20VC：Mag7财报：谷歌与亚马逊胜出——Meta与微软疲软｜Anthropic融资500亿美元及其对潜在IPO的意义｜Atlassian、Twilio和Five9超预期：SaaS末日来临？｜Sierra估值150亿美元：巅峰还是潜力？

他们并不是想跟谷歌竞争，去为Anthropic提供计算资源，也不该这么做。

They're not competing with Google to provide compute to Anthropic, nor should they.

所以他们这么做的原因有两个。

So the two reasons they use.

其中一个原因，是最近几轮讨论中他们提到的：天啊，我们用这些模型大幅优化了广告效果，数据明显变好了。

One is the reason they started to use a little bit over the last couple of calls is, Oh my gosh, we're optimizing our ad performance a whole ton using these models, and it's making the numbers better.

咱们具体看看这个，毫无疑问，数据确实很亮眼。

So let's dive into that a little, is that there's no doubt the numbers are amazing.

算一算就知道了，他们说过去几个季度，广告效果提升了10%到15%，表现确实更好了。

If you do the math, they're saying they're getting 10%, 15% lift in the last couple of quarters, the performance is better.

但除非看到A/B测试的结果，否则很难确认，这在任何效果提升分析里都是常态。

But it's super hard unless you see an AB test, and this is always the case with lift analysis.

你必须看到A/B测试——一半广告用上Meta的新模型，比如Llama 5，不管叫什么名字，反正就是新模型；另一半不用。

You need to see the AB test, you need to see half of the ads optimized using, let's call it the Meta Lama five or whatever the new model is called, I can't remember, whatever, and then half not.

你真的能观察到效果提升吗？

Are you really getting lift?

你确实可能获得了一些提升。

And you probably are getting some lift.

但每年花一千亿美元，值得吗？

Is that worth 10,000,000,000 a year?

也许值十五亿，但他们花了一千五百亿。我的意思是，他们一直用这个理由来解释LLM支出：说它能优化现有业务。

Maybe 15, But they're spending $150 So my point is they've been using this justification, the LLM spend, as being it'll optimize our existing business.

我一直觉得这纯属扯淡——如果要花一千五百亿资本支出，才换来两千亿美元业务中10%的提升，那很可能是个错误，尤其是当部分技术本可以外包获取的时候。

And I've always felt that's bullshit because if it takes $150,000,000,000 in CapEx to give a 10% lift on a $200,000,000,000 business, it's probably a mistake, especially when some of that technology might be available third party.

这次通话中明显多了些这样的说法：我们要打造下一代体验，简单说，如果人们从跟真人交流、看新闻，转向跟聊天机器人互动，那Facebook，我希望我们能占住这个位置。

So it was noticeable in this call, there was a little bit more of the, We're just going to build next generation experiences, qualitative comments on, Simply put, if people go from talking to other humans and looking at news to talking to chatbots, I, Facebook, want to be there when that happens.

我觉得这才是他们真正的理由。

And I think that's the justification.

所以市场现在明白了：哦，原来如此。

So the market is looking at that and going, I got it.

这根本不是一门生意。

This is not a business.

我只是想在事情发生时能参与其中。

This is a, I want to show up if something happens.

这相当于押了1500亿美元在一件还说不清楚的未来上，所以你自然得给它打个更高的折扣。

So it's $150,000,000,000 bet on the future that's not quite articulated, so you just put a slightly higher discount on that.

事情其实就是这么回事。

That's all that's happening here.

真搞不懂他们为啥要花这笔钱。

It's like, Don't know why they're spending it.

他会自己去做的。

He's going to do it.

他根本不需要征得谁的同意。

He doesn't have to ask any permission.

也许他会像在Instagram或WhatsApp上那样对了，也可能像在Meta上那样错了。

And maybe he'll be right, like he was on Instagram or WhatsApp, or maybe he'll be wrong, like he was on Meta.

但这跟谷歌不一样，人家那事儿你一听就明白。

But it's not like Google where you can go, Oh, I get it.

他们花了1500亿，但回笼了6万亿，而且每年还以80%的速度增长。

They're spending 150, but they're getting $60,000,000,000 back and it's growing 80% year on year.

再过两年，就能实现正现金流了。

In two years, it'll be cash flow positive.

这完全是另一回事。

It's just a different thing.

我觉得华尔街的看法也跟罗伊说的一样。

I do think that Wall Street is also to Roy's point.

华尔街的人整天在做表格，算GPU折旧、资本支出，钱都花到哪儿去了。

I Wall Street is all building these spreadsheets, GPU depreciation, CapEx, where it's going.

Meta的模式没法套用这套逻辑。

The Meta model doesn't support that.

这事儿太间接了，根本没法用那种方式衡量。

It just it's it's indirect and it just doesn't support it.

数学上也对不上，确实让人糊涂。

And nor does the math tie, but it's confusing.

他们都埋头做表格，争论资本投入的输入输出，光在那里吵来吵去。

They're all running spreadsheets and arguing over the inputs and the outputs of these capital investments to the outputs, and they just debate it.

我喜欢你这么说，杰森，其实你没明说，但事实就是，马克根本不会做这些表格，也不在乎这些玩意儿。

I like that, Jason, because effectively what you didn't say but is true is, And Mark isn't running those spreadsheets and doesn't give a shit about those spreadsheets.

谢谢你们这么用心。

Like, thank you very much, guys.

你们尽管去折腾吧。

Knock yourselves out.

我要花钱，只为保持竞争力。

I'm going to spend to be relevant.

他们一做这些表格，脑袋就疼，因为根本摸不着边。

Their little heads are hurting when they run the spreadsheets because they can't touch it.

给别人做倒是可以，但在这儿就不行。

And they can do it for the other guys, and they can't do it here.

你说得一点都没错。

I think you're exactly right.

他可不是在哈佛靠跑表格赚了两千亿美元的。

He didn't make $200,000,000,000 by running a spreadsheet at Harvard.

他就是把产品做出来，然后迅速打开了市场。

He just built the product and got the traction.

所以现在这情况也不会变了。

So it ain't ain't gonna change now.

在 Alphabet、亚马逊、Meta 和微软这四家公司里，你可以买一家，卖一家。

Of the four, Alphabet, Amazon, Meta, and Microsoft, you can buy one and sell one.

你会买哪一家？

What do you buy?

你会卖哪一家？

What do you sell?

买亚马逊，卖微软。

You buy Amazon and you sell Microsoft.

为什么选亚马逊？

Why do you buy Amazon?

这个我们之前还没提到过。

It's the one we didn't touch on.

对所有听众来说，亚马逊的营收是1810亿美元。

And for for everyone listening, Amazon, 181,000,000,000 revenue.

AWS是370亿美元。

AWS, 37,000,000,000.

连续15个季度增速最快。

Fastest growth in 15 quarters.

我的意思是，你看，现在市场上有两家净卖出，两家净买入。

You know, I mean, look, it's it's weekly held in the sense of, look, there's there's two net sellers, two net buyers.

谷歌和亚马逊表现不错。

Google and Amazon are doing well.

Meta和微软就不那么明确了。

Meta and Microsoft, not so clear.

所以你显然得在每一边各选一个。

So you obviously pick on one each side.

我选微软，是因为他们那句‘不含AI业务营收持平’让人真的停顿了一下，而Meta虽然干的事儿挺疯狂，但就算它停手，依然是全球最牛的广告网络业务。

And I picked on Microsoft because that sentence that they're flat excluding AI just made you really pause was Meta, even though what they're doing is crazy, they can always stop and they'll still be the best ad network business on the planet.

所以这是看空的一方。

So that's on the sell.

