Anthropic's $2T IPO, Zuck's AI Manifesto, Nvidia's $500B AI Bet, Grok's Comeback

频道: All-In with Chamath, Jason, Sacks & Friedberg
视频: https://allinchamathjason.libsyn.com/anthropics-2t-ipo-zucks-ai-manifesto-nvidias-500b-ai-bet-groks-comeback
原文语言: en
统计: 共 117 轮 · Jason Calacanis 34 · Gavin Baker 34 · David Sacks 26

[0:00] Jason Calacanis

All right, everybody, welcome back to the All In Podcast, the number one podcast in the world.David Sachs and Gavin Baker are with us this week.We got a short crew, but a long docket, and we are going to rock it.How are you doing, Gavin? Nice to see you.Fantastic.You're still in that SpaceX afterglow?

好，各位，欢迎回到 All-In 播客，全世界排名第一的播客。本周和我们在一起的是 David Sacks 和 Gavin Baker。人手不多，但议题单很长，我们照样把它拿下。Gavin，你怎么样？很高兴见到你。好极了。你还沉浸在 SpaceX 那波余温里吗？

[0:19] Gavin Baker

This is the biggest win of your career, correct?You know, SpaceX was a magical moment.And there's only one SpaceX, but to quote Bill Belichick, you know, it's on to Cincinnati.Got it.And there's a lot I think ahead for SpaceX, so the SpaceX story isn't over.Yeah, and you're only as good as your last investment in our business, I guess.Everybody's like, what now? Sachs, you're back. How are you doing, brother?

这是你职业生涯最大的一笔胜仗，对吧？你知道，SpaceX 确实是个神奇的时刻。世上只有一个 SpaceX，但套用 Bill Belichick 的那句话——「下一站辛辛那提」。明白。而且我觉得 SpaceX 前面还有很多故事，所以 SpaceX 这一章还没完。是啊，在我们这行，你的水平就只等于你最近一笔投资的水平。所有人都在问：那接下来呢？Sacks，你回来了，兄弟，最近怎么样？

[0:47] Jason Calacanis

How's the summer wrapping up for you?Good. I'm looking forward to talking to Gavin.Oh, fantastic.Well, I'll try to stay happy and not step on you.You missed a really great discussion we had on the all-in interview show with your guy, Rahm Emanuel.Rahm Emanuel.How'd that go? I heard, actually, it got kind of spicy, you know?

这个夏天收尾收得如何？挺好，我很期待跟 Gavin 聊。太好了。那我尽量保持愉快，不抢你的话。你错过了我们在 All-In 访谈节目里跟你的老熟人 Rahm Emanuel 的一场非常精彩的对话。Rahm Emanuel。聊得怎么样？我听说……场面还挺辣的？

[1:06] Jason Calacanis

It got super spicy.Rahm came in like full guns blazing, but it's a really great interview, and we'll take anybody who wants to do the interview show who's running in 2028.Did you catch it, Gavin, by chance? Did you catch it?

辣极了。Rahm 一上来就火力全开，但那是一场非常棒的访谈。任何打算 2028 年参选的人，我们都欢迎来上访谈节目。Gavin，你碰巧看了吗？看了吗？

[1:18] Jason Calacanis

I haven't watched it yet.Okay, got it in the queue? Beautiful.Of course.And based on the verdict on Rahm is that he's kind of a throwback to the Clinton-Obama DNC.I think if you were thinking of a Clinton moderate.Which means he has no chance whatsoever in the Mondani DSA Democrat Party.This is the great irony.And it's a very similar moment, I think, to what happened in your Republican Party where you had this sort of civil war internally, and I'm hoping the mods, the moderates, have a chance here.But, you know, who knows what's going to happen with these lunatic democratic socialists.We don't have too much of that.Did he have a take on that?

