微软下调AI销售增长目标，AI对风险投资与旅行的影响 | 2025年12月3日

欢迎各位收看Information的TI电视节目。

Welcome everyone to the Information's TI TV.

我叫阿卡什·巴斯里查。

My name is Akash Basricha.

今天是12月3日，星期三。

It is Wednesday, December 3.

今天我们为大家准备了一场精彩的节目。

We have got a great show lined up for you today.

首先，我们的微软记者将做客节目，谈谈他在公司发现的销售策略变化——随着AI产品难以吸引客户，这些策略正在调整。

First up, our Microsoft reporter joins the show to talk about the changing sales strategies that he's found at the company as AI products struggle to attract customers.

接着，我们将深入探讨我们的每周财经专栏，分析信用评级机构如何在AI狂热中充当现实的校准者。

We'll then dive deep into our weekly finance column, which looks at how credit agencies are acting as a reality check-in the middle of the AI euphoria.

我们还为您带来对风险投资界的最新洞察，了解投资者如何看待AI和科技领域的投资。

We've also got a pulse check for you on the VC world and how investors are thinking about AI and tech investing.

最后，我们将以在线旅游行业为结尾，并邀请一位欧洲创始人，谈谈2025年在该地区运营科技公司的真实体验。

And we will wrap up the show with a look at the online travel sector and bring on a European founder to talk a little bit about what it's like to operate a tech company in that region in 2025.

这是一期重磅节目，我们马上进入正题。

It is a big show, so let's get right on into things.

微软正在调整其销售策略，以应对向客户推销智能代理时面临的挑战。

Microsoft is making changes to its sales strategy as it tries to deal with challenges that it's facing to sell agents to customers.

这一消息来自《The Information》的最新报道。

That is according to a new report from The Information.

我想邀请我们的微软记者亚伦·霍姆斯，他撰写了这篇报道，来和我们详细聊聊。

And I want to bring on Aaron Holmes, our Microsoft reporter who wrote that story to talk all about it.

嘿，亚伦，最近怎么样？

Hey, Aaron, how's it going?

早上好。

Good morning.

早上好。

Morning.

谢谢你的邀请。

Thank you for having me.

那我们来谈谈你写的这篇报道。

So let's talk about the story you wrote.

微软在试图让更多的智能代理落入客户手中时，是如何改变其销售策略的？

How is it that Microsoft is changing its sales strategies as it tries to get more agents in the hands of its customers?

是的。

Yeah.

根据我从消息源了解到的情况，微软已经降低了其Azure销售人员向企业销售的AI智能代理软件的部分增长目标。

So essentially what I learned from sources is that Microsoft has lowered some of the growth targets for specifically AI agent software that its Azure salespeople are selling to enterprises.

也就是说，他们仍然希望今年的销量比去年有所增长。

And, you know, that means that they still do want to sell more year over year than they did last year.

但增长率的目标略有下调。

But the percentage growth targets have come down slightly.

我听说，这是在美国不同地区的多个Azure部门未能实现去年设定的目标之后发生的，这可能意味着目标设定得过于雄心勃勃，或者客户对这些产品的需求没有微软预期的那么高。

And I heard that that happened after you know, several different Azure units in different geographies in The US had trouble meeting the goals that were set last year, which could mean that the goals were simply too ambitious, or that customer demand for these products wasn't quite as high as Microsoft was expecting.

所以，如果他们在前线降低了销售配额，那么理论上这是否意味着整体收入的增长也会放缓？

Okay, so if they're lowering the sales quotas on the ground, then conceivably doesn't that mean that overall revenue would also slow in terms of growth?

是的，这是个好问题。

Yeah, it's a good question.

这意味着，具体来说，他们对这些产品销售增长的百分比目标已经较去年有所下调，也就是说，他们仍然计划每年销售更多，但增长速度没有去年初计划的那么快了。

So this means that specifically their growth targets, so the percent that they want these product sales to grow, has moderated from last year, which means, you know, they still plan to sell more year over year, but just not at quite as steep of a rate as they were originally planning as of last year.

对。

Right.

你认为造成这种情况的核心原因是什么？

And what's your sense about what the core reason is here for this?

我的意思是，这是销售方式的问题吗？

I mean, is this a problem with the way that people are selling the product?

是定价问题吗？

Is it pricing?

是价格太高了吗？

Are they too expensive?

我认为，向企业推广生成式AI代理的核心卖点是，这些工具可以帮助你自动化公司运营和员工工作的很大一部分。

You know, I think the pitch with generative AI agents to enterprises specifically is these are tools that you can use to automate large parts of how your companies run, how people do their jobs.

我认为，像微软、OpenAI、谷歌和亚马逊等销售方发现，对企业客户来说，这个门槛相当高，尤其是因为很多时候，这些软件的购买者并不确信AI能够完美运行而不出现错误。

And I think that what sellers like Microsoft, but also OpenAI and Google and Amazon and others are finding is that that's a pretty high bar for cut for enterprise customers to spend on, especially because, know, a lot of the time, the buyers of this software don't feel confident that the AI can function perfectly without making mistakes.