至于看多的一方，说实话，现在买谷歌太无趣了。

And on the buy, honestly, it's too boring to buy Google at this point.

一年前我们说谷歌很棒的时候，确实挺对的。

It was great a year ago when we were saying it's great.

但现在它的市盈率在这四家公司里已经是最高了。

Now it's slightly the highest valued on a PE basis of those four.

所以某种程度上，当时押注亚马逊是个临时的逆向操作——毕竟现在亚马逊已经能用上Anthropic的模型，能从中受益了。

So to some extent, it was an on the fly contrarian bet in that Amazon just has now access to the Anthropic models and is going to get that lift.

他们有AWS的分销渠道。

They have the AWS distribution.

现在又有了Anthropic的模型，双方的契合度更高了。

They now have the Anthropic models, so they're more aligned there.

有一点我们一直在谈，就是人工智能对收入的影响，以及对利润的次要影响。

One thing, we talk constantly about how AI impacts the top line and, to a lesser extent, the bottom line.

还有一种间接效应，对所有人都有好处，但对Meta一点好处都没有——我们不只是用AI来优化应用。

There's a derivative effect that benefits everyone and does not benefit Meta at all, which is we're not just using AI to make applications better.

我们正在创建海量的新应用。

We are building many, many more applications.

现在正迎来一波应用开发热潮。

There is an application boom.

这是有史以来最盛大的应用开发浪潮。

This is the greatest boom in application building in the history of our lives.

数量已经呈爆炸式增长。

The amount has exploded.

现在，卖AI计算令牌赚的钱，甚至比卖基础的AWS或GCP服务还要多。

Right now you make even more money selling tokens than you do selling basic AWS or GCP services.

但这里其实有双重热潮在同时发生。

But there's a double boom going on here.

所以我得把Meta放在最后，因为他们也没从这场应用热潮中受益。

So I got to put Meta last because they're not benefiting from the application boom either.

他们根本没从中获益。

They're not benefiting from it.

也许还会出现一种交叉情况，就是AI变得成熟了——但我完全不觉得这会发生。

And there may even be a crossover where the AI gets a little mature, which I do not remotely see happening.

明白了。

Okay.

但就算真发生了，应用热潮才刚刚开始。

But even if it did, the application boom has just begun.

各种应用百花齐放，说不定这一季度正是SaaS从崩溃中反弹的开始。

The thousand flowers of applications and, you know, maybe maybe this was the quarter SaaS bounce back from the SaaS apocalypse.

但我相信再过几年，我们回头一看，会笑谈什么SaaS末日，说：天啊，B2B应用的爆发程度，真是前所未见。

But I think in a couple of years, we'll look back and we'll laugh at the SaaS apocalypse and say, my God, the explosion of B2B applications were like nothing we've seen before.

那些老一辈的人，学会了新舞步，却没倒下。

There were just old guys, old guys that learned a new dance and didn't die.

但谁在乎呢？

But who care?

几年后我们回头看就会明白。

We'll look back in a couple of years.

那时候我们根本不会在意，因为应用爆发得太厉害了。

We won't care because of this app explosion.

人人都在开发各种各样的东西。

Everyone's building everything.

这真是太棒了。

It's so wonderful.

这可能会让亚马逊迎来一次超越以往的复兴。

And so like that could lead to a renaissance for Amazon beyond what we've seen.

没错。

Right.

凭借它传统的优势，它确实让微软和企业级市场受益，但再往下的领域就不是了。

Because of its traditional strength, It benefits Microsoft and the enterprise, but not below that.

对吧？

Right?

GCP 和 Gemini 这个组合挺特别的，成本低得惊人。

And GCP is in Gemini are this weird thing where it's very cost effective.

对开发者特别友好。

It's developer friendly.

我没法预测未来，但我确定 Meta 在这场应用爆发中会吃亏。

I can't predict, but I know Meta loses in the app explosion.

在我们聊 SaaS 门户之前，你刚才提到的 Palantir，我们是临时加进去的，就在最后一刻。

Before we move to the SaaS portal, which you mentioned there, Palantir, obviously, very, very recently, we added it last minute.

性能表现简直太棒了。

Home run-in terms of performance.

递延收入达到了 44.5 亿美元，增长了 134%。

RPO is up a 134% at 4,450,000,000.

Rule of 40 指标高达 145%。

Rule of 40, they're a 145%.

杰森，我引用了你的推文，那封信写得非常直接。

Jason, borrowing from your tweet here, cops letter was blunt.

这个数字只有AI基础设施公司，比如英伟达、美光和SK海力士才达到过。

This number has only been matched by AI infra companies, NVIDIA, Micron, and SK Hynix.

我的天，我真的不知道该说什么了，太惊人了。

I mean, guys, I don't know what to say other than holy fuck.

这些数据真的太棒了。

These numbers are really fucking good.

他还提到，虽然这和未来的指引不太吻合，但他明确说公司正在翻倍。

He also said and it this didn't quite tie to the forward guidance, but he also said that they were they were doubling.

也就是说，他们的增速还在加快。

So they're accelerating from there.

他说得很清楚。

He was clear.

不过，我实在没法把这些和华尔街的预期对上。

Now, I couldn't quite get that to tie to the Wall Street guidance.

所以有时候你凭直觉说话，但很明显，他们的规模正在翻倍。

So it happens when you shoot from the hip, but it was very clear that they're that they're doubling.

他们的表现非常出色。

So they are excel.

我觉得他只是凭直觉说了RPO落地的时间，没太细想。

I don't I just think he was shooting from the hip in terms of the timing of when the RPO would land.

哈里明确说了，欧洲市场才刚刚起步。

And Harry, he was very clear that Europe's just started.

欧洲的进展远远落后，不是从零开始，而是从采购层面才刚刚启动，商业领域才刚开动。

It's so far behind, not at the startup level, but at the buying level, that that's just started on the commercial side.

我理解的是，他们现在处于一个非常有利的位置，对吧？

The way I was thinking about it is they're just in a very enviable position, right?

因为如果你是一家大公司，卖应用软件的时候，你总希望成为客户最高层决策者最重视的两个项目之一。

Because if you're a large corporation, I always think that whenever you're selling application software, you want to be one of the top two initiatives for the most senior person you're selling to.

这才是赚钱的关键。

That's how you make money.

现在他们直接在向首席执行官推销，而对美国每一家大公司的CEO来说，最重要的两件事之一就是搞人工智能。

And right now, they're selling to the CEO, and the most senior initiative, the top two initiatives for every CEO in corporate America is do something in AI.

所以你的董事会也在这么告诉你。

So that's what your board is telling you.

那你打算怎么办？

So what are you going to do?

你仔细想想，其实能选的路没多少。

If you think about it, you don't have a large number choices.

是的，你可以跟Anthropic或者OpenAI签个协议，买Copilot，或者给每个员工买Claude，这样员工个人是能用上AI了，但对公司整体没什么实质影响。

Yeah, you can sign up something with Anthropic and Opening, you can buy Copilot or you can buy a coworker, you can buy Claude for every one of your employees, and that gets individuals using AI but doesn't move the needle.

你真正该做的，是搞一个覆盖全公司的大动作。

What you really want to do is some corporate wide initiative that's big.

我们投资的那些AI应用公司，大多都只是解决某个具体问题，因为它们本来就是这么起步的。

And all the AI apps companies that we all fund, they tend to be fairly point solution because they've come from that.

它们都是过去两三年才成立的。

They've been founded in the last two or three years.

他们特别擅长说：咱们拿Sierra举例，现在大家都在谈，干脆先解决你们的客户服务问题。

And they're really good at saying, let's take Sierra, we're talking about it, say, we'll deal with your customer support.