我还没看。好的，收进待看清单了？漂亮。当然。对 Rahm 的评判是：他有点像克林顿—奥巴马时代民主党全国委员会（DNC）的回归。如果你脑子里想的是一个克林顿式的温和派。那也就意味着，在 Mamdani 那种民主社会主义者（DSA）主导的民主党里，他一点机会都没有。这就是最大的讽刺。而且我觉得这和你们共和党当年发生的事非常像——党内打了一场内战。我倒是希望温和派这次能有机会。但谁知道这帮疯了的民主社会主义者会闹出什么。我们那边这种人倒不多。他对这个有什么看法吗？

[1:56] Jason Calacanis

Yeah, I mean, he...That they're taking over the Democrat Party?I'd say his...First of all, he's got incredible amount of experience having worked for Obama, Clinton, been mayor of Chicago, and ambassador to Japan.So he was awesome on foreign relations, etc.Yeah, and to your point, he thinks hard no on democratic socialists and the socialist policies are a road to nowhere.So he just went totally hard at it.As hard as you would, Sachs, in fact.I hope he has...Nothing would be greater than to have a moderate or more moderate candidates.I think we can all agree on that.So we'll see.We'll see.It's worth checking out.All right, let's get to the docket breaking news at the time of this recording.For those of you who don't know, we record on Thursdays, published on Fridays.Anthropic targeting a $2 trillion IPO.According to the Financial Times, this would obviously break the record-setting $1.75 trillion IPO by SpaceX.FT reported Wednesday night that investors expect the company to go public in October,which is like six to eight weeks away.This lines up with the Polymarket.The Sharps at Polymarket saying 80% chance of Anthropic IPO-ing this year.And Brad Gerstner, a fifth bestie, he also was signaling that this would happen sooner rather than later.

有啊，他……说他们正在接管民主党？我觉得他的……首先，他的经验多得惊人——给奥巴马干过、给克林顿干过、当过芝加哥市长、还当过驻日本大使。所以他讲外交那部分特别厉害。对，而且按你说的，他对民主社会主义者是坚决说不的，认为社会主义政策是一条死路。他就是硬刚到底。Sacks，你有多硬他就有多硬。我希望他……没有什么比出现一个温和派、或者更温和一些的候选人更好的了。这点我想我们都同意。所以走着瞧吧。走着瞧。值得一看。好，进正题——录制这期时的突发新闻。不知道的朋友说明一下：我们周四录，周五发。Anthropic 正在瞄准一个 2 万亿美元估值的 IPO。据《金融时报》报道，这显然会打破 SpaceX 创下的 1.75 万亿美元 IPO 纪录。FT 周三晚间报道称，投资人预计该公司会在 10 月上市，也就是六到八周之后。这跟 Polymarket 上的赔率对得上——Polymarket 上的老手给出 80% 的概率认为 Anthropic 今年会 IPO。而第五位「铁哥们」Brad Gerstner 此前也放过风，说这事会比预期来得早。

[3:24] Jason Calacanis

Annualized run rate.We talked about this, Sachs, a bunch.We'll end the year between $100 and $120 billion.This is an extraordinary ramp-up of revenue that we have never seen in Silicon Valley.At a $2 trillion valuation, obviously, you all can do the math, but it's 16 to 20 times sales,which is a fraction of what SpaceX and Palantir and some of these highly valued stocks went out at.So pretty interesting revenue growth over the last year.Anathropics still has AI's top model, but Grok, and we're going to get into it today, is catching up quickly.And OpenAI, not too shabby as well.Polymarket says 67% chance that at the end of the year, Anthropic will have the top model.Breaking on Thursday, Anthropics talks to buy an AI startup called Descartes for $6 billion.Their software lowers the cost of AI training and inference by making chips more efficient.And I'll stop here.Gavin, if you could comment, and I don't know if you're a shareholder, Anthropic, or not,so it'd be great to get that out of the way.But what's your take on just straight up the revenue ramp we're seeing here,and then what to expect from an IPO?

年化运营收入（annualized run rate）。Sacks，这个我们聊过好多次了。今年收尾会落在 1000 亿到 1200 亿美元之间。这是硅谷从来没见过的一种收入爬坡速度。按 2 万亿美元估值算，大家自己做除法——大约是 16 到 20 倍市销率（PS），这只是 SpaceX、Palantir 等一批高估值股票上市时倍数的一个零头。所以过去一年的收入增长相当有看头。Anthropic 仍然握着 AI 最强的模型，但 Grok——今天我们会展开讲——正在快速追上来。OpenAI 也不含糊。Polymarket 给出 67% 的概率，认为到年底 Anthropic 仍会拥有最强模型。周四还有突发消息：Anthropic 正在洽谈以 60 亿美元收购一家名叫 Descartes 的 AI 初创公司。他们的软件通过提升芯片效率来降低 AI 训练和推理的成本。我先说到这。Gavin，你能不能先讲讲——我也不知道你是不是 Anthropic 的股东，所以先把这个说清楚——但你对我们看到的这个收入爬坡直觉上怎么看？以及对这次 IPO 该有什么预期？