我认为，这还因为许多大型企业必须彻底改变他们使用软件或员工开展工作的方式，才能接纳这些AI代理工具。

I think it's also just, know, a function of the fact that a lot of large enterprises have to, you know, kind of overhaul the way that they use software or the way that people do their jobs in order to embrace these AI agent tools.

而这一过程所花的时间，可能比销售方原本预期的要长一些。

And that process is taking maybe a bit longer than the sellers had hoped.

对。

Right.

所以问题更多出在客户这边。

So it's more on the customer side.

并不是产品本身有问题。

It's not actually a problem with the product.

我的意思是，这些产品被承诺为能够实质性地替代人类大量工作的工具。

I mean, I think, you know, these products have been promised as something that can meaningfully replace the work that a human does a lot of the time.

而据我接触的客户反馈，并非所有人都相信当前的产品能够兑现这一承诺。

And you know, from the customers I've spoken to, not all of them believe that the products currently live up to that promise.

同时，许多客户对很快就能实现这一目标持乐观态度，也许明年年初就能做到。

At the same time, lot of customers are optimistic that you know, we're going to get there soon, maybe early next year.

但我认为，目前要合理地在企业中为AI代理投入大量资金还是有点困难。

But I think for now, it can be a little hard to justify spending meaningfully on AI agents in the enterprise.

微软在这一策略调整的故事中是唯一的吗？

Is Microsoft alone here in this shifting strategy story?

其实并不是。

You know, it's really not.

我的同事施里上个月报告说，OpenAI也下调了其向企业客户销售AI代理的目标数量。

I mean, we we reported my colleague Shri reported last month that OpenAI has also scaled down its own targets for how many AI agents it can sell to corporate customers.

我认为我们还看到其他厂商如Salesforce、谷歌、亚马逊等也出现了类似情况，总体来看，这些公司都在进行调整，试图弄清楚市场对这些产品的需求究竟有多大——坦率地说，这些产品在市场上还是全新的、未经充分验证的。

And I think we've also seen some similar things play out at other vendors like Salesforce, but also Google, Amazon, I think across the board, you know, we're seeing a readjustment in these firms, trying to figure out how much demand there is for these products, which to be fair, are, kind of a brand new proposition and somewhat untested in the market.

因此，出现一些调整是合情合理的。

So it makes sense there would be some adjustments taking place.

我想知道，对于销售团队来说，AI的表现如何？

I am curious for sales teams how good AI has been for them.

一方面，正如你所指出的，他们面临一些难以完成销售配额的挑战。

Because on one hand, as you pointed out, they've had some challenges reaching their quotas.

另一方面，人人都在购买AI，所以作为一名销售人员，这仍然是一个不错的行业，尽管难度很大。

On the other hand, mean, everyone's buying AI, so it's a good business to be in as a salesperson despite how difficult it is.

你对那些靠销售这些AI产品和服务获得提成的人的收入情况有多少了解？

Do you have any sense how lucrative this has been in terms of compensation for people who get paid on commission of these AI products and services?

这是否与2021年人们争相购买软件时的情况类似？

Is it at all familiar to what we saw in 2021 when people were were rushing to buy software?

这是否是一个根本不同的故事？

Is it a fundamentally different story?

我认为情况是相似的。

You know, I think it is similar.

但如果说有什么不同的话，那就是情况更加参差不齐。

But if anything, it's just been a lot more uneven.

我认为，一方面，企业级通用AI聊天机器人，比如ChatGPT或微软的Copilot，这类任何人都能随手用来查询信息的工具，需求非常强劲。

I think, you know, on the one hand, there is extremely strong demand for generalized AI chatbots, you know, things like ChatGPT for enterprise or Microsoft's Copilot that anyone can kind of just use to look things up on the side.

我认为，更棘手的是销售那些旨在为企业大规模自动化工作流程的AI代理工具。

I think where it gets a little bit more tricky is selling these AI agent tools that are meant to meaningfully automate large swaps of work at these companies.

据我所知，销售人员要达成微软或其他公司设定的目标确实更困难，这也可能是因为微软去年希望大量销售AI代理产品，因此给销售人员设定了非常高的目标。

And, you know, from what I've heard, it has been trickier for salespeople to hit the goals that you know, Microsoft or other companies have set, which could also just be a function of the fact that, you know, Microsoft wants to sell a lot of AI agent products and set really high goals for salespeople last year.

对。

Right.

很好。

Great.

好了，亚伦，感谢你前来做客。

Well, Aaron, thanks for coming on.

我们很感激。

We appreciate it.

我们很快再和你聊。

And we'll talk to you again very soon.

谢谢。

Thank you.

好的。

Okay.

如今，科技界的所有人都在更密切地关注债券市场，因为大型科技公司正在发行债务以资助其巨额资本支出，而信用评级机构正是这一故事中至关重要却常被忽视的群体。

Everyone in tech is watching the bond market more closely these days as big tech companies issue debt to fund their extraordinary capital expenditures, and the credit rating agencies are a group that are central to that story that sometimes get overlooked.

