我辍学后从零开始打造了一家价值 36 亿美元的公司

你可以用你的品牌语气，结合真实的客户数据，起草活动文案、博客文章和邮件。

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

所以你做的不只是内容。

So you don't create just content.

你做的是能带来转化的内容。

You create content that converts.

去了解一下 hubspot.com，这是为成长型企业提供支持的智能客户平台。

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

你还有没有其他比较反直觉的观点？比如，彼得·蒂尔问过，你相信什么但别人大多不认同？

You you have a couple other kind of contrarian you know, you have a good answer to the Peter Thiel question of like, what do you believe that few others would agree with you on?

我念几个给你听。

I'll read you a couple of them.

你基本上对就业市场持乐观态度。

You know, you're basically, like, bullish on the job market.

大多数人觉得有了人工智能，我们的工作都要没了。

I think most people think with AI, all our jobs are gone.

你却说，不会的。

You're like, no.

对就业市场持乐观态度。

Bullish on the job market.

其他人觉得有了AI，我们都会越来越少工作。

Other people think with AI, we're all gonna be working less.

埃隆说过类似的话，还有几个人也这么认为，你知道的，他们说——

I think Elon has said this and a few others, like, you know, you know, hey.

说什么每周都是四天周末之类的。

Here here comes the four day weekend every weekend or something.

你却说，不会的。

You're like, no.

我们会工作得更多，而像SaaS这样的软件公司会发展得很好。

We're gonna be working more, and that software companies, like the SaaS companies, are gonna do well.

所以，就业、努力工作、软件公司——这三样东西，大多数人我觉得都挺看衰的。

So, you know, jobs, work hard work, and and software companies, three things that most people, I think, feel pretty bearish on.

你有不一样的看法。

You have a different opinion.

你能说说你对这几点的看法吗？

You want to give us your take on each one of those?

很多人觉得工作会消失，或者我们会干得更少，其实这种想法背后，是低估了人类的创造力和想象力，以为我们对新事物的需求是有限的。

You know, so much of the idea that jobs go away or that we do less, you know, has to come from a place of effectively your short, just human creativity and ingenuity, and the idea that there's sort of, we don't have an insatiable appetite for more and new things.

但我看到的证据很少，表明我们不再想去寻找下一个疾病的疗法，不再想体验新的娱乐形式，不再想推出新的消费品，或者不再想制作新的播客。

And I've just seen very limited evidence that suggests that we don't want to go and discover the next cure for the next, you know, niche problem that people have, or the next new form of entertainment people want to experience, or the next new consumer product that people want to go and sell, or the next, new podcast that wants to be created.

如果你不相信这些会继续发生，那确实会认为，这种富足是用失业和减少劳动为代价换来的。

And so if you don't believe that that's going to happen, then sure, then you would basically believe we have that abundance sort of comes at the expense of jobs and of us doing things.

这种观点的前提是，AI代理会包揽所有真正有用的工作，让我们闲到没什么事可做。

And that theory has to be that basically, the agents are going to do all of the effectively the useful work, which then frees us up so much that there's really not much else left.

但我相信，不管好坏，我们总会找到办法给自己创造更多工作。

And I just think we will find a way to create a ton more work for ourselves, for better or worse.

我的看法其实也没那么理想化。

Like, it's not obvious that my view is particularly like Utopian.

说实话，某种程度上，马斯克的愿景，说句公道话，其实比末日论者更理想化。

Like, think, to some extent, Elon's vision, to his credit, is actually a far more Utopian one than the Doomers.

挺讽刺的是，他们其实信的是同一种技术趋势。

It's actually funny, like, because they believe in the same underlying technology trend.

他们俩都认为，如果把AGI的发展趋势推到极致，末日论者觉得这会毁灭人类，而马斯克则觉得我们会迎来一个一切都被替我们安排好的乌托邦。

So, they basically both believe in if you extrapolate out to AGI, then the doomers believe that destroys us, and Elon believes that we get this utopia where everything is just done for us.

我嘛，暂时还没法下定论，可能站在第三种立场——既不是完全认同他们中的任何一个，而是觉得最终结果会更务实：我们会用这项技术创造出一批新的需求，然后去一一满足它们。

And I kind of am more just until proven otherwise, maybe just in a third camp, which is like it's like the same progress of maybe both of those two, but with more of a pragmatic outcome, which is like we use that technology to just then create a new set of needs that we have to go in all kind of support and fulfill.

这世界依然会看重一大堆事情，清单长得不得了。

It's just like a very long list of things that the world is still going to effectively value.

比如，人们还是会看重孩子面对面的教育。

Like the world is going to still value in person education for children.

也会看重托儿服务。

They're going to value childcare.

还会看重去餐厅吃饭，享受人与人之间的互动。

They're going value going to a restaurant and having human interaction.

人们还是会看重去看一场演出。

They're going to value going to a show.

人们会看重找一位财务顾问，他不仅对市场和你的需求有理解，还能参考其他十位客户的状况来帮你做判断。

They're going to value talking to a financial advisor that appears to sort of have some sense of the market and your set of needs, and also has 10 other clients that they can kind of triangulate with.

或者找一位税务专家，出了错他得担责任，饭碗都可能不保，而不是一个随时能关掉的智能代理。

Or a tax professional that you can just like, you know is like accountable for if they get the thing wrong, like their jobs on the line versus like an agent that can be just shut off.

所以你根本不清楚这些AI代理到底负不负责。

So, you're not really sure what accountability they have.

正因如此，人类依然会留在决策环中。

So, like for all these reasons, like humans just stay in the loop.

我觉得，我们每个人未来要做的工作只会更多，而不是更少。

So, I just think we end up having still a lot more work for everybody to do.

至于每周工作四小时、四天的说法，其实暗含了一个前提：你得相信，只要行业内有家公司不实行四天工作制，它靠AI就能比你多产出20%到25%。

I would say then the four hour the four day work week thing also kind of suppose something different, is you basically have to believe that assuming that anybody in your sector decides not to do a four day work week, then that company with the power of AI will get 20% more or 25% more output than you will.

那么，到底哪个行业会达成共识，规定所有公司都只能每周工作四天？

And so which market is going to basically have some kind of like collective agreement that says, no, our category, everybody in our industry is only going to work four days a week.

所以这根本需要所有人达成一种共识，不可能有哪个角色突然决定：算了，我就多开发点软件吧。

So it's just like, it requires such a collective sort of like agreement on the part of everybody that you wouldn't then just like have some actor in the system decide, no, I'm just going to like, I will just ship more software.

我就比你多拉点客户，结果整个行业又回到每周工作五天。

I will sell to more customers than you do, which then gets everything back to five days a week.

所以，这种结果根本不太可能真的出现。

So, that's why it's like just very implausible for that outcome to really exist.

你听这个。

So, check this out.

有本书，Shaan，你一定会喜欢这个书名。

There's this book, Shaan, you'll like the name of this title.

书名叫《如何活好每一天二十四小时》，是1908年写的。

It's called How to Live on twenty four Hours a Day, and it's a book written in 1908.

你们真该读一读。

You guys should read it.

