人工智能隐藏着债务问题

本节目由BCX赞助，这是私营科技公司的公开股票代码。

Support for the show comes from BCX, the public ticker for private tech.

美国股市开启了历史上规模最大的财富创造浪潮，从底特律的工人到奥马哈的农民，人人都受益。

The US stock market started history's greatest wave of wealth creation from factory workers in Detroit to farmers in Omaha.

任何人都能拥有这些伟大美国公司的股份。

Anyone can own a piece of the great American companies.

但如今，最具创新力的公司长期保持私有，导致普通美国人一直无缘参与——直到现在。

But today, our most innovative companies are staying private longer, which means everyday Americans are missing out until now.

推出VCX，私营科技公司的公开股票代码，现已在所有股票交易平台上线。

Introducing VCX, a public ticker for private tech, now available wherever you buy stocks.

了解更多，请访问getvcx.com。

Visit getvcx.com for more info.

就是getvcx.com。

That's getvcx.com.

投资前请仔细阅读相关材料，包括投资目标、风险、费用和支出。

Carefully consider the investment material before investing, including objectives, risks, charges, and expenses.

更多相关信息请访问 getvcx.com 查看基金招募说明书。

This and other information can be found in the fund's prospectus at getvcx.com.

本内容为付费赞助。

This is a paid sponsorship.

乌克兰军方这次人事变动背后有什么原因？

What's behind Ukraine's military shakeup?

乌克兰军队在如何运用自主系统进行作战和战术整合方面，已经超越了美国军队以及全球其他任何一支武装力量，我们真该向他们学习。

The Ukrainians have now surpassed The United States Military and basically every other military on the planet in terms of doctrine for how to fight and integrate these autonomous systems on the battlefield, and we should be learning from them.

我是约翰·费纳。

I'm John Feiner.

我是杰克·沙利文，我们是《长远博弈》这档每周国家安全播客的主持人。

And I'm Jake Sullivan, and we're the hosts of the long game, a weekly national security podcast.

这周，我们来深入分析乌克兰国防部长被解职的事，他可是乌军无人机战略的总设计师。

This week, we're breaking down the dismissal of Ukraine's defense minister, the architect of their drone strategy.

这期节目已经上线了。

The episode's out now.

在你常用的播客平台搜索并关注《The Long Game》。

Search for and follow the long game wherever you get your your podcasts.

一万七千。

17,000.

我来给个数字。

I'll give a size.

我的货币市场波动很大。

My money market's mad.

如果钱是邪恶的，那栋楼就是地狱。

If money is evil, then that building is hell.

行情上涨了。

Show goes up.

专业人士根本没在关注行情下跌。

The pros are not watching show sell.

欢迎来到Prof G市场。

Welcome to Prof G Markets.

我是埃德·埃尔森。

I'm Ed Elson.

今天是7月23号。

It is July 23.

咱们来看看昨天的市场关键数据。

Let's check-in on yesterday's market vitals.

标普500和纳斯达克都下跌了。

The S and P five hundred and the Nasdaq declined.

道指基本持平。

The Dow was flat.

由于美国和伊朗的袭击升级，原油价格涨至每桶95美元以上，这是近六周来的首次。

Print crude climbed above $95 per barrel for the first time in nearly six weeks as both The US and Iran escalated their attacks.

十年期国债收益率上升了。

The yield on ten year treasuries increased.

最后，谷歌和特斯拉的股价在公布财报后都下跌了。

And finally, Google and Tesla shares fell after both companies reported earnings.

这些报告我们待会儿再细说。

We will get into those reports later.

明白了。

Okay.

还有哪些大事发生？

What else is happening?

大型人工智能公司正越来越多地依赖债务融资，而且很多债务其实藏了起来。

The big AI spenders are increasingly turning to debt, and a lot of it turns out to be hidden.

《日经亚洲》的一项调查发现， Alphabet、微软、亚马逊、Meta 和甲骨文这五家科技巨头，背负了大约1.65万亿美元的表外债务。

A Nikkei Asia investigation found that five tech giants, Alphabet, Microsoft, Amazon, Meta, and Oracle, are carrying roughly $1,650,000,000,000 in debt that doesn't show up on their balance sheets.

这个数字超过了它们实际披露的1.35万亿美元，也就是说，我们看不到的债务，现在已经比看得见的还多了。

That's more than the $1,350,000,000,000 they actually report, meaning the debt we can't see is now bigger than the debt we can.

Meta的表外债务大约是其披露债务的三倍，而甲骨文的债务在四年里暴涨了三十倍。

Meta's off balance sheet debt is roughly three times its reported debt, and Oracle's has ballooned about 30 fold in four years.

这些操作虽然合法，但却引发了一个关键问题。

It's all legal, but it raises one big question.

如果人工智能的需求没有投资者预期的那么强，会怎么样？

What happens if AI demand isn't as strong as investors are betting?

今天我们邀请到爱德·齐特隆做客，他是《Where's Your Ed At》通讯的作者，也是《Better Offline》播客的主持人。

Joining us to discuss this question, we're speaking with Ed Zitron, author of the Where's Your Ed At newsletter and host of the Better Offline podcast.

爱德，感谢你参与我们的节目。

Ed, thank you for joining us.

我想先给你一些背景信息。

I just want to give you some context.

昨天在这档节目里，我谈到甲骨文，说他们越来越依赖借款来建设数据中心，债务负担已经快到失控的地步。

Yesterday, I was talking about Oracle on this show and I was talking about how they're increasingly relying on debt to finance their data centers and how it's becoming kind of borderline unmanageable.

所以他们的信用评级被下调，借贷成本也一路飙升。

That's why their credit rating is getting downgraded and their borrowing costs are exploding.

他们正被拖入一个恶性循环。

They're kind of being sent into this downward spiral.