换句话说，在新AI领域里，Sierra可能是最能接触到高层的公司了。

And that's probably the most call high company in AI land, in new AI land.

也就是说，布雷特·泰勒的影响力比谁都大。

In other words, Bret Taylor can call higher than anyone else.

但大多数才25岁的创始人根本没这个能耐。

Most of these 25 year old founders can't.

所以对他们来说，能接触到的最高层就到这儿了。

So that's as high as they can call.

你可以花两百万美元用Sierra，也可以花两百美元用别的东西，这问题就来了。

So you can spend $2,000,000 with Sierra, you can spend $200 with whatever, and that's an issue.

如果你在管理一家美国大企业，而你的首要任务就是搞AI，那你怎么可能只花两百美元？这点钱根本解决不了问题。

If you're running corporate America, if your number one task is to do AI, you don't spend $200 because that doesn't solve the problem.

还有一家公司，过去二十年一直在跟美国政府做生意，动不动就是几千万甚至上亿美元的单子。

And then there's this one company that's been selling to the US government for twenty years that can move in 20 and $100,000,000 chunks.

他们真的可以说：你想彻底重构你的市场推广体系？想用AI全面升级你的商业智能？我们都能搞定。

They can literally say, yeah, you want to redo your entire go to market infrastructure, you want to redo your entire business intelligence with AI, we can do that.

我们给美国政府做过，给国土安全部做过，也给摩根大通做过。

We did it for the US government, we did it for the DHS, we've done it for JPMorgan.

他们真正能处理上千万美元的项目。

They are credibly moving in $10,000,000 chunks.

所以现在，这家企业的CEO可以说：今年我有两个重点任务。

So now this corporate CEO can say, I have two initiatives this year.

一个是推出新产品，另一个是让公司全面转向AI驱动。

Launch that new product and make this company AI first.

我刚签了Palantir，金额一千万美元，他们会派团队来，真正把事情推起来。

I just signed Palantir for $10,000,000 They're going to bring in people, they're going to make shit happen.

任务搞定，打个勾，向董事会汇报，现在是六月三十号，全年目标进展顺利。

Initiative done, tick it off, report to board, it's June 30, and I'm on track for the year.

听起来可能有点傻，但大公司要想办大事，就得花大钱。

And it sounds stupid, but big companies have to spend big money to do big things.

这就是为什么IBM能比它该有的寿命多撑了三十年，因为他们能搞定这种大单子、大项目。

It's why IBM has existed for thirty years longer than it should, because they can do those kind of big deals, the big initiatives.

所以EDS当外包服务商时才赚得盆满钵满。

That's why EDS made a lot of money as an outsourcer.

目前只有Palantir能有底气对人说：嘿，各位企业界的朋友，

Palantir right now is the only business that can credibly say, Hey, Mr.

你们这些财富500强，想用AI项目真正推动业务进展，带来可衡量、看得见的成果，实现全企业范围的转型？我们能办到。

Corporate America, Fortune 500, you want to move your needle with an AI initiative that has measurable, demonstrable results, that is about enterprise wide transformation, we can deliver you that.

而他们的竞争对手，不过是Accenture那种靠卖一堆云许可证的公司。

The alternative to them is Accenture with a bunch of cloud licenses.

这有什么了不起的。

Whoop de do.

我当然选那些正在赢的人。

I'd go with the guys who are winning the war.

Palantir是推动AI变革的通用催化剂，尤其是在机构投资领域。

Palantir is a general catalyst for AI transformation, you know, for the LP world.

你要是能推动一亿美元的项目，大概率不会被炒鱿鱼。

You need to move $100,000,000 You probably won't get fired.

没错。

Cool.

当然。

Absolutely.

而且它的价值也更清晰、更实在。

And has more clearly demonstrated value.

我知道这听起来太简单了，但你就从CEO的角度想想吧，杰森，你也清楚这一点。

No, I know that sounds really simplistic, but you just think of it from and Jason, you know this, but you think it's from the perspective of the CEO.

你不可能把公司最重要的三个项目，只押在二十万美元的预算上。

You can't have your top three initiative be a $200,000 spend.

这些AI项目，说白了就是小打小闹。

With any of these AI, it's just like, Oh, that's a of a weenie bet.

你回头就跟他们说：我们找的是那些为整个企业界和政府都干过活的团队。

You come back in and you say, We hired the guys that have been doing this for all corporate America and for the government.

我们花了五百万，这是三年的规划。

We spent $5,000,000 Here's the three year plan.

这事就这么定了。

You're done.

补充两点吧。

To add on to that, just maybe two things.

第一，数字明显涨了，可能是因为我年纪大了。

One, I'm not the numbers have obviously gone up because I'm dating myself.

我当年在Adobe当副总裁的时候，一次就两三个大项目，就像罗里说的那样。

When I was a VP at Adobe, there were only two or three big initiatives at a time, to Rory's point.

就这些。

That's it.

但每个项目每年至少两千万。

But they were all 20,000,000 a year and up.

所以现在可能涨到三千万了。

So maybe they're 30 today.

在财富500强的Adobe这样的规模下，要支撑一个对数万名员工至关重要的项目，成本就是这么高。

That's just what it costs at Adobe scale in the Fortune 500 to have something that was critically important for, you know, 5 figures of employees.

当时的预算就是这么定的。

And it was just that's what it was budgeted.

我记得以前Salesforce每年的预算高达2600万美元，Workday则是2400万。

I believe at back in the day Salesforce is 26,000,000 a year and Workday were 24.

那些才是真正的大型项目。

Those were the big projects.

部署起来要花五年时间。

They took five years to deploy.

现在Palantir不到一年就能搞定，简直不可思议。

Now Palantir can deploy it in less than a year, which is magical.

不过我还想补充一点，我做B2B这么久了，有件事挺有意思的。

But the other thing I would add, it was funny having been doing B2B for a while.

Alex Karp在财报电话会上提到过。

Alex Karp on the earnings call.

一开始他谈到了国防业务，说支持国家至关重要，甚至比客户还重要。

At first, talked about the defense business and he talked about how important it is to support our country and that like the only thing more important than our customers is supporting our country.

很明显，他们在西方世界的国防领域已经深度渗透。

And clearly they're deeply embedded in in the defense in the Western world.

接着他转到了商业业务这边。

Then he went to the commercial side of the business.

他说的一番话让我特别意外，虽然其实挺明显的——这跟我过去一年的经历完全不一样。

What he said to me was super interesting, even though it's obvious he's like this is different than in my entire year.

去年，我在商业这边，每次都是由某个利益相关方——比如市场部或者营收团队——把我请进来，然后我还得去说服另一个利益相关方，再说服下一个。

The last year, He said, my entire year in the commercial side, I get brought in by a stakeholder, marketing, revenue, someone, and then I got to sell to even a pounder, I got to sell to another stakeholder and another stakeholder.

整个过程要花上好几年。

And it takes a couple of years.

他说，就在去年，每次开会，所有相关方全都到场了。

He's like, in the last year, every stakeholder shows up to the meeting.

所有人都在。

Everyone is there.

所以，正如哈里所说，我们不仅是前两三个推动者，更重要的是，CEO和CFO现在都在说：所有人，现在都来参与进来。

So not only are we a top two or three drive, to Harry's point, drive the corporate change, the CEO and the CFO importantly are saying, now, everyone come to the table now.

如果Palantir是答案，或者至少是我们押注的方向，那我们不会再花两年时间去评估了。

If Palantir is the answer, or at least the bet, we're not going to evaluate for two years.