特别有意思，我刚拿到这本书。

It's really cool, I just got it.

这本书讲的是工业革命之后发生的事。

And it's all about what happened after the industrial revolution.

当时出现了大量白领岗位的激增。

And there's this huge burst of white collar jobs.

如今，数以百万计的美国人和欧洲人不再在工厂干活，而是从事这些白领工作。

And there's now millions of Americans, as well as Europeans who now are not in a factory anymore, and they're doing these white collar jobs.

书里还提到很多家庭主妇，她们说：‘现在我不用手洗衣服了，用机器就行，空出好多时间。’

And then there's all these like housewives, this is in the book, and they're like, Well, now I don't launder our clothing with my hands, I use a machine, and I have all this time.

白领们也觉得：‘我们确实多了不少空闲时间。’

And the white collar workers are like, you know, we have extra time.

于是大家都开始问自己：既然现在有这么多空闲时间，为什么反而觉得比以前更忙了？

And they're all asking themselves, Well, if we have all this extra time now, why do we feel busier than ever?

整本书的核心就是教你如何理解并充分利用这二十四小时，让每一天过得更充实。

And the whole book is how to make sense of, like, to, like, make your twenty four hour day, how to get everything you can out of it.

它有点像一本关于效率提升的书。

And it's a little bit of like a productivity book.

在

On the

关于‘比以往更忙’这点，你认识的每一个初创公司创始人，现在都忙得不可开交。

busier than ever thing, I mean, it's every startup founder you've ever met right now absolutely is busier than ever.

他们比以前没有AI的时候忙多了。

Like they're way busier than we were when before AI.

原因在于AI是一种具有欺骗性的技术，它让你能轻而易举地同时启动无数个项目。

And the reason for that is because AI it's sort of like this deceptive technology because it like it lets you get started on so many things so easily.

但你还是得把所有启动的事情做完。

But then you still have to complete all the things you started.

你可能会想，我就先部署一堆智能代理吧。

And so you think that like, I'm just going to deploy all these agents.

然后我就去酒吧放松一下，或者出去玩。

And then I'm going to like go to the bar or go hang out.

可等这些代理完成任务后，总得有人来决定：接下来该怎么处理这些信息？

But like when the agents are then done, somebody still has to be responsible for like, what do I do next with that information?

接下来我该怎么处理这个软件？

What do I do next with that piece of software?

那个生成的视频片段，我又该拿它怎么办？

What do do next with that video clip that got created?

所有这些最终还是得靠人来完成。

All of that becomes human work again.

我觉得现在最沉迷AI的人，其实都被工作淹没了，因为我们给自己开了太多头。

So I think every single person that is the most AI pilled right now is just we're just drowning in work because we're kicking off way more work for ourselves.

我们根本停不下来，因为现在启动这些任务实在太容易了。

And we can't ever get off that treadmill because of how easy it has become to just create this work.

不知道在这通电话前一小时，我启动了两个流程，其实根本没必要我亲自动手，但我还是做了。

Don't know, an hour before this call, I kicked off two kind of processes that now I didn't even need to start, but I started them.

现在我得再花一小时，去处理那些AI生成的结果。

And now I'm going to absolutely add another hour to my day because I'm going to go and do whatever the agent produced.

我得去跟进所有这些后续工作。

I'm going go and follow-up with all that work.

其实我根本没必要做。

Like and I didn't even have to.

但就是因为太容易启动了，结果我自己又多了一堆活儿。

But but it was so easy to kick it off that now I've created more work for myself.

所以以后每件事我们都这么干了。

So we're just going to do that for everything.

你漏掉了一个关键点。

Here's the one thing you're missing.

你没给它命名。

You didn't name it.

这就是你的杰文斯悖论。

This is your Jevan's paradox.

这正是你留下百年影响的机会。

This is your chance to live on for the next hundred years.

我们需要‘莱维悖论’——事情越容易做，你干的活儿就越多，每天结束时也越累。

We need Levie's paradox as basically the easier it is to do work, the more work you'll do, and the more tired you'll be at the end of every day.

是的

Yeah.

我的意思是，如果我们真想这么搞下去，那就叫Levie法则吧，老兄。

I I I mean, if we wanna run with that, we can we can we can Levie's law, dog.

别开玩笑了。

Come on.

搞砸了。

Fumble down.

我们不妨加点押头韵吧

We might as well get some alliteration in

对。

there.

是的

Yeah.

所以我们这是在立个法则。

So We're making a law.

你现在是不是在拼命工作？

Are you are you so you're working your ass off right now?

是的，我在拼命。

I am.

对。

Yeah.

简直疯狂。

It's insane.

我们最近请了一位Replit的CEO，他特别棒。

We had a Replitz CEO on recently, and he was amazing.

他讲了个故事。

And he told told the story.

Shaan说，这是他听过最真实的话。

Shaan's like, that's the realest thing anyone has ever said.

还挺有意思的。

It was kind of funny.

但他讲过一个故事，说在公司刚起步之前，他们曾整整好几年处在无人问津甚至濒临失败的状态。

But he told the story about how before they kind of took off, they're kind of in no man's land or even failing for like a handful of years.

大家都听过那条员工发的Slack消息或者短信，上面写着：‘我们能聊聊吗？’

And everyone knows this Slack message or this text message from an employee that says, Hey, can we talk?

他说，那时候我每天都能收到这么一条。

He was like, I got like one a day.

所有人都在辞职，我的神经简直都快崩溃了。

And so everyone was quitting and like my nervous system was just crashed.

是的。

Yes.

我们每个人都会经历这样的周期，但你是怎么坚持了二十年，还能让自己的神经撑得住的？

And we all go through cycles of that, but how has your nervous system like kept up doing this for twenty years?

因为你经历过太多疯狂的事了。

Because you've had some crazy shit happen.

我记得好像还遭遇过一次恶意收购的企图。

I think that you had like a hostile takeover attempt.

你提到过，曾经经历过几次过渡性的融资。

You said you've had like these bridge rounds happen.

我的意思是，你经历的风浪实在太多了，按理说你早就不需要再这么拼了。

I mean, you've like been through so so much shit, and I would assume you don't need to be doing this anymore.

你的身体是怎么扛住这些的？

How has your body handled this?

我不确定身体有没有扛住，但从大脑的角度来说，这真的非常非常有压力。

Well, I don't know that my body has handled it, but but I'd say my from a brain standpoint, it's very, very stressful.

我会定期去看心理治疗师，就为了帮自己平复焦虑的情绪。

I see a therapist just, like, to help me, like, calm myself down from an anxiety standpoint.

就像阿马德说的，那些是最让人崩溃的Slack消息。

You know, to Amjad's, you know, kind of example, like, those are, like, the worst Slack messages.

如果你要给各种压力事件排个序，关键人员离职绝对是顶级压力源——因为一个人走，马上就得应对四十个连锁反应。

There's, like, if you just ranked, of the stressful things, because you're just like there's like 40 implications when a key person leaves that you then have to instantly kind of cycle through.