到目前为止，我一直以为人工智能领域的债务主要集中在甲骨文这类公司，还没波及到那些更大的科技巨头。正因如此，当我看到《日经亚洲》的报道，发现有高达1.7万亿美元的隐藏债务一直被忽视时，我才感到如此震惊。

Up until this point, though, it has been my understanding that the debt in the AI ecosystem has been relatively contained to a handful of companies like Oracle, but that it hadn't infected the bigger names, which is why I was so alarmed to see this reporting from Nikkei, which says there's $1,700,000,000,000 of hidden debt that we haven't been seeing, that we haven't been looking at.

你刚写了一篇关于这个的长文。

You've just written a huge piece on this.

你深入调查了这些数据。

You investigated these numbers.

你对整个情况做了全面分析。

You did a sweeping analysis of all of this.

我们对人工智能领域的债务到底了解多少？为什么有这么多债务藏得这么深？

What do we know about debt in AI, and how is it that so much of it appears to be hidden?

这里涉及好几个因素。

So there are several factors here.

对于这些超大规模云服务商来说，他们之所以能做到这一点，是因为找到了一种特殊的会计处理方式。

With the hyperscalers, the reason they're able to do that is because they found this interesting accounting treatment.

以Meta为例，安永指出的问题在于，当你建造数据中心时，并不像公司自己说的那样：我要买下所有设备，然后为此举债，就这么简单。

The Ernst and Young claim is a red flag in Meta's case where when you build a data center, it's not like you as the company say, I'm gonna buy all the stuff.

事情没那么简单。

I'm gonna get the debt for this, and here we go.

他们设立了一个特殊目的实体或可变利益实体，也就是所谓的SPV，但持股比例并不高。

They make a special purpose vehicle or a variable interest entity, which is basically an SPV where you don't own as much.

比如Hyperion这个项目，像PIMCO和Blue Owl这样的投资者持有80%，而Meta只占20%。

So for example, with Hyperion, I think investors like PIMCO and Blue Owl own 80% of it, and Meta only owns 20% of it.

尽管Meta是唯一客户，而且未来还会往里面塞满GPU，但情况就是这样。

Despite bet Meta being the only client, the person that's gonna fill it full of GPUs as well, I think.

他们把这部分资产从账面上移走了，尽管Meta已经公开宣布这是他们的数据中心，可因为没完全拥有，就不用记在账上。

I they're think moving some across their assets despite them being the obvious despite Meta announcing it and saying this is our data center because they don't have full ownership of it.

所以这部分不会出现在他们的资产负债表里。

It's not counted as something they put on their balance sheet.

等开始产生收益后，这些资产会被归入经营性租赁的范畴。

They will have it, I think, within operating leases when it starts paying.

但问题是，整个AI数据中心行业都是这么操作的。

But the thing is, this is the entire AI data center industry.

每一个项目都是这种SPV，投资者的钱就是这么被吸引进来的。

Every one of them are these SPVs that people pull money into.

这个特殊目的实体持有债务。

The SPV owns the debt.

不过，由于是非追索权贷款，这个实体通常承担了大部分风险，也就是说，理论上你得先追索这个实体的资产，才能去找母公司。

The SPV often owns a lot of the risk, though, through nonrecourse loans, which means that, technically, you have to go after the assets of the SPV before you go after the company.

但归根结底，这些风险都是由它们承担的。

But, nevertheless, these things own the risk.

它们拥有那些显卡。

They own the GPUs.

它们支付给投资者的钱，完全来自数据中心的收入，这就引出了一个关键问题：如果数据中心的收入没到位，怎么办？

They pay investors not out of anything other than the revenues from the data centers, which leads to the important point of what happens if the revenue for the data center isn't there?

投资者怎么拿回本金？

How do investors get made whole?

答案是：拿不回来。

And the answer is they do not.

目前这还不被视为问题，只是因为这种情况还没真正发生过。

And the only reason this is not considered a problem yet is it hasn't actually happened.

实际上，数据中心还没建到那种规模，还没到收入无法流入的地步。

Like, the data centers have not been built at the scale that would have to happen for the revenue not to flow in.

那到底什么是SPV？

What actually is an SPV?

这一点很重要，因为你提到这些数据中心的SPV，相当于AI泡沫时期的CDO，而CDO正是引发全球金融危机的金融工具。

And this is important because you say that these data center SPVs are the AI bubbles equivalent of CDOs, which, of course, were the financial instruments that basically sparked the great financial crisis.

你说，今天它们就是我们的同类东西。

You say that that is our equivalent today.

所以，SPV到底是什么？为什么它们这么受欢迎？

So what actually is an SPV, and why are they so popular?

要明确一点，CDO其实不太一样。

So to be clear, a CDO is slightly different.

它持有的是抵押债券之类的东西，但它们和SPV一样，都是那种会给市场带来真正伤害的金融工具。

It's holding mortgage bonds and stuff, but they are the same kind of financial instrument that will cause genuine harm to the markets in the world.

所以，SPV本质上就是一个公司。

So an SPV is basically a company.

它就是一个公司。

It's a company.

名义上是有董事会成员的，但其实……

It has people people that technically are on the board of directors.

你可以这么操作，但本质上它就是个用来管钱和相关资产的载体。

You can do that, but, really, it's just a holding entity for money and stuff.

所以当他们建数据中心时，这个SPV会去借债。

And so when they build a data center, that SPV raise raises the debt.

债务由这个SPV承担。

That SPV holds the debt.

GPU也是由这个SPV来购买。

That SPV buys the GPUs.

数据中心也是它来建造的。

It builds the data center.

等资金进来后，所有付款——给承包商的、日常运营的开销——都由这个SPV来处理，资金流向也有明确的分配顺序。

And when the money flows into it, the SPV is the one that pays out to contractors, that pays the OpEx costs, and there's a cash waterfall even where it goes.