我们必须现在就行动。

We have to do this now.

亚历克斯在电话里表现得特别热情，带着点古怪又迷人的劲头，但这次采购节奏的加快还是让他吃了一惊。

And Alex was like, you know, he's so agro in his somewhat charismatic, weird, goofy way on the call, but he was taken aback by the compression of the buying.

对我来说，这感觉就像疫情期间的采购潮，因为他以前在商业板块从来没见到过这种场面。

It felt to me like the COVID buying cycle today because he never saw folks on the commercial side.

像Adobe这样的大单子他们以前也做过，但那时候从来不是所有人一起出现。

They did deals this big like Adobe, but instead of everyone's there.

这次，所有人都出席了TAMP会议。

Everyone shows up to the TAMP meeting.

你看，AI激发的想象力，是数字转型、数据库、客户端-服务器、SaaS这些都比不上的。

Look, AI sparks the imagination in the way that digital transformation, our databases, our client server, our SaaS just doesn't.

它有种让人惊呼‘天啊，这是真的吗？’的魔力。

It's got that spark of, Oh my God, is it live?

它有意识吗？

Is it sentient?

如今，美国企业界真的相信，这是改变公司的唯一途径。

Corporate America now believes it is the way to transform their company.

你说得对，杰森，每个董事会都在催促旗下公司赶紧跟上。

And you're right, Jason, every board is telling every company to get on top of this.

谁会站出来，说‘我不参加这个最重要战略的会议’？

Who's going to be the C level executive that says, No, I don't want to join the meeting on the most important court of initiative?

别指望这能行得通。

Don't think it's going to work.

现在的情况是，反对的声音几乎完全消失了，所以你肯定觉得，接下来会有很多资源错配。

What it means now is that there's literally the naysayer voice has gone to zero, so you do think there's going to be some pretty interesting misallocation.

就像我说的，你现在要是卖五百万美元的解决方案，而且正好是企业最重视的议题，那你可真是赶上了好时候，真不错。

As I say, if you're selling $5,000,000 solutions right now for the biggest corporate imperative, you're in a golden place, and good for them.

市值三千四百九十亿美元，这价格是已经完全反映预期了，还是仍有上涨空间？

At 349,000,000,000 market cap, though, is it priced to perfection, or is that upside?

价格已经完全透支了。

It's priced to perfection.

甚至比完全透支还要高。

It's priced to more than perfection.

我得承认，看到这一季财报时我松了口气。之前你几周前逼我列出我看好且敢公开说的股票，我提了Atlassian，待会儿再聊，但在高价股里，我说Palantir是最有优势的。

I will admit I exhaled when I saw this quarter because when you pushed me in again few weeks ago to name stocks I'm proud that I named Atlassian, we'll talk about that in a second, but on the expensive side, I said Palantir is the best situated.

没错，它确实贵得离谱。

And yes, it's wildly expensive.

你得连续两三年高速增长，才能配得上这个估值。

You really have to grow for two or three years to get into this valuation.

但如果这波热潮能持续，它是最有可能追上来的。

But if the boom has legs, they're the most likely to grow into it.

说实话，我做完那个播客后的晚上，醒来还吓了一跳，心想：天啊，这风险也太大了。

Look, it's still a terrifying I woke up the night after I did that podcast, and I'm like, Oh, that's a risky one.

Atlassian 这个算是价值型投资。

Atlassian, it's a value play.

这根本不用想，直接上就行。

It's a no brainer.

但这个不一样，你已经入场了，而这一季的业绩基本证明了你的选择是对的。

But this one, wow, you're in, but this quarter kind of justified that.

你算一算，要是连续两年翻倍，这价格就一点都不贵了。

Because, you know, you do the math and two years two years of doubling and it looks cheap.

但亚历克斯之前预测的是整整一年翻倍。

But Alex predicted basically a year of doubling.

所以如果你觉得这价格便宜，那也不是完全没道理，我还没完全想透，但确实没那么离谱。

So there's an argument, if you think that's cheap, I haven't fully thought if you think it's cheap, it's not the craziest place to get to.

对吧？

Right?

因为一百亿的规模，CEO 早就说了，他可能之前一两个季度稍微调整了一下数据，但明确表示好日子就在后头。

Because 100, we've already the CEO's already said, you know, he probably fudged a quarter or two, but he said it's coming.

其实吧，我说的这个可能听起来挺显而易见的，但每次想到罗里讲的帕兰蒂尔公司是怎么运作的，都让我重新意识到这一点。

So the thing the thing just just I know it, maybe it's Captain obvious, but just going to Rory's narrative about how corporations work in Palantir, it's just a reminder.

我每次跟人聊天，都会再次确认：没有哪家公司内部真有这种能力。

Every conversation I have is a reminder that no one has this expertise in house.

一家都没有。

No one.

这简直是咱们这一代人见过的最悬殊的内外能力差距。

It's so it's the worst gap between in house and external expertise in our lifetimes.

所以未来几年，帕兰蒂尔肯定会持续受益。

And so for years, this is going to benefit Palantir.

以前我们还纳闷，为什么Anthropic和OpenAI要搞这些咨询业务，觉得挺奇怪的，对吧？

It's going to mean like even if we were head scratching, why Anthropic and Open AR setting up these consulting entities, seem goofy, right?

其实一点都不奇怪，因为钱会越来越多，需求也会越来越大，但企业内部根本找不到能真正落地的人。

They're not because the dollars are going to go up, the initiative is going go up, but the inability to anyone in house that can remotely execute, remotely execute.

总会有像HubSpot那样的代理机构慢慢摸清门道。

And there'll be folks down some HubSpot agencies will figure this out.

一些Shopify开发工作室，还有整个生态里的人都以为，大部分HubSpot和Shopify代理商才是真正部署这些系统的主要力量，但其实这些代理商主要服务的并不是中小企业。

Some Shopify dev shops will everyone in these ecosystems thinks the majority of HubSpot and Shopify agencies who are who'd actually deploy most of them, most is not direct at these SMBs.

要用HubSpot，你得找代理商来部署。

You need an agency to deploy HubSpot.

那些没跟上节奏的，都已经出局了。

Of them are gone.

他们完全没有AI方面的布局。

They have no AI play.

但那些已经布局的，会面临无限的需求，因为不管是大企业还是中小企业，根本没人具备这种能力。

But the ones that do are gonna have infinite demand because these the whether it's enterprise or SMB, they no one has this expertise.

居然有这么少的人掌握这项技能，简直让人不好意思。

It's it's almost embarrassing how few people have this expertise.

就连Coinbase今天也在裁员15%。

Even Coinbase today, I'm laying off 15%.

我手底下还有经理管着经理。

I don't I have managers managing managers.

我不需要他们。

I don't need them.

我需要的是真正能做AI工作的人。

I need folks that will that will actually do the work in AI.

对吧？

Right?

如果连Brian都找不到合适的人，那其他人还有希望吗？

If coin if Brian doesn't have the people, what hope is there for the rest of the world?

这么说听起来好像有点贬低人，好像没人懂这行。

And I think, and when you say it in one sense, it sounds almost derogatory, no one has the expertise.

但说实话，我现在是以微工具专家的身份在说话。

But the real truth is this, I'm donning my micro icon hat here.

从逻辑上讲，你希望有人花六个月时间把这本事练出来，然后卖给我，这样我只需要付他们一周的工时，我自己也只花一周时间。

Logically, what you want is someone to take the time, six months, to develop the expertise and then sell it to me, and I can only have to pay them for a week of my time and their time.

这完全说得通。

It totally makes sense.