我可能一直坚持下来的原因，是觉得回报依然大于焦虑、压力和时间成本。

Probably the reason I keep doing it is because the upside still exceeds anxiety, and the stress, and kind of time cost, and

那你的动力到底是什么？

But what's your upside?

到这一步，应该不是为了钱了吧。

It's not money at this point, I would have to imagine.

那到底是什么？

What is it?

说白了，动力就是那种智力上的好奇和创造的兴奋感——亲手做出点东西，然后看到它真的被人们用起来，知道又能再往前推一步。

The upside is the, you know, for lack of any better explanation, is just like the intellectual curiosity and excitement of building something, and then having that thing be used in the real world, and knowing that like, you get to just move that forward another step.

而且现在这种感觉更强了，因为别人开发的大部分新技术，我们都能拿来当基础继续往上搭。

And then right now, I'd say it's even amplified because most of the technology that is being built by everybody else ends up being something we can also build on top of.

所以总有无穷无尽的新东西等着我们去探索、去参与。

So it's like there's an unending amount of things that we get to go and kind of play with and be a part of.

要是每天干的都是同一件事，纯粹是熬日子，我可能撑个五年还行。

You know, if we were doing exactly the same thing every single day, and it was totally a grind over like, you know, I could probably pull that off for maybe five years.

但要我连续干十年这种苦活，我真不确定自己能不能撑得住，不过五年大概还行。

I don't know that I'd be able to do like a decade of that, but like I could probably put in five years of just like total grind.

但这不仅仅是枯燥的重复，还充满了刺激，因为新模型一出来，马上就来了。

But this is a grind plus just sheer adrenaline because boom, new model drops.

这背后意味着什么？

What's the implication?

它到底能做什么？

What can it do?

它怎么发挥作用？

How does it touch?

对我们来说，这特别吸引人，因为所有这些都离不开非结构化数据，而我们正好负责存储和管理这些信息。

Well, us, we love it because it all needs unstructured data and the information that we get to store and manage.

不管是新模型、新的智能代理，还是整个领域的新动向，总有无穷无尽的东西等着你去探索、去深入研究。

And so whether it's new models, new agent work, what's happening in the landscape, there's just an unlimited amount of things that you can kind of bite into.

正因如此，这件事才让人如此兴奋。

And that makes it very exciting.

所以

So

你差点就放弃了吗？

Have you almost quit?

你最想退出的时候是什么时候？

When was the time you were closest to bailing?

我本人从来不会选择退出。

I would never, like, personally bail.

所谓的退出，可能就是你把公司卖了，或者被解雇了，但你也没去争。

So the bails that could exist would be like, you know, you kind of sell the company, or you like get fired, but you don't fight it.

你从来没想过辞去CEO的职位吗？

You never thought about resigning as CEO?

大概十九年前，准确说是十八年前，我一度怀疑：我真的是个CEO吗？

There was a moment like nineteen years ago, where eighteen years ago, where, you know, am I like a CEO?

我还是更像个产品人？

Or am I like a product person?

那是不是非得把自己逼成CEO不可？

And then do you have to get in a CEO?

后来我们请了一位首席运营官，问题就解决了。

And then we just solved that by getting a COO.

那时候我就想，天啊，居然有人专门设计了这么一个角色，就是给那些喜欢搞运营的人准备的。

And then I was like, Oh God, this isn't like, there's some like God created a role for people like me, where like somebody who like wants to do operational stuff gets to do that.

我可以继续做产品，同时又当CEO。

And then I get to still do product stuff, but also be CEO.

我心想，天哪，谁想出这个点子的，真是天才。

I was like, Holy crap, whoever came up with this idea, like, is brilliant.

从那以后，我对自己运营能力的疑虑就基本消失了。

So that from that point forward, that sort of solved any kind of like, you know, self doubt I had around like my operational skills.

剩下的问题就只剩下：公司能不能成？

And then the rest has just been like, is the company going to work?

还有，我们是不是得调整方向？

And then do we need to veer the company in a different direction or not?

你说你偶尔会去接受心理咨询，还挺有帮助的。

You said you go to therapy sometimes and it's been helpful.

什么是最关键的突破？

What's been an unlock?

是某种顿悟，还是有什么值得分享的进展？

Either maybe realization or is there a win that you could share?

刚开始做心理咨询时，我们只是发现，我不确定这词是不是人人都用，还是因为我之前就一直做心理咨询，比如Chatuche BT。

Early on in therapy, I kind of it was we just, like, identified I don't even know if it's, like, a word that everybody uses or only because, like, I've been going to therapy before, like, Chatuche, BT.

所以我也没去查过所有别人告诉我的那些术语。

So I didn't research everything that you were ever told.

但她用了‘灾难化’这个词。

But she used this term catastrophization or catastrophize.

也许这词挺常见的。

And so maybe that's a well known term.

我也不太清楚。

I have no idea.

但她的观点是，我总是把事情往最坏的方向想。

But the theory being that I catastrophize things.

我一听到一点消息，马上就想到最坏的结果。

So I get one piece of news, and then I instantly extrapolate out to the worst possible outcomes.

比如有个人走了，我就觉得整个公司要倒闭了。

Like this one person leaves, which means the entire company's out of business.

因为他一走，就会错过某件事，那件事一出问题，整个系统就崩了，最后完蛋。

Because they leave, and then they were going to miss this one thing, and that's going to stop working, and then that's going to break, and then doom.

顺便说一句，我觉得大多数AI末日论者其实都该去看看心理医生，这纯粹就是灾难化思维。

And by the way, I think actually like most people, most AI doomers should probably see a therapist because it's all just catastrophization.

对我而言，当我终于能意识到、能给这种状态命名的时候，我就开始能察觉到它什么时候在发生。

So for me, like what I basically just started to once I could like, kind of like, maybe like understand it and like name it as something, you can then feel when it's happening.

然后我就想，哎，我知道这是什么了。

And then you're like, you know what?

我知道这情况怎么回事。

I know what this is.

这种事我已经遇到过二十次、五十次了。

I've seen this 20 other times or 50 other times in this category.

你猜怎么着？

And guess what?

这根本不是世界末日。

It doesn't mean the end of the world.

事情并不会就此崩盘，也不会毁掉一切。

The thing doesn't end up blowing up and doesn't break everything.

你最终还是会恢复过来的。

You do recover.

这样一来，焦虑发作的周期就缩短了。

And so that sort of shortens the cycles of the anxiety pangs.

以前呢，你可能会被击垮整整三天。

Because previously, it would be kind of like you might be knocked out for three days.

因为你心里就想着：天啊，这下完了。

Because you're just like, Oh my gosh, this is the end.

整个事情都要完蛋了。

This is the end of the whole thing.

然后你经历得多了，就会觉得，嗯，这根本死不了人。

And then you go through it enough times and you're like, Okay, this is like totally survivable.

有时候我甚至有点儿两极分化了，因为一有坏事发生，我老是习惯性地轻描淡写，但我自己也拿不准什么时候该压低反应，什么时候该认真对待。

And then sometimes I almost like, and now I'm probably a little bit bipolar on it because half the time I will just downplay then when something bad happens because I don't have 100% intuition on when to what level to toggle it.