我同事肯·布朗今天在我们的周度财经通讯中发表了一篇专栏文章，评估了评级机构和标普在多大程度上尽职尽责——正如他所指出的，由于首先支付他们费用的是谁，这一点有时颇具争议。

My colleague Ken Brown's column out today in our weekly finance newsletter assesses the extent to which the rating agencies and S and P has been doing their job properly, which, as he points out, is sometimes a contentious point given who has to pay them in the first place.

肯，欢迎再次做客我们的节目。

Ken, welcome back to the show.

很高兴你再次回来。

It is great to have you back.

嗨，阿卡什。

Hi, Akash.

那我们来谈谈评级机构吧。

So let's talk about the ratings agencies.

我有个关于你的meta问题。

I have a bit of a meta question for you.

好的？

Okay?

小写的'meta'，不是大写的'META'。

Lowercase meta, not uppercase meta.

如果你要给评级机构打个分数，按照它们在当前AI债务故事中履行职责的有效性来评定一个字母等级。

If you were to give a rating to the ratings agencies, a letter grade in terms of how effectively they have been doing their job in the context of this AI debt story.

你会给他们打什么字母等级？

What letter grade might you give them?

嗯，到目前为止还不错。

Well, I mean, so far so good.

我可能会给个A，这已经是相当不错的评级了。

I'd probably put an A, which is a pretty good rating.

但他们总是说，当他们感到有些担忧时，会把公司列入观察名单，以备可能降级。

But what they always say is they put companies on watch for possible downgrades when they're getting a little nervous.

所以我认为，鉴于现在还处于早期阶段，而且他们只评级了一笔复杂的大额债券，之后一直在观察这些大型债务发行人，我建议把他们列入观察名单，以备可能降级。

So I think I would keep them on watch for a possible downgrade just because it's so early in this and they've really only rated one big complicated bond and then they've been watching these companies like all the big debt issuers.

所以我认为现在判断他们实际表现还为时过早，但初步迹象相当不错。

And so I'd say it's a little early to see how they're actually doing, but the early signs are pretty good.

对。

Right.

而你提到的那笔债券评级，这次是大写的META。

And and and the the one bond rating you're talking about is is meta, uppercase meta this time.

对吧？

Right?

对。

Right.

对。

Right.

所以，是的，Meta在路易斯安那州完成了这笔交易，为这个高达270亿美元的数据中心融资。

So, yeah, meta did this deal in Louisiana to fund this mega $27,000,000,000 data center.

他们这么做是因为不想把新的债务记在账上。

And they did it because they didn't want the new more debt on their books.

他们通过一个特殊目的实体完成这笔交易，这意味着这笔债务基本上不会体现在他们的资产负债表上，但他们确实借了钱。

They did it through a special purpose vehicle, which meant it basically doesn't sit on their books, but they borrowed the money.

这是一笔复杂的交易。

It was a complicated deal.

标普花了大量时间来评定这笔评级。

S and P spent a ton of time rating it.

这确实是第一个大型案例，也是他们介入并加入其中的地方。

And that's really been the first big one, and that's where they've stepped in and joined the party.

对。

Right.

总的来说，我很感兴趣，我知道科技公司发行的债务比其他行业要少，但总体而言，科技公司的信用评级应该还不错，我想。

Broadly speaking, I'm curious, and I say this knowing that tech companies haven't issued as much debt as they have in other industries, But broadly speaking, credit ratings around tech companies, they're pretty good, I imagine.

这些科技巨头，比如微软和谷歌等，产生大量现金，借款相对较少，因此它们的信用状况非常优异。

These tech giants, Microsoft and Google and all that, they generate so much cash and they borrow relatively little that they are pristine credits.

风险非常非常低。

There's very, very little risk.

我们之前谈到过，微软拥有最高的AAA评级。

We were talking earlier, Microsoft has a AAA rating, which is the highest possible.

这比美国政府的评级还要高。

It's higher than the US government.

所以不可能再好了。

So it doesn't get any better than that.

因此，信用评级机构和债券市场对科技行业通常并不那么重要。

And so, the credit rating agencies and the bond market generally haven't been that important to tech.

我的意思是，科技行业的一些部分，比如芯片公司确实会借款，但大型软件公司都心知肚明。

I mean, some parts of tech, chip companies borrow and such like that, but the big software companies know.

所以这是新情况，这也是一次考验。

And so this is new and so this is a test.

对。

Right.

当这些科技公司开始发行更多债务时，我很想知道你会关注哪些方面，如果他们确实计划通过这种方式为所有资本支出融资的话。

And I'm curious for what you might be watching for then as these tech companies move to issue more debt, if that's the way they plan to finance all these capital expenditures.

我的意思是，哪些因素会导致他们的评级下调？在这种情况下，你会关注哪些警示信号？毕竟评级机构一开始就是由这些公司付费的。

I mean, what sorts of things would bring their ratings down and what sorts of flags might you be looking for in this scenario where we know ratings agencies, they get paid by these guys in the first place.

所以他们必须认真履行自己的职责。

So they got to be doing their job properly.

没错。

Right.

这里存在一种固有的利益冲突：评级机构给出评级，但却是由发行债券的公司支付费用。

There's an inherent conflict of interest, which is these rating agencies give a rating, but they get paid by the companies that issue the bonds.

