如，你具体做了什么？

Like, what what did you do?

你是怎么做到的？

How did you do that?

就我而言，这是一个通用版本，但你可以随心所欲地定制。

In my case, I it's a gen it's a generic version, but you could do whatever you want.

对吧？

Right?

如果你想，比如，用特拉维斯

If you wanted to to to have, like, Travis

什么

What does

不过，“通用版本”是什么意思？

generic version mean, though?

你刚才说你就是战时首席运营官。

You just said you be wartime COO.

我说的是，这是战时首席运营官的特质。

Traits of I said this is the traits of a wartime COO.

而且，就像我说的，我实际上做的是，我说，我想要一个战时首席运营官。

And it, like I said actually, what I did was said, I want a wartime COO.

你知道，生成那个角色。

You know, generate that persona.

然后我看了看它生成的内容，做了一些调整，比如……

And then I looked at what it created and just made some, like

你平时是不是挺放松、挺乐观、挺积极的？

Do you have, like, a pretty chill, generally happy, positive outlook?

比如……

Like

你现在看到的是我内心的那个竞争者。

This is this is you're seeing the, like, the the the internal competitor inside of me.

对吧？

Right?

就像科比·布莱恩特的曼巴精神一样，那种感觉涌上心头。

Like, this is the the Kobe Bryant Mamba mentality, like, you know, like, coming through.

我喜欢有这样一个思考伙伴，它有点更冷酷无情，因为

I like having this, like, you know, this sort of thought partner that is a little bit more, like, ruthless because

对。

Yeah.

那不是我默认的模式。

That's not my that's not my default mode.

对吧？

Right?

我希望在那里有那种制衡。

I want to have that kind of checks and balance there.

作为CEO，你身边不一定总有那种既自在又可用，还具备这种能力的人。

And as a CEO, you're not always gonna have people who are comfortable, available, skilled in that way around you.

对吧？

Right?

就像，这就是你希望团队里有的那种人，当你需要他们快速启动的时候。

Like, this is this is, like, who you would want on your team when when you want them to spin up.

问个傻问题。

Quick stupid question.

你说我用Cursor来做AI。

You said I use cursor for my AI.

我以为Cursor是个编程工具，但听起来你是在用它来聊天和做决策。

I thought cursor is a coding tool, but you're it sounds like you're using it to chat and make decisions.

是不是我对Cursor的理解错了？

Am I just wrong about what cursor is?

还是说

Or

对。

Yeah.

我的意思是，Cursor你可以用Cursor、ClaudeCode、Codex。

I mean, cursor you can use cursor, ClaudeCode, Codecs.

这些表面上都是编程代理，但它们都有这种代理模式，你可以直接跟它对话。

Like, these are all ostensibly coding agents, but they have this, like, agent mode where you can just talk to it.

关键是，你跟它对话，因为它能写代码，所以它能帮你做事。

And the thing is it talks to it, but because it can write code, it can do stuff for you.

所以我跟非工程师的人说的时候，我会说，嘿。

And so the way the way I try and talk to people who are not engineers and say, hey.

你应该试试这个，别把它当成工程师用来开发工具的东西。

You should try using this is don't think of it as a tool for, you know, engineers to build stuff.

把它想成你雇了Cursor或ClaudeCode或Codex来当工程师，你把事情交给它做。

Think of it as, like, you're hiring Cursor or ClogCode or Codex to be the engineer you delegate things to.

就像，你肯定跟一些工程师合作过。

Where it's like, you you probably have engineers you've worked with.

你就跟那个工程师说，帮我建这个，或者帮我建那个。

You like to tell that engineer, go build this for me or go build that for me.

这就是我用Cursor的原因，因为它能帮我建东西，而不是我自己去写一堆代码。

Like, that's why I use Cursor is because it goes and builds stuff for me, not because I'm using it to actually go build a bunch of stuff.

我其实不怎么写代码。

I don't really write any code.

等等。

Wait.

所以，当你需要跟AI聊天或者问问题的时候，你直接就打开Cursor了？

So so your default, when you just need to, like, talk to an AI or chat or ask a question, you go straight to Cursor?

你不用ChatGPT或者Claude，还是说……

You don't do ChatGPT or Clot or

……都不用？

anything?

基本上，尤其是在电脑前的时候，我大部分时间都在跟Cursor聊。

I mean, mostly, especially when I'm on my desktop, I'm mostly talking to Cursor most of the time.

它里面有几种不同的模型，你可以来回切换，这点我很喜欢。

It has, like, different models that you can sort of toggle back and forth on, which I like.

所以我可以试试各种不同的模型之类的。

So I can, you know, test a bunch of different models and things like that.

你是打字还是说话？

Are you a typer or a talker?

我最近很多都是用语音转文字。

I I'm I do a lot of voice to text these days.

所以我用这个应用，叫 Monologue。

So I use this, this app Monologue.

为什么用 Monologue？

Why Monologue?

我也不知道。

I don't know.

我就是喜欢它。

I like it.

我试过好几个这样的应用。

Like, I've tried a few of these.