所以对别人，我就会说，这完全没问题。

So then for other people, I'll just be like, this is totally fine.

咱们肯定能挺过去。

We're gonna be totally fine.

这真不算什么事。

This not a big deal.

原因就是，我已经提前把最坏的后果给消化掉了。

And it's just because, again, I've like, I've sort of pre mitigated the catastrophe.

但有时候，我还是会稍微让情绪释放一下。

And then other times, you know, I then still let it loose a little bit.

不过这大概是我用过最管用的办法了。

But that's probably one of the best tools I've been able to have.

上周我们跟瑞·达利欧做了一次活动，他让我们做了些性格测试，这算是他的一个招牌做法。

We got to do a thing with Ray Dalio last week, and he had us do like these personality tests, because that's one of his kind of shticks.

我在神经质这一项上得了99分，满分是100。

I'm like a 99 out of 100 on being neurotic.

你介意自己不是满分吗？

And it bothers you you're not 100?

是啊，我整晚都没睡着。

Yeah, stayed up all night.

我当时就想，

I'm like,

都问了些什么问题？

what questions are I

我漏掉了哪些？

missed?

我的那些问题是

My questions are

已经够神经质了。

neurotic enough.

但我觉得，我真正纠结的事情其实挺零散的。

But I think it's like I'm it's probably jagged what I'm actually neurotic about.

我觉得就那么五件事而已。

I think there's only like five things.

我有些特别在意的小事，比如最常见的就是，你知道的，Slack上的一点小问题。

I have some like like, you know, my the most common Slack is usually just like this.

有个地方差了三个像素，我刚把网站翻了一遍，结果这事就卡在我脑子里了，挥之不去。

Something's three pixels off, and I was just like going through our website, and I and I just like like I it got stuck in my head.

今天的播客由我们的朋友Mercury赞助。

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

他们推出了世上最好的银行产品。

They make the world's best banking product.

我想你早就知道了。

I think you know this already.

我所有的生意都用Mercury。

I use Mercury for all of my businesses.

我觉得我大概有七八个生意吧。

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

我们所有生意的对公账户都用Mercury。

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

他们最近刚推出了个人账户服务。

And now they actually just launched a personal banking account.

所以我的个人账户也放在那儿了。

So I have my personal account there.

我已经把账户从富国银行和摩根大通转过来了。

I moved off of Wells Fargo and Chase.

我现在完全用Mercury了。

I'm just all in on Mercury.

为什么呢？

Why?

我喜欢用起来简单的产品。

I like products that are easy to use.

我喜欢那些能懂我、懂我麻烦的产品。

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

比如，和我老婆一起开个联名账户，特别方便。

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

申请虚拟卡片也是一点就成。

Very easy to spin up virtual cards.

点一下就能开通储蓄收益。

One click, and I get savings yield.

所有我需要的功能，全都在一个地方解决了。

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

如果你在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.

谁要是不同意这点，我敢跟他打一架。

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

前往 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，并非 FDIC 保险银行。

Mercury is a fintech, not an FDIC insured bank.

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

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

A，成员受 FDIC 保障。

A, members FDIC.

你相信这类性格测试吗？达莉亚请过不少非常成功的人来分享，他们几乎都对这类性格测试深信不疑。我以前一直觉得这跟星座没啥区别。

Are you a believer in these personality tests the way Dahlia we've had actually a bunch of really impressive successful people come on and almost and very much swear by the kind of personality test stuff, which I had always just thought is horoscope.

挺可爱的。

Like, that's cute.

要不要再送你一块水晶？

Do you want a crystal too?

我以前真不怎么信这个。

Like, I didn't I didn't really believe.

现在我觉得，有这么多聪明人都跟我这么说，我只好信了。

And now I'm like, I get I think enough smart people have have have told me this where I'm like, okay.

我可能得重新想想我以前那些笑话了。

I should probably reassess my my my jokes here.

除了雷，还有谁对他感兴趣？

Who else is into him besides Ray?

我们之前还请过另一个谁来着。

We had who was the other was on.

他是个投资人。

He he's, you know, an investor.

我觉得他特别特别聪明。

I think he's he's very, very smart.

他讲过，有一次性格测评彻底改变了他的人生，就是那种测试。

He was he's he described how his life changed from a, you know, an assessment that that told him, like Yeah.

嘿。

Hey.

你之所以有这种感觉，是因为你在玩一种管理客户和人际交往的游戏，但你其实更适合独自作战的竞技类型。

The reason you feel the way you feel is because you're playing a game of managing clients and people, but you actually thrive in solo player numbers based competitive games.

当他转去做投资时，一下子就顺了，因为这正是那种能让他发挥优势的游戏。

And, you know, he when he switched to investing, he, like, thrived because that's exactly the type of game that that rewards.

对。

Yeah.

你相信他吗？

Do you do you believe in him?

在遇到雷之前，我更倾向于支持肖恩的观点。

I probably veer more on Sean's end prior to to his Ray interaction.

我觉得这挺有意思的，通常在公司团建时，都是个不错的破冰话题。

I think I think it's, like, fun as, like, it's usually, like, always, like, a good icebreaker at a at a corporate off-site.

但我很少能从中做出什么实际的行动。

I've I've rarely left being able to do anything actionable about it.

是啊，我们知道你是红方。

Like, yeah, we know you're red.

我们都知道你会很主动。

Like, we we we know you're gonna be aggressive.

那针对这一点，我们该怎么办？

What what are we gonna do about that information?

我们知道你很擅长合作。

We know you're collaborative.

很明显，你是个喜欢合作的人。

Like, it's very obvious you're collaborative.

有些人会用它来招聘。

Well, some people use it for hiring.

他们会让你做类似这样的测试，是的。

They they make you do, like Yeah.

有一家公司叫Culture Amp，你只要输入要招聘的岗位，它就会告诉你需要什么样的性格类型，真厉害。

There's a whole company called Culture Amp, I think it's called, where you where you enter you enter the job that you need to hire for, and it tells you what personality type, and then you Wow.

有

Have

他们的优势识别测试每年能赚上亿美金。

your StrengthsFinders does, like, a 100,000,000 a year on their their

是的。

Yeah.

但说实话，我最后还是靠直觉做事，别人说我这样也没问题。

Well, really so but, like, I so I I end up somewhere just in this, like, I'm told I'm fine with it.

我可不会拿这个来管理公司。

I'm also like, I I don't run the business on it.

这就跟我谈宗教人士时说的一样。

It's like how I talk about people who are religious.

没错。

Yes.

确实如此。

It is.

对。

Yeah.

如果你想要有自己的信仰，比如说。

If you wanna have your religion, like,

我们没问题。

we're good.

没错。

Yes.

看起来确实有帮助。

Seems like it helps.

另一个类似的问题，也属于了解自己的范畴。

Different question kind of in the in the same vein of kind of know thyself.

据我了解，你是个特别痴迷商业战略的人。

From what I understand, you're a pretty big, like, business strategy nerd.

我听说你经常读到深夜，而且已经这样好多年了。

I've seen I've heard that you read books until late at night, you've you've been doing this for a long time.

