本集简介
双语字幕
仅展示文本字幕，不包含中文音频；想边听边看，请使用 Bayt 播客 App。
OpenAI收购了TBPN。
OpenAI buys TBPN.
这会值得吗？
Will it pay off?
SpaceX申请了IPO，而且规模不小。
SpaceX files for an IPO, and it's a big one.
伊朗袭击了波斯湾地区的亚马逊目标，导致至少几个设施无法运行。
And Iran has hit Amazon targets in The Gulf, rendering at least a few inoperable.
这将在本节目周五特别版中紧随其后播出。
That's coming up on a Big Technology Podcast Friday edition right after this.
欢迎收看Big Technology Podcast周五特别版，我们将以我们一贯冷静而细致的方式解析新闻。
Welcome to Big Technology Podcast Friday edition where we break down the news in our traditional cool headed and nuanced format.
今天为大家准备了一场精彩的节目。
We have a great show for you today.
我们将深入探讨OpenAI收购TBPN这一争议性举措，以及这是否值得。
We're gonna talk all about OpenAI's controversial buy of TBPN and whether it will pay off.
此外，SpaceX终于在筹备首次公开募股。
Also, SpaceX has an IPO in the works finally.
他们计划以2万亿美元的估值筹集400亿至800亿美元的资金。
They're looking to raise maybe 40 to $80,000,000,000 at a $2,000,000,000,000 valuation.
我们还将讨论私人信贷领域的一些问题，以及伊朗袭击了亚马逊的部分基础设施。
We're also gonna talk about some issues in the private credit world, and Iran has hit some Amazon infrastructure.
我们会探讨这意味什么。
We'll talk about what that means.
今天加入我们的是莉兹·霍夫曼。
Joining us is Liz Hoffman.
她是Semafor的商业与金融编辑，今天作为特别嘉宾来到我们节目。
She's Semafor's business and finance editor and a special guest here with us today.
莉兹，欢迎来到节目。
Liz, welcome to the show.
嘿，亚历克斯。
Hey, Alex.
谢谢邀请我来。
Thanks for having me.
感谢你能来做客。
Thanks for being here.
我们先从本周的重磅消息说起吧。
Let's start with big news of the week.
我当时确实对着这条新闻的标题反复确认了两次，就怕看错了日期，那天明明是4月2日，我还以为是4月1日愚人节呢。
Definitely had to check the headline twice on this to make sure that it was actually April 2 and not April 1.
这个消息来自《华尔街日报》。
This is from the Wall Street Journal.
OpenAI收购了科技行业脱口秀节目TBPN。
OpenAI buys tech industry talk show, TBPN.
OpenAI要进军每日新闻领域了。
OpenAI is getting into the business of daily news.
这家开发了ChatGPT的公司表示，它已收购线上脱口秀节目TBPN。该节目主打科技新闻分钟级分析和高管专访，目标是与彭博社和CNBC展开竞争。
The maker of ChatGPT said it has acquired TBPN, an online talk show that aims to compete with Bloomberg and CNBC in by the minute analysis of technology news and executive interviews.
尽管TBPN的观众规模仍然较小，平均每集在各个在线平台上的观看人数约为7万人，但该节目在硅谷的权力人士中颇受欢迎，他们认为它比传统新闻媒体更支持科技行业。曾作为嘉宾出席的首席执行官包括Meta Platforms的马克·扎克伯格、微软的萨提亚·纳德拉，以及OpenAI的萨姆·阿尔特曼。
While TBPN's audience remains modest, averaging around 70,000 viewers per episode across various online platforms, the show has become popular among Silicon Valley power players who consider it more supportive of their industry than traditional news outlets chief executives who've appeared as guests include Meta Platforms, Mark Zuckerberg, Microsoft Satya Nadella, and OpenAI's Sam Altman.
好吧。
Alright.
我对这笔收购是否明智有一些看法，但莉兹，我很想听听你的观点。
I have some thoughts about whether this is a good purchase or not, But Liz, I'm curious to hear your perspective.
你能解释一下TBPN是什么吗？
Can you explain what TBPN is?
因为还有一些人并不了解。
Because there are some people that still don't know.
还有，你认为OpenAI收购他们是一个正确的决定吗？
And whether you think OpenAI is making the right move in buying them.
是的。
Yeah.
这是一个非常受欢迎的科技访谈节目。
It's a really popular tech talk show.
我的意思是，这个节目每天直播好几个小时，在硅谷真有一批忠实观众，确实有真正的高管来参加。
I mean, it's on air live for, like, a lot hours every day and has has actually has a real following in Silicon Valley, has real executives come on.
我不确定他们是否会称自己为记者。
I don't know whether, you know, they would call themselves journalists.
问问他们这个问题会很有趣。
It'd be an interesting question to ask them.
他们明显站在哪一边，这点毫无疑问。
They are clearly there's no doubt about what side they're on.
他们希望上节目的人成功，而且采访通常都比较轻松。
They they want the people on their show to succeed and and it's generally a fairly light interview.
但他们确实占据了科技媒体认知空间的很大一部分。
But they've they've, like, captured a real portion of the sort of tech media brain space.
你觉得这是个好决定吗？
Do I think it's a good move?
我不知道。
I don't know.
我的意思是，萨姆·阿尔特曼说他买下它是因为他喜欢。
I mean, Sam Altman said he bought it because he likes it.
而且，换个角度看，亿万富翁购买媒体公司是因为它们有影响力，而且他们喜欢拥有一个扩音器。
And and, you know, one way to look at it is that, you know, billionaires buy media companies because they're influential and they like having a a megaphone.
你知道的？
You know?
鲁珀特·默多克拥有很多媒体。
Rupert Murdoch owns a lot of media.
巴克·贝尼奥夫收购了《时代》杂志。
Barc Benioff bought Time Magazine.
显然，杰夫·贝佐斯收购了《邮报》。
Obviously, Jeff Bezos bought The Post.
杰米·戴蒙最近接受了一次采访，他说他退休后可能会创办一家媒体公司。
Jamie Dimon gave an interview recently where he was saying maybe what he'll do after he he retires is start a media company.
你知道，拥有媒体是一件很酷、很有趣的事。
You know, it is a sexy, fun thing to own.
你知道，但我也觉得科技界，尤其是人工智能领域，正在环顾四周，意识到这个国家并不支持他们，他们面临一些需要回答的问题，而且真正的反对声音已经开始出现。
You know, but but also I think tech and and particularly AI is sort of looking around and realizes that the country is not with them and that they have some questions that they need to answer and some real pushback is starting to emerge.
我不确定，购买一个主要面向那些已经认同你观点的人的媒体公司，是不是解决这个问题的办法。
I don't know that buying a real industry inside media company that is mostly talking to people who already agree with you is is gonna be the way to do that.
但控制叙事这个想法并不新鲜。
But, you know, controlling the narrative is not a new idea.
好的。
Alright.
首先，我要指出这里有一个区别。
So first of all, I'll point out there was a difference.
这里的不同在于，你提到的贝佐斯、贝尼奥夫这些例子，对他们来说几乎是个人的副业。
There's a difference here, which is that these examples that you gave of Bezos, Benioff, they're almost like personal side projects for these billionaires.
而这是OpenAI收购TBPN，作为直接的内容营销工具。
This is OpenAI acquiring TBPN as a, direct content marketing vehicle.
所以这非常不同。
So that's very different.
所以他们期望这不仅仅是一个面子工程，还能为公司的利润做出贡献。
So they expect that it will not just be a vanity project, but contribute to the company's bottom line.
有趣的是，他们已经停止了所有广告投放。
And interestingly, they're stopping all advertising.
那么问题来了：为什么要做这笔交易？
So then the question is: Why do the deal?
我来读一下我听过的关于这笔交易的最佳论点，然后我会解释一下，其实我已经透露了答案。
I'm gonna read the best argument that I've heard for the deal, and then I'm gonna explain, I'll give it away.
我不喜欢这笔交易，但我会在读完之后再解释原因，然后我们可以一起讨论。
I don't like the deal, but I'll explain why after I read this and we can talk through it.
所以有个人提到，OpenAI 去年的收入达到了 37 亿美元。
So this is someone who talked about OpenAI did 3,700,000,000.0 in revenue last year.
顺便说一下，最新的数据显示，他们现在每月的收入大约是 20 亿美元。
By the way, the latest numbers are they're doing like 2,000,000,000 a month now.
他们只花了其中的一小部分，就掌控了每一位创业者在决定开发什么时的‘客厅’。
They spent just a fraction of that to own the living room of every founder deciding what to build on.
大多数人把TBPN收购当作一个媒体新闻，但实际上它是一个渠道故事。
Most people are treating TBPN as an the TBPN acquisition as a media story, it's actually a distribution story.
以下是他们的思路。
So here's the thinking.
如果OpenAI打算在未来几年投入大约1.3万亿美元——可能略少一点，也可能略多一点——来赌它的AI能成功的话。
If OpenAI has know, it's gonna spend, what, 1.3 something trillion dollars in the neighborhood, maybe a little bit less, maybe a little bit more over the coming years to, you know, make this bet that its AI is going to work.
它正在进入企业市场。
It's getting into the enterprise world.
所有将要向它花钱的公司都在关注TBPN，因为它是这个行业的重要出版物。
All of the companies that are gonna be spending money with it are they are watching TBPN because it is this industry publication.
就像你提到的，这些片段在X平台上会迅速走红。
And like you mentioned, the clips go viral on X.
许多高管在关键时刻前来，分享他们对世界的看法，很可能是因为他们认为会得到一次友好的采访，而他们往往确实如此。
They have all these executives coming in at important moments, sharing their perspective on the world, probably again because they think they'll get a friendly interview and often they do.
哦，他们有时候也会问一些很好的问题。
Oh, they do ask some good questions sometimes.
如果你是OpenAI并且拥有这个网络，你就能接触到所有做决策的人。
And if you are OpenAI and you own that network, you can reach all the people who are making decisions.
我该选择OpenAI吗？
Do I wanna go with OpenAI?
我该选择Anthropic的传统内容营销吗？
Do I wanna go with anthropic traditional content marketing?
你知道，也许这听起来合理，但在一个大型行业中，你需要大胆的举措。
You know, maybe that makes sense, but this is a big swing, and in a big industry, you need big swings.
这有道理吗？
Is that logical?
我不确定是否更好，如果你真的想要进入企业市场，那就选Salesforce。
I don't know that that I think that there's better like, if you really want distribution and enterprises, by Salesforce.
对吧？
Right?
找到进入这些大公司的最后一公里渠道。
Like, find find whatever that last mile is into these big companies.
我认为，很多支出将决定任何人工智能是否能真正收回成本，但这场竞赛的赢家不会来自那些高管在TBPN上露面的公司。
And I think, you know, a lot of the spending that's gonna determine whether any AI really pays for itself, but, you know, who wins this race isn't gonna come from the the companies whose executives are on TBPN.
赢家会来自高盛和沃尔玛这样的公司。
It's gonna come from, like, Goldman Sachs and Walmart.
我的意思是，人工智能必须脱离科技经济，才能拥有真正来自其他行业的企业客户，而这些人根本不会关注TBPN。
I mean, it it's gotta it's gotta escape the tech economy to have real enterprise customers that are just, like, out in other industries and, like, those people are not watching TBPN.
我不确定。
I don't know.