我甚至会想我是怎么学会人工智能的。

I even think about how I learn AI.

很大一部分是靠自己动手做出来的。

A good slug of it is doing it myself.

但一旦卡住，我就直接打电话给我们的首席数据科学家，说：别让我在这儿耗一两天了，你直接告诉我答案吧，我好继续往下走。

But when I get stuck, I just ring our chief data scientists and say, Look, I could spend a whole day or two slogging through this point, but just tell me the freaking answer so I can keep moving here.

这其实就是一个缩影，每个公司都该这样。

That, in a microcosm, is what you need in every organization.

你是市场部门的主管，上面让你推这个项目，你就得找懂行的人来干。

You're a line manager in marketing, you've been told you want to roll this shit out, you just got to hire the expert who knows.

所以，杰森，你说得对，这其实是在考验每一个顾问，甚至每个求职者——这也正好扯到大学毕业生失业的问题。

Which is why, to your point, Jason, this is a Darwin test for every consultant, and frankly, anyone looking for a job too, which kind of gets to the undergraduate unemployment discussion.

如果你不提升自己的能力，那真是活该。

If you don't develop the skills you have, you deserve it.

我真心觉得，只要你掌握了别人需要的技能，专心致志地打磨真正有用的东西，这些公司里头，到处都有人抢着要。

I truly believe if you develop the skills that people want and focus like a laser on the things they actually want, there's a real demand for that skill in all these companies.

我管这叫二十五万美元的SDR。

I call it the $250,000 SDR.

现在市场上出现了一种新现象，SDR不再只值六万或八万美元，而是出现了少数年薪二十五万的顶尖SDR。

There is now a market, instead of SDRs being worth $60 in The US or $80, there is a market for a small number of $250,000 a year SDRs.

至于那些只上过半年社区大学、连‘11个实验室’都拼不对的六万美元SDR，基本没人需要了。

But there's almost no need for a $60,000 SDR with six months of junior college research that can't spell 11 labs.

现在真的不需要这样的人了。

You just don't need that person anymore.

三年前可能还需要。

You needed them three years ago.

但今天已经完全用不上了。

You don't need them today.

你花二十五万雇一个人，他能顶二十个普通SDR的活儿。

But you can spend 251 for for someone that's as productive as 20 human SDRs.

这才是你真正该掌握的本事。

Those are the skills you have to have.

接下来我们讲讲SaaS的末日，但这事儿真搞笑。

We're going to move on to the SaaS apocalypse next, but it's so funny.

我们马上要干一件特别逗的事。

We are about to do something incredibly funny.

我们打算跳过议程里的这一项。

We're about to skip the one line item in the agenda.

顺便说一句，全面碾压。

Just to point out, beat across the board.

蒂姆收官打得漂亮。

Tim punches out on a high.

结果太惊人了。

Stunning results.

这个季度表现太棒了。

Great quarter.

还没什么真正的AI故事。

No real AI story yet.

谢谢大家被这场狂热牵扯进来。

Thank you everyone else for getting caught in hysteria.

与此同时，我们正在打造全球最大的公司之一，而且表现非常出色。

Meanwhile, we're just building the largest you know, one of the two largest companies on the planet and doing really well here.

我们没把钱浪费在资本支出上，继续回购股票，一切为了股东利益。

Not wasting our money on, you know, CapEx, continuing to do buybacks, just working for the stockholders.

我只是顺带提一下这件事。

I just felt the need to call that out in passing.

我觉得内存芯片的供应限制挺有意思的。

I thought the memory chip supply constraints were interesting.

确实有意思，而且这不光是苹果的问题，我刚看到一个有趣的数字。

It is interesting, and it's not just for Apple, but just one interesting statistic I saw.

大家都在说，这个季度资本支出预算普遍上调了10%、20%甚至30%。

Everyone's talking about how the CapEx budgets have all been raised this quarter, 10% or 20%, 30%.

其中很大一部分预算上涨，并不是因为增加了设备投入，而是因为内存成本暴涨，同样的硬件现在贵多了。

A significant slug of that CapEx raise is for the same physical amount of CapEx, just at a higher price, because the memory portion of anything you're doing has exploded in cost.

所以我同意，我觉得明年iPhone的价格会直接反映出这一点。

So I do agree, and I think that you're going to see some impact of that right down to the price of your iPhone next year.

目前看来还没影响到需求，但这种影响迟早会传到消费者这边。

It doesn't appear to have impacted demand, but you probably will see that filter through.

到时候你买不起的iPhone，可能连运到美国都成问题。

That would be the iPhone you can't afford transported to The U.

S型号。

S.

坐的飞机连远东的航油都加不上。

On a plane that can't have jet fuel from the Far East.

不过，事情就是这样了。

But yeah, there you go.

接下来还有更多苦头要吃。

Much pain to come.

但没错，我觉得内存价格，还有由此带动的内存股，不只是对苹果，而是对整个资本开支的格局都有影响。

But yes, I think memory prices and obviously, thus memory stocks is a factor not just for Apple, but across the whole CapEx story.

这周Mac Mini的价格从599美元涨到了799美元，全是内存涨价闹的。

Well, mean, basically increased this week the price of the Mac Mini from $5.99 to $7.99 because of memory.

他们把599美元的型号下架了，因为卖光了，全怪这些该死的芯片短缺。

They dropped the $5.99 because it's sold out because of these damn open cars.

可能是16GB和8GB的区别吧。

Maybe it's the 16 meg versus the eight.

我记不清具体差多少了，但就算按现在的价格，差距也没到200美元。

I forget what the difference is, but it's not $200 even at current prices.

这种隐形通胀迟早会引发严重的社会经济压力——当普通人根本不在乎多花200块买Mac Mini的时候，真正受伤的是普通老百姓。

So all this stealth inflation, it's going to lead to a lot of socioeconomic stress when folks don't care whether there's a $7.99 or $5.99 Mac Mini and the average individual is hurt.

就拿旧金山的租房市场来说，现在竞争激烈到难以想象，跟我们刚做这个节目那会儿比，那时候住旧金山可便宜多了。

It's like the rental market in San Francisco is is almost unimaginably competitive versus even when we started the show, it was cheap to live in San Francisco.

现在根本租不到房了。

Now now you can't rent anything.

房源压根儿就不存在。

They don't even exist.

没人会走。

No one's going to leave.

我看到收购价比标价五百万还高出三百万。

I see buy prices like 3,000,000 higher than the 5,000,000 listed.

但实际情况并不是这样。

But consistently, it's not

对。

Yeah.

我刚从Anthropic的股权兑现中赚了三千万。

I just made 30,000,000 on my Anthropic vesting.

我伴侣想买房子。

My partner wants a house.

我能怎么办？

What am I gonna do?

跟她说这房子不值五百万？

Argue that it's not worth 5 mil?

我的意思是，都过去六个月了。

I mean, I and it's been six months.

我就直接买了。

I'm just gonna buy it.

我其实不在乎。

I don't really care.

明白了。

Okay.

上周我们还处在低迷期，正盼着点好消息呢。

Last week, we were in a bit of doldrums, and we wanted more positive news.

SaaS的末日可能要结束了。

The SaaS pocalypse could be over.

Atlassian涨了29%。

Atlassian up 29%.

Twilio涨了20%。

Twilio up 20%.

Five Nine 涨了百分之二十三。

Five nine up 23%.

伙计们，阳光终于出来了吗？

Boys, is the sunshine coming out?

SaaS 的末日真的结束了吗，还是只是这三家表现好的公司拉开了差距？

Is the SaaS pocalypse over, or is this a case of three good performing companies pulling away?