[4:40] Gavin Baker

So a few things.I'm not currently an Anthropic shareholder.I'd say the Anthropic SpaceX detente probably came a little too late.I don't generally invest in Elon's competitors.I don't think it's a wise course of action.That tracks.And the detente was obviously,Elon decided when he had so much extra compute with Colossus,that he would rent it to Dario and Anthropics.Absolutely.And I think he, almost probably more importantly,got comfortable with Dario's morals,and that Dario is a good person.And like all humans, to be human is to err.So maybe he's made some mistakes in the way he's handled government relations.But he got comfortable that his morals were in the good place.As Elon said, they don't set off my evil detector.And he has a pretty good evil detector.He famously cut his conversation with the SPF short.So a couple of things on Anthropic.The rumored numbers are exceptional.We've really never seen anything like this.I still think those moments in April and Maywere some of the most extraordinary moments in the history of capitalism.And what's wild to me is that probably on the margin,Anthropic is losing share to OpenAI, to open source, and to Grok.And they're still growing so fast.

几点。我目前不是 Anthropic 的股东。我会说，Anthropic 和 SpaceX 之间的和解（detente）大概来得有点晚了。我一般不投 Elon 的竞争对手，我觉得那不是明智之举。合理。而这次和解显然是因为，Elon 发现 Colossus 上有那么多富余算力，就决定把它租给 Dario 和 Anthropic。没错。而且我觉得，可能更重要的是，他对 Dario 的道德水准放心了，觉得 Dario 是个好人。跟所有人一样，是人就会犯错。所以也许他在处理政府关系的方式上犯过一些错。但 Elon 认可的是他的道德底子是正的。用 Elon 的话说，他们没有触发我的「作恶探测器」。而他的作恶探测器相当准——他有个著名的例子，跟 SBF 聊到一半就把对话掐了。所以关于 Anthropic 有几点。传出来的这些数字是极其罕见的，我们真的从没见过这种东西。我到现在仍然认为，4 月和 5 月的那几个瞬间，是资本主义历史上最不可思议的时刻之一。而让我觉得离谱的是，在边际上，Anthropic 大概率正在把份额输给 OpenAI、输给开源、输给 Grok，可它还是长得这么快。

[6:07] Jason Calacanis

The numbers are still exceptional.So I think that speaks to the strength of it.And why is that?This is an important thing for people to understand.The pie is getting ginormous.So even if on a percentage basis, Anthropic is losing on a percentage basis,the real number, and we talked about open source being dark tokens that aren't tracked anywhere,this is a major, major pie-growing moment.That's what you mean there, Gavin?

数字依然是罕见的。所以我觉得这恰恰说明了它的强度。为什么会这样？这一点很重要，大家要理解：蛋糕正在变得巨大无比。所以即便按百分比算 Anthropic 在丢份额，绝对数字——我们之前也聊过开源那部分是「暗 token」，哪儿都统计不到——这是一个非常非常大的「把饼做大」的时刻。Gavin，你说的是这个意思吧？

[6:39] Gavin Baker

That's the point.I think AI broadly accelerated in the month of July,which was kind of crazy to contrast with the public stock market action.Now we have a higher resolution view of what was happening.But a few things on the Anthropic IPO.Listen, it's an exceptional business.They've executed really well.And they haven't just executed on the model.Their products are amazing.I always think about Eric Vistria.I don't think you've ever had on the pod, but I think he'd be a great guest.He's a good friend.He said, has a thought experiment nearly two years ago that he thought OpenAI would still,and Anthropic, would still have a lot of value, even if they lost at the model layer.Because the product, the harness, the user familiarity, it's all so important.And I think there's an element of truth to that.But as far as the IPO, listen, the first thing I would just say is the $2 trillion.I'm a little skeptical.We saw this with SpaceX.If you're, everybody's jockeying to be lead left for Anthropic.And if you lose that, first thing you want to do is make whoever's lead left look bad.So you link to the press and you say, hey, we're in the room.You know, the bar is a $2 trillion IPO.You said on a previous episode, $3 trillion to $4 trillion.