我其实一直觉得这只是一个面子工程，而你刚才说，那些其他高管、其他亿万富翁是个人购买了这些东西。
I actually just sort of thought it a visit of vanity play, and you were saying, yeah, that those those other executives, those other billionaires bought those things personally.
但像萨姆，我假设这是OpenAI的股票吧？
But like Sam, this I assume this is an OpenAI stock.
实际上，我也不完全确定。
Actually, I'm not a 100% sure.
但那是萨姆·阿尔特曼的，那是他的投资工具。
But that is Sam Altman's that is his vehicle.
对吧？
Right?
他个人和财务上都与OpenAI深度交织。
He he is so intertwined personally, financially with OpenAI.
我猜，那就是那里正在发生的事。
I assume, you know, that that that's what's going on there.
我对这一点持怀疑态度，原因稍微有点不同：我记得十年前，WeWork和亚当·纽曼如日中天时，亚当迷上了冲浪。
I I'm skeptical of it, I guess, for, like, a slightly different reason, which is, you know, I remember go back, like, ten years when WeWork and Adam Newman were on top of the world, and and Adam got really into surfing.
于是他收购了那家人工浪池公司。
And so then he bought that wave pool company.
一般来说，你只要给科技创始人足够的风投资金，他们就开始沉迷于自己的个人爱好项目。
Like, in general, you know, you you give tech tech founders enough VC money, and they do they do start to indulge their personal sort of pet projects.
而那个案例中，这至少可以说是WeWork的巅峰时期，或许也是科技圈里最夸张的案例之一。
And that, I guess, least in that case, was sort of, like, absolute peak WeWork, perhaps a a top tech that one a bit.
我愿意从几个角度跟你分析一下看空的理由。
I'll make the bear case with you, for a couple of reasons.
我认为核心问题是，即使TBPN是科技行业的爱好者，一旦这笔收购完成，你就失去了公信力。
And I think the the core of it is even if TBPN is something that is a fan of the tech industry, the moment that acquisition is made, you lose your credibility.
对吧？
Right?
你仍然可以提供有信息量且有趣的内容，但当你所隶属的公司如此显赫地拥有OpenAI这样的科技巨头，而你又在评论科技领域时，人们不可能不怀疑你的深度访谈：他们是不是在为OpenAI的利益服务？
Like, you're you could still be informative and entertaining, but when you have a company that owns you as prominent as OpenAI in in the tech space and you're commenting on the tech space, there's no way people will look at your important interviews and say, how, you know, are they trying to advance OpenAI's interests in doing this?
因为最终，OpenAI的利益就是他们的利益，尤其是正如我们俩都怀疑的那样，他们很可能拿到了大量股票作为报酬。
Because ultimately, OpenAI's interest is their interest, especially as we both suspect if they got paid in a lot of stock.
问题是，他们怎么会出现在像Meta最近那场大型开发者大会那样最前沿的场合？他们现在去参加这类活动时，作为OpenAI的分支，还能否获得与高管接触的权限？
And the question is, how do you end up showing up at events like they were front and center at Meta's, big developer event recently, and can are they gonna go now to that event with the access to the executives as a branch of OpenAI?
他们还向公司的首席游说官克里斯·拉恩汇报工作。
They're also reporting into Chris Lahain, who's the company's chief lobbyist.
所以我怀疑他们会失去相当一部分观众。
So I suspect that they will lose a good chunk of their audience.
而且再次强调，这并不是一个庞大的观众群体。
And again, it's not a massive audience.
每集大约有七万人观看，这不算少，但也不算庞大。
It's about 70,000 per people per episode, which is nothing to scoff at, but it's not huge.
这显然是针对科技领域的，但我怀疑，在这次收购之后，他们开展工作会比以前困难得多。
It's obviously tech specific, but I I suspect that it'll be much much more difficult for them to do their job than it was before this.
我觉得这是对的。
I think that's right.
我的意思是，我在说自己的书。
I mean, I'm talking my own book.
你在说你的书。
You're talking to yours.
独立媒体的价值很大程度上在于它的独立性。
Independent media is, you know, a lot of its value is in being independent.
这并不是全新的事情。
This isn't totally new.
几年前，我觉得霍罗威茨曾试图推出自己的媒体公司。
Couple years ago, I think it was interesting Horowitz, like, had launched kind of tried to launch their own kind of media company.
那是开始，它是
It was the beginning It's of
叫做Future。
called Future.
Future。
Future.
是的。
Yes.
他们打算绕过传统媒体。
They were gonna disintermediate the press.
没错。
Exactly.
对。
Right.
他们没能
They couldn't
他们根本跟不上新闻的数量。
They they couldn't even keep up with the story volume.
他们根本没怎么发布过。
Like, they never published.
没错。
No.
还有那个大型在线时尚杂志。
And there's also that big online glossy.
我觉得它叫Colossus，总是做那些大型的、闪闪发光的专访。
I think it's called Colossus that's always doing these, like, big glowy profiles.
他们知道自己有《名利场》那种感觉，但根本不会去质疑这些科技巨头。
They know they have the feel of like a vanity fair, but but are are not there to poke holes in any of of these tech giants.
你知道的。
You know?
但你现在看到越来越多像《All In》这样的播客，你知道的，我觉得杰森最近还在推特上说，直接联系，直接联系，直接联系。
But but you're seeing a lot more like the all in pod, you know, and I think Jason was was tweeting about this saying go direct, go direct, go direct.
你知道，媒体生态已经发生了很大变化，像X这样的平台，以及碎片化媒体的长尾部分，使得直接传播变得容易多了。
You know, the media ecosystem has changed a lot and and places like X and and, you know, the sort of long tail of fragmented media make it a lot easier to to go direct.
所以我们拭目以待吧。
So we'll see.
我觉得你说得对。
I think you're right.
我想，如果他们真的非常在意自己的公信力，就不会把公司卖给一个他们所报道的、如此基础的媒体公司。
I guess if they really care a lot about their credibility, they wouldn't have sold to a company that they cover that sort of basic media.
没错。
Correct.
所以这是第一点。
So that's that's number one.
而且，我觉得在这方面有人可能会反驳我，就像你刚才说的，你是个怀旧派，坚持那些已经过时的价值观，而现在人们只想获得信息，不需要那些总爱挑刺的人；TBPN的两位主持人很友好，他们依然会在AI领域拥有影响力。
And, you know, I think there's an argument to be made against me on that front, which is just like, you know, you're like like you said, you're you're you're salty and one of the people who believes in values that are long gone, and people just wanna be, you know, informed without these people who are who are, salty about things and the two hosts of TBPN are friendly and they're gonna still have purchase in the, AI world.
我会接受这个反驳意见，虽然我觉得我对自己有点太苛刻了。
And and I'll accept that as a counter argument, although I sort of I've I think I was just a little too harsh on myself.
我对很多技术确实持乐观态度，但毫无疑问，过去有很多怪人，他们总觉得新闻报道只能有一种方式，TBPN从来就不是新闻，他们也会在OpenAI上搞砸。
I do have optimism about a lot of technology, but there are certainly a lot of well, we can we can both agree that there have been a lot of cranks out there who've just sort of like, you know, journalism is done one way, TBPN was never journalism, and they're gonna suck at OpenAI too.
我对这种观点有点不同意。
And I sort of disagree with that notion.
但对我来说，更大的问题是，这可能是用错误的方法解决正确的问题。
But I I I think to me the bigger note, the bigger problem here is, this might be a solution, the wrong solution to the right problem.
我的意思是，莉兹，你说得对，你之前提到了这一点。
And by that I mean, you're right, Liz, and you brought this up.
目前，人工智能面临着一个非常严重的认知问题。
AI has a very big perception problem right now.
这一点，人工智能领域的人士是清楚的。
And this is something that AI, you know, they are mindful.
就连你在《华尔街日报》上看到的报道也表明，邀请他们加入正是出于这一目的。
And even the reporting that you're seeing in come out in the information in the Wall Street Journal tells us that this was, you know, part of the aim of bringing them on.
这是Fidji Simo，OpenAI AGI部署负责人。
This is Fidji Simo, OpenAI's head of AGI deployment.
她在一份内部备忘录中写道。
She wrote in an internal memo.
当我思考我们在OpenAI的沟通未来时，有一点变得很清楚：标准的沟通手册根本不适用于我们。
As I've been thinking about the future of how we communicate at OpenAI, one thing that's become clear is that the standard communication playbook just doesn't apply to us.
我们不是一家典型的公司。
We're not a typical company.
我们推动着一场重大的技术变革。
We drive a really big techno technological shift.
西莫希望探索新的方式，向公众传达OpenAI的使命及其技术的影响和应用。
Simo Simo wanted to explore new ways to communicate OpenAI's mission and the impact and applications of its technology to the public.
针对周四OpenAI员工关于收购的提问，西莫表示，她认为OpenAI在公开沟通其产品优势方面可以做得更好。
In response to questions from OpenAI staffers about the acquisition on Thursday, Simo said an internal message she believes OpenAI can do better at communicating publicly about the benefits of its products.
好的。
Okay.
所以，除了OpenAI内部人员在西莫明确告诉他们不要做额外任务后，还认为这是一场愚人节玩笑之外。
So aside from the fact that the people inside OpenAI thought that this was an April fool's joke after Simo explicitly told them no side quests.
这个理由，我不太明白，我也在听。
This rationale, I don't I and listen.
我理解你想改变叙事方式。
I I I get wanting to change the narrative.
但我认为这是错误的解决方案。
I just think it's the wrong solution.
因为如果你看看YouGov的民意调查，我们在这个节目中讨论过，哪些人不喜欢人工智能？
Because if you look at the YouGov polling, and we've talked about it on this show, who are the people that do not like, AI?
对人工智能最反感的人群，是那些从未使用过、也没见过别人使用过它的人。
The most negative group about AI are people who have not used it or seen anyone else use it.
他们对人工智能持正面看法。
They're only positive about AI.
数据显示，22%的时间是正面的，60%的时间是负面的。
It looks like 22% of the time and negative 60% of the time.
那些见过人工智能使用但自己没用过的人，虽然有一些正面看法，但在纯粹的负面情绪中，他们占比最高，达到62%。
People that have have seen it used but haven't used it for themselves are the most, they have some positive views, but in terms of pure negative, they're the most 62% of them are there.
但那些经常使用AI的人群中，只有26%的人对AI持完全负面或负面多于正面的态度。
But the people who regularly lose use AI compared to that sixty two percent, only twenty six percent of them are entirely negative or more negative than positive on AI.
而这正是TBPN的受众。
And that is TBPN's audience.
那些经常使用AI的人。
The people that regularly use AI.
如果你想想需要听到这个信息的公众群体，他们并不是目标人群。
You think about the cross section of the public that needs to hear this message, it's not them.
目标是另一群人，而TBPN并不针对他们。
It's the other people, and TBPN doesn't cater to them.
这就是为什么我认为这次收购是错误的。
And that's why I think this acquisition is the wrong one.
是的。
Yeah.
嗯，不是这样的。
Well, no.
这正是我认为这只是萨姆想做的一件事的原因。
Which is why I actually just think this is a this is just a thing Sam wanted to do.
在我看来，最简单的解决方案就是最容易的那个。
It seems to me that the simplest solution is is the easiest one.
而且，你知道，他们必须采取更广泛的公众沟通策略，来说服那些日益憎恨这项技术、害怕它、不喜欢其创始人、并担心自己会丢掉工作的大批公众。
And, you know, they are gonna have to, you know, follow a a much broader public discourse strategy to convince the huge swaths of the public that really just increasingly by the day hates this technology and is terrified of it and doesn't like the people founding it and, you know, is afraid of losing their jobs.