我试过像Whisper Flow和Super Whisper。

I've tried, like, Whisper Flow and Super Whisper.

至于Monologue，我也不太清楚。

And Monologue, I don't know.

它就是，怎么说呢，就是很懂我。

It just, like, it just gets me.

就像，我不知道，我找不到更好的词来形容。

Like, I I don't know, like, I don't know of a better way to describe it.

顺便提一句。

Quick side note.

我觉得这会是未来软件的一个有趣趋势。

I think this is gonna be, like, an interesting thing about software for the future.

当软件开发成本变得这么低的时候，我觉得你会看到更多有个性的软件。

Like, when it is so cheap to build software, I think you're gonna see more software that has, like, a personality to it.

你知道吗，想想那些消费品。

You know, you think about, like, consumer goods.

就像牛仔裤，你能从多少家店买到？或者别的东西，你能从多少地方弄到？

It's like how many places can you buy jeans from or how many places can you get, you know, whatever from.

其实差别不大，但你会选这个牌子，我可能选另一个牌子，就是因为我觉得它更好，但功能上其实一样。

There's, like, not that many differences, but you choose like, you might choose one brand and I might choose a different brand, and it's just because, like, I like it better, but it functionally does the same thing.

我觉得以后会有更多软件给人这种感觉。

I think there's gonna be more software that kinda, like, feels that way.

对。

Yeah.

确实。

It's yeah.

不。

No.

我觉得这点说得很好。

I mean, I think that's a good point.

就像阿迪达斯和耐克的鞋子，功能上其实完全一样，对吧？或者像T恤品牌，你们还记得我昨晚在看那个Blake 102的纪录片吗？

It's like the Adidas and the Nike shoes do the exact same thing, like, you know, or like t shirt brands, like, do you remember I was watching this Blake one hundred two documentary last night.

你们还记得Hurley吗？

Do you guys remember Hurley?

Shaan，你还记得Hurley的T恤吗？

Shaan, you remember Hurley shirts?

那时候到处都是。

They're everywhere.

我昨晚在看那个节目。

And I and I was watching this show last night.

我就想，我们当时为什么那么喜欢那些T恤？

I'm like, why did we love those shirts so much?

我们居然会为那个傻乎乎的logo着迷，是不是挺疯狂的？

Isn't that crazy that we got obsessed with it was just that stupid logo.

而且为什么那个logo对我们来说那么重要？

And why did that matter to us so much?

我觉得这点说得很好。

And that's I think that's a good point.

这就是我们看待B2B的方式，即使是无聊的B2B软件也一样。

That's how we're gonna think about B2B, even B2B boring software.

嗯，我觉得这真的很有意思，因为以前总是靠功能或者价格来区分，对吧？

Well, I think it's a really fascinating point because it always was differentiated on either features or price, right?

这两个基本上就是竞争的主要维度。

Those were kind of the two main axes of competition.

然后设计开始被加入进来。

And then design got layered in.

我觉得有一阵子，更好的设计、更好的UI/UX成了差异化因素，设计也获得了话语权。

And I feel like there was like a wave where better design, better UIUX became a differentiator, and design got a seat at the table.

但如果你看看其他产品，不是软件类的，要么靠品牌，要么靠地位，比如奢侈品，它是在传达你的某种身份。

But if you look at like other products, non software, it'll be either kind of brand or status, like luxury, it's signaling something about you.

这在软件领域并不存在。

That doesn't really exist in software.

听起来你说的意思是，竞争会有一个新的维度，那就是个性。

And it sounds like what you're saying is there's gonna be a new axis of competition, which is personality.

我更喜欢哪个模型给我的感觉，而不是哪个模型参数更多，或者评估基准分数更高，人们不会这样来选择这些东西。

Who do I like the feel of when I when I interact with it versus which model has more parameters and, you know, a higher, you know, benchmark score on the evals is not gonna be how people are gonna choose these things.

因为我已经在这样做了。

Because I'm already doing that.

我觉得你说得完全对。

You're I think you're absolutely right.

嗯，我确实有这种感觉，你知道，作为一个经常玩各种模型的人，我开始发现自己对不同任务有不同的偏好。

Well, and I certainly feel it like, you know, as someone who plays around with a lot of the models, I I'm starting to find myself have some preferences for, like, different tasks.

比如，如果我让它去执行一个工程任务，不需要跟它对话，我有喜欢的模型来做这个。

Like, if I'm asking it to actually go do an engineering task where I don't need to talk to it, there's models I like for that.

你懂吗？

You know?

比如，Fable就非常棒。

Like, Fable is fantastic.

比如说，如果你给它一个非常复杂的工程任务，我会觉得，太好了。

Like, if you're, like, you're gonna give it a really complex, like, you know, engineering task, I'm like, great.

让Fable去试试看。

Have Fable go take a take a swing at this.

它肯定能搞定。

It's gonna nail the thing.

但是，天哪，我可不想跟Fable聊天。

But, man, I do not wanna talk to Fable.

就是，跟它聊天一点都不好玩。

Like, it is it is not fun to talk to

完全不好玩。

at all.