这个问题一直存在已久。

And so this has been a problem forever.

从来没有人解决过这种利益冲突。

No one's ever solved this conflict of interest.

所以你看，这些大型科技公司可以发行的债务远远超过它们目前发行的规模。

So look, the giant tech companies, they can issue way, way, way more debt than they have issued.

那里不会有任何问题。

There is not gonna be anything there.

我关注的是规模较小的公司。

What I'm looking for is smaller companies.

甲骨文就是一家，目前处于信用观察名单上。

Oracle is one, they're on credit watch.

他们已经发行了债务，并且有承诺义务。

They've issued debt and they have commitments.

所以这些会得到评级，对吧？

And so that'll be rated, right?

但所有这些问题中最大的疑问是，过去十年中迅猛发展的私人信贷和私人债务市场。

But the biggest question in all of this is all the private credit, the private debt market, which has boomed in the last decade.

他们已经发行了大量债务，而这些债务既没有评级，也不透明。

And they have come in with a lot of They've issued a lot of debt and that is not rated and that is not transparent.

所以我们根本不知道市场上到底有什么，以及具体的条款是什么。

So we just don't know what's out there and what the terms are.

然后我们还有像OpenAI这样的公司，它们承诺了大量债务，基本上就是说：我们将会产生天文数字般的收入。

And then we have companies like OpenAI, which is committing to a lot of debt and a lot of basically say, We're going to generate a gazillion dollars in revenue.

我们能够偿还这笔债务。

We'll be able to pay off this debt.

但目前它们的收入虽然很多，却远未达到那个水平。

But right now they generate a lot, but not nearly that much.

因此，这些正是我们关注的领域。

So those are the areas that we're focusing on.

我们应该关注哪些私人信贷领域的大公司？

Who are the big names in private credit that we should be watching?

主要是像阿波罗和黑石这样的大型私募股权公司，但也包括像Blue Owl这样的公司，它参与了元宇宙交易并完成了其他一些交易。

Well, it's a big private equity firms like Apollo and Blackstone, but it's also firms like Blue Owl, which it was involved in the meta deal and has done some other deals.

此外，还有许多其他专门从事这一领域的参与者，它们非常不透明，很难准确理解它们的具体操作。

And so there's a bunch of other players out there that specialize in And they're very opaque and it's hard to understand exactly what they're doing.

因此，我们不断追问，但这将是一个挑战。

And so, we keep asking, but it's gonna be a challenge.

你还写到了评级机构如何应对加密货币的问题。

Now, you also wrote about how the ratings agencies have been dealing with crypto.

你能稍微谈谈他们到底在评级什么吗？

And talk a little bit about what what are they even rating?

你知道的？

You know?

你写了一点关于泰达币的内容。

You you wrote a little bit about Tether.

我们知道，这些稳定币 supposedly 是由储备资产支持的。

And, you know, we we know there's the idea that these stablecoins are supposed to be backed by reserves.

但据我所知，这并不是债务，对吧？

But that's, in my experience, it's not debt, right?

那么他们到底在评级什么呢？

So what exactly are they rating?

它们不是真正的债券评级，也不是信用评级。

So they're not real, they're not bond ratings, they're not credit ratings.

信用评级机构关注的是这些稳定币的财务稳定性。

The credit rating agencies look at financial stability for these stable coins.

所以，Tether、Circle 和其他一些项目都是稳定币，意味着它们与美元保持一比一的锚定。

So, Tether and Circle and some of these others are stable coins, which means they are one to one against the dollar.

它们不应该具有波动性。

They should not be volatile.

它们几乎就像加密现金，对吧？

They're almost like cash, crypto cash, right?

因此，信用评级机构介入进来，评估这种资产的价值与美元保持一比一的可能性有多大。

And so the credit rating agencies have gone in and said, well, how likely is it that the value of this thing is gonna be one to one with the dollar?

于是他们设定了一个评级，从一到五级，五级为最差。

And so they had a rating, they have five rating, one to five, five being the worst.

Tether 当时是四级。

Tether was at four.

他们将其评级下调至五级。

They downgraded it to five.

Tether 是全球规模最大的稳定币。

Tether is by far the world's biggest stablecoin.

他们这样做是因为Tether持有大量非常安全的资产来支撑稳定币，但也持有一些高风险资产。

And they did it because Tether owns a lot of very safe stuff to back the stablecoins, but they also own some risky stuff.

他们持有比特币、贵金属、贷款，这些都是波动性较大的资产。

They own Bitcoin, they own precious metals, they own loans, and these are things that are volatile.

因此，他们持有的这部分高风险资产占总持仓的比例，从去年的17%上升到今年上个季度的25%。

And so the portion of that that they own, the portion of their overall holdings in this risky stuff rose from 17% a year ago to 25% this year in the last quarter, most recent quarter.

所以他们认为这个比例太高了。

And so they said that's too high.

这个稳定币的稳定性已经不再有保障了。

The stability of this stablecoin is not necessarily there anymore.

因此，他们将其评级下调至最低级别，即‘薄弱’。

So, load it to their lowest possible rating, which is weak.