你读书。

That you read books.

是的。

Yeah.

你就是那种人。

You're one of those guys.

你知道吗，当阿姆贾德上节目时，他提了两次，说的是那个叫什么来着？

You know, when Amjad came on, he's he twice, he's referenced, like, the what's it called?

就是关于企业护城河的那几个优势，七种还是五种，反正就那么几种。

Like, seven powers or five powers or however however many powers there are of defensibility.

只有七种。

Only seven.

明白了。

Alright.

七种。

Seven.

对。

Yeah.

如果我去问那些你曾经指导过的创始人，或者你一起合作过的人，问他们——

Are you if if I was to ask people who kind of either founders you advise or people you've worked with, if I was like, yo.

亚伦最常提到、最推崇的那些框架是什么？他总想让大家去采纳哪些方法？

What are Aaron's kind of, like, the frameworks he really, like, pulls a lot or tries to, like, get people to adopt?

你能给我们举几个这样的例子吗？

What what are some of those that that you could help teach us?

是的。

Yeah.

我读过他所有的书。

I've read every book.

所以我对这些书非常熟悉，我觉得现在我手上这套书是最全的。

So I I have a pretty good I have a pretty good I I believe I have the best set of books at this point.

这可是少数几件让我特别有自信的事情之一。

This is something that I'm it's, like, one of the rare things I'm, like, overconfident about.

如果创始人只读一本书，那就只读《七种力量》。

If founders only read seven powers, just do seven powers.

这些书都明显很优秀。

It's all that's like obviously good.

但如果你再加一本，读读《定位》。

But if you add to it, you read positioning.

你听说过这本吗？

Have you you know this one?

是的，我确实有《定位》这本书。

Is that Yes, by still have Positioning.

没人读《定位》，结果就搞砸了自己的市场定位策略。

Nobody reads Positioning, and then they up their whole market positioning strategy.

《七种力量》把另外七本书的核心内容提炼得特别精彩，每个人都该读一读。

So what Seven Powers does is that he's abstracted basically seven other books in a very compelling way, and everybody should read Seven Powers.

但如果你没深入理解《创新者的窘境》和另一本《创新者的解答》，那就像是缺了搭档一样。

But if you don't deeply understand Innovator's Dilemma and this other book Innovator's Solution, it's like this great tandem.

对，它们主要帮你认清问题所在。

Yeah, they only do give you the problem.

他们没有，我从来没真正

They don't I've never really

对，我压根不知道还有《创新者解决方案》这本书。

Yeah, didn't even know there's an Innovator's Solution.

我的意思是，他真的很会推销这些续作。

I mean, he he knows how to sell the the, you know, sequels.

所以《创新者反击》就这么来了。

And so so Innovator strikes back

这简直是个三部曲。

this the trilogy.

大家都坚持不下去了。

Well, everybody everybody gives up.

没人会去读解决方案，因为读完《创新者困境》这300页，他们已经快崩溃了。

They never read the solution because they're already, like, 300 pages into the dilemma, and they're like, oh my god.

所以你最好把《困境》和《解决方案》这两本书连着读，一前一后。

So you you want to read both The Dilemma and The Solution in tandem, like, back to back.

你得读一读《七种权力》。

You want to read Seven Powers.

你得读一读《定位》。

You want to read Positioning.

你还可以读读《蓝海战略》，稍微轻松一点，但内容很扎实。

You want to read you know, for it's a little bit more on the fun side, but blue ocean strategy.

这本书挺不错的，学术性为主，又带点趣味性。

It's good kind of mostly academic plus a little bit infotainment.

然后就是《跨越鸿沟》或者《龙卷风》。

And then maybe crossing the chasm or inside the tornado.

如果你有时间，能把自己关在房间里把这六本书全读完，那你就能百分之百预测科技领域里发生的每一件事，准得离谱。

So if you had the time, and you could be locked in a room and read six books, if you read that, you will be able to predict 100% of things that happen in technology, like without fail.

你完全能猜到别人会出什么招。

Like you'll know every competitive move that people are going to make.

你也会明白市场为什么会这样运转。

You'll know why markets do the things that they do.

你会比盲目猜测下一步会发生什么理解得深得多。

You'll understand so much more than just trying to wing it and guess what's going to happen next.

你有没有一份反推荐书单？就是那些很火、大家都读，但你觉得其实没必要读的书？

Do you have, like, an anti read list where you think these are popular and people read them, but they're you don't think people should?

有。

Yeah.

有些领导力类的经典书籍，我确实没怎么被吸引进去。

Well, there are some kind of, like, your your kind of, like, classics in, like, leadership books that I I have not found myself getting as into.

但我明白为什么别人会喜欢这些书。

But I do appreciate why people get into them.

它们没能像作者希望的那样，激起我那种强烈的共鸣。

It doesn't, like, trigger my same sort of visceral, you know, kind of reaction I think that they're intending.

你是那种会读《领导者最后吃》或《从为什么开始》的人吗？

You're a leaders eat last or a start to eat why type of guy?

《整理床铺》那本呢？

Make your bed?

是的。

Yeah.

我不会去点名批评哪本书，毕竟我尊重这个行业，也不想特意挑出哪一本来说。

I'm I'm I'm not gonna I you know, I got because I respect the the trade, I don't wanna I'm not gonna call out anything specifically.

私下里我可能会提那么一两本。

Maybe offline, I'll I'll mention a couple.

有几本书我翻到五十页左右就觉得，唉，这内容也太老生常谈了。

But there are there are few where I'm, I'm, like, 50 pages in, and I'm like, I think this is kind of, like, a little bit too trite.

其实现在回头看这些书挺容易的，因为它们讲的都是十五年、二十年前的案例。

So Well, let let's use this because right now, it's easy in hindsight when you read the books because they're, like, giving you a case study from fifteen years ago, twenty years ago.

你就像马尔科姆·格拉德威尔那样，用事后诸葛亮的方式，把历史重新解读一遍。

And you can sort of Malcolm Gladwell, like, revisionist history your way into, like Yep.

想得出什么结论都行。

Any conclusion you want.

那些特别有意思的是像九三年写的书，里面还说数字设备公司将来会成为全球最大的公司，就因为……

Those are the fun ones of, you know, the ones that were written in, like, you know, '93, and it was like, you know, Digital Equipment Corporation will be the largest company on the planet based on you know?

结果两年后就倒闭了。

And it's like, it died two years later.

对啊。

Right.

但现在AI领域一片迷雾，确实如此。

But but right now, there's this fog of war with AI, and it's Yeah.

简直就像《权力的游戏》。

There's basically Game of Thrones.

埃隆就是北境之王。

You've got, like, Elon, the king of the North.

他正率军南下。

He's coming down.

他想让大家开心起来。

He's trying to make it happy.

还有Anthropic那边。

You got the Anthropics.

我觉得吧，当你考虑这七大势力的布局时，你觉得事情会怎么发展？

I guess, how do you see this playing out once once you think about, like, seven powers positioning?

你有没有一些自己的分析框架？能不能预测一下，未来七年里，哪些事情你会猜得特别准，或者完全猜错？

When you think about some of the frameworks you have, do you have any predictions for us that you can look really smart on or dumb on in the next seven years?