我们之前几次讨论时，是把它们归为一类的。

Where we categorized these companies the last few times a day.

让我来拆开说清楚。

Let me just break it up.

首先，Five Nine，我根本不关心。

First of all, Five nine, I'm not interested in.

他们重新加速增长了百分之九，我也不在乎。

Reaccelerated 9%, I don't care.

我不是说他们没受益，但我觉得他们还不配跟那些公司混在一起。

Okay, I'm not saying that they're not benefiting, but like, I don't think that deserve the to be to be with the friends.

首先，看到Atlassian和Twilio重新加速，确实让人感到欣慰。

First of all, it is heartening to see Atlassian and Twilio reaccelerate.

这两家公司都是老公司了，对吧？

These are older companies, right?

Atlassian到现在还是创始人在领导。

Atlassian is still founder led.

Mike之前上过我们的节目，记得吗？

Mike was on the show, right?

他真的很棒。

He's great.

但之前大家都对Atlassian挺担心的，对吧？

But everyone was worried about Atlassian, right?

Twilio呢，他们把Jeff给赶走了，就是那个我们特别喜欢的嘉宾。

Twilio, they, you know, they pushed Jeff out who was on the show that we love.

结果那些老掉牙的产品，居然也能重新加速了。

And somehow moldy oldy products were able to reaccelerate.

所以这是两个很棒的故事。

So two great stories.

至少它们证明了一点：对于那些受益于人工智能的公司来说，低谷已经过去了。

And if nothing else, they support the idea that for folks that are benefiting from AI, we're past the bottom.

对于那些受益的公司，接下来几个季度的数据可能不会那么亮眼，但之后的节目里应该会好转。

For folks that are benefiting, we're going to see some more quarters come out that maybe aren't as good, right, before the next show.

不过我想强调一下，这是我的观点和偏见——华尔街忽略了一个微妙之处，我相信你们也会认同。

But I do just want to this is my bias and my perspective that Wall Street, a subtlety that I think you guys will agree with the Wall Street didn't see.

Atlassian 在推广它的AI产品方面做得非常出色，我们之前也聊过这个。

Atlassian got really good at and we've talked about this monetizing its AI product.

产品叫 robo.ai。

It's robo.ai.

上个季度卖得特别火，客户都愿意掏钱买。

It sold the F out of it last quarter and people were happy to pay for it.

不过，它的新增客户数量仍在放缓。

Okay, its net new customer account is still slowing.

他们的新增客户数量还在放缓。

Its net new customer account.

Twilio 的情况正好相反。

Twilio did the opposite.

我们刚开始做这个播客的时候，这家公司几乎已经不行了。

This company was dead when we started this podcast.

现在它重新加速了，增长率达到了20%，而且还在提速。

Now it's reaccelerated to 20 but accelerated.

他们的公开信息没那么清晰，但过去一年里，由于AI和其他初创公司的影响，他们的新增客户数量可能增长了40%。

Their disclosure is less clear, but their net new customer account may have grown 40% in the last year because of AI and other startups and other companies.

不过，每个客户的年经常性收入并没有涨太多，对吧？

Now the ACV per customer hasn't gone up quite as much, right?

但确实出现了一波爆发式增长。

But there is an explosion.

Eleven Labs 在用 Twilio 的服务。

Eleven Labs uses Twilio.

所有这些做AI的人都在用Twilio。

All these AI folks use Twilio.

所以Twilio在两个方面都赢了。

So Twilio won on two points.

大家用它来做AI，而且新增客户数量也在增长。

Folks use it sort of for AI and they had net new customer growth.

Atlassian通过AI把现有客户群变现了。

Atlassian monetized its base with AI.

这表现不错，对吧？

Thumbs up, right?

通过了考验。

Pass the test.

但不清楚它是不是在吸引新客户。

But it's not clear it's attracting new customers.

所以，如果你的AI故事只是从现有客户身上挖出更多收入，却没有扩大客户群，那本质上你可能是在推迟坏消息，对吧？

And so you might be deferring bad news ultimately if your AI story gets more revenue from your base but does not expand it, right?

还有待验证。

To be proven.

还有待验证。

To be proven.

你说的事实没错。

You're right on the facts.

那回到你刚才说的，SaaS的末日真的结束了吗？

So going back to the you said, Is the SaaS apocalypse all over?

还没。

No.

三个月前，我们就说Teams值得买入。

Three months ago, we said, I think Teams are buy.

Atlassian也值得买入。

Atlassian is a buy.

只要你别老盯着那些细枝末节，从大方向看，当人们把这类股票的价格压得跟要归零一样时，反而可能迎来20%到30%的反弹。

If you just avoid going from guardrail to guardrail, the big picture is when people price these stocks like it's all going to zero, you've got this potential for a 20% or 30% bounce.

你现在也看到了，对吧？

And you saw that right now, right?

以Atlassian为例，我觉得它从现在开始不可能涨到十倍。

Taking Atlassian for example, I don't think it's like it's not going to 10x from here.

它不会，也永远不会成为一家以AI为核心、增长五倍或十倍的公司。

It's not, and it's never going to be an AI first, 5x, 10x growth company.

它真正会成为一家管理得非常好的公司，就像你说的，杰森，如果你没法新增客户，那稳定增长可能就在20%到30%之间。

What it is, is going to be a really well run company that, as you say, Jason, if you can't add new customers, maybe your steady state growth is 20% or 30%.

如果还能再拉来新客户，增长空间还能更高。

If you can add new customers on top, maybe you can go higher.

这些公司都会实现正现金流，年增长达到30%。

These are going to be cash flow positive companies growing 30%.

它们会妥善解决股票期权的问题。

They're going to deal with the SBC issue.

到那时，它们的估值不再是三倍，而是六倍。

Then they're going to be, instead of being worth three times, they're going to be worth six times.

我的意思是，你就得在这个范围内操作。

My point is this, you're to move in that bounded range.

当大家恐慌到觉得一切都完了，股价跌到三倍估值，但公司基本面其实还不错的时候，你就能买入，轻松拿两倍回报。

And when people's heads start exploding and saying it's all going to shit and they trade at three times, but the company's fundamentals are decent, you can buy and get a two x.

可当大家被各种 hype 洗脑，股价还像 ServiceNow 上次发布前那样高高在上，卖到六七倍营收的时候，你就特别容易被回调打中。

And when people start believing all the bullshit and they're still pretty lofty like ServiceNow was before the last announcement, when you're trading at six, seven times revenues, you're very vulnerable to a correction.

你说得对，低端市场是五个九。

And you're right, the low end is five nines.

是的，它增长了百分之九。

Yeah, it grew at 9%.

Atlassian 的收入增长相当亮眼，GAAP 收入达到了百分之三十二。

Atlassian's growth rate was pretty impressive, 32 in GAAP revenue.

我对新客户增长乏力并不意外，因为我一直觉得，就算没有 AI，这些公司的增长也会放缓，因为这些市场已经相当成熟了。

And I'm not surprised at the lack of new customers, because I always felt these stocks would slow down independent of AI because these markets are fairly well served.

十五年来，Atlassian 在满足这类产品需求方面做得非常出色。

Atlassian has done a great job over fifteen years of meeting the need for this product.

现在，任何额外的客户增长——无论是新公司成立，还是从别人那里抢来的客户——

And now, whenever additional customer growth, either as new company formation or takeaways from someone else.

所以这并不意外。

So it's not surprising.

这是一家边界清晰、执行得非常好的公司。

It's a bounded, well executed company.

所以我为他们感到高兴。

So I'm delighted for them.

我早就说过，两个月前有人问起的时候我就这么说了。

I'm delighted I said by it two months ago when we were asked.