就是这个意思。我认为整个 AI 在 7 月出现了加速，这和公开股票市场当时的走势形成了很怪的反差。现在我们对当时发生了什么有了更高分辨率的视角。不过关于 Anthropic 的 IPO 有几点。首先，这是一门极为出色的生意，他们执行得非常好。而且他们不只是把模型做好了，他们的产品也非常棒。我经常想到 Eric Vishria——我觉得你们从没请他上过节目，但他会是个很棒的嘉宾，他是我的好朋友。他大约两年前提出过一个思想实验：他认为即便 OpenAI 和 Anthropic 在模型层输了，它们仍然会有很大的价值。因为产品、外壳（harness）、用户的使用习惯，这些都太重要了。我认为这里面确实有几分真理。但说回 IPO，我第一句想说的是这个 2 万亿——我有点怀疑。SpaceX 那次我们也见过：大家都在抢 Anthropic 的主承销商（lead left）位置，一旦你没抢到，第一件想做的事就是让拿到主承销的那家难看。于是你去给媒体放风：嘿，我们也在那个房间里，门槛是 2 万亿美元的 IPO。你在之前一期节目里说过是 3 万亿到 4 万亿。

[7:55] Gavin Baker

Absolutely.But I think that's probably where it might trade.So they're sandbagging now.Yeah, we haven't seen the S1.You know, I do think like if you're bringing a company out responsibly, like you do, you want to price it in such a way that you can absorb the lockup.Because you don't want to price it really high, create a lot of volatility.That's like very disconcerting for employees.You do not want anyone distracted at this moment.So I think that they will price it smartly.Maybe that is $2 trillion.Maybe it is $2 trillion.And it trades to $3 trillion.But you just got to remember these leaks are always coming from the bankers who probably lost lead left.And they want to make lead left look bad.So if it comes out of $2 trillion, it probably means that the testing the waters and the roadshow just went incredibly well.So people, they do want to own this.We've never, you know, these numbers are unprecedented.So are you saying the roadshow probably occurred already or quietly occurred?

没错。但我认为那大概是它最终会交易到的区间。所以他们现在是在压低预期（sandbagging）。是的，我们还没看到招股书（S-1）。你知道，我确实觉得，如果你是在负责任地把一家公司带上市——像你做的那样——你会希望定的价格能吃得下解禁（lockup）。因为你不想把价格定得太高、制造一堆波动，那对员工来说是非常让人心神不宁的。在这个节骨眼上你绝对不想有任何人分心。所以我觉得他们会聪明地定价。也许那个价就是 2 万亿。也许定在 2 万亿，然后交易到 3 万亿。但你得记住，这些泄露出来的消息，往往来自那些多半没拿到主承销位置的投行，他们就是想让主承销难看。所以如果真是按 2 万亿出来，那多半意味着预路演（testing the waters）和路演（roadshow）都进行得极其顺利。人们是真的想要持有它。我们从没见过这样的数字。所以你是说路演其实已经进行过、或者悄悄进行过了？

[8:51]

No, definitely not.I'm saying if they price it at $2, it means like the testing the waters went well, the roadshow went well, and they've got a lot of confidence about where it's going to trade.Sachs, in the history of Silicon Valley, you pointed out multiple times here during this podcast in 2026 that we've never seen a revenue ramp like this.Can it continue?

不，绝对没有。我是说，如果他们定在 2 万亿，那意味着预路演顺利、路演顺利，而且他们对上市后会交易到哪儿很有信心。Sacks，在硅谷历史上，你在 2026 年这档节目里多次指出，我们从没见过这样的收入爬坡。它还能持续吗？

[9:13] David Sacks

And if it does continue, what would that say about the value of AI to corporations and enterprises where Claude is the definitive leader?Well, I think it can continue, although it is a legitimate question.Because if you're ending the year at, let's say, $100 billion run rate up 10x plus year over year.And by the way, they've grown 10x year over year for the last three years.So it's growing exponentially.If that rate of growth were to continue, that hit a trillion dollars of ARR by the end of next year.And then the question you have to ask is, well, is the TAM big enough for that?