你问得正好，基于我对那六本书的信心，我本该能给你一个明确的答案。

I'm glad you asked because it does I should I should based on my confidence on the power of those those six books, I should be able to tell you the answer.

但它们都没能预料到AI市场的出现。

They did not anticipate the AI market.

我得澄清一下，你刚才说，只要读了这些书，就能百分之百准确预测任何事，对吧？

Just for the record, I think you said if you read these books, you could predict anything with a 100% hit rate all the time.

那你说吧，我们听着。

So go ahead.

没错。

Yes.

现在

Now

你刚说，我不知道。

you just said, I don't know.

完全猜不出谁会赢。

No idea who's gonna win.

不过也好，因为这里还有其他因素，是AI三定律没提到的。

Well, good partly because there's other factors here that that Al Al Trite didn't write about.

他根本没预料到中国会在开放权重模型上胜出。

So he didn't know whether China would win in open weights models.

我是说，真的还有别的因素，比如政府，它在这件事上是个变量。

No, I mean, there are literally other factors because we have government like, government is such an X factor in this.

中国本身就是一个变量。

China is an X factor.

所以这些事我完全没法预测。

So no idea on all these things.

这些书真正厉害的地方是，如果你是个创业者，它能帮你判断你的点子有没有可能成事。

More what these books are good at is like, it'll be if you're an entrepreneur, it'll tell you if your idea is going to be remotely, going to work or not.

所以它在早期阶段更管用，比如这家公司能不能找到一个能切入的细分市场？

And so it works better in kind of like early stage, will this company find a category that it can wedge into?

还是说现有的巨头更可能直接拿下这个市场？

Or will the incumbent more likely take the category?

我常拿《创新者的窘境》和《创新者的解答》当例子，它们基本上能告诉你75%的情况下，你的创业点子有没有机会

So I use Innovative Dilemma and Innovative Solution as an example will basically tell you 75% of the time whether you have a

作为一个新创公司能否站稳脚跟。

shot as a new startup.

你能举个例子吗？说得更明白一点。

Can you give an example, make it a little more obvious?

是的，其实《创新者的窘境》这本书的核心，很多人都误以为讲的是科技颠覆。

Yeah, I mean, like, the whole point of Innovator's Dilemma was everybody kind of thinks it's like a tech disruption book.

觉得无非就是被什么新技术给干掉了。

It's like, oh, got disrupted by a tech or something.

但这种理解太简单了。

But that's like it's too simplistic.

关键在于，《创新者的窘境》告诉我们：如果某个商业模式对现有巨头没有吸引力，他们就不会去碰。

The key is what Innovator's Dilemma tells you is if the business model is not something that the incumbent wants to pursue, because the business model is unattractive to the incumbent.

如果你用这个视角来看，就能很快判断：作为一个新创公司，现有巨头会不会觉得这个模式没意思？

So if you look at it through that lens, it will very quickly tell you, like, you have a new startup, like, does the incumbent is the incumbent going to find that business model unattractive or not?

只要他们觉得没吸引力，就根本不会跟进。

And if it's unattractive, they won't pursue it.

但如果他们觉得有吸引力，那你就得默认他们会全力跟你竞争。

And if it is, then you very much, you know, need to assume that that incumbent is going to try and compete with you.

这时候你得想清楚：这个技术是不是一种现有巨头本来就擅长的持续性技术？

Then you have to decide, is that a technology for whatever reason is like a sustaining technology that the incumbent is going to be classically good at?

还是说，这个技术太难搞了，他们根本搞不定？

Or is it like so hard for them to figure out that they're not?

这就解释了为什么谷歌在AI领域一定会做得很好——他们不会放任消费市场溜走，因为这个商业模式太有吸引力了。

And then that tells you things like Google is going to obviously get really good at AI and they're going to like not like they're not going to like let the consumer market just disappear because it's actually an attractive business model.

毕竟，用谷歌来回答AI问题，对商业化一点负面影响都没有。

As like, there's nothing about having an AI answer from the Google experience that would be bad for monetization.

所以三年前那些说谷歌要完的人，现在一看就明白了，谷歌铁了心要全力做这件事。

And so like everybody that wrote Google off three years ago was like, it's very obvious that Google wants to go do this one fully.

反过来，这些年我们看到不少商业模式，大公司都不愿意把基础设施迁到云端。

Conversely, there's a lot of business models where over the years, we saw there were a lot of incumbents that didn't want to move infrastructure to the cloud.

因为一旦迁到云端，客户从上万个变成三四个，完全不一样了。

Because if they moved it to the cloud, instead of having 10,000 customers, they would only have three or four customers.

这对某些软件公司或基础设施服务商来说，是完全不同的商业模式。

And that was a totally different business model for certain software providers or certain infrastructure providers.

所以当云服务兴起时，谁会面临压力，其实一眼就能看出来。

And so you could see who was going to be under pressure as the cloud grew.

我之所以用这些框架，就是因为它能帮你预测：大公司会怎么反应？

So I just use these frameworks because they kind of help you predict again, how is an incumbent going to respond?

他们是会选择封闭式应对，还是……

Are they going to respond in like a way that is sort of like closed?

他们会用对的一套方法来应对吗？

Are they going to respond with the right set of mechanisms?

这种事情天天都在发生。

And that just happens all the time.

AI领域的情况也差不多，老牌企业会怎么应对，其实早就能猜到个大概。

AI generally is kind of playing out with not that different of response mechanisms from the incumbents that that you would also again kind of expect.

谁会进入每个市场？

Like, who's gonna go and and kind of enter each each market?

他们打算怎么竞争，诸如此类的问题。

Who's gonna how are they gonna compete, etcetera?

你只对这种商业策略感兴趣吗？

Are you only interested in that in this business strategy stuff?

我每次读《蓝海战略》这类书，都会想，这讲的分明是像Box这样的公司。

Like, whenever I read, like, Blue Ocean Strategy, lot of times I think, dude, this is for, like, a business that is this is like Box.

这种策略适合那种动不动就能从几十亿变成几百亿的大企业，对小企业来说不太适用。

This is like a, you know, multibillion dollar company who can can swing up and become a tens of billions or hundreds of billions of dollar company over the next decade, not as much, like, from a SMB lens.

比如我住的纽约市，就有一种叫‘快闪贝果’的东西。

Like, for example, where I live in in New York City, we have this thing called pop up bagels.

这种做贝果的方式还挺有意思的。

And like, it's like a kind of an interesting take on a bagel play.

我觉得他们其实刚融了资，不过呢。

I think they've actually just raised VC, but like.

这立刻就反驳了你的观点。

Instantly disproving your your

问题。

question.

不过，你有没有想过，比如你有没有特别着迷于一些商业案例？我们之前聊过，有个叫纳特·弗里德曼的人，收购了那家棒球卡公司。

But, like, do you ever think about do you ever, like, nerd out on, like, you know, we were talking about I forget the guy's name, Nat Friedman, who bought, you know, the baseball card company.

纳特·特纳。

Nat Turner.