对。

Right.

最后一个问题。

Last question for you.

当你思考所有这些正在发行的债务时，尤其是在AI这个话题上，如果我们回到这一点，从一两年后的角度来看，这个故事的看涨和看跌情形分别是什么？

As you think about all this debt that's being issued, certainly in the AI story, if we come back to that, If you think about the bull case and the bear case of this story, thinking out a year or two from now, what is the bull case and then what's the bear case?

看涨的情况是，这些大型公司拥有大量现金。

Well, the bull case is these giant companies have a ton of cash.

另一个看涨的因素是，现在还处于早期阶段。

The other bull case is it's very early.

这些公司未来几年的利息支付，完全可以从零花钱里轻松支付。

These guys can write the interest payment checks for the next couple of years out of petty cash.

这对他们来说根本不是大事。

This is not a big deal for them.

所以在未来几年内，可能不会出现任何坏账问题。

So, nothing probably will happen in the next couple of years in terms of bad debt.

真正可能发生的是，市场会变得紧张，从而停止为这些公司提供资金。

What would happen is the market would get nervous and stop funding some of these companies.

比如说，如果市场停止为OpenAI提供资金，那对这家公司来说将是灾难性的。

So, say they stopped funding OpenAI, that would be terrible for that company.

这对信贷市场来说不会太糟，但对那家公司来说会很糟糕。

It wouldn't be terrible for the credit markets, but it would be terrible for that company.

所以，这就是看涨和看跌两种情况。

So, that's both the bull and the bear case.

我的意思是，从长远来看，真正的看跌情景是我们看到OpenAI的增长速度没有它预期的那么快。

I mean, the real bear case over time is we see OpenAI doesn't grow as fast as it thinks.

此外，私人信贷市场崩盘，这会是一个真正的意外。

Also private credit market tanks and it's a real surprise.

所以，这些就是看跌情景的可能性。

So those are sort of the bear case scenarios.

很好。

Great.

肯，谢谢你能来。

Well, Ken, I wanna thank you for coming on.

我觉得还有很多内容可以留到下周你的下一篇文章中继续探讨。

I I think there's lots there to to maybe pick up with next week in your next column.

那我们就到这里，但很快还会再和你聊。

So we'll end it there, but we'll we'll talk to you again very soon.

谢谢。

Thank you.

好的。

Okay.

随着人工智能领域以及整个科技行业对泡沫言论的持续，风险投资家们仍面临着艰难的任务：试图做出长期投资并预测未来十年的发展。

Well, as bubble talk persists in the AI sector and the tech sector at large, venture capitalists have the difficult job of still trying to make long term bets and predicting the future ten years out.

风险投资行业目前也正处于一个引人入胜的时刻。

The business of venture capital is also in a fascinating moment right now.

为了探讨这一切，我想邀请六四五风投的管理合伙人兼联合创始人南迪·奥基克。

And to talk about all of that, I want to bring on Nnamdi Okike, Managing Partner and Co Founder at six forty five Ventures.

南迪，欢迎来到节目。

Nnamdi, welcome to the show.

很高兴你来到这里。

It's great to have you here.

很高兴能来这里。

It's great to be here.

谢谢邀请我。

Thanks for having me.

非常感谢。

Appreciate it.

我想和你聊聊几个月前我读到的一些数据，来自PitchBook和美国风险投资协会，我知道你是该协会的董事会成员。

So I want to talk to you a little bit about some data that I was reading a few months ago, PitchBook and the National Venture Capital Association, where I know you're a board member.

他们发布了关于融资数据的季度报告。

They put out their quarterly report on funding data.

现在是第四季度。

Know, we're in Q4.

我们正在等待第四季度的报告发布。

We're waiting for the Q4 report to come out.

但至少就前三个季度而言，让我感到着迷的一点是，关于有限合伙人（LPs）对风险投资公司的募资情况，我们正趋于接近近十年来的最低水平。

But at least as it relates to the first three quarters, I mean, one of the points that fascinated me is, as they said, with respect to LPs and fundraising from venture capital firms, we are sort of trending towards what could be nearly a decade low.

我想了解一下，你认为现在为什么会这样。

And, you know, I wanted to get your sense on why that's the case right now.

我的意思是，为什么风险投资基金现在从有限合伙人那里募资这么困难？

I mean, you know, why are venture funds having so much trouble fundraising from their LPs?

是的，这是个很好的问题。

Yeah, it's a great question.

我认为有几个原因。

I think there's a few reasons.

我认为其中一个原因是，过去几年向有限合伙人分配的资金一直很少。

I think one is there's been a lack of distributions going back to LPs in the past couple of years.

IPO和并购交易都减少了，而有限合伙人需要构建投资组合，他们希望在风投领域保持合理的比例，但又不能过高。

There's been fewer IPOs, fewer M and As, and obviously LPs have portfolio construction, and they want to have a reasonable amount in venture, but not too high an amount in venture.

因此，我认为由于缺乏分配，他们的投资组合配置有些失衡。

So I think their portfolio construction is a little bit thrown off due to the lack of distributions.

我认为这实际上意味着他们无法向基金投入那么多资金。

I think that really means they can't put as much into funds.