如果真能持续，那对于「AI 对企业和大公司的价值」意味着什么——毕竟 Claude 在这块是毫无争议的领跑者。我觉得是能持续的，尽管这是个合理的问题。因为如果你今年收尾在，比方说，1000 亿美元的运营收入，同比涨 10 倍以上。顺带一提，他们过去三年每年都是同比 10 倍。所以这是指数级增长。如果这个增速继续下去，明年年底就会摸到 1 万亿美元的年度经常性收入（ARR）。那接下来你必须问：总可寻址市场（TAM）撑得住那么大吗？

[9:50] David Sacks

But also, is there enough compute?Is there enough energy?I think you start to get into physical constraints.I do think the TAM is big enough.I think the demand is going to be there.There's so many new applications of AI.Agents are just taking off now.You saw that with the launch of GrokBot.All these other companies are going to keep extending the uses of AI into more and more context.So the demand for tokens is going to keep growing exponentially.I think the question is whether they can physically meet that demand.And then also Gavin's point about market share.Other companies are seeing what they're doing and they're competing more effectively to take share.Remember, Anthropic made this huge bet on coding.And that ended up being a very prescient bet.I don't know if it was because of ideology.They thought that was the way to get to recursive self-improvement first.I don't know if it was because they saw Cursor.That's the reason.Building on top of them.They watched Cursor's utilization and they said, we're going to take that business.And that's such an important note for you to make.So they've started moving into verticals where they see their own users building successfully on top of their platform.

还有，算力够吗？能源够吗？我觉得这时候你就开始撞上物理约束了。我确实认为 TAM 是够大的，我认为需求会在那里。AI 的新应用太多了，agent 现在才刚刚起飞，你从 GrokBot 的发布就能看到。所有这些公司都会不断把 AI 的用途延伸到更多场景里。所以对 token 的需求会继续指数级增长。我觉得问题在于他们能不能在物理上满足这个需求。然后还有 Gavin 说的市场份额问题：其他公司看到他们在干什么，正在更有效地竞争、抢份额。别忘了，Anthropic 当年在编程上下了那个巨大的赌注，那最后被证明是极有先见之明的一注。我不知道那是不是出于意识形态——他们认为那是最先抵达递归自我改进（recursive self-improvement）的路径。我也不知道是不是因为他们看到了 Cursor。就是因为这个。因为 Cursor 建在他们上面。他们盯着 Cursor 的用量，然后说：这门生意我们自己做。你这个点说得太重要了。所以他们开始进入那些「自己的用户在自家平台上做成功了」的垂直领域。

[10:58] David Sacks

So I think that definitely was a signal to them to get into coding.Nonetheless, they got into it first.They got into it very successfully.That led to a boom.We hear that OpenAI now, they've basically pivoted into coding.They're doing a better job with that now with GPT 5.6.We've heard their growth rate is accelerating.They used to be growing at 3 to 4x a year.I've heard that over the past two months, that growth rate is now over 20% month over month, which if you extrapolate it out over 12 months, would be a 10x growth rate.So you're now seeing OpenAI growing as fast as Anthropic having made that pivot.We can talk about whether we think SpaceX AI may be able to get there as well.But in any event, look, I think that regardless of whether Anthropic is able to grow 10x next year, I think they're going to grow very fast.And then the question is just, what does the market share look like relative to their competitors?

所以我认为这肯定是促使他们进入编程领域的一个信号。不管怎样，他们先进去了，而且做得非常成功，带来了一波繁荣。我们听说 OpenAI 现在基本上也全力转向编程了，他们靠 GPT-5.6 在这块做得更好了。我们听说他们的增速在加快。他们以前是一年 3 到 4 倍，我听说过去两个月，这个增速已经超过每月 20%，如果外推到 12 个月，就相当于 10 倍的年增速。所以你现在看到的是，OpenAI 完成那次转向之后，长得跟 Anthropic 一样快。我们可以再聊聊 SpaceX 的 AI 是不是也能追到那个位置。但不管怎样，我认为无论 Anthropic 明年能不能再涨 10 倍，他们都会长得非常快。剩下的问题只是：相对于竞争对手，他们的市场份额会长什么样？

[11:56] Jason Calacanis

And I guess, Gavin, it would be worth double clicking on two points there.The first point, is there going to be a market for these tokens?And can the growth keep up?In other words, is there a demand?I think all three of us agree.Yeah, sure, of course there's demand.But the next two pieces, which Sachs mentioned, were energy data centers.And that is a question I want you to answer.Can they get to $1 trillion in revenue?

Gavin，我觉得这里有两点值得深挖。第一点：这些 token 到底有没有市场？增长跟得上吗？换句话说，需求在不在？我想我们三个人都同意——需求当然在。但接下来那两块，也就是 Sacks 提到的能源和数据中心，那才是我想让你回答的问题。他们能做到 1 万亿美元收入吗？

[12:22] Jason Calacanis

And let's slow down here and actually try to do some back-of-the-envelope math.And this is your wheelhouse.What would it take in terms of infrastructure for them to go from $100 billion a year in revenue and 10x it?