纳特·特纳。

Nat Turner.

抱歉。

Sorry.

你有没有从这个角度想过这个问题？

Do you ever think of it from that perspective?

有啊。

I do.

我确实会特别投入，但前提是得跟我感兴趣的事沾上边。

I definitely do nerd out, but only if it like crosses my universe.

我有个朋友做过一个在线卖气球的网站，他既做批发，也卖零售。

So I had a friend actually that had an online balloon website, And he was selling kind of balloons to some wholesale, some consumer.

那特别有意思，我们经常一起头脑风暴，比如怎么把派对用品的批发和零售业务做大做规模化。

And that was really fun because we could go and brainstorm like, how would you do a consumer or wholesale kind of like party supplies business at scale?

所以呢，说实话，现在我很少有时间这么玩了，但一想到这些，还是觉得特别有意思。

And so, yeah, I mean, it's like, I don't find myself being able to as much, but like, it is always fun to get your arms around.

我们那时候老在一块儿瞎琢磨。

We were always brainstorming.

怎么才能跟Party City竞争呢？

It's like, how do you compete with Party City?

他们有个挺复杂的问题，因为有实体基础设施，所以零售成本特别高。

Like, okay, so like, they've got this one complex thing because they have physical infrastructure, which means they have a high retail kind of cost.

这就让他们很难完全转向数字化。

So it's hard for them to go as full kind of digital.

这其实就是传统大公司常遇到的那些困境。

So there was a lot of like classic incumbent dilemmas.

我觉得，任何两个人合伙、卖实体产品的初创公司，都会碰到和那些拿风投的软件公司一模一样的市场挑战。

I think every, you know, two person startups that are selling physical things in the real world run into the exact same, you know, market factors that a software business, you know, with VC run into.

如果时光倒流，你还是个大学生。

If you were, rewind the clock, you're a college student.

你刚创办Box的时候，互联网正好让这类想法成为可能。

You know, when you started Box, the Internet enabled ideas like that.

如果你现在年轻、自由、充满干劲，你会想去做什么？

If you were free, young, hungry to do something now, what do you think you would want to go build?

就因为我能忍得住折腾，我大概会一头扎进AI的狂热里，反正也得试试看。

Just because of my tolerance for pain, I would probably end up somewhere right in the center of the AI craziness, just because I'd to give it a shot.

你大概和我们一样，晚上九点到十点之间，刷着Reddit或者别的地方，看那些技术宅们分享的酷玩意儿。

You're probably doing what we're all doing, which is at like nine to 10PM, you're like toying around on Reddit or whatever, and like looking at all the nerdy cool stuff.

最近几周，有什么东西特别吸引你吗？

What what what is catching your eye in the last couple weeks?

BrightBerry？没什么特别的。

BrightBerry, nothing

过去几周没什么新变化，说实话，我的工具组合也没啥意外的。

has changed sadly in the past couple weeks, but my my stack is not surprising.

我所有打开的标签页，或者每个应用图标，不是Codex就是Cursor、Perplexity、Claude、Figma。

It's it's like a like every tab is is one of or every app icon is one of Codex, Cursor, Perplexity, Claude, Figma.

我啥都用，全都有。

Like, I have everything.

比如Perplexity，如果你想让电脑联网查资料、逐页读网页内容，我就会直接切到Perplexity Computer。

And I'm like like like, Perplexity is like if you want like cloud based computer use that's going to like really go to the website and read each line of text, I'll click off to Perplexity Computer.

要是只是做点基础调研，其实选择挺多的。

If you're just doing basic research, you have a number of options.

如果你在搭建一个原型网站，我会试几个不同的工具。

If you're building a prototype website, I play with a few different tools.

这方面也没什么出乎意料的。

So nothing surprising on that front.

你对当前软件行业在公开市场上的表现怎么看？

What do you think about what's going on with the public markets in terms of software?

因为这个现在挺让我们感兴趣的。

Because that interests us right now.

我觉得肖恩之前好像写过什么。

Like, I think Sean was was he had written down here.

你说啥了，肖恩？

What did you say, Sean?

你觉得这是代际性的买入机会吗？

You think this is a generational buy?

我说是的。

I said, yeah.

就是说，有资格跟你自己的系统对话。

Like, you know, permission to talk to your book.

现在的软件，到底是处于一个世代性的买入机会，还是说你得再论证一下？

Either is is software right now in a in a generational buy spot or, you know, make the case?

你可以谈谈你自己，或者别的公司。

Can talk about yourself or other companies.

我对这个话题的投资建议非常谨慎，因为你完全受制于其他各种因素，比如现在是不是芯片周——一到这种时候，不管怎样软件股价都会跌。

I'm very nervous about any investment advice on this topic, simply because you're at the mercy of many other factors of like, is it chip trade week, which just means software goes down no matter what.

所以我只能说，我不会给出具体的投资建议，因为我不清楚眼下这些资产到底该用什么样的估值倍数。

And so like, I say, I'll make separate investment advice because I don't know what the right kind of multiples are at any given moment for this stuff.

我只想说，大概九个月前，人们第一次开始在推特上讨论什么‘系统性记录软件’之类的东西。

I would just say, people probably for the first time ever started tweeting things like system of record nine months ago or whatever.

但如果你抛开这些一时的潮流氛围，回归到‘系统性记录软件’的本质，这些系统其实就是你财务数据、客户数据，或者在我们这里，合同和财务文件的权威存储中心。

But if you kind of take out any of the temporary zeitgeist nature of any of that, and you just go back to the core of literally system of record software, These systems are used as like the authoritative place where your accounting data goes or your customer data goes or in our case, your contracts and financial documents go.

所以这类系统，根本不会有人轻易想着要重新搞一套全新的、完全不同的系统，还非得自己来开发。

So these are not the things that like are high on the list of I'm going to go and just like try and build a totally alternative different system for, and I want to build it myself.

我只是想省下几十万甚至上百万美元。

And I just want because I want to go and save a few $100,000 or a million dollars.

这些系统可是公司核心运作的关键部分。

Like these systems are in the kind of core guts of these companies.

所以很多人说‘我用一周就写了个原型’，原因就在这儿。

So that sort of is why a lot of the software that people say, oh, I vibe coded it in a week.

但这并不意味着福特会拿这个原型去替换他们的ERP系统。

That doesn't necessarily equate to well, like Ford is going to go and replace their ERP system with that vibe coded thing.

没错，你确实快速搭出了一个能用的原型，但那和支撑一家受证交会监管、牵涉全球供应链的企业级系统完全是两码事。

Like, yes, you were able to stand up a prototype that was functional, but that's just like totally different from like running your enterprise that is held accountable to the SEC and a global supply chain on powering that.

这就是为什么很多软件不会像人们以为的那样轻易被取代。

So that's like why a lot of software won't go away in the same way people think.

但更令人兴奋的前景是，当有智能代理在企业内部四处行动、需要完成各种实际任务时，会发生什么？

But then the upside, which is much more exciting is what happens when you have agents that are running around and they need to go do all this useful work in your enterprise?