我认为基金的融资速度也在加快。

I think there's also been an increasing pace of fundraising by funds.

我认为这现在正在系统中逐渐显现，即LPs能向基金投入多少资本。

And I think that is now kind of working itself out in the system in terms of how much capital LPs can put into funds.

我认为有诸多因素在起作用。

I think there's a myriad of factors.

我还觉得LPs在选择投资标的时更加审慎，也更关注各类资产长期回报的表现。

I also think LPs are being thoughtful in terms of where they want to place bets and thoughtful in terms of what the returns of the asset classes are going to look like over time.

我认为之前的基金批次，尤其是2021年和2022年的批次，回报表现并不出色，这让我觉得一些LPs开始犹豫了。

And I think the previous vintages, especially the '21, '22 vintage, haven't been very stellar in terms of returns, which I think for some LPs has given them some pause.

现在，我想回到你刚才提到的关于风险投资基金融资速度加快这一点。

Now, I just want to come back to that point you made about the pace having increasing for these venture funds.

如果LPs提供给他们的资本总量在减少，为什么这些融资活动的节奏反而在加快呢？

Why is it that the pace of these fundraising events is increasing if the total amount of capital that LPs are giving them is decreasing?

这怎么可能说得通呢？

How does that make any sense?

所以我认为你需要回顾过去几年的情况。

So I think you got to go back the past couple of years.

在过去的五六年里，你看到的是基金之间的间隔时间大大缩短了。

So what you saw past five, six years is that the amount of time between funds became much shorter.

这正好发生在2021年、2022年左右的时间段。

This is right around 'twenty one, 'twenty two timeframe.

因此，基金筹集得更快，更多资本涌入了系统。

So funds were being raised faster and more capital was going into the system.

我认为，如果你看看四五年前筹集的那些基金，它们的回报一直滞后。

And I think if you look at those funds that were raised, say, four or five years ago, the returns have been lagging.

DPI（分配与已投入资本比率）非常低。

There is very little DPI.

我认为，2021年筹集的基金中，顶级四分位基金的分配与已投入资本比率远低于1倍。

I think the top quartile distribution to paid in ratio for funds raised in '21 is much less than 1X.

即使是最好的基金，其回报也仅为所筹资金的25%左右。

It's something for even the best funds, it's maybe 25% of what they raised.

因此，分配非常有限。

So there's been very limited distributions.

现在你开始看到这种情况产生影响，有限合伙人再次表示：如果我们从这些基金中拿不回资金，为什么还要继续再投资或往这些基金里投入更多钱呢？

And you start to see that working itself out now in terms of LPs again saying, Well, if we're not getting capital back from these funds, why should we be reinvesting or putting more money into these funds?

因此，风险投资是一种周期性资产类别。

And so, venture is a cyclical asset class.

它会随着浪潮起伏，有限合伙人需要看到回报和分配，才能有信心将资本投入其中。

It does move in waves and LPs need to see returns and distributions to have confidence that it's a great place to be putting capital into.

我认为，这正是导致今年有限合伙人出资情况如此的原因之一。

And I think some of that is driving what you're seeing in terms of LP contributions this year.

因此，我们看到IPO活动略有回升。

And so we've seen IPOs pick up slightly.

但这并没有达到人们原本期望的那种井喷式增长。

It's not the avalanche that people, I think, were hoping for.

我的意思是，IPO活动确实有所回升。

I mean, have been picking up.

现在我们先不谈这些即将上市公司的表现。

Now we'll put the performance aside of these new companies that are going public.

这完全没有帮助到分配吗？

Is it not helping distributions at all?

我的意思是，情况至少对有限合伙人来说有好转吗？

I mean, is the situation at all getting better, at least for LPs?

我认为隧道尽头有了一些光亮。

I think there's some light at the end of the tunnel.

正如你所说，今年你已经看到了一些IPO。

To your point, this year you've seen IPOs.

我主要投资于金融科技领域。

So I do primarily invest in fintech.

今年你看到了像Chime和Clarina这样的公司，正如你所说，它们在IPO后的表现并不出色，但至少为投资这些基金的普通合伙人和有限合伙人提供了流动性。

And this year you've seen companies like Chime and Clarina and others to your point, they haven't performed great post IPO, but at least they provided liquidity to the GPs and the LPs invested in those funds.

所以这是好事。

So that's a good thing.

我认为，如果你观察私募市场，通常会发现从早期风险投资（这是我们所做的）到退出的时间周期被拉长了。

I think generally what you find if you look at private markets is that there has been an elongation of time from say first early investment from early stage venture, which is what we do.

我们在种子轮和A轮进行投资，直到退出。

We invest at seed and series A to exit.

我认为，从公司创立或首次投资到上市的中位数时间，现在即使是最好的公司也达到了十三四年之久。

I think median period of time between initial investment or a company being founded and IPO is now upwards of thirteen, fourteen years for even the best companies.

这与十五年前的情况大不相同，那时周期要短得多。

That's very different than what it might have looked like fifteen years ago, where periods were much shorter.

而且你还会看到后期投资的快速增长。

And you also have this kind of growth of late stage investment.