我的意思是，我不认为这会改变舆论走向。
I mean, I I don't think, I don't think this is gonna change the narrative.
如果OpenAI内部真正聪明的人认为这能改变什么，我会感到惊讶。
And, I guess I'd be surprised if really smart people inside OpenAI thought it would.
这在我看来就是萨姆喜欢，所以他的公司就买了它。
This just seems to me like this is a thing Sam likes, so his company bought it.
但我下周要去TBPN，我想我会问问他们对这一切的看法，看看能不能稍微扭转一下采访的方向。
But I'm going on TBPN next week, I think, so I will ask them what they what they make of all of this, See if we could flip the interview a little.
这才是该做的事。
That that's the way to do it.
你为什么认为人们对人工智能这么消极？
Why do you think people are so negative on AI?
因为他们害怕它。
Well, they're afraid of it.
我的意思是，他们真的害怕人工智能会夺走他们的工作，这是一种非常正常且在方向上至少在这里并不 inaccurate 的恐惧。
I mean, just literally afraid that it will take their jobs, which is a very normal fear and not inaccurate in directionally, at least here.
但我还认为这纯粹是一种反乌托邦的想象。
But I also think it's just dystopian.
就像，你看。
Like, it is look.
我喜欢人工智能。
I like AI.
我的意思是，这显然是一种神奇的技术。
Like, I mean, this is obviously a magical technology.
它具有巨大的经济效益。
It has huge economic potential benefits.
你知道吗，如果你跟医疗和生物技术领域的人聊聊，AI在药物发现方面能做什么。
It you know, you talk to people in, for example, health care and biotech about what it could do on the drug discovery side.
这简直就像魔法一样。
And this is like magic stuff.
但感觉好像没人真正掌控局面。
But it feels like it doesn't have anyone in the driver's seat.
我认为硅谷整体上并没有很好地证明自己值得信赖。
And I don't think that Silicon Valley in general has done a good job showing that they can be trusted with much of anything.
甚至连他们自己都说并不真正理解它。
And even they say they don't really understand it.
你正在谈论一种强大的技术，而开发它的人，即便从最善意的角度看，也跟普通美国人截然不同——行为方式不一样，富有得难以想象，还有一堆怪癖，比如他们一个个都像卡通人物一样。
You're talking about a really powerful technology being developed by people who charitably do not look like the average American, do not behave like them, are unimaginably wealthy, like have weird habits about like, you know, they're all like they're all slightly cartoonish.
然后你看到它出现在你的社区里，以那些巨大的、面目模糊的数据中心形式出现，这些中心在不同程度上——我们或许可以争论这是否完全属实——正在消耗你的能源、占用你的水资源，它们没有面孔，令人恐惧。
And then you see it showing up in your communities in these big faceless data centers that are to varying degrees, we can argue about how true this is, taking your energy, taking your water, like putting they're faceless and they are really scary.
所以人们为什么讨厌这些东西，对我来说并不难理解。
And this is not a huge mystery to me why people hate this stuff.
你说得对。
And you're right.
使用它的人似乎很喜欢。
People who use it seem to like it.
你能看到它的优势。
You see the benefits of it.
是的，这面临着巨大的形象问题，你已经开始在本地社区看到真正的反弹了。
Yeah, this has a huge, huge image problem, and you're starting to see real backlash in local communities.
我不认为这些大公司曾经认真思考过这个问题，或者他们现在才开始意识到，自己根本无力讲好这个故事。
And I don't think that's something with these big companies have thought a lot about or they're starting to now and realizing that they are woefully ill equipped to tell that story.
对。
Yeah.
他们已经开始注意到了。
They're noticing it.
我的意思是，他们意识到自己在叙事上存在问题，但TBPN似乎并不会成为解决这个问题的答案。
I mean, they noticed that they have a narrative issue, but it does not seem like TBPN will be the answer to that.
有趣的是，就在我们刚要进入录音室的时候，收到了《华尔街日报》的一些新闻。
Interestingly, just as we started to, head to the recording booth, so to speak, some news from the Wall Street Journal.
OpenAI的高级执行官菲吉·西莫将休医疗假离开公司。
OpenAI top executive Fidji Simo to take medical leave from the company.
OpenAI的产品与业务负责人菲吉·西莫——顺便说一下，她现在是AGI部署的负责人——因健康原因将离开公司数周，这个缺席……抱歉，我说重复了。
OpenAI's product and business chief Fidji Simo, who, by the way, is now the head of AGI deployment, She's taking medical leave from the company for several weeks, an absent that that that sorry.
这一缺席正值这家初创公司正在重组领导层并准备公开上市之际。
An absence that comes as the startup revamps its leadership ranks and prepares to make its public market debut.
在给员工的内部备忘录中，西莫表示，她的神经免疫状况恶化了，自去年八月入职以来，她一直推迟了医疗检查和新疗法，以便专注于工作。
In an internal note to staff, Simo said a neuroimmune condition had worsened and that she had postponed medical tests and new therapies to stay focused at work since starting her job last August.
我不确定是不是自己逼得太紧了，现在真的需要尝试新的干预手段来稳定我的健康状况。
It's not clear that I pushed a little too far and that I really need to try new interventions to stabilize my health.
OpenAI内部正在发生更多人事变动。
There's more shake ups happening within OpenAI.
OpenAI正在对高管团队进行其他调整，将原本由运营主管布拉德·莱特卡普负责的商业职责，交给了前Slack高管兼营收负责人丹妮丝·德雷瑟。
OpenAI is making other changes to executive team to its executive team handing former Slack executive and revenue chief Denise Dresser commercial responsibilities that were pervy previously the purview of operating chief Brad Lightcap.
莱特卡普现在将专注于特殊项目。
Lightcap will now focus on special projects.
OpenAI表示，目前没有立即任命新首席运营官的计划，而这是莱特卡普之前担任的职位。
OpenAI said it doesn't have any immediate plans to appoint any new COO, which is the role that Lightcap held previously.
OpenAI内部的持续变动仍在继续。
Constant change within OpenAI, and it continues.
是的。
Yeah.
我的意思是，这些公司——撇开从事技术工作的人员不谈，我认为这些公司最关键的角色是首席商务官。
I mean, these companies, the the arguably, setting aside the people working on the technology, I think the key role at these companies is chief commercial officer.
你如何将这些东西商业化并实现盈利？
How do you commercialize this stuff and monetize it?
回到我们最初讨论的问题：企业业务在哪里？
And getting back to where we started on where is the enterprise business?
你如何获得这些订阅？
How do you get those subscriptions?
我经常使用OpenAI，但我从没给过他们一分钱。
I use OpenAI a lot, but I don't think I've ever given them a dollar.
那么，你如何将这项技术从实验室带入真正的经济体系呢？
So how do you take this from the lab, you know, into into the real economy?
关于这一点，我还有一个最后的观点，之后我们就可以继续了。
I have one last point on this, then we can move on.
OpenAI，尤其是在Fidji的领导下，最近几周变得更加专注了，不再搞旁支项目，关闭了Sora，因此我认为存在一种懒惰的论点，说确实如此。
OpenAI, especially under Fidji's leadership, has been, in recent weeks, much more focused, no more side quests, shutting down Sora, and so there's been, I think, a lazy argument that that yeah.
一种懒惰的论点认为购买TBPN是旁支项目，这确实通常是，但我认为这种说法误解了‘不搞旁支’的指令，这个指令的核心其实是计算资源。
A lazy argument that buying TBPN is a side quest, which it invariably is, but I think it misunderstands this no side quest directive, which was all about compute.
你不希望把计算资源浪费在旁支项目上。
You don't wanna waste compute on side quests.
你应当专注于核心事务。
You want to focus on the core thing.
是的，这会占用一些注意力，但我认为TBPN并不算真正违背了关于旁支项目的指令。
Yes, it will take some attention, but I don't think TBPN is exactly going back on the on the command about side quests here.
是的
Yeah.
我觉得那个项目并没有消耗大量计算资源。
Don't think that show is sucking up a ton of compute.
不。
No.
制作这些人工生成的视频所需的计算量，肯定远少于AI生成的视频，这在某种程度上是一个有趣的时刻，尤其是对那些认为AI会立即取代人类工作的人来说。
Definitely less to make these human generated videos than the AI generated videos, which is kind of an interesting moment in time for those, again, going back to the job loss thing, for those who think AI will immediately take over.
要做好这些事情需要大量计算资源，不能简单地说‘让AI去做就行了’。
Well, it takes a lot of compute to do a lot of these things, and it's not easy as easy as saying, let's just allow the AI to do it.
不。
No.
而且你在另一端也能看到，这些令牌非常昂贵，对吧？
And you're seeing it on the other side, these two, these tokens are expensive, right?
我的意思是，使用AI服务的人开始累积的账单确实让人有些震惊。
I mean, the bills that are starting to rack up for people who use it, it's a bit of a sticker shock there too.
是的。
Yeah.
所以你有令牌价格，还有视频价格。
So you have the token prices, you have the video prices.
因此，AI将取代所有人工作的这种说法，我认为有点被夸大了。
And so this moment where AI will take over everybody's jobs, I think, is a little bit overblown.
但你和黑石集团首席执行官拉里·芬克进行了一次非常有趣的对话，谈到了几件事，首先是他说的AI相关工作。
But you had a very interesting conversation with Larry Finks, CEO of BlackRock, about a few things, and first is AI jobs that he talks about.
他说，AI将创造许多工作岗位，而我们社会尚未做好准备来填补这些岗位。
He says AI is going to create many jobs, and we're not prepared as a society to fulfill those jobs.
这是一场危机。
That's a crisis.
这是一个有趣的视角。
That's an interesting perspective.
你能再多谈谈这个吗？
Can you talk a little bit more about that?
他其实说的是蓝领工作，这些工作本来就严重缺乏足够的人手来填补。
He was really talking about blue collar jobs that you know, which were already really in in short supply of the the people to fill them.
比如焊工和电工，这些工作某种程度上是机器人最后才会取代的，还有暖通空调和管道工作。
Talking about welders and electricians and you know, which is in some ways, like, the last jobs that the robots will take that HVAC plumbing.
这类工作在物理世界中似乎很难由人工智能完成。
That stuff seems hard for AI to do in the physical world.
而如今，建造所有由黑石集团及其他机构融资并快速兴建的数据中心，以满足OpenAI的计算需求，这些工作急需人手。
And it's like acutely needed now to build all the data centers that BlackRock and others are financing and throwing up to service OpenAI's compute needs and all of that.
但在更广泛的就业领域，或许更容易看到AI带来的影响，这可能也适用于过去的工业革命。
But on the broader jobs front, I guess it is a lot easier to see and maybe this was true of past industrial revolutions.
人们把这称为第四次工业革命。
People talk about this as the fourth industrial revolution.
相比之下，更容易看到AI会消灭哪些工作，而不是它会创造哪些工作。
It's a lot easier to see the jobs that AI will kill than the jobs that it will create.
尤其是在知识经济领域，我过去十五年一直在报道华尔街，这些行业大多是金字塔型的大型企业，大量吸纳来自常春藤盟校和商学院的年轻毕业生，让他们花上数年时间从事一些枯燥的基础工作，比如咨询、投资银行，做PPT、处理大量电子表格。
And particularly in the knowledge economy, these are I've spent the last fifteen years covering Wall Street, and these are big pyramid businesses where a lot of They hoover up these kind of young graduates you know, from the IVs and from business schools, and they spend a bunch of years doing a kind of drudge work, like this consulting, investment banking, that kind of thing, a PowerPoint presentations, a lot of spreadsheets.