这些代理要做的实际工作，离不开访问这些系统里的数据。

Well, the useful work they're going do is going to require access to data that's inside these systems.

而且通常需要一些限制机制，让代理在参与时能确保不会擅自改动你ERP系统、CRM系统或核心工作流程中的关键部分。

And it's going to often require kind of guardrails that the agent is sort of participating and ensuring that the agent doesn't go off the rails and completely change out fundamental parts of your ERP data or your CRM data or kind of core workflow.

所以它们需要的是可预测的软件，这些软件具备正确的边界、数据访问权限、操作权限和工作流设计。

So, they need deterministic software that they are kind of participating in that have the right walls, the right data access, the right permissions, the right workflow design.

这些功能大多会来自现有软件，因为大多数企业的流程本来就已经在这些系统里搭建好了。

That's largely going to come from existing software simply because that's where the workflows have already been built out in most enterprises.

因此，其实有很多理由说明，一旦代理能够融入这些工作流程，某些软件类别反而会释放出更大的价值，因为你现在能用这些软件做更多事情。

So there's a lot of actually ways to argue that there's more upside to certain software categories once agents can participate in those workflows, because you can just do now way more with that software.

就我们的情况而言，我们确实看到使用量在上升，因为代理现在可以自由访问所有这些数据。

So in our case, we actually see an increase in usage because agents are now roaming around accessing all of this data.

你希望它们访问的，正是用户有权访问的数据，这就意味着你需要一套可靠的身份权限和访问控制机制。

And you want them to access the same data that the user has access to, which means you want something that has reliable permissions and access controls and whatnot.

那么问题就变成了：这些现有的软件公司该如何把代理带来的新价值变现呢？

So then it really just becomes a question of like, how do some of these incumbent software companies monetize that agentic upside?

我觉得你会看到，主要还是以使用量为基础的收费模式。

And I think you're going to see, you know, mostly it's mostly like a consumption oriented model.

这会更偏向于无头架构的方式。

It'll be on this more of this headless approach.

但我觉得，由于智能代理的普及，软件的使用量会大幅增加。

But I think there's going to be a ton of usage of software as a result of the agent, you know, kind of adoption piece.

不过，要据此判断该买或卖哪些软件，还是很难说。

But again, hard to then say like, okay, so what should you buy or sell based on that?

每个人都得亲自去研究，判断哪些软件会因为智能代理而用得更多，哪些又会因此用得更少。

You have to, everybody has to kind of go and do the work, and sort of try and make a judgment call of like, what software will get used more because of agents, which is what software gets used less in the future because of that.

是的。

Yeah.

我们跟HubSpot合作很多，跟达梅什、布莱恩他们也都是朋友。

We were, you know, we work a lot with HubSpot, and we are friends with Darmesh and Brian and and those guys.

这简直让人难以置信。

It it's kind of insane.

它的市值是营收的两倍半，而营收每年还在以30%左右的速度增长。

The the market cap is like two and a half times the revenue, and the revenue is growing 30% a year or something like that.

这太疯狂了。

It's crazy.

简直疯狂到极点。

It's crazy, crazy, crazy.

我倾向于认为它会上涨。

I tend to believe it will go up.

我不觉得密苏里州的水管工会自己去开发一个客户关系管理系统。

I just don't think that like a plumber in Missouri is going to make their own CRM.

是的。

Yeah.

对。

Yeah.

我觉得，出于合理的原因，我们往往采取一种过于简单的二元看法。

I think the, for good reason, like we tend to have a, you know, kind of a simplistic binary approach.

但你看看像 vibe coding 这样的东西，就会想，它是不是会取代我们现在已经用的那些软件。

But like you look at vibe coding and you say, well, vibe coding must then replace the software that we already use.

而真正的答案很可能是否定的，它大概只是在我们现有的软件基础上再加一层功能。

And probably the real answer is no, it'll probably just be built on top of the software that we already use.

所以，IT人员会去进一步定制自己的工作流程，但用的是他们信任的、稳定可靠的数据平台，这样工作效率会非常高。

And so it'll be the IT person going and customize their workflow even further, but on a data stack that they trust is reliable and you're going work very effectively.

你看到的这种趋势，其实挺有意思的。

It's kind of interesting, the signal that you see.

过去一个月里，Anthropic除了发布Fable之外，最重要的发布就是这个叫Claude Tag的东西，你可以和Claude像同事一样协作工作。

So, Anthropics biggest announcement other than Fable in like the past month is this thing called Claude Tag, where you work with a Claude kind of colleague in a shared way.

他们是在Slack里推出的这个系统。

Well, what system they launched in Slack.

他们为什么要这么做？

Why did they do that?

因为用户本来就在Slack里，而Slack本身就具备合适的权限体系，能支持这种多人协作的智能代理。

Because the users are already in Slack and Slack has the right effectively permission boundaries to be able to have a shared collaborative agent that you would work with.

为什么Claude Tag这么有威力？

Why is Claude Tag so powerful?

这是因为它可以访问你授权它使用的软件系统中的数据。

It's because it accesses your software systems that you can give it access to data.

比如 Box 就是这样一个数据源。

So Box is one of those data sources as an example.

所以它并不是像有人说的那样，Claude 赢了，SaaS 就输了。

So instead of it sort of being like, Claude wins, so SaaS loses.

实际上，这更像是一个智能基础层。

You actually can be like, no, actually this is this intelligence substrate.

它能支持一系列非常实用的使用场景。

It offers some set of kind of very useful use cases.

但同时，它也会嵌入在那些本身就有大量应用场景、能稳定创造价值的确定性软件里。

But then it's probably going to also exist within deterministic software that also has a bunch of use cases that kind of create value.

我觉得，一旦你不再执着于每天靠不断发指令来操控软件，而是接受有某些软件始终稳定可靠地在那里，再让智能能力叠加在上面，去处理那些非确定性的任务，这才是更合理的方向。

So I think once you kind of move on from the zero sum nature of like, okay, I'm going to go prompt my way into software every single day to, no, I'm going to like have some software that is always there that is reliable and deterministic, and I'm going to have intelligence kind of get added to that, that does more non deterministic things.

这大概就是未来更自然的平衡方式。

That's probably like a more logical balance that you'd expect in the future.

兄弟，你太牛了。

Dude, you're awesome.

你聪明得没话说。

You're smart as shit.

我们特别喜欢和你聊天

We love talking with

我的意思是，你聊的都是我懂的东西。

I mean, you only talk about the things that I know.

所以，要是我给你一堆我不愿谈的话题，你也能应付得来。

So if you I can give you lots of topics that I'm not prepared to to discuss.

企业CEO对企业的看法

Enterprise CEO has take on enterprise

没错。

and Exactly.

真的，太厉害了。

And, like, wow.

这还用说。

No kidding.

企业级软件。

Enterprise software.

兄弟，谢谢你来参加节目。

Well, dude, thanks for coming on, man.

我们一直关注你很久了。

We've we've enjoyed following you for a long time.

今天能和你聊一会儿，真的很开心。

It's it's been fun getting to hang out with you here for a little bit.

就这样。

That's it.

这期播客结束了。

That's the pod.

好的。

Alright.