我们都有这样的同事。

We all have coworkers like that.

他们某些任务做得特别棒，但跟他们相处不太愉快。

They crush it at some tasks, but you're not not a great hang.

对。

Yeah.

没错。

Exactly.

但是呢，

But then

还有一些模型，你会觉得，嘿。

there's other, like, models where you're like, hey.

如果我想来回讨论、辩论一下我们在做什么，我会觉得，好吧。

If I wanna sort of, like, go back and forth and have, like, a discussion and a debate around, like, what we're doing, I'm like, okay.

我觉得我能更好地忍受那种讨论，而不是跟Fable聊天，跟它聊的时候我就觉得，

I I feel like I can, you know, tolerate that discussion a lot better than, like, trying to talk to Fable where I'm just like,

天哪。

oh my god.

我一直在想，比如Claude跟我说话的时候，它常用的那些话是什么？

I are all the phrases like, when Claude talks to me, it's like, what are the what are its phrases?

就像是

It's like

不是这个。

not this.

是那个。

It's that.

或者，这里有个真正的问题，虽然不太明显。

Or here's here's the thing that's quietly really the problem.

对。

Yeah.

说实话。

The honest truth.

关键的是，

The load bearing,

你知道，对。

you know Yeah.

好，那咱们开始吧。

Like, alright.

咱们直接进入核心。

Here's let's get to the spine.

对。

Yep.

这些是AI特有的，还是模型特有的？

Are those specific to AI or to the model?

肯定是模型特有的。

To the model, for sure.

我觉得大多数模型都衍生自Anthropic系列或GPT系列。

Like, you can tell most models, I think, are derivative of either either the, you know, like, Anthropic series or GPT series.

所以这两个系列有各自鲜明的个性。

And so, like, you you know, those two kinda have two distinct personalities.

但如果你去用那些开源工具之类的，就能看出它们主要是在模仿谁。

But if you go use, you know, any of the, like, open source tools or things like that, you can kinda see, like, who they've been distilling a lot.

就是，它会继承那个模型的性格，你知道的，

Like, it sort of inherits the personality of, you know,

就是上面那个。

the one above.

哥们，我注意到我的同事开始对Claude说，比如前几天有人说，咱们说到承重部分吧，或者类似这样的话。

Dude, I've noticed that my coworkers are starting to say, like, to Claude, like, someone the other day said, like, let's get to the load bearing part or, like, something like of like

就是他们口头说的时候，不是写下来的时候？

Like when they're talking out loud, not in writing?

对。

Yeah.

对。

Yeah.

对。

Yeah.

对。

Yeah.

我也注意到了。

I've noticed that too.

我也注意到了。

I've that.

当他们开始说那些话的时候，我就想，那是克劳德。

When they start saying things, I'm like, that's Claude.

然后，就是有点像……

And these and and it's sort of like

把他们给带进去了。

plugged them.

对。

Yeah.

就像，或者他们试着那样说话，这样他们就能用克劳德写的东西，然后说，不。

Like or or they're trying to talk that way so then they can, like, use Claude's writing and be like, no.

真的是我写的。

It's really me.

到底是谁在训练谁？

Who is training who?

对吧？

Right?

是我们训练AI，还是AI在训练我们？

Are we training the AIs or the AIs training us?

而且我觉得，其实还可以更进一步。让我给你看看另一个用例，这个也挺有意思的。

And and I think, like, I think you can go a step further where actually, let me show you this other use case because this is interesting too.

这是一堆合成数据，不是真实的。

So this is a bunch of synthetic data, so it's not real.

不过，我手头有真实版本的数据。

But, you know, I have a real version of it.

里面有一列账户，你可以点进去看看每个账户的情况。

And it's got a list of accounts, and you can kinda click into the accounts and see what's going on with it.

比如我们点开这个Acme物流，就能看到很多相关信息。

So if we click into Acme Logistics here, you can see a bunch of stuff about it.

就像，哦，使用量下降了34%。

It's like, oh, usage is down 34%.

四十五天后有一个续约。

There's a renewal in forty five days.

这是他们使用的所有应用。

Here's all the apps they're using.

这是你需要关注的组织里的人，等等。

Here's the people in the organization that you need to care about, etcetera.

但接下来，有趣的部分来了。

But then, like, here's the interesting part.

它给出了建议的CEO外联，比如你应该给Jordan发邮件，问候一下他，看看情况怎么样。

It's got the suggested CEO outreach, where it's like, you should email Jordan and, you know, check-in on him, see what's up.

所以这就是我觉得AI变得非常有趣的地方，因为从某些方面来看，现在谁是老板？

And so this is where, you know, I think AI gets really interesting because in some respects, who's the boss now?

比如，我是老板，还是AI是老板？

Like, am I the boss, or is the AI the boss?

你知道吗，现在每个周日早上，我醒来都会有一份名单，上面列着10个我想联系的客户。

You know, every Sunday morning now, I wake up with a list of, you know, 10 customers that I wanna get in touch with.