AI如今实际上几乎能完成所有这些工作。
AI can actually today do almost all of that.
所以，拉里在采访中还提到，你会看到这一代人。
And so, you know, the other thing Larry said in the interview is that, you know, you're gonna see this cohort.
他基本上是在说，哇，2026届的毕业生。
He was basically saying like, woof, like class of 2026.
对吧？
Right?
在未来几年里，技能缺口会变得非常非常大，我认为我们会看到大规模的青年失业。
Like, there's gonna be a couple years where the skills gaps are just really, really wide, and we're gonna see I and I think we're gonna see massive youth unemployment.
你知道，接下来几年里，22到25岁这个年龄段的人群。
You know, that, you know, 25 to or 22 to 25 year old cohort for the next couple of years.
目前他们的失业率已经远高于全国平均水平，我认为这会急剧飙升，情况会非常糟糕。
It's already, you know, elevated beyond the the national unemployment, and I think you are just gonna see that skyrocket, and that's really bad.
我们还没有解决这个问题的办法。
We don't have a solution for it.
是的。
Yeah.
我的意思是，你们都明白，公司可能会说，我们有AI来做这些初级工作。
I mean, you all like, I could see the firm saying we have AI to do that entry level work.
另一方面，我可能得赶紧跑到大学校园里，试图招到那些懂AI的22岁年轻人，让他们为我的公司工作。
On the other hand, I might be scrambling to university campuses to try to get these 22 year olds who understand AI into my organization and working with it for my company.
我觉得你其实是在担心那两三年的缺口。
I think you're actually worried about that gap like couple of years.
我觉得，你知道的，我算是互联网原住民。
Like, I think, you know, I'm I was sort of Internet native.
对吧？
Right?
也许我并不是移动原住民。
And maybe and but was the probably not mobile native.
我2000年就参加工作了，我觉得我上班的时候还没用上第一台iPhone，这么说你能明白吧。
Like, I got my first job in 2000 and I think I had my first job before I had my first iPhone, if that helps you.
那么你就拥有了一个移动原生的员工群体。
And then so you had then you had a mobile native employee base.
我不确定这有多大帮助。
I'm not sure how helpful that was.
移动设备显然是我们生活中一个重要的平台，但它相当直观。
Mobile is obviously a big platform for our lives, but it's fairly intuitive.
但我认为你可能会面临一代工人被跳过的情况。
But I think you may end up with a generation of workers that just get skipped.
你知道，人们都在说什么？
You know, what's what's Like, what are people saying?
我不可能会被AI取代我的工作。
Like, I'm not gonna lose my job to AI.
我会因为比不上懂AI的人而丢掉工作，所以我正努力学习如何使用AI。
I'm gonna lose it to someone who knows how to use AI, and so I am endeavoring to learn how to use AI.
但我真的很担心那些高中生。
But I'm, like, afraid of high schoolers.
我现在并不害怕25岁的人，因为他们并不比那些懂互联网但不是AI原生的稍年长者处境更好。
Not I'm not afraid of 25 year olds right now because I don't I don't think they're in a much better boat than than slightly older people who kind of understand the Internet but are not AI native.
我认为这里也有同样的论点。
And there's the same argument, I think.
我其实一直在思考这个问题，比如在SaaSpocalypse和科技公司的大洗牌中，我很好奇成为一家AI原生公司到底能有多大优势。
I've been thinking a lot about this actually, like, with the SaaSpocalypse and the tech companies and the shakeout there that I wonder how much of an advantage it is gonna be to be an AI native company.
对吧？
Right?
有很多问题，比如Salesforce能否通过代理功能，充分构建自己的AI系统来保持相关性和用户粘性？
There's a lot of questions of, well, can Salesforce just do enough with agents and build enough of its own AI to remain relevant and sticky?
但对我来说，真正的问题是那些直接建立在这些模型之上的AI原生公司。
But but the real question to me is the AI native companies that are being built directly on top of these models.
就像在DNA里是否有什么东西，让它们能像当年互联网原生公司那样胜出——除了微软之外，那些前移动时代的科技公司完全被那些直接建立在互联网之上的公司超越了。
Like, is there something in the DNA there that makes that makes them win in the way that, you know, really with the exception of of Microsoft, that that pre mobile generation of tech companies just got got totally overshadowed and overtaken by the ones that were built right on top of the Internet.
你对这个问题最好的猜测是什么？
And what's your best guess as the answer to that one?
我不知道。
I don't know.
我不知道。
I don't know.
但我认为，这比我们自互联网以来所见过的任何技术飞跃都要大。
But I think I suspect that this is a bigger technological leap than anything we've seen certainly since the internet.
它比移动互联网更大。
It is bigger than mobile.
我认为，AI原生公司的基因中可能有一些东西会帮助它们取得成功。
And I think there's probably something in the DNA of AI native companies that will help them succeed.
其实，我最近刚去了一趟，和一群律师交谈过。
Like, was just I've been on a tour talking to bunch of lawyers recently.
他们正在使用Harvey，也在使用Logota。
They're using Harvey, they're using Logota.
你知道，汤森路透在AI上投入了大量资金，他们还推出了CoCounsel。
You know, Thomson Reuters has spent a lot of money on on AI, and and they've seen called co counsel.
这实际上是一个完美的案例研究。
And that's actually, like, a perfect case study.
像汤森路透这样的公司，有着一百多年历史，是一家大型专业服务和媒体机构，它能构建足够的AI来保持相关性吗？
Like, can Thomson Reuters, you know, 100 and something year old, you know, big professional services and media organization, build enough AI to remain relevant?
还是Harvey直接过来说，我们全部都是在Anthropic上构建的？
Or does Harvey just come and say, we just built this all in Anthropic?
顺便说一句，我们只有大约50名员工，而且不再招聘了。
And by the way, we have like 50 employees, and we're not hiring anymore.
当经济模型开始对传统公司不利时，问题就出现了。
Is where the economics start to go sideways for legacy companies.
当然。
Definitely.
我的意思是，去年年底我问萨姆·阿尔特曼关于与谷歌的竞争时，他向我提出的观点就是如此，他说基本上会存在两种公司。
I mean, this was the argument that Sam Altman made to me at the end of the year last year, where I asked him about the competition with Google and he goes, basically, they're going be two types of companies.
一种是试图将AI附加到现有系统上的公司，另一种是从零开始构建AI的公司。
Ones that try to bolt AI on and one that build ones that build AI from the ground up.
他说，我们是从零开始构建AI的。
He goes, we're building AI from the ground up.
他们只是在附加AI。
They're bolting it on.
这就是我认为我们有机会胜出的原因。
That's why I think we have the chance to win.
他可能是对的，但这确实是现在对科技巨头们最该问的问题。
I he might be right, but that's obviously that is the right question to ask about about big tech right now.
我们刚公布了新的就业数据。
We just got new jobs numbers in.
在最新的数据中，你看到AI导致的失业现象了吗？
Are you seeing anything in terms of AI job loss in the new in the new numbers?
没有。
Not really.
三月的就业数据其实很不错，紧接在之前几个月非常糟糕的数据之后。
The jobs numbers came out today for March pretty good, actually, and on the on the heels of some couple of months of, like, very, very bad ones.
我认为昨天有一份来自私人数据追踪机构Challenger Gray的报告，它确实追踪了所谓的与AI相关的裁员，而这类裁员显然正在迅速增加。
I think there was a report yesterday from a private data tracker called Challenger Gray that it does track sort of what it counts as AI related job losses that are obviously growing quickly.
我对这个问题的看法是，AI只是公司裁员的借口。
My theory on this is that AI is an excuse for companies to do layoffs.
就像技术还不够好一样。
Just like the tech isn't good enough.
它还没有足够深入地融入到工作流程中，尤其是在科技公司之外。
It isn't embedded in enough workflows to, like, maybe outside of tech companies.
你知道，马克·扎克伯格谈了很多关于Meta工程师现在到底做什么、不做什么的内容。
You know, Mark Zuckerberg has talked a lot about kind of what what a a meta engineer is and isn't doing right now.
所以也许他们会放慢招聘速度。
And so maybe there, like, they will hire more slowly.
我认为这不过是掩护罢了。
I think it's air cover.
这些公司在2010年代末到疫情期间大规模超量招聘，当时又不想在疫情中期裁员，因为那样会显得很糟糕。
These are companies that massively overhired, you know, in the late twenty tens through the pandemic, didn't wanna fire people in the middle of a of a pandemic because it makes you look bad.
然后我们就像是被冻结了几年，一直搞不清楚该怎么做。
And then couldn't then we're kind of frozen for a couple of years and couldn't quite figure out what to do.
我觉得AI只是个幌子。
I think AI is just air cover.
是的。
Yeah.
这周我在达拉斯的一个会议上看到了Indeed提供的惊人数据，是关于软件开发者的。
I saw unbelievable data from Indeed this week, actually, at a conference down in Dallas where you looked at software developers.
我只想描述一下这个图表的大致趋势，因为我当时在问我们是否看到Indeed上开发者职位的招聘或发布数量有所增加，确实增加了。
And I'm just gonna describe the contours of the chart because I was asking about whether we've seen an increase in developer hires or developer listings on Indeed, and we have.
但只是轻微上升，而如果你再往回看，到2020年、2021年，也就是疫情中期，就会发现职位发布量像一座高山一样陡增。
But it's this small tick up, and then you scroll back to about 2020, 2021, right, mid COVID, and there's a mountain of new listings.
所以显然，当时经历了一轮大规模招聘，而到现在，人们还在琢磨这对他们公司2026年意味着什么。
So clearly, there was just a massive hiring phase and, you know, people are trying to figure out what it means for their companies in 2026 still.
当他们意识到自己招人过多时，说是因为AI就方便多了，也更前瞻，总比说‘天啊，我们公司管理得一塌糊涂’要好。
And when they realize they've over hired, say AI because it's a lot more convenient and forward looking to say that than to say, oh shit, we mismanaged our company.
展开剩余字幕（还有 382 条）
完全正确。
Totally.
杰克·多西可以裁员一大波人，然后说这是AI的原因。
And Jack Dorsey can lay off a bunch of people and say it's AI.
也许人们会相信他。
Maybe people believe him.
但我认为这完全是个借口。
But I think it's totally an excuse.
软件开发者这个案例很有趣。
And the software developer is an interesting one.
前几天我跟Anthropic的首席商务官聊过。
I was talking the other day to the chief commercial officer Anthropic.
他说，你有这么多开发者，但他们本质上只是管理一群编写代码的机器人。
And he was saying that you have all of these developers, but they are basically managers of a bunch of bots that are writing the code.
我认为Claude的代码99%都是由Claude自己写的。
I think 99% of the code for Claude is is written by Claude.
而且，这也没问题。
And, like, that's fine.
那个人的工作内容略有不同，但那些没有被聘为初级开发人员的人就惨了。
That person has a job that's a slightly different job, but it's the it's the people who aren't hired to be those junior developers.
我的意思是，这对我来说是一个指数级趋势，而不是线性趋势，因为受到冲击最大的是这些知识经济岗位，而它们在大公司中的层级结构是最庞大的。
I mean, this is a this is a a like, an exponential trend, not a not a linear one to me because of the the industries that are being so hit by this are these knowledge economy jobs, and they have the widest pyramids of of any big companies.