支付运营成本，支付投资者，理论上如果有多余的钱，再分给所有者。

Paying OpEx, paying investors, and then paying paying the people that own it in theory if there's more money than there should be.

这种情况根本不会发生。

That is not gonna happen.

但问题在于，这些项目却被当成稳定的基础设施来推销，比如商业地产或住宅，还说这只是实体土地。

But, nevertheless, these things are so dangerous because they are being sold as stable infrastructure, like commercial real estate or residentials and just saying, oh, it's physical land.

这还不错。

It's good.

这很好。

It's great.

但问题在于，关键区别是，数据中心和普通房贷、商业地产完全不一样。

But the problem is and the critical difference is that data centers are nothing like a regular mortgage, and they're nothing like commercial real estate.

无论是建设过程还是持续运营，这些都属于复杂的金融操作。

These are complex financial operations, both in the construction and the ongoing costs.

更重要的是，你得有收入，才能从中获得回报。

And then there's the important thing of you need revenue to actually get paid out of them.

而且这些全是项目融资，说白了就是专门为某个项目筹钱，除非像CoreWeave或Oracle这样的客户承担了付款责任，否则根本没多余的钱来支付给其他人。

And because they're all project financing, which just means they exist to finance a project, there is no other money for them to pay people out unless, like, the company on the other end, a CoreWeave and Oracle, what have you, isn't on the hook for those payments.

说白了，责任在客户身上。

Basically, client is.

如果客户是像OpenAI这样的公司，结果付不起钱，那投资这些SPV的投资者就彻底倒霉了。

And if the client is, say, I don't know, OpenAI and can't afford to pay it, then investors who have invested in these SPVs, well, they're kind of shit out of luck.

我这里说的投资者，其实指的是现在所有参与私人信贷的人。

And the problem is when I say investors, I mean anyone who's involved in private credit right now.

私人信贷的资金非常充裕，像养老基金、退休金、保险基金、教师养老金、警察养老金这些，都会投进去，这点我们待会儿再细聊。

Private credit is well funded, as I'm sure we'll get into, by pensions, retirement funds, insurance funds, teachers' pensions, cops' pensions, all this different stuff.

不管我们愿不愿意，其实我们都已经卷入了私人信贷。

We are all involved in private credit whether we want to or not.

我想念一段阿曼达·亚科内从彭博社发的原话，你文章里也引用了。

I just wanna read you a quote from Amanda Yakone of of Bloomberg, and you quoted this in your piece.

她把这事儿跟安然公司相提并论。

She made this comparison to Enron.

她说，恩罗公司利用美国的会计规则，把高达数亿美元的债务藏在资产负债表外的实体里，瞒着投资者和贷款方，最终导致了美国历史上最严重的公司倒闭之一。

She said, quote, Enron exploited US accounting rules to hide from investors and lenders hundreds of millions in debt it had bundled into off balance sheet entities, obligations that contributed to one of the biggest corporate collapses in US history.

而你刚才描述的情况是，像Meta、微软、亚马逊、谷歌、甲骨文这些大科技公司，正在设立一堆空壳实体、也就是SPV，表面上是为了融资建数据中心，实际上却在悄悄隐藏数千亿美元的债务。

And what you are describing right now is that there is a trend among the big tech companies, Meta, Microsoft, Amazon, Google, Oracle, they are creating these shell entities, these SPVs, seemingly to hide literally hundreds of billions of dollars of debt that they are issuing in order to finance their data centers.

换句话说，他们在财报里大肆炫耀自己建了哪些设施，可那些设施背后借的巨额贷款，却全都被藏起来了，压根儿不提。

In other words, they're they're showing us in their reporting and in their earnings, here's here's everything we we've built, but then they're hiding all of the borrowing that they took on to build all those things that they're kind of bragging about in their in their, reporting.

这确实和当年恩罗的做法非常相似。

That does seem to be a very, very similar link.

你能再深入说说吗？这种做法是怎么一步步发展成这样的？将来又可能怎么崩盘？

I guess, could you expand on that, on how this has come to be and potentially how this might unravel?

我得澄清一点，除了甲骨文，我觉得其他公司还不至于垮掉。

So I wanna be clear that other than Oracle, I don't think

这些公司不会倒闭。

these companies die.

但说实话，尼克·APs的说法有点模糊，搞不清楚他指的是3000亿美元还是1.6万亿美元的隐藏债务。

But I will say, it's kind of hard that Nick APs, it wasn't clear if they were saying there was 300,000,000,000 or $1,600,000,000,000 of hidden debt.

那篇文章写得挺怪的，但我们就假设吧，有大约一万亿的表外债务。

It was a very weirdly written piece, but let's say it's a trillion or so of off balance sheet debt.

这意味着有整整一万亿的贷款，我们根本没法准确统计，只能靠媒体报道来推测。

That means that there's a trillion dollars in loans that we just don't we we actually cannot really quantify outside of media reporting.

但我得说清楚。

But I must be clear.

这不光是超大规模云服务商在这么做。

It's not just hyperscalers doing it.

CoreWeave也在做。

It's CoreWeave.

Iron也在做。

It's Iron.

Nebius也在做。

It's Nebius.

有些情况下，英伟达也在这么做。

It's NVIDIA in some cases.

任何建造数据中心的人，都在利用这些空壳公司来转移所有权，同时也把风险转嫁给投资者。

It's whoever is building a data center is using these shell corporations as a mean of of offiscating ownership and also offiscating risk as in basically handing off the risk to investors.

尽管有些贷款是有追索权的，理论上你往上追责，最终还能起诉公司本身。

And even though there are some that are recourse loans, even if there's a up the chain, you could eventually sue the company.

但到那时候，所有GPU早就被卖光了，你怎么都逃不掉。

At that point, you would have sold off all the GPUs, you'd be screwed either way.