我常在这里说，事情的发展完全符合预期：这些公司根基扎实、增长迅速、现金流强劲，如果增长率能冲到30%以上，估值接近六倍甚至更高才合理——这正是疫情前十几年它们的正常估值水平，事实就是这样。

As I often say here, things are proving out to be exactly what they should be, which is these are solid, high growth, cash flow generating companies, probably worth closer to six times if they're going north to 30%, maybe even a little higher, which is what they traded at for a decade and a half before COVID, and there you go.

但当然你说得对，对我来说，Twilio更吸引人，因为它也在获得全新的客户增长。

But of course you're right, but to me, Twilio is more interesting because it's also benefiting from net new customer growth.

杰夫·劳森上节目那会儿，Twilio还……其实也没过去多久。

When Jeff Lawson was on the show, Twilio was in the I mean, it wasn't even that long ago.

那时候它可是被冷落了，对吧？

It was in the doghouse, right?

看起来像死了一样，产品也过时了。

Looked dead, a dated product.

哈里问杰夫这事，杰夫说：其实我有一阵子没管这摊事了。

And Harry asked Jeff about this and Jeff's like, well, I actually haven't been in the game for a little while.

我在搞核聚变，想办法利用太阳的能量。

I'm working on nuclear fusion and harnessing the sun's energy.

但要是非让我想想的话，Twilio会从中受益，因为API正是AI和智能代理需要的东西。

But if I had to think about it, Twilio is going to be a beneficiary because APIs are what AIs and the agents need.

我们占着主导地位。

We have the dominant service.

随着这些智能产品规模扩大，它们都会用上我们，Twilio自然会表现不错。

And as these agentic products scale, they will use us and Twilio will do well.

可能不是Segment，对吧？

Maybe not segment, right?

还有一些其他事情，但核心业务肯定会。

And some other stuff, but the core business will.

他说得对。

And he was right.

它变成了这样，你知道的，就是只有一个案例。

It become, you know, it's just n equals one.

如果你去看Replit或者Lovable，它们也会默认选择这个方案。

If you go into replant or lovable, it's also the default choice.

这其实回到了你最初的那个说法。

It goes back to your early framing.

你有没有为现有客户带来新产品的价值？

Are you getting value for new products to your existing customers?

你有没有吸引到新客户？

And are you getting new customers?

为什么Lovable或者Replit要费劲重新搭建这套系统？

And why would Lovable or Replit bother rebuilding this stack?

尤其是如果你能通过API按需购买，那就直接用它好了。

Especially if you can buy it on an API basis, just call it a day.

风险在于，如果底层有人做出了比Twilio更好的替代品，你可能会换掉它。

Well, the risk would be if at the infra layer there someone had built a better Twilio that was that was better, you might switch.

对吧？

Right?

但Sierra这家公司，150亿美元的交易，都是跑在Twilio上的。

But Sierra, dollars 15,000,000,000 deal run on run on Twilio.

对吧？

Right?

我不算专家，但很明显，可靠性是关键。

There's a lot of I'm not a total expert, but obviously there's reliability, right?

在基础设施之下，还有更底层的支撑。

There's infrastructure under the infrastructure.

所以我的体会是，对很多人来说，没人能超越Twilio。

And so my learning is for a lot of folks, nobody beat Twilio.

这不仅仅是软件的问题。

It's not just software.

他们本可以的。

They could.

我的意思是，这完全有可能，对吧？

Like, it's possible, right?

但现在他们成了受益者，因为这套基础设施足够好，就像苹果那样。

But if now they're a beneficiary because this infrastructure was good enough like Apple.

这么说有点间接，但确实有那么点意思——他们正好搭上了这波趋势的顺风车。

I mean, it's a little attenuated, but there's something it's good enough to benefit from all the trend happening.

够用了就行。

It's good enough.

还有哪些传统的SaaS公司被困在‘SaaS牢笼’里，像Atlassian和Twilio那样有望被释放出来？

Who are the other traditional SaaS companies in SaaS jail that should be released in the same way that hopefully Atlassian and Twilio are being released?

这里我和Rory的看法就不一样了。

Well, here's where Rory and I diverge.

这就是我担心Atlassian不够格成为AI双重受益者的原因。

I I that's why I'm worried Atlassian isn't a two pronged AI beneficiary.

要成为AI的双重受益者，你得既能靠AI赚钱，又能吸引新客户。

Be a two pronged AI beneficiary, you have to be able to monetize your AI and you have to attract new customers.

这有两个方面。

There's two prongs.

到目前为止能做到的，都是跟基础设施靠得近的公司，对吧？

And the ones that have done it so far are close to infra, right?

比如Cloudflare、Twilio、Mongo、Datadog，连DigitalOcean都做到了，这说明谁都有可能。

Cloudflare, Twilio, Mongo, Datadog, and even DigitalOcean, which proves anybody can do it, Right.

如果你是第十一个云服务商，还能把股价涨三倍多，那只要有基础，谁都能做到。

If you're the eleventh cloud provider and you can grow three fifty two percent stock price, anyone that has a can do it.

但我最想看的是哪一个。

But I'll tell you the one I'm waiting to see.

这周HubSpot宣布了，他们即将发布的版本里，智能代理将和人类员工享有同等地位。

The HubSpot this this week announced that HubSpot will put agents on parity with humans in their coming release.

他们的平台将完全对智能代理开放，并确保 HubSpot 的代理版本至少与人工版本持平。

It'll be their platform will be completely open to agents and they'll make sure that the agentic version of HubSpot is at least at parity with the human version.

这才是正确的方向。

That's the right vision.

不过，他们会因此涨价吗？

Now, are they going to overcharge for it?

说实话，我现在也不确定了。

I mean, I'm not even sure anymore.

这在中小企业之外也很重要。

It matters outside of SMBs.

我想看看，虽然有点晚了，但还不算太晚。

I want to see if that it's a little late, but it's not too late.

确实有点晚了。

It's a little late.

但还不算太晚。

It's not too late.

我想看看这能不能行得通。

I want to see if that works.

我想看看它能不能成功。

I want to see if it works.

某种程度上，这正是马克·贝尼奥夫之前说过的无头架构理念。

In a way, it's Mark Benioff's headless vision that he talked about.

对吧？

Right?

如果HubSpot能成为中小企业和销售运营领域所有智能代理的中心平台，它的增长肯定会再次加速。

If HubSpot can become the hub for agents in all of its categories for SMBs, for GTM, it should reaccelerate dramatically.

咱们等着瞧吧，看它能不能成。

Let's see if it works.

如果这点都做不成，那其他的一切也就不用指望了。

If it doesn't work there, I think we can write all the rest off.

它所有的竞争对手都一样。

All of its peers.

我的意思是，应该能行，HubSpot 的客户群体太广泛了。

I mean, should work, HubSpot is such a broad customer base.

相比 Monday，它技术导向更强，对吧？

It's relatively more tech centric than Monday, right?

所以这个无头架构，我不是说它能让 HubSpot 翻倍。

So I think this headless thing, I'm not saying it should lead HubSpot to double.

对吧？

Right?

这在数学上根本不可能。

It's not mathematically possible.

我相信在未来十二个月里，如果这个策略真的落地，应该能给 HubSpot 带来真正的重新加速。

I believe over the next twelve months, this strategy, if it's real, should lead to genuine reacceleration at HubSpot.

否则，这些传统领域就没救了，因为它们会彻底向所有代理商开放。

Otherwise, it's there's no hope for these classic categories because they're going to make it completely open to all agents.