而且它还会给我建议，告诉我该说什么、该做什么、为什么要这么做，诸如此类的东西。

And it got proposals for what I should say and what I should do and why I should do it and all that sort of stuff.

说实话，我特别喜欢这样。

And I candidly love it.

这些对我来说以前真的很难。

Like, this is stuff that was really hard for me.

我天生就不是那种特别擅长社交的人，不像你们那样，总是想着要跟每个人保持联系。

Like, I'm not naturally a person who's just, like, you know, ultra networker like you guys are, where I'm just like, oh, I gotta keep in touch with everybody.

所以现在我就让一个智能助手参与进来，跟它说，帮帮我，让我在这方面做得更好。

And so now I've kinda put an agent in the loop and said, like, hate me help me be better at this stuff.

对，我就是这样，每天早上醒来，周日早上就会收到简报。

And, yeah, I just I literally wake up in my Sunday morning briefing.

上面写着，这些是你需要关注的账户。

It says, here's all the accounts you need.

我已经把邮件草稿放在你的收件箱里了。

I've already drafted the emails in your inbox.

你只需要审阅一下，做点修改，或者直接点发送就行。

You just need to review them and make edits or press send.

我觉得这里面有点门道，如果你把这些工具搭建得好，它们慢慢就会开始告诉你该怎么做。

And I think there's, like, something to this where, you know, if you if you do a good job of, like, building these tools up, they'll kinda start telling you what to do.

不过你还是得保持清醒，因为如果你不动脑子，有时候可能会说些傻话。

And you do have to keep your brain on because if you turn your brain off, you'll probably sound dumb from time to time.

但总的来说，我觉得在某些场景下，这些东西能当挺好的老板。

But by and large, I think these things make pretty good bosses for certain use cases.

嗯，Brian Halligan，HubSpot的创始人之一，他要么是自己想出这个主意的，要么就是在Brian的播客上，Twitter的Jack Dorsey提出了这个想法。

Well, Brian Halligan, the one of the founders of HubSpot, he either came up with this idea or it was in an interview where Jack Dorsey from Twitter came up with the idea and it on Brian's podcast.

但他们基本上说的是，很多人感到困惑。

But they basically said, a lot of people are confused.

他们以为会有一个像人脑或CEO一样的东西在中心，然后周围有一堆代理跑来跑去干活。

They think it's gonna be a human brain or a CEO in the center, and then all these agents running around doing stuff.

他说，不能这么看。

He was like, that's not the way to look at it.

AI才是大脑。

AI is the brain.

AI将成为CEO。

AI is going to be the CEO.

而人类的职责就是给它提供信息，然后在它是否听从AI的问题上做一些判断。

And the humans, their job is just giving it information and then making some judgment calls if it wants to listen to the AI.

但总的来说，AI应该真正做出那些艰难的决定，因为它很可能做得更好。

But in general, the AI should be actually be be making the hard decisions because it's likely gonna do a better job.

所以这改变了我的看法。

And so that kind of, like, changed how I thought about it.

听起来你也是这个意思，只是没这么说而已。

And it sounds like that's sort of what you're saying without saying it that way.

嗯，我觉得是这样。

Well, I think so.

对吧？

Right?

如果你能把公司所有的机构知识都接入AI，它能处理的信息量会比人类CEO多得多。

If you are able to hook up all of your company's institutional knowledge to the AI, it's gonna be able to reason over way more information than a human CEO.

就是，它确实能做到。

Like, it just can.

比如，它能查看代码库里的每一次提交。

Like, it can it can look at every single commit into the code base.

它能读完每一段客户对话。

It can read through every single customer conversation.

它能扫描Reddit、X或LinkedIn上关于你的每一条评论。

It can scan every comment about you on Reddit or X or LinkedIn or whatever.

所以它知道的肯定比你作为人类能知道的要多得多。

So it just knows way more than you as a human is ever gonna know.

因此，在某些类型的决策上，它会更有用。

And so that's gonna make it way more useful for certain types of decisions.

比如说，现在刚开始的时候，你把它当作一个助手来帮你做这些决定。

Like, at like, to start right now, it's like you use it as an assistant to sort of help you make these decisions.

但我觉得很有可能你会开始说，嘿。

But it seems pretty plausible to me that you might start saying, hey.

在某些情况下，我可能会更经常采纳它的建议，对吧。

I I'm gonna take a suggestion more often than not in certain Right.

某些地方。

Places.

对。

Yeah.

我觉得确实如此。

I think that's so true.

我在想，AI会不会有那种，因为你知道，好处是AI不会有像人类那样的情绪起伏，或者那些会阻碍它做正确事情的恐惧，或者避免冲突，或者人类所拥有的各种偏见。

I wonder I wonder if AI is gonna have, like because, you know, the upside would be AI is not gonna have the emotional roller coaster that a human would have or fears that would hold it back from doing the right thing or avoid, you know, avoiding conflict or, or different things that humans have as their biases.

然后我在想，我们是不是得留意AI的认知偏见。

And then I wonder if we're gonna have to come up with the cognitive biases of AI to watch out for.