但更重要的是，这确实是一场企业丑闻。

But the larger point is, yeah, this is a corporate scandal.

这是一场巨大的丑闻。

This is a huge scandal.

这本该是全球各大媒体的头版新闻。

This should be front page news everywhere.

这本该引发股东们的集体抗议。

This should be a shareholder riot.

但正如你所说，我们生活在一个崇拜富人的环境里，所以大家只是把它搁在一边，视而不见。

But because we live, as you put it, well, my show, in this cult of worshiping the wealthy, we just kind of put it to the side.

不过，这对这些公司来说是个不小的风险。

This is a significant risk to these companies, though.

这正好赶在谷歌刚公布财报的时候，说的就是谷歌。

This is a real like, Google we're talking just as Google's earnings came out.

他们的搜索收入刚刚没达到预期。

They narrowly missed on search revenue.

其他业务也在放缓。

Their other businesses are slowing down.

整个公司都在出现这种情况。

It's it's happening across the board.

就在这些巨额债务逐渐逼近的时候，他们的业务却要开始减速了。

So their businesses are gonna slow down just as these massive debts creep up.

就算他们有办法脱身，

And let's say they have ways of getting out of these things.

投资者照样会亏得一塌糊涂。

The investors will still be screwed.

参与其中的投资者还是会血本无归。

The investors involved will still be screwed.

如果他们取消这些项目，我猜他们可能会部分偿还投资者，但所有相关方都免不了要亏钱。

And if they cancel these projects, I imagine they'll pay off investors to a point, but anyone involved is going to take a loss.

然后这种影响会扩散开来。

And then this spreads out.

这些才是所谓的好消息。

Like, these are the this is the good news.

大型云服务商手里的这些，才是优质的特殊目的载体。

The fact that the hyperscalers have these, these are the good SPVs.

我觉得这些才有可能挺过来。

These are the ones that I think might survive.

问题是，每个数据中心都是通过特殊目的载体运作的，每个投资数据中心的人，投的都不是数据中心本身，而是它未来能带来的收益。

The problem is every single data center is through an SPV, and every single investor who's invested in a data center has invested not in the data center itself, but in the potential revenue it brings in.

而问题在于，这种预期完全依赖于人工智能的算力需求要比现在高出十五倍以上。

And the problem there is that is dependent on there being 15 or more times the AI compute demand than currently exist.

一旦你开始认真算这笔账，真的会吓一跳，可大家对这事都太轻描淡写了。

It's act it's genuinely terrifying when you start quantifying it, and people are way too flippant about this stuff.

他们说，没那么严重。

They say, oh, it's not as bad.

没比2008年金融危机更糟。

Oh, it's not as bad as the great financial crisis.

可这本身就是一场巨大的金融危机。

It's still a great financial crisis that this happens.

这事儿太可怕了，而且无处不在。

It's still horrifying, and it's everywhere.

你说得对。

You're right.

就像次级抵押贷款一样，AI数据中心的债务审批极其松散，几乎没有任何抵押，全靠空洞的预测和炒作就给项目放贷，还款可能性极低。

Quote, much like a subprime mortgage, AI data center debt is being poorly underwritten, virtually uncollateralized, and issued to projects that have extremely low likelihoods of repayment all based on flimsy information and hype driven mania, end quote.

听起来挺尖锐，但我真觉得这话说得一点都没错，基本就是事实。

Sounds like a provocative statement, but I actually think it it's pretty much factual and and correct.

我觉得在拿这次情况和次贷危机做对比时，那个说法里漏掉了一点：很多那些承保糟糕、几乎没抵押的债务，都被藏得严严实实，连评级机构、投资者和交易员都没意识到发生了什么，因为它们在市场里被掩盖得太好了。

I think one thing that that isn't in that statement when we make the comparison to the great financial crisis is the fact that a lot of that poorly underwritten debt, that virtually uncollateralized debt was obfuscated and hidden away to the point where ratings agencies and investors and traders didn't even know what was happening because it had been so well hidden in the markets.

它们被搞得特别复杂。

It had been so well kind of complexified.

我想问问你，你觉得现在的情况和那时候一样吗？我是说，投资者和评级机构，比如标普刚刚下调了甲骨文债务的评级，他们应该都清楚。

I guess my question to you, do you think that this is the same situation where investors, ratings agencies I mean, know that S and P just downrated Oracle's debt.

所以很明显，他们对这里可能出了什么问题还是有一定了解的。

So clearly, they have some semblance of an understanding of what might be going wrong here.

不过，你是不是觉得市场并没有把这一点反映在价格里，也就是说，投资者其实根本不清楚这些公司正在发行的债务情况？

But is it your view that this is not being priced in that investors literally don't know what is happening in terms of the debt that is being issued among many of these companies?

我来给你讲讲私募信贷这回事。

Let me tell you the tale of private credit.

过去几年，私募信贷扩张得很快，开始和保险公司以及退休基金合并或捆绑在一起。

So private credit in the last few years has gone on a tear merging with or requiring insurance companies and retirement funds.

阿波罗收购了阿提娜。

Apollo bought Athene.

他们是直接合并的。

Well, they merged with them.

Blue Owl 收购了 Cover。

Blue Al bought Cover.

其他人我记不清了，但黑石的基础设施基金确实是靠养老基金注资的。

They're I forget who the others are, but, nevertheless, Blackstone's infrastructure fund is funded by retirement funds.

问题在于，很多特殊目的载体都是靠私人信贷来融资的。

The problem is is a lot of the SPVs are funded by private credit.

这个私人部分才是最危险的。

That private part is the lethal part.

标普不评级私人债务。

S and P doesn't rate private debt.

除了私人信贷基金，没人会评。

No one does other than the private credit fund.