我们把节奏放慢，真的来做一遍粗略估算（back-of-the-envelope math），这是你的看家本领。要让他们从每年 1000 亿美元收入再翻 10 倍，基础设施上需要付出什么？

[12:36] Gavin Baker

And is that humanly possible in the next year?So a few thoughts.I think it's the right questions.First, internally, Anthropic is very confident.I have been told by multiple people I trust that Dario has said that Anthropic might be the only private company in the world at some point.Think about that.And in this vision, an Anthropic maximalist vision, there's Anthropic and then there are governments.And that's it.So there's a lot of confidence.I'm sure they have more advanced checkpoints than Fable up their sleeve.They've executed really, really well.I would probably take the under on them being the only private company in the world.I think it's going to be a long time before they land a rocket.Are they deliver your lunch?

而这在未来一年里，人类做得到吗？几点想法。我觉得这些问题问得对。首先，在内部，Anthropic 非常有信心。我从多个我信任的人那里听说，Dario 说过：Anthropic 有朝一日可能会是全世界唯一的私营公司。你品品这句话。在这个「Anthropic 极大主义」的愿景里，世界上就只有 Anthropic 和各国政府，就这两样。所以他们信心很足。我相信他们手上还压着比 Fable 更先进的模型检查点（checkpoint）。他们执行得非常非常好。但如果让我下注「他们会成为世界上唯一的私营公司」，我大概率押「不会」。我觉得他们要学会让火箭着陆，还得等很久。还是说他们要给你送外卖？

[13:31] Gavin Baker

I don't know.There might be other businesses in the world.Zipline might want to bring you a burrito.Uber might want to take you somewhere.A Waymo might want to drive you home.But sure, Dario.I might interpret that as a negative signal because it's so hubristic.You know?

我不知道。世界上应该还会有别的生意。Zipline 也许还想给你送个墨西哥卷饼，Uber 也许还想载你一程，Waymo 也许还想把你开回家。但行吧，Dario。我可能会把这句话解读成一个负面信号，因为它太狂了，你懂吗？

[13:46]

Yeah.This is getting into SBF land a little bit.Yes.I would certainly discourage Dario from saying that ever again to anyone.So you don't.So AI Jesus should not claim that he is AI Jesus.Okay.Pride cometh before the fall.Yes.And that will be a long fall.If they, I mean.But look, I have heard from other people at the company.I heard that when they were at about 60 billion of ARR, that they thought they could get to600 billion in one year.So meaning they thought they were on that 10X ramp from the middle of this year.Now they're at over 80.Doesn't seem to be slowing down.So I think, look, I think this is the big question is, you know, how fast can they grownext year?

对。这有点开始往 SBF 的方向走了。是的。我肯定会劝 Dario 别再对任何人说这种话了。所以你不……所以「AI 耶稣」不该宣称自己是 AI 耶稣。好吧。骄兵必败。是的。而且那会是很长的一跌。如果他们，我是说……但你看，我也从这家公司别的人那里听说过：他们在大约 600 亿美元 ARR 的时候，认为一年内能做到 6000 亿。也就是说，他们从今年年中开始就认为自己在那条 10 倍的爬坡线上。现在他们已经超过 800 亿了，看起来没有放缓的迹象。所以我觉得，最大的问题就是：他们明年到底能长多快？

[14:31] David Sacks

Does it go all the way from 100 or 120 to a trillion dollars?Look, if they merely got to four to 500 billion, that would make them the biggest softwarecompany.I mean, if they get to 200 billion, that's an amazing outcome.So where does that put them versus Microsoft's office franchise?

是从 1000 亿或 1200 亿一路走到 1 万亿美元吗？你看，就算他们只做到 4000 亿到 5000 亿，那也已经是全世界最大的软件公司了。我是说，哪怕他们做到 2000 亿，那都是一个了不起的结果。那这跟微软的 Office 业务比又在什么位置？

[14:47] Gavin Baker

Like where or Google's search franchise?Like we're starting to become a peer to those two companies.A hundred.Absolutely.And so just to break the question down, it's a good question.And to demand and supply.So demand, you know, depending on how you count it, there's, you know, 25 to, you know, 65trillion in knowledge work.25 is kind of a number I've used.I was told by, you know, the head of the AI Institute at one of the three largest investmentbanks that it was way low.So this is the TAM that they're going after today before we have robotics.And then it's a question of, is it labor substitution or accelerating growth?