以Stripe这样的公司为例，它们在私募市场中保持私有状态的时间很长，估值也很高，但依然没有实现退出事件。

So if you think about a company like Stripe, for example, how long they've been private and how high they're valued in the private markets, but again, haven't had the exit event.

所有这些因素，再次影响了有限合伙人从IRR角度所看到的情况。

And so all of those things, again, impact the LP perspective in terms of their IRR.

有限合伙人既关注IRR，也关注总投资回报倍数，但你只是延长了投资所经历的时间周期。

LPs think about both IRR as well as total value to paid in, total multiple, but you're just elongating the period of time in which you're investing in that.

这是LP们会考虑的因素。

That's something the LPs take into account.

对。

Right.

现在我想问你关于那份十月报告中的另一个数据点。

Now I want to ask you about another data point from that report in October.

风险债务达到了几年来的最高水平。

Venture debt is at its highest level in quite a few years.

目前风险债务的情况如何？

What's the story around venture debt right now?

风险债务市场一直非常有趣。

The venture debt market's always very interesting.

我们显然在几年前看到了硅谷银行的倒闭，这在很大程度上是由风险贷款实践驱动的。

We obviously saw the collapse of Silicon Valley Bank a couple of years ago, kind of largely driven by venture lending practices.

风险债务市场很有趣。

The venture debt market is interesting.

我们对投资组合公司的看法是，它可以作为股权融资的补充。

The way we think about it for our portfolio companies is it can be a compliment to equity financing.

它可以帮助公司筹集更多的资金。

It can enable companies to raise a bit more money.

风险投资融资、股权融资通常是风险债务的后盾。

The venture capital funding, equity funding is typically the backstop of venture debt.

但当风险债务水平非常高时，正如你所提到的，我总会感到有些担忧，因为这些风险债务提供给了那些通常不盈利、没有正现金流、严重依赖持续融资的公司。

But I always get a little bit wary when venture debt, to your point, levels are very high because venture debt is being provided to companies that typically aren't profitable, don't have positive cash flows, and are really relying on continual fundraising.

所以，比如以人工智能为例，对吧？

So if, for example, take AI, right?

如果未来几年人工智能公司的融资市场下降或枯竭，这些风险债务提供方将面临困境，因为这些公司难以从股权市场筹集更多资金。

If the fundraising market for AI companies drops or dries up somewhat over the next couple of years, those venture debt providers are gonna be in tough shape in terms of those companies being able to raise more money from the equity markets.

对。

Right.

通常情况下，公司并没有太多实物资产，嗯，也许在数据中心公司的情况下它们有资产，但大多数软件公司都没有资产。

And typically companies don't have a lot of physical Well, I guess in the case of maybe data center companies, they have assets, most software companies don't have assets.

所以你提到当水平过高时会感到担忧。

So you talked about getting weary when levels are too high.

你现在感到担忧吗？

Are you weary right now?

我并没有仔细查看过这些倍数，但你提供的数据——它们已达到历史高位——确实令人有些担忧，是的。

I haven't looked at those multiples really closely, but the data point you're providing in terms of them being at historical highs, that is a little bit concerning Yeah.

为了

To

嗯。

yeah.

今年的规模有望成为十年来最大的一年。

It's it's pacing to be the biggest year in in ten years.

至少报告追溯到了这个时间范围。

That's as far back as the report went, at least.

我的笔记就到这里了。

So that's as far as my my notes go.

这让我感到担忧。

That is that is concerning to me.

确实如此。

Definitely.

我们简单聊聊估值吧。

Let's talk briefly about valuations.

你知道，作为风险投资领域的观察者，我想听听你的看法：大家都说估值太高了，这确实反映了对交易的竞争，人人都在追逐同样的项目。

You know, one of the things that I I I wanna get a sense from you on, as a student of the venture capital sector is everybody talks about valuations being high, and that is certainly a reflection of the competition for deals and everybody trying to go after the same deals.

是的。

Yes.

从机制上讲，需要发生什么变化才能让估值下降？

Mechanically, what has to change for those valuations to come down?

你觉得所谓的音乐什么时候会停止？

You know, when do you think the music, quote unquote, stops?

我的意思是，股市是一回事。

I mean, you know, stock market is one thing.

但你知道，是不是要等股市的表现反映到初创公司的估值上？

But, you know, is it waiting for the stock market to sort of translate into startup valuations?

你知道，要让这些交易的竞争减少，战术上必须发生什么变化？

You know, what tactically has to happen for competition for these deals to decrease?

这是个非常好的问题。

That's a really good question.

推动估值的因素有几个，尤其是在风险投资的早期阶段，甚至成长阶段。

There's a couple of factors that drive valuations, especially at the early stages of venture, even growth stage.

对吧？

Right?

首先是资本的供给量。

The first is just the pure supply of capital.

有多少资本在争夺这些交易？

How much capital is competing for deals?

这是个供需问题。

That's supply and demand question.

对。

Right.

对吧？

Right?

所以当市场上资本过剩时，估值通常会上升。

So when there's a plethora of capital in the market, typically valuations go up.

只是有更多的资金。

There's just more Right.