几个月前有报道称，Blue Owl 在只看了十分钟之后，就决定向 Stargate Abilene 投资高达一百亿美元。

So there was a story in the information a few months ago that said that Blue Owl decided to invest, I think, up to $10,000,000,000 in Stargate Abilene after ten minutes.

你认为十分钟内能完成多少尽职调查？

How much due diligence do you think we can get done in ten minutes?

估计也做不了多少。

Probably not very much.

这对特殊目的载体来说太常见了。

This is standard for the SPVs.

所以当人们说评级机构在这里失职时，

So while people will say, oh, well, the ratings agencies failed here.

其实不是这么回事。

Actually, no.

是证交会在这里失职了。

The SEC failed here.

真正支撑这一切的，是那些精英们对私人信贷的疯狂追捧——这本质上是一个数万亿美元的影子银行体系。

The elite like, the insanity of private credit, which is basically a multitrillion dollar shadow banking system, is what's propping this up.

这些特殊目的载体所募集的债务，绝大部分都来自私人信贷。

For the most part, the debt raised for these SPVs is coming from private credit.

你看看Core Wave的情况。

You look at Core Wave.

资金来源全是私人信贷。

It's coming from private credit.

如果是银行提供的，比如日本的三井住友和三菱日联，它们其实就是直接把贷款往里灌。

When it comes from the banks, the banks themselves, SMBC and MUFJ out of Japan, they're there's like, they're just funneling the debt straight in.

有很大一部分资金来自通过私人信贷注资的机构投资者，评级机构根本插不上手。

So much of this is coming through institutional investors that are funding money through private credit that the ratings agencies aren't even involved.

评级机构之所以参与甲骨文，是因为甲骨文主要靠发债融资，但它的很多私人项目都是SPV结构，债务是由SPV来筹措的。

And the ratings agencies don't the only reason they did it with Oracle is Oracle has predominantly raised its money via bonds except for the fact that a lot of their private projects are SPVs, and the SPV raises the debt.

理论上讲，比如他们在密歇根建的数据中心，属于有追索权贷款，你可以直接找甲骨文追责。

And it's this situation where, in theory, like, think the Michigan data center they're building is a recourse loan, so you can go directly after Oracle.

但问题就在这儿。

But that's the thing.

即便在这种情况下，也存在各种法律障碍，一旦数据中心项目大规模出问题，要追责就变得极其困难，甚至根本搞不清到底该由谁负责。

Even in that situation, there are these legal barriers that if every data center is exploding, will make it hard to litigate at scale or at least hard to pull down who's actually responsible.

如果你碰巧和Athene、Couvera或者其他任何保险公司有关系，最好查清楚他们的钱到底投到哪儿去了。

And we don't like, if you are anything touching an Athene or a Couvera or any other insurance company, it is worth checking where their money's going.

我觉得Lions实际上投资了CyrusOne的债券。

I think the lions actually invested in the CyrusOne bond.

这些债券虽然被评了级，但明明是垃圾级，可现在保险公司基本由私募信贷基金操控，他们根本不在乎。

And the thing is, these bonds get rated somehow they get rated as junk, but because insurance companies are now effectively run by private credit funds, they don't give a crap.

是的。

It's like, yeah.

当然。

Sure.

因为这个巨大的谎言。

Because of this giant lie.

这个谎言就是，数据中心就跟发电厂和工厂一样，是所谓的AI工厂。

And this giant lie is that data centers are the equivalent of investing in power plants and factories, that they're AI factories.

但实际上，你投资的只是一个规模庞大的建筑群，或者说是整个园区，里面堆满了快速贬值的GPU，而这个行业根本还没证明自己真有那么大的需求。

When in fact, what you're investing in is a very large, like, town sized building or siri a campus that is full of depreciating GPUs for an industry that has not proven it has the demand.

我的估算显示，真正的AI算力需求每年大约是10万亿美元。

And, I mean, my estimate for actual AI compute demand is about $100,120,000,000,000 a year.

我算过一笔账：如果他们真把计划中的130吉瓦IT负载全部建成，也就是那些宣称正在规划中的实际运行GPU，光是支付这些设备的费用，每年就需要16800亿美元的算力收入，而且必须持续支付——一旦中断，SPV的贷款契约就会违约。

The I did the math, and it's something like if they build all a 130 gigawatts of IT loads, so the actual operational GPUs that they say is in planning based on Sightline Climate, we will need $1,680,000,000,000 of annual compute revenue just to pay for them all, and they'll have to pay consistently because if you stop paying consistently, the loan covenants with the SPVs break.

最荒谬的是，全球软件产业的总收入还不到8万亿美元，我们却指望凭空再变出一个软件产业来。

And the crazy part about this is the global software industry is less than $800,000,000,000, So we're just gonna magic up another software industry.

我要说明的是，我这话听起来可能很激进。

And to be clear, what I am saying sounds radical.

我觉得你之前其实也提到过类似的观点。

I think you kind of said this already.

听起来像是个让人害怕的说法。

It sounds like a, oh, scary thing.

这确实有点危言耸听。

Oh, it's alarmist.

但这只是最基础的数学计算而已。

This is just very basic maths.

Sightline Climate提到，目前有190吉瓦的容量正在规划或建设中，PUE为1.35，也就是能效水平，其中130吉瓦，每兆瓦成本一万二千美元。

This is Sightline climate said a 190 gigawatts of of capacity was being was in planning or under construction, PUE of 1.35, which is just the energy efficiency, 130 gigawatts, $12,000,000 a megawatt.

这其实是非常基础的计算，任何人都能算出来。

It really it is a simple math that anyone could do.

关键是，哪怕实际需求只有这一半，我们依然远远不够。

And the thing is, even if I even if it's just half of that, we still don't have enough demand.