或者跟谷歌的搜索业务比？我们开始变成那两家公司的同量级选手了。一千亿。完全正确。所以把问题拆开来看，这是个好问题——拆成需求和供给。需求这边，取决于你怎么算，知识工作（knowledge work）的总盘子大概是 25 万亿到 65 万亿美元。25 万亿是我常用的数字。全球三大投行之一的 AI 研究院负责人告诉我，这个数字远远低估了。所以这就是他们今天要吃的 TAM——还没算上机器人。接下来的问题就是：这到底是劳动力替代，还是加速增长？

[15:29] Gavin Baker

Thus far, it really does look like it's accelerating growth.There are, you know, there are more job openings for software coders today than there were a year ago.There's more demand for coders.And that is kind of in the epicenter of the blast zone.There is some evidence that for very young people coming, you know, right out of college.Yeah.For, you know, that maybe there's been a negative impact.But broadly speaking, it has to come from one of those two.And I just think it's, you know, we, it's better for everyone if it comes from accelerating growth.And these companies are now at a scale where it's going to, if it's going to come from accelerating growth,we will start to see that in the national accounts over the next 18 months.I am going to, okay.You want to go to the energy part?

到目前为止，看起来真的更像是加速增长。今天软件工程师的职位空缺比一年前还多，对码农的需求变多了，而这可是爆炸中心的震中位置。确实有一些证据表明，对刚出校门的非常年轻的人来说，可能出现了负面影响。但总体上讲，它必然出自这两种情形之一。而且我就是觉得，如果结果来自加速增长，那对所有人都更好。这些公司现在的体量已经到了这样一个程度：如果真的是加速增长，未来 18 个月我们就会在国民经济核算数据（national accounts）里看到它。我打算，好吧。你想进能源那部分了？

[16:19] Gavin Baker

Supply.Yeah.And listen, this is, this is, you know, where kind of the rubber meets the proverbial road.Like I, um...You know, the people who build these data centers, you know, like the best way to understand it is go watch that series Landman.And it's like, you need like Billy Bob Thorntons out there, you know.Permian Basin time.The Permian Basin.You're in the patch.You know, you're actually worried about narcos.You know, it is, it's the tip of the spear.It's like 110 degree heat.Um, you have to orchestrate thousands of people in these remote locations.It's atoms.It's atoms, not bits.It's atoms, not bits.And it's really, really hard.Now, America, we're actually really good once we gear up for something like that.America and capitalism.We're very good at solving those problems.Everybody talks about like the ultimate constraint.Blade are these turbine blades.And, you know, they're only made in like two facilities in America and Europe.And like each facility is like, you know, a giant building and there's some 40 ton press.But you know what?

供给。对。听着，这才是所谓真刀真枪的地方。就是，嗯……那些真正在建这些数据中心的人，最好的理解方式是去看那部剧《Landman（土地人）》。你需要的是那种 Billy Bob Thornton 式的硬汉在外面干。二叠纪盆地（Permian Basin）时间。二叠纪盆地。你人就在油田里。你要真的担心毒贩。这是矛尖上的活儿。华氏 110 度的高温。你得在这些偏远地区调度成千上万的人。这是原子的世界。是原子，不是比特。是原子，不是比特。而且真的非常非常难。不过美国一旦为这种事开动起来，我们其实是很擅长的。美国加上资本主义，我们非常擅长解决这类问题。大家都在说最终的瓶颈是那些涡轮机叶片。这东西全美国和欧洲加起来只有两家工厂能做，每家工厂就是一栋巨大的厂房，里面有个 40 吨的压机。但你猜怎么着？

[17:34] Gavin Baker

Now they're running, you know, 24 hour shifts.They're getting, you know...Necessity is in fact the mother of innovation.And nothing like capitalism will make that happen.In fact, boom, supersonic making a new jet for, you know, supersonic.I know Sachs has one on deposit.Those jets.He's like, you know what?

现在他们已经在跑 24 小时三班倒了，他们正在……需要确实是创新之母。而没有什么比资本主义更能把这事推成。事实上，Boom Supersonic 在造一架新的超音速客机——我知道 Sacks 订了一架。那些飞机。他们说：你猜怎么着？

[17:53] Gavin Baker

We're going to build a turbine business first because we have so much demand.And then Elon, I understand there was a news report that he bought a turbine company.And he did it personally because that speaks to me that he just wanted to go fast, right?