突然间，有限合伙人募资的稀缺性开始产生实际影响。

Suddenly all the scarcity in LP fundraising, it starts to actually have an impact.

是的。

Yeah.

所以随着时间推移，正如你所说，你会开始看到进入基金的资本减少的影响。

So over time, to your point, you'll start to see the impact of less capital going into funds.

进入公司的资本也会减少。

Less capital will go into companies.

这就是风险投资在估值方面的一个制衡机制。

And that is one of the checks and balances of venture in terms of investing in valuation.

我认为第二点是风险投资者如何看待回报？

I think the second thing is how do the venture investors perceive the returns?

所以过去几年在人工智能领域，你看到一种竞相投资于那些被认为最终会产生巨大成果的公司的现象。

So what you've seen in the past couple of years in AI is that there's been this race to invest in companies where the belief is that at the end of the day, the outcomes are going to be significantly large.

因此，历史上规模巨大的融资轮次，现在比以往更早地完成了。

And so you've historically large rounds getting done much earlier than historically was possible.

所以在共同创立我们的公司June之前，我在Insight Partners担任增长型投资者大约八年。

So before co founding our firm June, I was a growth investor for about eight years at Insight Partners.

历史上，增长型投资非常注重业务的基本面：业务模式的质量、增长，以及盈利路径、底层利润率和业务本身有多好。

And historically growth investing was very much based upon fundamentals of performance, quality of business model, yes, growth, but also path to profitability, underlying margins, how good is your business?

在这个时期，投资者在一定程度上偏离了这一点，尤其是在人工智能公司中，人们只关注极快的增长，而忽视了或根本不清楚企业的基本面。

Think investors have gotten away from that a bit in this kind of period of time, especially for AI companies where it's all about extremely rapid growth, but where the fundamentals of businesses are getting lost a bit or are just unknown.

现实是，在人工智能领域，新的商业模式正在被试验，但我们实际上还不清楚最终什么样的人工智能企业才算优秀。

The reality is in AI, there's new business models being experimented with, but we don't really know per se what a great AI business looks like at the end the day.

我们缺乏像传统SaaS公司或消费类公司那样的底层业务质量的衡量指标。

We don't have the metrics in the underlying business quality that we might have for a traditional SaaS company or even a consumer company.

所以我认为，最终结果很可能是，当市场明朗后，投资者会转向高质量的商业模式。

So I think the result will likely be that as things shake out, there'll be a flight to quality in terms of high quality business models.

我同意。

I do

快速想想，这种情况会多快发生？

think quickly how quickly does that happen?

我们说的是2026年上半年吗？

Are are we talking first half of twenty six?

我觉得未来两年内会这样。

I I would say over the next two

两年。

years.

好的。

Okay.

嗯

Well

我认为你已经开始看到一些情况出现了，正如你所提到的，资本供给的变化已经开始显现。

I think you're start I so I think you're start so I you're starting to see a few of these things starting to take place to your point about supply of capital that's starting to manifest itself.

我认为公开市场有时是一个信号。

I think the public markets sometimes are a signal.

如果估值出现显著下滑，现在公开的AI公司并不多，所以很难比较。

So if there were meaningful declines in valuations Now there aren't a lot of public AI companies, so it's harder to comp.

但如果你看到某些公司股价下跌，比如英伟达和其他公司，可能会对私募市场产生一些影响。

But if you saw certain companies drop, so Nvidia as an example, others, there might be a little bit of an effect on the private markets.

但归根结底，通常还是取决于商业模式质量的逐步明朗化。

But at the end of the day, typically it's based upon things starting to shake out in terms of business model quality.

而且我认为，如果你再看看这些AI公司，它们筹集资金的规模和速度，必然会出现一次洗牌，因为目前给予它们的估值倍数太过极端。

And I do think if you look at, again, some of these AI companies, how much they're raising and how fast they're raising, there has to be a shakeout because the multiples that are being applied to them are so extreme.

是的。

Yeah.

那些公司筹集了这么多资金

Mean companies raise that

有时候，是的。

sometimes Yeah.

他们甚至还没有收入，更不用说产品了，有些公司就是这样。

They don't even have revenue, let alone products, some of them.

所以，你看，我们喜欢谈收入倍数，但在某些情况下，根本没有收入。

So So see, like I mean, we like to talk about revenue multiple, but in some cases, there's no revenue.

所以，是的。

So there's Yeah.

不。

No

那么，你已经雇了多少人？

Well, how much is how many people you've managed to hire?

我觉得这些倍数也相当疯狂。

I think those multiples are also pretty crazy.

Nnamdi，感谢你前来做客。

Nnamdi, I wanna thank you for coming on.

这是一次很棒的对话，我们一定会持续关注。

It's a great conversation, we certainly will be watching.

感谢你抽出时间，我们很快再聊。

We appreciate the time, and we'll talk to you again very soon.

谢谢你们邀请我。

Thanks for having me.

我真的很享受这次交流。

Really enjoyed it.

谢谢。

Thank you.

好的。

Okay.

让我们转向旅行和晚期科技的未来。

Let's turn to the future of travel and late stage tech.