所以我们现在陷入一种奇怪的状况，这种风险已经分散到各个角落，因为都是私下交易，我们根本不清楚到底蔓延了多广，尤其是保险公司的部分。

And so we're in this weird situation where this risk is now spread everywhere to the point that because it's private, we actually do not know how far, and it's within the insurance companies.

这跟AIG倒闭的情况不一样。

It's different to how AIG collapsed.

规模要小得多。

It's just smaller.

但另一个问题是，AI数据中心实在太贵了。

It's smaller little bits, but the other problem is is that AI data centers are so expensive.

最低也要五亿美元，很可能高达几十亿。

They cost 500,000,000 minimum, probably several billion.

所以，不是成千上万笔次级抵押贷款慢慢崩盘，而是一下子就崩掉五亿、六十亿、二十亿、十亿的大项目。

So instead of having millions of or hundreds of thousands of subprime mortgages that collapse in a kind of slow boil over time, it's gonna be a 500,000,000, a 6,000,000,000, a 2,000,000,000, a 1,000,000,000 collapses.

每次这种事发生，私募信贷基金就得大幅减记，还有一堆投资者血本无归，除了把显卡全扔进市场，根本没别的办法——可那市场很快就会被显卡淹没。

And each time one of these happens, that's a massive markdown with a private credit fund, but also a bunch of investors who have lost money and have no recourse beyond, I don't know, selling all the GPUs in a market that will become saturated with them.

真正让我们陷入危险境地的，正是这种追索权的问题。

It's the recourse point that is getting us into such dangerous territory here.

投资股票去赌未来，这种事大家心里都有数，风险自己承担，这没什么。

I mean, it's one thing to make a to sell equity on a speculative bet about the future, in which case, you know, the people know what they're buying into.

我们清楚自己在冒什么险。

We know what the risks are.

但要是用债务去融资一个极度投机的生意，还把债务藏起来，那就完全是另一回事了。

It's another thing to finance an extremely speculative, business through debt and then to hide that debt.

因为如果这事儿黄了，问题可不只是股价下跌那么简单。

Because, I mean, if this doesn't work out, we're not just talking about equity going down here.

这会直接引发破产。

We're talking about bankruptcies.

我们谈的是违约问题。

We're talking about defaults.

还有一个问题，就是保险基金和养老基金之所以投资私募信贷，是因为它们需要高收益。

There's also one one other problem, which is the reason that insurance fund insurance and, retirement funds invest in private credit is because they need yield.

它们需要持续的收益回报。

They need ongoing yields.

如果这些数据中心无法支付回报，养老基金和保险公司的保费就拿不到钱。

So if these data centers do not pay out, we have retirement funds and insurance premiums that cannot get paid.

我肯定它们有一些缓冲资金。

And I'm sure they have some buffer.

我肯定它们还有其他资产，但就算达不到AIG商业票据那样的崩盘程度，也足以对整个系统造成严重冲击。

I'm sure they have other assets, but it I mean, it doesn't have to be an AIG commercial paper level collapse for this to be systemically damaging.

到时候，除了国债，恐怕再也找不到其他高收益投资了，而国债现在利率正在上升。

And after this, there is no more yield to be found, I guess, other than treasury bonds, which are going up now.

挺好。

Yay.

但当它们下跌时，谁也说不准。

But when they go down, don't know.

真正糟糕的是，大多数人根本不知道这事。

But so it really is and what sucks is most people have no idea about this.

大多数人根本没想到，加州养老基金居然投到了这些数据中心里。

Most people don't realize that, like, the California pension fund is invested in blue out.

这事儿无处不在，而且不只是在美国。

Like, the these are it's everywhere, and it's not just in America.

全球都是如此。

It's across the board.

比如荷兰的养老基金、魁北克的CDPQ养老基金，都投了数据中心。

You've got, I think, the Dutch pension funds in data centers, CDPQ, which is the Quebec pension fund.

他们投资的还是核心数据中心。

They invested, I think, in a core data center.

这事儿到处都是，我气的是，这么普遍的问题却几乎没人谈，大家还总是一听就轻描淡写地说：‘没那么严重’。

Like, this is everywhere, and I'm I'm furious at the fact that it's so ill it's so rarely discussed, but also people are so quick to being like, oh, it's not as bad.

没那么糟。

Oh, it's not as bad.

没那么糟也还是糟。

Not as bad is still bad.

好的。

Alright.

我们得改天再继续聊这个。

Well, we're gonna have to continue this conversation, another time.

今天就先到这里，不过这内容真有意思。

We're gonna have to let you go, but it is fascinating stuff.

我也建议我们的听众去读读你的文章。

And I do encourage our listeners to go read your article.

写得非常严谨，直击我们刚才讨论的很多问题。

It's extremely rigorous, and gets at a lot of the issues that we're talking about.

埃德·齐特隆是《Where's Your Ed At》通讯的作者，也是《Better Offline》播客的主持人。

Ed Zitron is the author of the Where's Your Ed At newsletter and host of the Better Offline podcast.

埃德，非常感谢你抽出时间。

Ed, always appreciate your time.

谢谢。

Thank you.

谢谢邀请我参加。

Thanks for having me.

广告结束后，我们来详细分析特斯拉和谷歌的财报。

After the break, a breakdown of Tesla and Google's earnings.

如果你想了解更多市场动态，欢迎订阅我的每周通讯《Simply Put》，网址是 simplyput.profgmedia.com。

And for even more markets insights, you can subscribe to my weekly newsletter, Simply Put, at simplyput.profgmedia.com.

本节目由 Apple News+ 赞助。Apple News+ 汇集了你感兴趣的所有内容，一应俱全。

Support for the show comes from Apple News plus Apple News plus has everything you're into all in one place.

超过五百种刊物，覆盖你最关心的各类话题。

Over 500 publications covering the topics that matter most to you.