我觉得你说得对，杰森，它们当中会有更多无法重新加速。

I think to your point, Jason, far more of them are going to fall into the not reaccelerating.

重新加速会是例外，而不是常态。

Reacceleration will be exception, not the rule.

至于其他的，问题就在于，它是慢慢融化的冰块，还是迅速融化的冰块？

And then for the others, it's a question of, is it a slowly evaporating ice cube or a quick evaporating ice cube?

这场危机背后的根本问题并没有改变。

The underlying issues in this asspocalypse haven't changed.

要想让围绕它的争议减少，我觉得还得再有几个公司进入这个类别。

For the drama to reduce around it, we need a few more folks to be in that category, I think.

我不确定Atlassian能不能同时抓住这两点，但我支持它，因为它能带动整个行业。

I'm not sure Atlassian gets both prongs, but I'm here for it because it lifts all the tides.

如果我们能有三四个这样的公司，那虽然问题还在，但至少我们可以换个话题了。

If we could have three or four of these, then we could kind of The underlying issues are there, but we can move on and talk about other things.

我再稍微替Atlassian说句好话，因为你说过另一个原则是‘没有新客户’，这当然是最好的情况，但我觉得这不现实，就像我不认为Zoom还能再拉到新客户一样。

Just to push a little bit on the positive on Atlassian again, because I thought your other rule was not new customers, which obviously is the best of all, but I don't think it's realistic any more than I think Zoom will ever get a new customer again.

每个人都需要Zoom账号，而他们早就有了。

Everyone needs a Zoom account has one.

但你之前提的另一个标准我觉得很好，就是他们能不能向现有客户推出新产品。

But the other test you applied, which I thought was a good one, was can they sell new products to their customers?

Atlassian的日活跃用户和AI收入确实有明显增长，我觉得这里面确实有提升的空间。

And Atlassian's daily active users, their AI revenues did take a jump, so I do think there's some lift there that you can get.

要是有十到十五家公司都能这么做，那我们就清楚什么是SaaS公司的优秀标杆了，因为我觉得现在大家最困惑的就是这个。

Anyway, it would be great if 10 or 15 of the companies start doing that, then you have a sense of what good looks like for SaaS companies, because I think that's what people are struggling with.

这些公司根本不可能变成纯粹的LLM、Sierra、Harvey这类东西，这完全不可能。

There's no universe in which any of these companies become obviously not an LLM or a Sierra or a Harvey or anything like that.

这事儿压根不会发生。

That's just not going to happen.

从现状来看，这条路走不通。

You can't get there from here.

关键问题是，你能不能在保持自由现金流为正、股票薪酬可控、毛利和净留存率健康的情况下，重新回到30%的增长？让没人再担心你未来价值归零。

The question is, can you return to 30% growth with free cash flow positive and stock based comp under control and a gross and net retention such that no one's terrified that you have a zero terminal value?

只要你能做到这些，展现出真正的价值，你那些规模优势、几十亿美元的收入，才能真正发挥出来。

If you can do those things and demonstrate relevance, then the advantages you bring to the table in terms of scale, in terms of a couple of billion dollars in revenue, can all come to the fore.

你说得对。

And you're right.

有几个人已经做到了。

A couple of these guys have done it.

如果更多公司能做到，我们就知道什么是真正的成功了。

If more of them do it, then we'll know what winning looks like.

正如你提到的，杰森，那些没做到的，差距也会变得非常明显。

And to your point, Jason, then it'll become painfully clear what not winning looks like.

坦白说，我们看到Medallia的情况，就是投资者转身离开，觉得这公司根本没戏。

And let's be frank, what we saw in Medallia was an investor walking away saying, This thing isn't winning.

我实在看不出它还有翻盘的可能。

I just can't get that from here.

就算它EBITDA是利息的五到六倍，可还是没什么希望。

Even though it had positive EBITDA that was five or six times coverage, there's just nothing here.

看到Twilio和Atlassian的做法，会对正在努力的公司起到激励作用，也会彻底击垮那些没跟上的——因为大家会明白，成功就该是这样，而你根本没做到。

Seeing what Twilio and Atlassian have done will be a positive for the people who are on that journey and will be the nail in the coffin for the people who aren't because it'll be like, oh, that's what it takes to win and you're not doing it.

先生，您来填空吧。

Mister, fill in the blank.

接下来聊聊私营公司领域，Anthropic估值高达440亿美元。

Onwards to private company land, Anthropic at $44,000,000,000.

我觉得Jason提的这个问题，是你加进去的，问得特别到位。

I thought Jason's question here, which you put in, was was exactly the right one.

要维持这样的收入增长，有足够的开发者吗？

Are there enough developers for this level of revenue growth to continue?

当你看看开发者服务的整体市场规模时，马上就能回答‘有’。

And you can immediately say yes when you look at the TAM of, like, the service value of developers.

但你看这里，每天的用量是一亿次。

But when you look at this, it's a 100,000,000 per day.

你们当时看到这个数据，是怎么想的？

How did you guys react to this?

我觉得这个问题问得正好。

I think that it's the right question.

我觉得，如果非得挑一个数字来帮我形成对这件事的判断，我们现在也在做投资组合方面的研究，那最关键的问题是：一个完全成熟的AI优先型公司，每位工程师的日常token开支占其薪资的百分比是多少？

I think that if I was to pick one number that I'd like to know, which would give me an informed opinion on this, and we're trying to figure it out, we're doing some work in the portfolio, what is the steady state token spend as a percentage of salary dollars per engineer for a fully mature AI first organization?

因为我刚才正好在做这方面的调研，从底层推算知识型工作的总潜在市场。

Because I actually did a just was looking at this, doing the bottoms up knowledge work TAM.

除了那种太模糊的通用人工智能，其他大多数岗位都有任务自动化的潜力，但在我看来，这个比例不到10%。

In anything other than pure AGI, which is too arm wavy for me, most of the other jobs have a task automation potential, and my view, that's sub 10%.

营销、销售、会计这些领域，确实能自动化一部分，但不可能全盘替代。

You can do some of marketing, some of sales, some of accounting, but not all of it.

但写代码就不一样了，今天就能自动化掉海量的工作。

Whereas for coding, you can do a shit ton today.

别急着反驳我前面说的，哈里。

And don't argue the former yet, Harry.

我的意思是，写代码这块，自动化空间有多大，现在看得很清楚；其他领域还没这么明确。

The point is, coding, it's pretty clear that there's a huge amount of automation that can be done in a way that it's not as clear yet for the other areas.

咱们先就认这个前提吧。

Let's just go with that for now.

所以，编程才是真正的前沿。

Therefore, coding is the tip of the spear.

编程嘛，你想怎么说都行，就像是矿井里的金丝雀。

Coding is pick your cliche, it's the canary in the coal mine.

因此，如果你能知道长期稳定的自动化支出占工程师薪资的百分比，你就知道这事儿能有多大。

So therefore, if you know what the long term steady state automation token as a percentage of salary is, you know how big this can be.

如果这个比例达到百分之二十或三十，Anthropic 就有机会迈入千亿美金营收的行列，甚至可能接近五千亿。

And at 20% or 30%, Anthropic can grow into that multi $100,000,000,000 revenue category, maybe even half $1,000,000,000,000.

但如果只有百分之五，那就难多了。

At 5%, it gets a lot harder.

对吧？

Right?

所以在我看来，这才是关键问题。

So to me, that's the question.

当你看安德烈·卡帕西说的，他以前用它完成百分之二十，现在则靠它搞定最后那百分之二十。

When you look at Andre Kapathi saying, right, he used to use this for 20%, and now he helps it with the final 20%.