比如说，我们知道有沉没成本谬误，还有近因偏差。

Like, okay, we know we have this sort of sunk cost fallacy, and we have this bias for recency bias.

那AI会有什么样的偏差呢？

What is AIs gonna be?

因为我们越是把决策权交给AI，就越需要留意这些偏差。

Because we're gonna need to be aware of those the more we sort of turn over decision making to AI.

对。

Yeah.

今天的播客由我的朋友们在Mercury赞助播出。

Today's podcast is brought to you by my friends at Mercury.

他们做的是全球最好的银行产品。

They make the world's best banking product.

我想你已经知道了。

I think you know this already.

我所有的业务都在用Mercury。

I use Mercury for all of my businesses.

我觉得我大概有七八家公司吧。

I think I have, like, maybe seven or eight businesses.

我们所有公司的企业银行账户都用的是 Mercury。

We use Mercury as our business banking across all of them.

而且他们最近刚刚推出了个人银行账户。

And now they actually just launched a personal banking account.

所以我的个人账户也开在那里。

So I have my personal account there.

我已经把富国银行和摩根大通的账户都关了。

I moved off of Wells Fargo and Chase.

我现在完全只用 Mercury。

I'm just all in on Mercury.

为什么？

Why?

因为我喜欢用起来简单的产品。

I like products that are easy to use.

我喜欢那些能理解我和我遇到的问题的产品。

I like products that get me and the problems that I have.

所以，比如，和妻子开联名账户非常容易。

So, like, very easy to make a joint account with my wife.

创建虚拟卡也非常简单。

Very easy to spin up virtual cards.

一键就能获得储蓄收益。

One click, and I get savings yield.

我需要的一切都集中在一个地方。

It just has all the stuff that I need in one place.

所以，如果你在寻找市场上最好的银行产品，那绝对是Mercury。

So if you're looking for the best banking product on the market, it's definitely Mercury.

谁不同意这一点，我就跟谁干一架。

I will fist fight anybody who disagrees with me on that.

去mercury.com/personal了解更多。

Go to mercury.com/personal and learn more.

Mercury 是一家金融科技公司，不是 FDIC 承保的银行。

Mercury is a fintech, not an FDIC insured bank.

银行服务由 Choice Financial Group 和 Column, N. 提供。

Banking services are provided through Choice Financial Group and Column, N.

A，FDIC 成员。

A, members FDIC.

所以你经营着一家价值数十亿美元的公司。

So you run a company that's worth billions of dollars.

你每天做的决策中，有多少实际上是 Wade 和 AI 之间的较量？

How many decisions are you making a day that it's actually Wade versus AI?

嗯，我觉得，任何高风险的决策，现在 AI 都会参与意见。

Well, I I would say, like, any high stakes decision, the AI is weighing in on now.

你知道，就像在管理层会议和董事会会议上，AI 也有一席之地，我们会让它把这些事情提出来。

Like, you know, there's sort of a seat at the table and, you know, the exec meeting and the board meetings and all that sort of stuff where we're asking the AI to, like, surface those things.

有没有一个例子，AI 改变了你的想法，或者得出了你没想到的正确结论，然后你决定采纳它的建议？

Is there an example where AI kinda changed your mind or had got to the right answer that you weren't at and you decided to go with it?

我举个例子，说明我觉得AI特别有用的地方。

Here's an example of a place where I find AI to be really useful.

咱们回到招聘的例子。

Let's come back to the hiring example.

你知道，我们有一个“抬杠者”流程，我批准每一份工作邀请，但高管团队成员会参与招聘流程的最后阶段。

So, you know, we we run a bar raiser process where I approve every job offer that goes off, but, like, exec team member sort of is on the final part of the hiring loop.

这个流程的目的是审计招聘过程，同时也要弄清楚，嘿。

And it's intended to audit the hiring process in addition to sort of figuring out, hey.

这个人会不会提升Zapier的文化和工作水平等等。

Is this person gonna elevate the culture, the work, etcetera, here at Zapier?

在过去，没有AI的时候，当我处理这些审批时，我经常会——你知道，如果我遇到一个不太确定的候选人，我能感觉到有点不对劲。

Now in the past, pre AI, when I would go through those approvals, I would often you know, if I would come across a candidate where I was not so sure on, I could kinda smell something is wrong.

我面试的经验已经够多了。

I've just done enough interviews now.

我面试过几千人。

I've interviewed thousands of people.

我看了成千上万份申请，我的经验比大多数招聘经理都多。

I've looked over thousands of applications where I just have more reps than most hiring managers have.

大多数招聘经理一年也就招一两个人。

Like, most hiring managers hire hire one or two people in a year.

而我呢，我经手的量远远超过他们。

It's like, I'm I'm I've just looped over, like, way, way more volume.

所以有时候我说不清具体哪里不对，但我就是能感觉到有问题。

And so I don't always have the language to say, like, why it's wrong, but I can still smell it.

所以我经常去找他们，问他们：你确定吗？

I'm like and so I would often go to these folks and be like, are you sure?

感觉有点不对劲。

Something is off.