全球知名美食媒体的数千道食谱，来自全美五十个州的本地新闻，全球体育赛事报道，随时随地可听的音频故事，还有 Apple News+ 独家提供的每日谜题——所有内容都为你精心挑选。

Thousands of recipes from celebrated food publications around the world, local news from all 50 states, sports coverage from across the globe, audio stories you can take anywhere, and daily puzzles exclusive to Apple News plus All of it curated just for you.

新用户可免费试用一个月，访问 news.apple/learn 了解更多。

New subscribers try it free for one month at news.apple/learn.

条款适用。

Terms apply.

经营企业不该像应付一个软件小组作业那样累人。

Running a business shouldn't feel like surviving a software group project.

一个软件管财务，一个管库存，一个管销售，结果它们全都不互通。

One app for accounting, another for inventory, another for sales, and somehow none of them talk to each other.

这就是 Oto 的用武之地——一款一体化的商业管理软件，把销售、财务、库存、营销全都整合到一个强大的平台上。

That's where Oto comes in, an all in one business management software that brings every part of your business together, from sales and accounting to inventory and marketing, all in one powerful platform.

不用折腾集成，不用来回切标签，最重要的是，再也不用碰电子表格了。

No messy integrations, no bouncing between tabs, and best of all, no spreadsheets.

别再管软件了，用一个统一系统来管你的生意。

Stop managing software and start managing your business with one unified system.

立即免费试用，访问 odoo.com/vox。

Try for free today at odoo.com/vox.

就是 odooo.com/vox。

That's odooo.com/vox.

天哪。

Wow.

我真庆幸当年高中选了微积分，而不是理财课。

I'm so glad I took calculus in high school and not personal finance.

这话谁也没说过。

Said nobody ever.

刚毕业那会儿自己摸索理财，不仅让人头大，还可能真金白银地吃亏。

Figuring it all out for yourself when you're fresh out of school can not only be daunting, it can literally cost you.

我是Vivian Too，你的富裕闺蜜，这期《财富与闲聊》我们倒回十年前，那时我刚毕业，靠着一次次踩坑和摔跟头，才慢慢学会打理自己的财务。

I'm Vivian Too, your rich BFF, and on this episode of Net Worth and Chill, we're rewinding a decade to when I was first out of college and learning to care for my own finances the hard way with plenty of slip ups and face plants.

既然我早就踩过这些雷，那就把经验分享给你，让你少走弯路。

But since I already made those mistakes, I get to share them with you so you don't have to.

想避开我掉过的所有坑？欢迎在你常用的播客平台收听本集，或者去YouTube搜索 yourrich BFF 观看。

If you want to swerve all of the potholes I fell into, give the episode a listen wherever you get your podcasts or watch on youtube.com/yourrich BFF.

我们继续回到Prof G Markets。

We're back with Prof G Markets.

特斯拉刚刚公布了财报，华尔街感到失望。

Tesla just reported earnings, and Wall Street was disappointed.

在财报发布前，特斯拉公布了本季度令人印象深刻的交付数据，同比增长25%。

Ahead of earnings, Tesla had reported impressive delivery numbers for the quarter, up 25% year over year.

这让投资者觉得，特斯拉最艰难的时期已经过去了。

That gave investors the impression that Tesla's worst quarters were behind it.

去年这个时候，公司还在报告交付量下降13%，但今年第二季度的财报却讲了另一个故事。

This time last year, the company was reporting a 13% decrease in deliveries, but the company's second quarter earnings told a different story.

虽然营收同比增长了约26%，超过了分析师预期，但利润却下滑了5%，本季度的自由现金流更是亏损了约11亿美元。

While revenue beat analyst expectations up about 26% year over year, profits fell 5% over the same period, and the quarter's free cash flow came in about $1,100,000,000 in the red.

财报公布后，股价盘后下跌超过3%，目前相比高点已下跌约25%。

The stock fell more than 3% after hours and is now down roughly 25% from its peak.

今天我们请来Karim Bousta，他是DVx Ventures的联合创始人兼管理合伙人，也曾任特斯拉和Lyft的副总裁，来帮我们分析特斯拉这一季度的表现。

So here to break down Tesla's quarter, we are speaking with Karim Bousta, co founder and managing partner at DVx Ventures and former vice president at both Tesla and Lyft.

Karim，感谢你参与我们的节目。

Karim, thank you for joining us.

有意思的是，这次交付量还不错，尤其是跟去年相比。

The interesting story here, you got decent delivery numbers, especially compared to last year.

收入是增长了，但利润却下滑了。

Revenue was up, but profits down.

那利润端到底出了什么问题？

What went wrong here, on the bottom line?

我觉得这些公布的数据，对于过去一两年一直关注特斯拉动态的人来说，根本不会感到意外。

I think what we're seeing with these published results should not come as a surprise for anyone who's been understanding and following what's been going on with Tesla for the past, I'd say, eighteen to twenty four months.

实际上，过去几个季度，特斯拉的销量一直在下滑。

Essentially, what's been happening, we've seen declining sales for several quarters in a row.

但更重要的是，它的市场份额也在持续萎缩。

But more importantly, we've seen declining market shares.

即使在这个季度，特斯拉终于时隔几季再次实现了销量增长，你也得放在整个电动车市场快速扩张的背景下来看——毕竟，整个市场的增速远超特斯拉的销量增幅。

And even in this quarter, where for the first time in a few quarters, Tesla has been able to report increasing sales numbers, you have to look at this in the context of a growing EV market where the market overall has been growing faster than the sales number that Tesla has been reporting.

所以这是一个关键因素，印证了我们过去几年一直看到的情况：特斯拉在汽车业务上存在根本性问题，那就是产品线老化了。

So that's one factor, which confirms what we've been seeing for the past few years, which is Tesla has a fundamental issue on the automotive side, which is essentially the vehicle lineup, the product lineup, is becoming old.