想想那些初级助理、年轻的投行职员、年轻律师、年轻顾问吧。
Just you think about kind of the the junior associates, the the the baby investment bankers, the baby lawyers, the baby consultants.
这个金字塔将从这样变成这样，再变成这样，最后变成这样。
That pyramid is gonna go from this to, like, this to this to this.
我觉得，最终会变成
And I think, you know, it's gonna end up
一个钻石形状。
being like a diamond.
那些在听的人，解释一下你刚才用手比划的是什么意思。
Those who listening Explain what you're doing with your hands.
哦，抱歉。
Oh, sorry.
我刚才在比划那个金字塔，底部越来越窄，最终变成一根烟囱，甚至可能变成一个钻石形状——因为真正占最大比例的员工，像你我这样年纪的人，非常擅长管理一大批人工智能代理。
I was I was making the pyramid get get narrower at the base until it eventually turns into a chimney and then maybe even a diamond where, like, the actual the biggest chunk of the workers are, you know, people your age, my age, who are, like, pretty competent at managing a bunch of agents.
然后顶部有一些资深人士，负责开拓大量业务。
And then you have some senior people at the top, like generating a bunch of business.
这太疯狂了。
That is crazy.
好吧。
All right.
我真的很想聊聊SpaceX的IPO，因为埃隆·马斯克和他的团队已经提交了上市申请。
I definitely want to talk about the SpaceX IPO because Elon Musk and co have filed for that IPO.
我们还有一则关于海湾地区亚马逊数据中心的消息，或许也可以聊聊私人信贷和人工智能的现状。
And we also have some news about Amazon data centers in The Gulf, and potentially we can talk about what's going on with private credit and AI.
我们会在广告后聊这些，但首先，莉兹，我想请你宣传一下你的新播客，大家应该去找到并收听。
So we're gonna do that after the break, but first Liz, I'd love for you to shout out you have a new podcast and people should go and find it and listen to it.
我有。
I do.
谢谢。
Thank you.
是的。
Yeah.
你看起来在播客领域玩得很开心。
You look like you were having so much fun in podcast land.
我们决定加入你们的行列。
We decided to, to join your ranks.
对。
Yeah.
我们六周前推出了一档名为《复利》的节目。
We launched a show about six weeks ago called Compound Interests.
这是我们的商业与金融节目，每周采访那些正身处塑造企业运营和市场运作方式的宏大趋势中心的人。
It's our business and finance show, and it's weekly interviews with people who are, like, right at the middle of these big forces that are are shaping how businesses are run and how markets are are functioning.
得好好猜一猜。
Had to run a great guess.
我们请来了Uber的达拉，聊了聊司机的未来，按照他的说法，本质上就是每晚给自动驾驶出租车‘盖被子’，照顾和维护这些庞大的自动驾驶车队，这非常有趣。
We've had we had Dara from Uber on kind of talking about what a driver you know, the the future of Uber is is, he tells it, essentially tucking the robo taxis in at night, the kind of care and feeding of these big autonomous fleets was super interesting.
所以，是的，这过程非常有趣，但也让我对你所做的工作有了更多的敬意。
So, yeah, it's been it's been a lot of fun, but it's given me a lot of respect for for what you do.
在话筒的另一边，事情要难得多。
It's it's harder on the other side of the mic.
这需要花时间去学习。
It gotta take some time learning it.
这是完全不同的事情。
It's a different thing.
好的。
Alright.
所以我很想进入我们的第二组故事，我们稍后回来继续。
So I definitely wanna get to our second set of stories, we're gonna do that when we come back right after this.
我在这档节目中采访过许多优秀的科技创始人，其中有一个出人意料的普遍挑战反复出现：找到合适的域名。
I've interviewed a lot of great tech founders on this show, and one surprisingly universal challenge comes up again and again, finding the right domain name.
我在推出《大科技》时也遇到过这个问题。
It's something I ran into myself when launching Big Technology.
你想要的名字往往已经被注册了，因此很容易就想随便选一个了事。
The names you want are often taken, and it's tempting just to settle and move on.
但我最敬重的那些创始人，从不在基础问题上妥协，而域名正是其中之一。
But the founders I respect most don't settle on fundamentals, and your name is one of them.
它应该能立即传达出你实际打造的产品是什么。
It should immediately signal what you actually build.
这正是我欣赏 .tech 域名的地方。
That's what I appreciate about dot tech domain names.
它再合理不过了。
It just makes sense.
它向世界、你的客户、你的投资者，以及任何搜索你的人明确传达：你正在从事科技领域，简洁直接，无需任何修饰。
It tells the world, your customers, your investors, anyone googling you that you're building in technology, clean, direct, no qualifiers.
我看到越来越多认真的初创公司开始采用这种方式。
And I'm seeing more serious startups lean into it.
Nothing.tech、1x.tech、aurora.tech、ces.tech、ultra.tech、alice.tech、neon.tech、blaze.tech、pie.tech，还有更多。
Nothing.tech, 1x.tech, aurora.tech, ces.tech, ultra.tech, alice.tech, neon.tech, blaze.tech, pie.tech, and so many more.
如果你正在打造一个以技术为核心的产品，别将就。
If you're building something tech first, don't settle.
从你选择的任何注册商那里注册一个.tech域名，从第一天起就明确你的定位。
Secure your dot tech domain from any registrar of your choice and make your positioning obvious from day one.
我们回到《大科技》播客，和Liz Hoffman一起讨论本周的科技新闻。
And we're back here on Big Technology podcast talking about all the week's tech news with Liz Hoffman.
她是Semafor的商业与财经编辑。
She is Semafor's business and finance editor.
我们来聊聊SpaceX的IPO。
Let's talk SpaceX IPO.
SpaceX计划以2万亿美元的估值上市。
SpaceX, they're gonna try to go public at a valuation of 2,000,000,000,000.
我认为他们在此次IPO中可能会筹集400亿到800亿美元。
I think they're gonna look to raise something like 40 to $80,000,000,000 in this IPO.
你对此有什么看法？
What do you think about this?
这是否符合你之前的预期？
Is this sort of where you expected it to land?
我的意思是，据彭博社报道，他们的估值将超过2万亿美元。
I mean, it looks like they're going to this is according to Bloomberg at more than $2,000,000,000,000.
在短短几个月内，SpaceX的估值将增长近三分之二。
SpaceX's evaluation would increase by nearly two thirds in a matter of months.
该公司在二月收购了马斯克的xAI公司，使SpaceX的估值达到1.25万亿美元。
The company's acquisitions of of Musk's x AI company valued SpaceX at 1,250,000,000,000.00 in February.
一方面，作为一名财经记者，我觉得这某种程度上令人安心。
On the one hand, I find it kind of comforting as a finance reporter.
长期以来，我一直有一个观点，认为这些公司最终不得不上市，它们一直保持私有状态这么久。
Like, I've had this theory for a long time that these companies were eventually going to have to go public, that it stayed private for so long.
这其实就是2010年代末到2020年代初的主旋律——当时有这么多私人资金，根本不需要公开市场。
That that was really the story of kind of late twenty tens and early twenty twenties that there was just so much private money out there that, like, who needs the public markets?
但正如我们之前讨论的，这一代大型初创企业实际上需要极其庞大的资本投入。
But actually, what this generation of of big startups is doing, as we've been talking about, is super capital intensive.
你知道，造火箭真的非常昂贵。
They're you know, it's really expensive to build rockets.
我认为SpaceX实际上已经盈利了。
I think SpaceX actually is profitable.
所以从很多方面来看，它和2010年代上市的那些软件公司截然不同。
So and so in a bunch of ways, looks really different than that generation of companies, software companies that went public in the twenty tens.
我的关注点是，这笔钱数额实在太大了。
I mean, what I'm watching for is, that's a lot of money.
这笔钱总得有个来源。
It's gotta come from somewhere.
我的意思是，这真的很奇怪。
I mean, it's it's weird.
我们通常不会把IPO看作是需要大量资金从一组投资转移到另一组投资的市场轮动事件。
Like, we don't usually think of IPOs as, like, market rotational events that require a bunch of money to move out of one set of investments and into another.
但800亿美元，这将是有史以来最大IPO的三倍，而那个最大IPO本身又是平均IPO的10到20倍。
But, like, $80,000,000,000, this is that would be three times larger than the largest IPO ever, which was itself 10 times larger, 20 times larger than the average IPO.
我指的是2019年沙特阿美上市，筹集了250亿到300亿美元，但过去25年平均每次IPO仅筹集约2亿美元，这对投资银行来说是个轻松的活儿。
I mean, that was that would have been the Saudi Aramco I b IPO in 2019 that raised $25 or $30,000,000,000 But the average IPO over the last twenty five years raises about $200,000,000 And so that is an easy job for investment bankers.
但前几天我写过，你将在今年五月和六月听到的市场里那阵巨大的吸吮声，其实是银行家们四处奔走，寻求800亿美元的资金——顺便说一句，我认为他们能筹到。
But I wrote the other day that that giant sucking sound you're going to hear in the markets you know, this May and June is is bankers running around asking for $80,000,000,000, which, by the way, I think they'll get.
这是一家盈利的公司，似乎把一件重要的事情做得非常好，而埃隆·马斯克是个极其出色的销售者。
This is a profitable company that, like, seems to do an important thing really well, and Elon Musk is an incredibly good salesman.
人们愿意投资他的公司，但这些资金实际上必须来自某个地方。
People want to to invest in his companies, but that's gonna come out of somewhere actually.
我认为，如果你手里的股票那时仍因战争而低迷，投资者会非常不愿意通过抛售来实现这些亏损。
And I think if you've got stocks, you know, still beaten down by the war by then, Investors are gonna be pretty reluctant to crystallize those losses by selling.
就像，我不知道。
Like, I don't know.
我认为，如此大规模的资金流动可能会对一些更流动的市场产生一些意想不到的连锁影响。
I think you could actually this is a big enough movement of money that you might see some weird knock on effects in kind of more liquid markets.
如果我是赌徒的话，我会密切关注国债。
You know, I'd have my eye on treasuries if I were I were a betting woman.
我的意思是，这规模太大了。
So, I mean, it's giant.
但我认为这是好事。
But I think it's I think it is good.
我认为，以Anthropic为例，它现在价值多少？
I think that, you know, anthropic, just to take an example, is as valuable at what is it?
它现在四五岁了，价值相当于谷歌15岁时、亚马逊25岁时的水平。
Four or five years old now as Google was when it was 15, as Amazon was when it was 25.
而它的全部财富都集中在私人市场，掌握在相对少数人手中，而谷歌上市时市值大约是200亿美元。
And all of its wealth has been created in the private market in in the hands of a relatively small number of people, whereas 99% I think Google was worth $20,000,000,000 when it went public.
因此，它4万亿美元价值中的99%都流向了公众手中。
So 99% of its $4,000,000,000,000 of value has accrued in public hands.
我认为这总体上是好事，回到我们之前讨论的内容。
And I think that is generally good to go back to what we were talking about earlier.
人工智能正威胁着让贫富差距急剧扩大。
There is like a the the wealth gap that AI is threatening to just massively explode.
我认为这是一个需要密切关注的危险政治因素。
I think it's like a really dangerous political force to keep an eye on.
因此，我鼓励这些公司上市。
And so, like, I would encourage these these companies to go public.