我觉得这个人缺这个、缺那个。

Like, I don't I don't think they have this, this, or that.

这挺有意思的。

And it was interesting.

我和员工的这种互动总是有点紧张。

That interaction with my staff was always a little bit fraught.

比如，他们会说，我支持这个人。

Like, folks would be like, well, I stand by this person.

等等。

Like, wait.

你是打算直接做招聘决定吗？

Are you just gonna make the hiring decision?

如果是的话，你如果不想雇他们，就别雇。

If so, like, you just if you don't wanna hire them, don't hire them.

然后我就会说，那其实不是我想要的。

And it sort of be like and I'd kinda be like, well, that's not really what I want.

对吧？

Right?

我希望你们来做决定。

Like, I want you to make the decision.

我只是想给你一些指导，希望你能自己推理出这个问题。

I'm just trying to, like, give you some guidance here, and I want you to sort of be able to, you know, reason through this problem.

如果你已经回答了，而且你有信心高效地回答了，那我就会支持你。

And if you've answered it, if you have confidence that you have, you know, efficiently answered it, then I will back you.

咱们去雇这个人吧。

Let's go hire this person.

但如果你还没回答，我希望你把这个考虑进去，然后去弄清楚，比如，你是否需要获取更多信息，或者需要调整你的方向，等等。

But if you haven't, then I want you to, like, take this into account and go figure out, like, you know, is there more information you need to get, or do you need to adjust your course, etcetera.

但实际情况往往不是这样。

But that's not often how it would play out.

它通常就会变成，嗯，韦德想让我雇这个人，还是不想让我雇这个人？

It would just sort of play out as like, well, does Wade want me to hire this person, does Wade not want me to hire that person?

然后我就，唉。

And I'm and I'm like, ugh.

所以，总之，这就是AI之前的情况。

Like so, anyway, that's kind of what it was happening pre AI.

有了AI之后，我会把作战会议或者招聘委员会交给这些家伙来主持，然后我会注意到，它比我更能说会道。

Post AI, I'd run the War Council or, like, my hiring committee over these folks, and then I would notice, like, it's just more articulate than me.

它能说，嘿。

It could sort of say, hey.

我注意到，这个人有这些特质，而评审小组没有仔细审查这一点。

I noticed that, you know, this person has these traits, and the panel did not scrutinize this.

所以你需要去仔细审查这个。

So you need to go scrutinize this.

然后它会给出一个评分，比如，和你在Zapier雇过的所有人相比，我觉得这个人成功的可能性有多大。

And it would, you know, it'd give, like, a score of, like, here compared to everyone you've hired at Zapier, here's how here's how likely I think they're gonna be successful or not.

所以现在我可以把这个带到招聘小组面前，说，嘿。

And so now I could take that to the hiring panel and say, like, hey.

AI说了这个。

The the AI said this.

我同意。

I agree.

这到底是怎么回事？

Like, what what what's going on here?

然后我突然注意到大家的行为发生了变化，人们会说，哦，有意思。

And I noticed all of a sudden the behavior shift where people would go like, oh, interesting.

我觉得AI在这点和这点上是对的，但在这点上它错了。

I think the AI did get this and this right, but I think it's wrong on this.

原因如下。

And here's why.

因为我有它没有的其他背景信息。

Because I have this other context that it doesn't have.

我就觉得，太好了。

And I'm like, great.

这正是我想知道的。

That's all I was wanting to know.

就像，我还想知道，我还漏掉了什么？

It's like, I want like, what else what else am I missing?

我在这里没看到什么？

What am I not seeing here?

而且我发现，可能是因为跟AI争论比跟CEO争论更容易，也可能是因为它更有说服力，说话方式让人更容易针对具体问题来讨论。

And I just found that, like, be maybe because it's not maybe because it feels like arguing with an AI is easier than arguing with the CEO or maybe because it's more eloquent and it, like, actually has a way of talking that makes it, like, easier to address the specific points.

我不太确定，但我确实注意到，我们在招聘方面的决策质量提高了，部分原因是我们把AI加入了决策流程。

I'm not exactly sure, but I did notice the way we made decisions around hiring got better in part because we added the the AI to the loop.

我不知道你们是否曾经相信过一些性格测试。

I don't know if you guys were ever, like, believers in some of these personality tests.

比如有个叫Culture Amp的东西，我想是这么叫的，它是一个软件，在招聘前让你做性格测试。

Like, there's this thing called Culture Amp, I think it's called, where it's like a software where you do personality tests before you hire someone.

雷·达里奥也有他自己的版本，随便吧。

Ray Dalio has his version of it, whatever.

而且通常有一个尺度，人们会落在上面，比如他们可能会说，我不相信任何涉及性格测试的东西。

And there's like tends to be it's like a scale where people fall in where they're like, I don't believe in anything involving, like, personality tests.

我注意到，有了AI之后，我更倾向于听从它的建议，因为我做过一些性格测试，并用它来训练我的AI，这对我帮助很大。

I've noticed with AI, I have fallen closer to the side of, like, I follow what it says because I've done some of these personality tests and I've used it to train my AI and it's helped me a lot.