他们一直没能更新产品。

It has not been able to renew it.

去年Model Y确实做过一次小改款，但本质上还是基于十年前初代Model 3和Model Y同款平台。

There's been a refresh of the Model Y last year, but it's still based on the same platform that launched almost ten years ago now with the Model three and then the Model Y.

这些年，无论是产品还是技术，都没有真正的创新，无法帮助特斯拉挽回市场份额的流失，更别提提升利润率了。

So there hasn't been real innovation, both from a product side and technology side, that would have allowed for the company to catch up on the market share losses that they've been seeing and at the same time continue improving the margins.

实际情况恰恰相反。

It's actually been the opposite.

利润率一直在下滑。

The margins have been eroding.

这季度的财报数据就清楚地说明了这一点。

We're seeing it in the results that have been reported this quarter.

归根结底，问题就在于过去几年里，特斯拉缺乏真正的创新、新产品，以及全新平台的推出。

So all this comes down to the lack of innovation and new product and fundamental platform launches that we haven't seen for the past few years.

而与此同时，竞争对手们已经迎头赶上，甚至在六七年前特斯拉领先的地方，它们现在推出了更出色的产品，拥有更好的利润率和更高效的生产体系。

And that's in the context of a competition that has been increasing with all competitors catching up to where Tesla was even six or seven years ago and creating great products with better margins, better production systems than what Tesla has been able to accomplish.

所以总的来说，我对这些财报结果感到意外。

So all this to say, I'm surprised by this by this results.

这一切其实都印证了我们过去几个季度看到的趋势——当放眼整个产品线，缺乏真正的创新和突破性新品，一切都变得清晰了。

It's all coming down to what we've been seeing in the past few quarters, and everything is coming into place when you look at the the lack of of innovation and new new breakthrough product on the product product lineup side.

特斯拉的看多者会问，那机器人出租车呢？

The Tesla bulls will say, what about the robotaxi?

他们还会问，那人形机器人Optimus呢？

They will say, what about Optimus, which is the humanoid robot?

你对这些说法怎么看？

What would you say to those people?

你真的相信这些产品能成事吗？

Are you convinced by those products?

如果特斯拉的核心业务——汽车本身——正在下滑，你真觉得这些新项目能救得了公司吗？

And do you think that that is gonna work for the company if the existing product being the car itself is deteriorating as a business?

这确实是马斯克和特斯拉支持者下的赌注——他们相信特斯拉会从一家电动车制造商，转型为一家以机器人、自动驾驶和人工智能为核心的公司。

That's indeed the the the bet that Elon and and Tesla bulls are making that there is gonna be an evolution, a transformation of the company that is gonna evolve from an EV car manufacturer to a robotaxi robotics company, AI driven company.

这确实是个了不起的战略和远见。

And that's a great strategy and a great vision.

但你得看看基本面：这些产品到底会是什么样子？

Now you have to look at fundamentals of, okay, what are these products going to look like?

就拿人形机器人Optimus来说，有很多问题值得追问。

If you take Optimus, the humanoid robot, for example, there are a number of questions that people should ask.

这个产品到底是什么？

What is really this product?

它要解决什么问题？

What problem is it addressing?

它真正的市场在哪里？

What market does it really have?

需求又会从哪里来？

Where is the demand going to come from?

即使你解决了这些问题，还有一个关键问题：你真能造出这个产品吗？

And even assuming that you solve these questions and these problems, then there's the other question that is, are you going to be able to make this product?

这是个全新的产品，全新的供应链，全新的设计。

It's a brand new product, brand new supply chain, brand new design.

特斯拉以前从没做过这种东西。

There's no history at Tesla.

也没有制造这类产品的经验。

There's no experience building such a product.

更没有过在如此复杂程度上，大规模提升供应链和生产能力的记录，还要兼顾质量与成本。

There's no track record of being able to to ramp a supply chain and a production system at that level, the level of complexity that is required and the quality and the cost, etc.

所以当你看埃隆押注的这些新业务时，得面对十几项根本性问题，每一项都必须得到肯定的答案，公司还得有能力全部解决。

So when you look at these new lines of business that Elon is betting the future of Tesla you look at a dozen fundamental questions that need to be answered positively with the ability of the company to address and solve all of them.

在我看来，这些号称要推动公司下一波增长的新业务，还带着一大堆问号。

So a lot of question marks, in my opinion, with regards to these new lines of business that are supposed to be the next wave of growth for the company.

今年到目前为止，股价已经跌了百分之十五。

I mean, the stock is down 15% year to date.

从高点算起，已经跌了将近百分之二十五。

It's down almost 25% from its peak.

但即便如此，它还是家价值一万四千亿美元的公司。

But still, it's a $1,400,000,000,000 company.

目前股价对应的市盈率接近350倍。

It's trading at nearly 350 times earnings.

所有这些乐观情绪，还是建立在对人形机器人和自动驾驶出租车的期待上吗？

Is all of that optimism still based in expectations around the humanoid robot and the robotaxi?

如果是的话，会不会太乐观了？

And and if so, is it too optimistic?

大家是不是太相信这家公司能完美解决所有这些问题了？

Are people putting too much faith, in this idea that all of those questions are going to be answered well by the company.

我还想补充一个常被低估、几乎没人提起的挑战。

So I will add one more challenge that Tesla has to overcome that is often underestimated or that we rarely talk about.

如果看看特斯拉过去成功的根源——顺便说一句，SpaceX也一样——本质上，都是埃隆的愿景，以及他把愿景转化为战略的方式。

If you look at what has made Tesla successful in the past and the same applies to SpaceX, by the way so fundamentally, there's Elon's vision, Elon's aspirational vision, the way he puts the vision into a strategy.