是的。
Yeah.
这将使SpaceX的规模超过标普500指数中仅有的五家公司，再次根据彭博社的数据。
So this would put SpaceX at bigger than this is, again, according to Bloomberg, all but five of the companies in the S and P 500.
因此，只有英伟达、苹果、字母表、微软和亚马逊会比它更大。
So only Nvidia, Apple, Alphabet, Microsoft, and Amazon would be bigger than it.
它将超过Meta和特斯拉。
It would be bigger than Meta and Tesla.
顺便说一下，我认为谈到资金可能流动的去向，很多特斯拉股东可能会说：‘如果我投资是为了押注马斯克的话。’
And by the way, I think speaking of places that we could see people move money from and to, I think a lot of, Tesla shareholders might say, oh, if I I'm doing this to bet on Musk.
如果我能入手SpaceX，它的前景在我看来更清晰——特斯拉现在是押注Optimus，而SpaceX则是押注太空和AI，我可能会卖出一些特斯拉股票，转而买入SpaceX，这会非常有意思。
If I could get a hold of SpaceX, which has, like, a clear I think a clearer trajectory, like Tesla right now is a bet on Optimus and SpaceX is, like, a bet on space and AI, I might just sell some shares from Tesla and buy in SpaceX, and that could be very interesting.
我觉得这是对的。
I think that's right.
不过我也想说，我认为他们让SpaceX上市的一个被低估的原因，最终是为了将其与特斯拉合并。
Though I would also say, I think one of the underappreciated reasons that that they are taking SpaceX public is to merge it with Tesla ultimately.
对吧？
Right?
就像埃隆·马斯克一直在整合他的商业帝国。
Like, Elon Musk has been collapsing his his empire.
从实际角度来看，一家私营公司，即使规模很大，也很难收购一家大型上市公司。
And just as a practical matter, it's pretty hard for a private company, even a big one, to buy a big public company.
这根本行不通。
It just doesn't work that well.
你得想出很多对策。
You gotta come up with a lot of catch.
你不能用你的股票。
You can't use your stock.
所以我假设特斯拉最终会被合并。
And so I I would assume that Tesla gets merged in at some point.
那么问题就变成了，你可以算一算，怎样以最便宜的方式获得合并后的特斯拉和SpaceX的股份？
And so then it's just a question of, like, you can do the math and what's the sort of cheapest way to to get yourself a share in the combined Tesla SpaceX?
我不知道。
I don't know.
这非常有趣。
That's very interesting.
是的。
Yeah.
也许这会发生。
Maybe that will happen.
那会很疯狂。
That will be crazy.
一个超级马斯克公司。
One super Musk company.
我觉得这就是发展方向。
I think that's where it's headed.
是的。
Yeah.
他以前就做过类似的事。
And he's done this before.
你知道，我们当初收购SolarCity时，和特斯拉合并，那是一笔非常复杂的交易，当时我们有个说法，认为他们会把太阳能板装到车上，但其实他们根本没怎么实现。
You know, we bought SolarCity with Tesla, sort of famously complicated deal, and had some theory of the case that they were gonna put solar panels on the cars, and they they never really did.
但没错，他一直在做的是xAI、x、SpaceX这些事情，显然就是这个方向。
But but, yeah, it's clearly what he's been up to with x AI and x and and SpaceX and all that.
好的。
Okay.
我想快速谈谈我今天在《Big Technology》上发表的报道：根据我看到的一些内部通讯，我们有消息称伊朗已经袭击了巴林和迪拜地区属于亚马逊的多个站点，导致这些地区的几个可用区完全瘫痪，并且预计会长时间无法恢复。
I want to talk quickly about the story that I just published today in Big Technology, and that is that we have some reporting that Iran has hit enough sites in Bahrain and Dubai that belong to Amazon that there are a couple of availability zones in those regions that are hard down according to some internal communication that I was able to see, and that they're gonna be unavailable for an extended period of time.
这是来自我获取到的内部备忘录的内容。
So this is from the internal memo that I was able to get ahold of.
这两个地区的服务仍持续受损，各服务不应期望能维持正常的冗余性和弹性水平。
The two regions continue to be impaired, and services should not expect to be operating with normal levels of redundancy and resiliency.
我们正在积极为该地区的客户释放和预留尽可能多的容量，各服务应缩减至支持客户迁移所需的最小规模。
We are actively working to free and reserve as much capacity as possible in the region for customers, and services should be scaled to the minimal footprint required to support customer migration.
因此，这些站点不仅遭到袭击，公司还要求亚马逊员工降低对这些地区的优先级。
So not only have these sites been hit, but the company is asking employees with in AWS to deprioritize them.
我觉得这很有趣。
And I think this is interesting.
也就是说，我们即将进入战争的第六周，而伊朗革命卫队现在正威胁要攻击其他多家美国科技巨头，包括微软、谷歌、苹果、英特尔以及其他多家公司。
I mean, we're we're about to enter week six of the war, and now the IRGC is threatening multiple other US tech giants including Microsoft, Google, Apple, Intel, and a bunch of others.
如果这场战争持续下去，我们可能会看到海湾地区大量大型科技基础设施遭到破坏——这些设施原本是因当地拥有资金和能源优势而建立的，且已投入了大量投资。
So if this war continues to keep going, we could actually see a lot of big tech infrastructure in The Gulf, which was obviously built because there's access to money, access to energy there, and big investments have been made there.
我们可能会看到它成为目标。
We could see it targeted.
是的。
Yeah.
我想问你一个问题。
Do you think I mean, I'll ask you a question.
你一直报道这些公司。
Like, you you cover these companies.
你觉得这场战争让他们变得如何？
Do you think this war has made them?
我可以从华尔街的角度说说我的看法，但你觉得这场战争有没有让他们重新思考这种大规模涌入该地区的做法？该地区过去五到十年一直动荡不安，却建了很多豪华酒店，吸引了大量资金，让人误以为它比实际更安全。
I'll give you my thoughts on the Wall Street side, but do you think this has made them rethink this massive rush into a region that has just been volatile and violent forever in the last five or ten years, built a lot of fancy hotels and had a lot of money and sort of people into thinking it was safer than maybe it is?
我想说一点。
I will say this.
我认为商界有很多乐观主义者。
I think there's a lot of optimists out there in the business community.
所以我认为他们的看法是，希望这场战争是最近一段时间内的最后一场，不会持续下去。
And so my read on their perspective is they are hoping that this is going to be the last war for a while, and it's not something that will continue to go on.
基本上，他们把这场战争看作是终结战争的战争，你几乎必须用这种乐观的视角来看待，因为你知道，想想我们现在从人工智能实验室听到的消息吧？
Basically, they're viewing it as a war to finish wars, and you almost have to view it through this optimistic lens because, you know, think about what we're hearing now from the AI labs, right?
他们基本上已经耗尽了在西方世界能获得的所有资金，现在他们就像是在首次公开募股前的最终Boss——海湾国家，之后才会进入公开市场。
Like, they have gone through basically all the funding they can get in the Western world, and now they're like sort of at the final boss before the IPO, which is The Gulf States, and then they're gonna go to the public markets.
因此，这是他们最后几个可用的资本来源之一。我想，我们之前也讨论过，硅谷在看待某些情况时常常戴着玫瑰色眼镜，这可能就是其中之一。
So it's like one of the last remaining sources of capital for them, And I think like we talked a little bit earlier about how Silicon Valley can sort of have this, you know, rose colored glasses on situations, and and this might be one of them.
所以，这就是我的看法。
So that that would be my perspective on it.
我的意思是，当风投们全都是一群充满梦想、向上乐观的人时，会发生什么。
I mean, is what happens when VCs are all like up into the right dreamers.
我不知道，说什么‘终结战争的战争’，但第一次世界大战时，每个人也是这么说的。
I don't know, said war to end wars, I just like, that is what everyone said about World War I too.
所以我不确定，这让我感觉很不好。
So I don't know, that gives me bad vibes.
这很有趣，因为华尔街也像大型科技公司一样，迅速涌入海湾地区，原因相似。
It's interesting because Wall Street has kind of made the same rush into The Gulf that Big Tech has for similar reasons.
我不知道。
And I don't know.
我的意思是，这从根本上说是一个关于人的生意。
I mean, I think it is a people business, fundamentally.
你在谈论的是有形资产。
You're talking about hard assets.
前几天我刚和艾奥尼克（Aon）这家保险经纪公司的一位高管聊过。
And actually, was just talking the other day to the guy who runs a big business at Aon, the insurance brokerage.
我当时在想，购买战争保险的成本发生了什么变化？这种保险能保障这些资产免受战争和恐怖主义破坏。
I was like, I wonder what's happened to the cost of literally, can buy war insurance, which, you know, insures these assets against war and terrorism damage.
他说，价格在几周内上涨了1900%，而单个资产的可获得保险额度从30亿美元骤降至1亿美元。
And he said prices have gone up 1900%, and the available coverage for any single asset has gone from 3,000,000,000 to a $100,000,000 in the matter of a couple weeks.
是的。
Yeah.
以前你只需要支付50个基点或25个基点的保费，现在 upfront 的保费已经涨到了5%。
You used to be able to get it for 50 basis or 25 basis points, now it's costing 5% in premiums upfront.
所以，那些以承保风险为生的人认为，这些资产发生糟糕事情的风险非常高。
So, you know, the people who underwrite risk for a living think there's a lot of risk of bad stuff happening to these assets.
但在人员方面，有一段时间，像在海湾地区工作，我这里用引号说一下。
But then on the people side, you know, for a while, like, being in The Gulf was I I say this in air quotes.
那曾经是个艰苦的外派岗位。
Like, it was a hardship posting.
对吧？
Right?
你会在海外派遣包里拿到更多钱。
You would get more money in an expat package.
除了在几年内不用缴税之外，高盛还会把你派到那里，那确实是个艰苦的外派岗位。
And on top of getting to pay no taxes for a couple years, your Goldman Sachs would send you out there, and and it was a it was a hardship posting.
我觉得它正处在不再被视为那样的边缘。
And I think it was just on the verge of becoming not that.
它当时正处在从艰苦外派地转变为这些跨国公司另一个全球办公室的边缘。
It was just on the verge of being another global office for these big multinational companies.
我不禁想知道，那些在战争期间离开的外派人员，是否想回去，以及回去要付出什么代价。
And I wonder whether, you know, the people who have the expats who've left during the war, like, want to go back and at what cost.
好的。
All right.
我接下来要提两个跟进问题。
I'm going to do two follow ups here.
首先，你认为无法获得保障会对公司在此地建设的能力造成什么影响？
First of all, what do you think the lack of the ability to ensure will do to companies' ability to build there?
我猜想科技巨头无论怎样都能继续建设，因为它们资金雄厚，但谁知道呢？
I would imagine that the tech giants will be able to keep building no matter what, because they have so much money, but who knows?
然后，如果人们不想回去，会发生什么？
And then, yeah, what happens if the people don't want to go back?
关于第一个问题，我会持相反的观点。
On the first one, I would take the other side of that.
他们有很多钱，但就像你如果没有房屋保险就无法获得房屋抵押贷款一样，这是银行的要求，因为抵押品是房子，没有保险政策就无法为建设这些数据中心获得融资。
They have a lot of money, but in the same way that you can't get a mortgage on your house if you don't have home insurance, like that's just a requirement, the bank requires it because the collateral is the house, you cannot get financing to build these data centers without an insurance policy.