但说到招聘，你们会相信这些性格测试吗？就是那种判断一个人适不适合这份工作的测试。

But when it comes to hiring, do you guys buy into this at all when it comes to personality tests and, like, figuring out if a person could be good for the job?

因为有了AI，这个过程会变得容易很多。

Because that makes with AI, it makes the process so much easier.

你觉得呢，

What

肖恩？

do you think, Sean?

我的想法正好相反，其实跟性格测试本身关系不大。我觉得就像自动驾驶一样，一种思路是把所有传感器都装到车上。

I I go the the other way, meaning it's not really about personality tests per se, but I find that, you know, like with self driving, one approach was let's put all the sensors on the car.

装摄像头。

Let's get cameras.

再多装几个摄像头。

Let's get more cameras.

再装激光雷达。

Let's get LIDAR.

我们装上雷达。

Let's get radar.

我们装上声呐。

Let's get sonar.

我们装上……还有什么能装的？

Let's put a What else do we do?

味觉测试。

Taste test.

所有能装在车上的不同传感器都算上。

Are all the different possible sensors you could put on the car?

而特斯拉的做法基本上就是，它只需要视觉。

And the Tesla approach was basically like, it needs vision.

只需要眼睛。

It needs eyes only.

人们认为其中一个原因是成本，我觉得激光雷达非常贵是一个因素。

And one of the reasons why people thought it for cost, which was I think a factor that LiDAR was very, very expensive.

所以从长远来看，这并不实际。

So it was not practical in the long term.

但埃隆说这么做是因为，当传感器太多时，你会得到太多相互矛盾的数据，不知道该信哪个。

But the reason that Elon says that he did it was because when you have too many sensors, you get too much conflicting data and you don't know which one to trust.

这样更容易出错，更容易出问题。

It makes more faulty, more error prone.

实际上你应该做的，是只保留所需的最少传感器，然后对它们进行充分训练。

And actually what you should do is take only the minimum number of sensors needed, but then train the hell out of them.

这有点像我的招聘思路，我现在会简化。

That's kind of my approach to hiring, where I now simplify.

我需要花在一个人身上的时间更少，我寻找的要素也更少，但这些要素必须是决定性的。

I need less time to spend with somebody, and I look for fewer things, but those have to be home runs.

否则，我就不会录用这个人。

Otherwise, I'm just not gonna hire the person.

你是什么意思？对。

What do you Yeah.

对。

By Yeah.

不过你得回答问题啊。

Answer the question though.

就是说，你明白吗？

Like, do you get okay.

你的意思是，不是看几百个问题，而是只看三个问题。

What you're saying is instead of looking at like hundreds of questions, you're looking at three questions.

你擅长这个、这个还是那个？

Are you good at this, this or that?

好，那你怎么收集这些信息呢？

Okay, so how do you gather that information?

所以，又是从埃隆那里偷学的。

So again, stealing from Elon.

他说，你要找的是有卓越能力的证据。

He talks about, you want evidence of exceptional ability.

所以我找的就是，不是那种——第一个筛选条件就是有没有非凡能力的证据。

So all I'm looking for, I'm not looking for like The first filter is just evidence of exceptional ability.

所以我现在不是要评估文化契合度。

So I'm not trying to assess culture fit right now.

我不是在问，我喜欢他们吗？

I'm not asking do I like them?

我不是在问，他们有没有相关经验？

I'm not asking like do they have the relevant experience?

我只想知道，你以前有没有干过什么特别牛的事？

All I wanna know is have you done some exceptional shit before?

然后把这个故事讲给我听。

And then tell me that story.

在那个故事里，我会试着搞清楚，是你自己干的吗？

And in that story, I'm gonna try to understand, did you do it?

还是你只是在一个团队里跟着干的？

Or were you on a team of people doing it?

还是说你是沾了别人的光才做成那件事的？

Or were you riding somebody else's coattails doing that thing?

如果你真的没做过什么特别出色的事，那指望这里成为你职业生涯中第一个做出非凡成就的地方，这个赌注我可不太想下。

And if you really haven't done anything exceptional, the odds of this being the place where you have the first exceptional thing of your career is not really a bet I'm excited to go take.

所以这就是我筛选的第一个人标准，而且这不是靠性格测试来决定的。

And so that's the very first filter that I'm looking for, and that's not personality test driven.

接下来就是某种性格测试了，比如我会和他们一起做点事，凭直觉感受一下，我们之间有没有那种创意上的默契。

Then after that is a kind of personality test of like, I try to do something with them where I get the gut feel of like, do we have creative chemistry?

当我在琢磨某个问题的时候，我会觉得这个人适合待在房间里一起讨论吗？

Do I feel like this person is good to have in the room when I'm trying to figure something out?

还是说，当这种情况发生时，我觉得他们并没有让对话变得更有深度？

Or do I feel like they're not elevating that conversation when that happens?

所以其实我的筛选就两步，就这么多。

And so I'm really just kind of like a two step filter and that's it.

我只需要一个信号就够了。

And that's all I want a signal.