正如你所说，这些公司有很多钱，但看看Meta在路易斯安那州进行的大型数据项目吧。
As you say, these companies have a lot of money, but they've required Look at what Meta was doing in Louisiana as that big data project.
他们从蓝猫和PIMCO等机构获得了大量私人资金，因为不想把这些负担记在资产负债表上。
And they took a lot of private money from Blue Owl and PIMCO and the market because they don't want it on their balance sheet.
所以，如果你无法为这些东西投保，就无法建设它们。
So you can't insure this stuff, you can't build it.
好的。
Okay.
那人们呢？
And the people?
哦。
Oh.
我认为人们会回去的。
I think the people will go back.
我认为他们只是被野心和金钱驱动。
I think they're just People are motivated by ambition and money.
他们会
They will
回去的。
go back.
我觉得这是对的。
I think that's right.
在海湾地区过一种免税的生活确实有吸引力。
And there is something to living a tax free lifestyle in The Gulf that is Right.
对合适的人而言非常有吸引力。
Is quite attractive to the right kind of person.
但我确实认为，这在一定程度上延缓了该地区全面融入全球高端金融体系的进程。
But but I do think it probably sets back the kind of full inclusion of that region into into the the global sort of high finance set, you know, a little bit for sure.
当然。
Definitely.
你去过那里吗？
Have you been out there?
你以前在那里待过吗？
Have you spent time out there before?
说来有趣。
You know, it's funny.
我去过。
I have.
去年我去了沙特阿拉伯。
I was in Saudi Arabia last year.
十二月底的时候我去了阿布扎比。
Was in Abu Dhabi the end of in December.
我记得当时飞抵阿布扎比，不知道为什么我之前没想到，飞机右侧是阿拉伯半岛和阿联酋。
And I remember flying in, and I don't know why it hadn't occurred to me, but you're flying into Abu Dhabi and at the right side of the plane is is the Arabian Peninsula and and The Emirates.
而左侧，隔着海湾就是伊朗。
And then the left side right across the across the Gulf is Iran.
我的意思是，当你去那里时，很容易忘记，因为你在阿布扎比，那里有购物中心，还有四季酒店。
I mean, is it is when you go, it's it's easy to forget because you're in Abu Dhabi and there's a Galleria and there's a Four Seasons.
感觉非常西方化，到处都是穿着西装的富人走来走去。
Like, it feels very Western and there's lot of rich people in suits walking around.
但我几周前才听说了这件事。
But it's like I heard about this a couple of weeks ago.
这就像一栋好房子坐落在一个糟糕的街区。
It is a nice house in a bad neighborhood.
是的，我今年夏天在卡塔尔待过。
Yeah, I was in Qatar over the summer.
我想我之前在节目里说过，可能不算新话题，但那里对我来说实在太热了。
And, I think I've said this on the show, so it might not be new ground, but just a little too hot for me.
我的意思是，也许是我去的时候正好是那个季节，但我真的不喜欢高温。
I mean, maybe it was the season that I was there for, but I didn't really like heat.
我记得晚上的温度都超过华氏100度。
Was, I think, above 100 degrees Fahrenheit at night.
我以前从未见过这种景象，除非我再回多哈，否则我觉得我再也不会看到了。
And I'd never seen this before, and I don't think I'll ever see this again unless I go back to Doha.
在多哈市中心，人行道是空调的，这意味着你走在人行道上时，会感受到一股冷风从脚下吹来，这是因为人行道下方有通风口不断吹出冷气，把原本100华氏度的夜晚变得像80华氏度一样舒适。
In Downtown Doha, the sidewalks are air conditioned, which means that as you walk down the sidewalks, you feel a cold breeze coming up, and that is because there are vents underneath the sidewalk blowing cold air and turning what is a 100 degree night into something that feels more like an 80 degree night.
不，但我觉得这有点像是对整个地方的隐喻。
No, but I think that's a little bit of a metaphor for the whole place.
对吧？
Right?
就像这是一个结构上存在问题的地区，但他们通过工程手段让它感觉比实际应该更舒适一些。
Like, this is a structurally problematic region of the world that they have engineered into feeling a little more comfortable than it probably ought to.
这个比喻很贴切。
That is that is a good metaphor.
好吧。
Okay.
在我们结束之前，我想聊的最后一件事是你提到的私人信贷，还有Meta如何有效利用Blue Owl来建设其数据中心，我们之前在节目中也聊过一些，不过我们可以再深入谈谈。
So last thing that I want to talk about before we go, you talked about private credit, you talked about how Meta, used Blue Owl effectively to build some of its data centers, and things have we've talked about this on the show a bit, but we can talk about a little bit more.
这些私人信贷公司的情况正变得越来越棘手。
Things are getting a little hairy for these private credit companies.
这周你的故事就是这个。
This is your story this week.
随着私人信贷市场的焦虑持续，Blue Owl面临赎回潮。
Blue Owl faces withdrawal flood as private credit jitters persist.
随着私人信贷市场的焦虑不断加剧，投资者纷纷撤出Blue Owl旗舰和科技导向的信贷基金。
Inventor investors rushed out of Blue Owl's flagship and technology focused credit funds as, as jitters continue to sharpen in private credit.
其规模最大的基金——一个360亿美元的企业贷款基金——收到了相当于其资产22%的赎回请求，而规模较小的60亿美元基金则出现了41%的撤资率，Blue Owl对这两只基金的赎回都设定了5%的上限。
Its largest fund, a 36,000,000,000 pot of corporate loans, received redemption requests for 22% of its assets, and a smaller $6,000,000,000 fund, saw 41% withdrawal rates, and Blue Owl capped both of those had 5% withdrawals.
所以，这实际上意味着，想从这些私人信贷公司取出资金的人，远多于它们愿意支付的金额。
So basically, this means a lot more people, wanna go and get their money out of these, these private credit companies, than it's willing to distribute.
大家都在问，这是否意味着一场大崩溃的开始？这场崩溃是否会拖垮整个经济，因为它们的投资——比如AI基础设施——可能无法带来回报。
And is this the big mean, everyone's wondering, is this the beginning of a crash that's going to take down the whole economy because they made these, you know, investments that might not pay off an AI infrastructure.
真相到底是什么？
What's the truth?
是的
Yeah.
好的
All right.
我认为你需要深入分析一下
I think you have to unpack it
一点。
a bit.
哦，明白了。
Oh, okay.
我们必须深入分析。
We have to unpack.
好的。
Alright.
我们实际上在这些赎回数据中看到的，是一场缓慢进行的银行挤兑。
What what we are actually seeing in in those redemption numbers, we are seeing a slow moving run on a bank.
所以回到几年前，你还记得硅谷银行在一个周末因为所有人都提取存款而倒闭吗？
So to go back a couple years, remember when Silicon Valley Bank, like, failed over a weekend because everyone took their deposits out?
这基本上就是私人信贷行业目前正在发生的情况。
That is basically what is happening in the private credit industry.
由于这些基金的结构方式，你提到它们设置了5%的限制，这没错，但这就是它们的设计目的。
Because of the way these funds are set up, you noted that they gated at 5%, and that is true, but that's what they're supposed to do.
这些基金被要求每季度提供回购机会。
These funds are required to make quarterly offers to buy back.
所以这些是非上市基金。
So these are non traded funds.
你持有其中的一份份额，但并不像在交易所那样可以自由交易。
So you own a share in it, but it's not like you can trade it on an exchange.
但每个季度，你都可以去找Blue Owl，说我想赎回我的份额，只要整体上不超过5%的人想这么做，他们就会给你退款。
But every quarter, you can go to Blue Owl and say, I'd like my money back for the share, and they'll give it to you as long as in aggregate, not more than 5% of people of you know, wanna do that.
这就是它们的设计初衷，因为记住，它们持有的是难以买卖的贷款。
So that's that's what they're designed to do because remember, they're holding loans that they can't buy and sell very easily.
所以他们会保留一些现金来应对这些提款请求，当一切正常时，这套机制是有效的。
So they keep some cash on hand to to fund those those withdrawal requests, and whenever things working fine, it's working fine.
但显然，这套机制的设计就是在恐慌发生时失效，而我们现在正是处于这种状况。
But obviously, this is designed to malfunction when there's panic, which is where we are now.
所以目前来看，这是一场银行挤兑，属于流动性问题。
So for the moment, it's a bank run, which is a liquidity problem.
他们持有的是流动性差的资产。
They have illiquid assets.
人们想要拿回自己的钱。
People want their money back.
这不是一个能轻易解决的问题。
That is not a fixable problem.
所以这方面的情况就是如此。
So that's where we are on that side.
另一个你所关注的问题是，是否存在信用质量的问题？
The other question which you're getting at is, is there a credit quality problem?
这些贷款本身有问题吗？
Are the loans themselves bad?
是的。
And Yeah.
为什么人们会这样挤兑？
Why are people making this run?
这只是行为问题吗？
Is it just behavior?
谈谈这个。
Talk about that.
这部分是行为问题。
It is partly behavior.
部分原因是看。
It's partly look.
比如，在蓝阿尔格成立的前六年里，它从专业的机构投资者那里筹集资金。
Like, for the first six years it was around, Blue Alge raised money from sophisticated institutional investors.
而且这不仅仅是Blue Alge的问题。
And it's not just Blue Alge.
还有黑石、KKR、阿波罗以及所有这些哈佛商学院的大型公司。
It's Blackstone and KKR and Apollo and all of these HBS big firms.
然后他们说，好吧，我们在那里有点资金用尽了，但还有很多个人投资者。
And then they said, well, we're kinda tapped out there, but there's a whole bunch of individual people.
如果我们也把它们卖给他们呢？
What if we sold it to them too?
因为人们的日常生活方式与机构不同，你必须为他们提供季度性的流动性。
And because people live their lives differently than institutions, you have to give them this quarterly liquidity.
所以这在某种程度上是他们自身行为和渴望持续增长、寻找新的融资渠道的结果。
So this is sort of like they were is a consequence of their own actions and their own desire to keep growing and to find new new places to raise money.
贷款本身是否糟糕，这是一个合理的问题。
The question of whether the loans themselves suck, that is a fair question.
我们其实并不知道。
We don't actually know.
其中很多确实来自2021年和2022年这一轮贷款。
A lot of them are certainly, there's this vintage of lending from, call it, 2021 and 2022.
那里可能有一些非常糟糕的贷款。
Probably some real bad loans there.
他们是不是筹了太多钱？
Did they raise too much money?
这可以说是软件行业的巅峰时期。
This was the the sort of peak software.
他们当时在发放这些贷款。
They were making these loans.
我不知道你是否在节目中提到过一种叫ARR贷款的东西。
I don't know if you've talked about them on the show yet called ARR loans.
通常，银行会根据公司的盈利能力来评估其借款能力。
Usually, banks underwrite a company's ability to borrow based on its profits.
但科技公司，尤其是快速增长的公司，并没有利润。
But tech companies, especially fast growing ones, don't have profits.
他们有收入。
They have revenue.
所以他们以年度经常性收入为依据。
And so they based it on annual recurring revenue.
说，这家公司的营收非常高。
Said, well, this company's making a lot of top line money.
让我们把这作为衡量标准吧。
Let's lend them let's let's make that the metric.
让我们改变，你知道的，上千年来的借贷经济模式。
Let's change, you know, ten thousand years of lending economics.
把这作为衡量标准。
Make that the metrics.
然后，大概三年后，当他们开始盈利时，我们就切换一下，让贷款开始像传统贷款那样运作。
And then, like, three years when they've got a profit, then we'll kind of flip we'll flip a switch and the loan will start behaving like a traditional loan.
结果到了三年时，这些公司没有一家盈利，他们只好说：‘哎呀。’
And then they got to three years and none of those companies had profits and they're like, oops.
所以很明显，这些贷款的信用审核相当糟糕。
So obviously, the credit underwriting was pretty bad on a bunch of that stuff.
但你知道，Blue Owl 几周前成功出售了一些资产。
But, know, Blue Owl was able to sell a couple weeks ago.
他们在另一个基金里遇到过类似问题，最终卖出了大约15亿美元的贷款，价格高达面值的99.7美分。
They had this problem in another fund, and they were able to sell, I think, like a billion and a half of loans at 99.7¢ on the dollar.
所以我不确定。
So I don't know.
我觉得大部分贷款可能都没问题。
I think, like, most of the loans are probably fine.
顺便说一句，我相信你的听众都明白资本结构，这些贷款属于有担保的优先债权，是任何企业财务结构中最安全的部分。
And by the way, not to like I'm sure your listeners understand a capital stack, like, the credit, this is senior secured credit, which is the safest place to be in any kind of corporate financial structure.
如果你觉得这里有问题，那下面的股权——主要是私募股权——几乎就等于零了。
If you think there's a problem there, the equity underneath it, whether that's it is mostly private equity, is like a zero.
所以这很奇怪，想象一下洪水来袭，顶层的人反而比底层的人更早恐慌，但现实恰恰就是这样。
So it is weird, you know, if you imagine this is like a a floodwaters coming in, it's weird to have the people on the Top Floor panicking first before the people on the Ground Floor, and yet that's what's happening.
这让我觉得，当前的情况更多是情绪性和技术性因素驱动的，而不是因为这些贷款本身有问题，但我并不认为这才是当前状况的真正原因。
And so that makes me think that this is sort of more emotional and technical and less about the the loans themselves might be bad, but I don't actually think that that is what is informing what's happening now.
好的。
Okay.
安德鲁·奥索金最近也谈过类似的观点，说私募股权会比私募信贷更早出问题。
Andrew Osorkin was recently on talking about the similar thing that private equity would have the problem before private credit.
让我试着理清一下问题出在哪里，你告诉我我理解得有多偏差，因为我肯定有偏差——是不是私募股权在软件公司上投了很多钱？
So let me just see if I can understand where the issue would be and you tell me how off I am because I'm gonna be off Is it that private equity has invested a lot of money in software companies.
而现在出现了一种叫人工智能的技术，让这些软件公司的价值以及它们的经常性收入（ARR）受到了质疑。
And now there's this thing that's called AI that sort of throws the value of those software companies and those companies ARR into question.
因此，那些押注于这些软件公司的私募股权公司开始出现问题。
And so therefore, the private equity companies that have bet in these soft bet on these software companies begin to fail.
当它们失败时，私募信贷也随之崩溃，这才引发了这种连锁式的经济灾难。
And then when they fail, private credit starts to fail, and that's how you get this case cascading economic disaster.
你担心的是这个吗
Is that the worry you
我是说，某种程度上是这样。
I mean, sort of.
这就是金融运作的方式。
That is how finance works.
对吧？
Right?
也就是说，如果一家公司陷入困境，最初的损失由股东承担。
Which is that if a company gets in trouble, the first losses are borne by the shareholders.
对吧？
Right?
然后债权人拿走剩下的部分。
And then the creditors get whatever is left.
所以从实际角度来看，这是真的。
So as a practical matter, that's true.
这些公司的收入会消失，利润也会消失，它们将无法偿还贷款，贷款方会迫使它们破产，股权归零。
These companies, their revenue would go away, their profits would go away, they would be unable to repay their loans, their lenders would force them into bankruptcy, the equity is zeroed out.
所以你需要关注像Toma Bravo、Vista这样的公司。
And so that's where you would look at firms like Toma Bravo, Vista.
对吧？
Right?
这些其实是典型的科技类公司，它们是大型私募股权公司，基本上只持有企业软件公司。
These, like, really tech like, these are big PE firms that only own enterprise software companies, basically.
它们拥有科技类企业。
They own tech businesses.
因此，问题就出在这里，债权人会拿到剩余的部分，而其中一些贷款会变成我们所说的‘减值贷款’，它们可能会蒙受损失。
So that's where you're actually looking for the problem, and then the creditors will get what's left, and some of those loans be what we call impaired, and they might lose some money.
也许它们只能收回每借出一美元中的90美分，但那些位于这些债权下方、价值数千亿美元的股权却变得一文不值。
Maybe they only get 90¢ back on the dollar that they lent, but that equity, which is hundreds of billions of dollars of equity sitting under sitting junior to this credit is worthless.
人们没有更多讨论这个问题的原因，我想说的是，我三个月前就开始写这个话题了，当时我就在想，为什么没有恐慌呢？
The reason that people aren't talking about this as much I will say I I started writing about this, like, three months ago, and I was like, where is the panic there?
答案是，这些基金其实并没有像私人信贷那样向个人投资者大力推广。
And the answer is that it's actually they just haven't marketed those funds to individual investors the same way that private credit has.
所以
So the
为什么股权会变得一文不值？
twitchiest Why people does the equity become worthless?
因为我想想办法解释清楚，又不至于太枯燥。
Because it's I'm trying to think how to explain this in a way that isn't gonna be super boring.
用金融术语来说，它是次级的。
It is in finance speak, it junior.
它是一种次级债权。
It is a junior claim.
如果你把建房子打个比方，债权人的基础是私募信贷，而股权则在上面。
So if you think about building a house the creditors foundation is private credit equity.
没错。
Gets Correct.
当洪水来临时，私募信贷就像住在顶层公寓里。
The private credit is in the penthouse when the flood comes in.
股权位于底层。
The equity is on the Ground Floor.
所以它们首先承担损失，然后依次是优先股权，再是次级债权。
So they're they just bear the losses first, and then it goes up to the, the preferred equity and then the junior credit.
但私人信贷属于优先债权，通常由资产作为担保。
But the private credit, these are senior claims that are secured mostly by assets.
这些人的投资至少还会剩下一些东西。
There's going to be something left for these guys.
软件公司倒闭时，私人信贷公司会优先获得剩余资产，是的。
Software company fails, actually the credit private credit companies get whatever's left of the assets first Yeah.
然后是私人
And then private
没错。
Exactly.
如果他们亏了钱，股权就血本无归了。
If they've lost money, the equity has lost all of their money.
有趣的是，私人信贷正在资助这些人工智能基础设施的建设，而这些人工智能基础设施随后实际上在压低私人股权公司所投资的那些公司的价值。
Isn't it interesting that private credit is funding the establishment of these AI infrastructure builds as those AI infrastructure builds then go to effectively undercut all these companies that private equity has invested
我们时时刻刻都在资助自己的灭亡。
We're all funding our own demise at all times.
我想指出，这是一个很好的技术性观点。
I would note this is a good, like, technical point.
蓝鹰公司中负责数据中心业务的部分，其实是他们收购的一家与信贷业务完全无关的公司。
The part of Blue Owl that is doing the data center stuff is like a business that they bought that's actually unrelated to their credit business.
但没错，黑石公司很可能就是一个很好的例子。
But yes, you know, Blackstone's probably a good example of this.
他们拥有一个大型的数据中心业务。
They got a big data center business.
他们拥有QTS以及一大批数据中心，这些设施正在创造技术，而这些技术正在严重冲击他们私人股权组合中的公司，进而又冲击着其他人的私人信贷组合。
They have QTS and a bunch of these data centers that are, you know, creating the technology that is sort of wreaking havoc in their private equity book that is then wreaking havoc in somebody else's private credit book.
所以
So
难以置信。
Unbelievable.
我们都在资助自己的灭亡。
We are all funding our own demise.
是的。
Yes.
除了我们这些独立媒体人。
Except us in independent media.
我们正在做上帝的工作。
We're doing the Lord's work.
你和
You and
我，亚历克斯。
me, Alex.
没错。
That's right.
是的。
Yeah.
当然，对于任何在重大科技领域做出重大财务决策的人来说，我们只会使用ARR，而不会用其他任何指标。
Well, of course, for anyone who's making a big finance decision about big technology, we'll use ARR and not anything else.
对吧？
Right?
来做这些决策。
To make those decisions.
显而易见。
Obviously.
是的。
Yeah.
但我并不想回到我们最初的话题，比如，这个播客到底卖什么？
But I didn't to go back to where we started, like, what do sell this podcast?
TBPN 刚刚为你提供了一个相当不错的参考案例。
Where TBPN just put a pretty good comp out there for you.
所以，显然从收入角度来看，我永远不会问你的利润。
So but obviously, on a revenue basis, I would never ask you your profits.
对于那些好奇的人，Big Technology并不出售。
Big Technology for those who are wondering, we're not for sale.
我们不会卖给OpenAI。
We're not going to OpenAI.
我们不会卖给Anthropic。
We're not going to Anthropic.
我们不会卖给XAI。
We're not going to XAI.
我们不会卖给Salesforce。
We're not going to Salesforce.
我们不会这么做。
We're not doing that.
这个节目，我写的这份通讯，将保持独立。
This, this show, this newsletter that I write is staying independent.
我们不会出售，我会继续做下去，再干个二三十年就退休。
We're not selling, and I'm gonna just keep doing this for next twenty or thirty years and then retire.
所以继续支持我们吧。
So stick with Hold us
坚守你的原则。
on to your integrity.
通过
With
没错。
That's right.
我的意思是，我就直说了。
I mean, also I'll just say it.
我喜欢做这件事。
I enjoy doing this.
我不想为别人工作。
I don't want to work for somebody else.
就像我一开始说的，你一旦进入一家公司，你就成了这家公司的内容营销部门。
Like I said at the beginning, you know, the second you come inside a company, you're now that company's content marketing arm.
你不再做你以前做的事了。
You're not doing what you did.
你在做完全不同的事情。
You're doing something completely different.
而这并不是我想做的事。
And that's that's not what I'm interested in doing.
如果你想在节目中做广告，那就是另一回事了。
If you want to advertise on the show, that's a different story.
你可以在这里做广告。
You can advertise for you can advertise here.
我们很乐意收到你的消息。
We're happy to hear from you.
查看节目说明中的邮箱地址，但不，这不对外出售。
Check out the email address in the show notes, but but no, not for sale.
好的。
Alright.
知道了，很好。
Good to know.
这个播客叫《复利效应》。
The podcast is called Compound Interest.
我们的嘉宾是莉兹·霍夫曼，莉兹。
Liz Hoffman has been our guest, Liz.
欢迎。
Welcome.
这是你第一次上我们的节目。
First time having you on the show.
这次对话真的很有趣。
This was really fun.
谢谢你来参加。
Thank you for coming on.
这是我的荣幸。
It was a pleasure.
随时欢迎。
Anytime.
谢谢你邀请我，亚历克斯。
Thanks for having me, Alex.
好了，各位。
Alright, everybody.
非常感谢你们的收听和观看。
Thank you so much for listening and watching.
我们下次再见，欢迎收听《大科技播客》。
We'll see you next time on Big Technology Podcast.
关于 Bayt 播客
Bayt 提供中文+原文双语音频和字幕，帮助你打破语言障碍，轻松听懂全球优质播客。