但显然，我管理的团队规模不是那样的。

But obviously, like I'm not running the size of org that way it is.

和我直接合作的人都是在一个非常非常小的团队里，每个人都必须是十倍效率或者更好。

And I'm working with people who are working directly with me in a very, very small team where everybody has to be, you know, 10 x or better.

否则，就不值得让他们留在团队里。

Otherwise, it's just not worth having them on the team.

所以这是我通过艰难的方式学到的，因为我以前的做法非常不同。

And so that's the I've learned that kind of the hard way because I used to do it very differently.

对。

Yeah.

我们在这方面做过一些工作。

We so we've done some work with this.

比如，我们内部有时会做迈尔斯-布里格斯类型测试。

So, like, we've done, like, Myers Briggs type stuff internally at times.

我们还做过一个叫伯克曼的测试。

We've done there's this one called Berkman that we've done at times.

我觉得这东西有点用，但也就一般般。

I have found it to be moderately useful.

我有个CEO朋友，在哥伦比亚经营一家上市公司，他给每个新员工都做DISC性格测试，而且特别坚持，比如招会计，就一定要这个性格类型，必须符合那个画像。

I have a CEO buddy who runs this public company in Colombia that does disk profiles for every hire, and he's, like, insistent where it's like, I have an accountant, they need to be this profile, this dotted it on.

就是说，如果不符合，那他们就不适合

It's like, if they're not, they're not gonna be

团队。

a good fit on the team.

我还没到那个程度。

I'm not going that far.

我的最终看法是，首先，你确实需要一个人非常出色，这是最重要的。

I think what I where I net out is that I mean, you do need somebody who is exceptional, first and foremost.

每个人都有性格类型，但这并不能说明你能力强不强。

Like, everybody's got a personality type, but it doesn't tell you if you're good or not.

比如，你测出来是什么结果就是什么结果，但你仍然得有能力才行。

Like, you know, you you can score whatever you're gonna score, but you still need to be good.

对。

Yeah.

但我确实觉得，团队很重要。

But I do think that, like, teams matter.

比如，当我审视自己的性格或工作方式时，有些方面我特别擅长，也有一些固定的习惯，我知道如果我能和一个有互补特质的人搭档，我会表现得好得多，因为他们能在我没注意到的地方帮我兜底。我不知道这背后有没有什么科学依据。

Like and when I look at my my personality or my way of working, there are things where I am exceptionally good at, and there are certain defaults that I have where I know that if I am partnered up with somebody who has, like, a complimentary traits, I'm gonna be way better because they're gonna have my backside in places that I I don't know if, like, there's any science behind this.

但你看我的行政助理和我，比如迈尔斯-布里格斯性格测试，她有三个字母和我不同。

But if you look at my EA versus mine, like Myers Briggs, she has three different letters than I do.

那个测试有四个字母，其中三个她的都不一样。

So there's four letters in that, and three of them of hers are different.

她确实喜欢做那些我不喜欢做的事，而且她擅长那些我不擅长的事。

And she does, like, have she, like, likes doing things that I don't like doing, and she's good at those things that I don't like doing.

结果这真的很有帮助，就像在体育里，比如篮球队，你需要有人去投篮，比如迈克尔·乔丹会投所有的球。

And it ends up being, like, pretty helpful because it gets kinda like in sports where it's like, you know, basketball team, it's like, well, you need somebody that's gonna you know, Michael Jordan's gonna shoot all the shots.

但你也需要丹尼斯·罗德曼去防守和抢篮板。

Well, you need Dennis Rodman who's gonna, like, lock down d and grab boards.

但你们招聘的时候会这么做吗？

But are you doing that for when you guys are hiring?

我们不会这么做。

We don't do it.

不会。

No.

我们实际上不会让人做测试，然后说，嘿。

We don't actually have people, like, take a test and say, like, hey.

你会怎么做？

What are what are you going to do?

不同的招聘经理可能多少会自然地这么做，但对我们来说这不是一个科学的方法。

Different hiring managers are probably doing it somewhat, like, by like, organically, but it is not a science for us.

我有个想法，关于公司如何用AI来改进招聘。

I have I have an idea for how I think companies could use AI to improve their hiring.

我不确定这是否完全合法。

I don't know if this is, like, fully, like, legal.

我不知道能不能完全按我说的这么做，但大概就是这么个练习。

I don't know if you could do this exactly the way I'm gonna describe it, but here here's the exercise.

我加入。

I'm in.

你回顾一下你的团队，把过去三年招进来的人都列出来。

You go through your team, so you have all the past hires you've made in the last, let's call it three years.

然后你可以给他们排个序，谁表现特别出色，谁表现一般，谁的表现时好时坏。

And you could stack rank who's exceptional, who's acceptable, and who's kind of like been a source of kind of hit or miss performance, let's say.

好。

Cool.

这样你就能给团队里的员工排序，然后说，太好了。

So you could stack rank employees in your team, and you could say, great.

你手头可能还有他们面试时的文字记录或者录音。

You also have maybe the the transcripts or the the recordings of all the interviews that they did.

对吧？

Right?