我们要先建一门涡轮机生意，因为需求实在太大了。然后 Elon，我看到有新闻报道说他买下了一家涡轮机公司，而且是他个人买的——这对我来说说明他就是想快，对吧？

[18:05]

And on top of that, you know, residuals for private jets are like extremely strong these days.And a big part of the reason is people are taking jet engines off old planesand repurposing them as turbines to power data centers.A hundred percent.That's insane.It's crazy.This is like a big business and growing really, really fast.And then, you know, you have Caterpillar, you have Cummins, you have GE Vanova,you have Siemens Energy.They're all expanding capacity fast.So I think the supply question, you know, the Billy Bob Thorntons here in America,you know, are going to work with those companies and get these turbines on the ground.It's politics and it's regulation.And the fact that Texas, you know, has kind of announced, you know, what I think is probably...An energy audit is what you're referring to.Governor Abbott announced this week he's going to do an energy audit.Not that he's going to slow things down.Not that there's a moratorium, but that everybody has to go through an energy audit.If you BYOE, you're going to be fine.But if you don't have a solution for that and you want to tap the Texas,the great state of Texas's grid, you're going to need to maybe pass some auditbecause we don't want the whole grid going down here,

而且在这之上，私人飞机的残值这些天强得离谱。很大一部分原因是人们把老飞机上的发动机拆下来，改成涡轮机去给数据中心供电。百分之百。太疯狂了。是很疯狂。这已经是一门大生意，而且长得非常非常快。然后你还有卡特彼勒（Caterpillar）、康明斯（Cummins）、GE Vernova、西门子能源（Siemens Energy），他们都在飞快扩产能。所以我觉得供给这个问题，美国这些「Billy Bob Thornton 们」会和这些公司一起把涡轮机装到地上。剩下的是政治和监管。而德州刚刚宣布的那个，我觉得多半是……你说的是能源审计吧。Abbott 州长本周宣布他要做一次能源审计。不是说要放慢速度，不是说要暂停，而是所有人都得过一遍能源审计。如果你自带电源（BYOE），那你没问题。但如果你没有自己的电力方案，又想接德州、这个伟大的德克萨斯州的电网，那你可能就得过个审计，因为我们不希望整张电网垮掉，

[19:24] Gavin Baker

which it does every January when things freeze over.Yeah. And it's like right now, the narrative on AI is not good.It's that a data center takes all the water because someone made a mistake in a book three years agoand they're off by, I think, a factor of 100,000 in the amount of water a data center uses.So it's just wrong.They raise your cost of electricity, which is just wrong.They lower your cost of electricity.The Wall Street Journal just read a story about a town called Ellendale,where it was a dying town, a data center got built, it's revitalized the town,tax revenue was 10xed.It's been the best thing that's ever happened.But nobody's telling those stories and people need to.And I think, you know, we've discussed on the show many times,the Chinese Communist Party, they do have a very active anti-AI, anti-data center campaignbecause that's what's good for them and bad for America.And someone needs to tell the truth about AI.One of the best phrases in the Bible is the truth shall set you free.Yes.Can't set you free if nobody tells the truth.Yes.Yeah.Yeah.I mean, the data center panic is a testament to the power of the media to create hoaxesand hysteria over literally nothing.Every single aspect of the data center story is wrong.

而它每年一月一到结冰的时候确实会垮。对。而现在关于 AI 的叙事很糟糕。说什么数据中心把水都抽走了——那是因为三年前有人在一本书里写错了，用水量算错了，我认为大概差了 10 万倍。所以那纯粹是错的。说它们推高你的电费，那也是错的，它们其实拉低你的电费。《华尔街日报》刚写过一个叫 Ellendale 的小镇的故事：一个正在凋敝的小镇，建了个数据中心，把整个镇盘活了，税收翻了 10 倍，这是这个镇有史以来发生过的最好的事。但没人讲这些故事，而这些故事是需要有人讲的。而且我们在节目上说过很多次，中国共产党确实在很积极地搞反 AI、反数据中心的舆论战，因为那对他们有利、对美国不利。得有人把关于 AI 的真相说出来。《圣经》里最好的一句话是「真理必叫你们得以自由」。对。但如果没人说真话，真理就没法让你自由。是的。对。对。我是说，数据中心恐慌就是媒体凭空制造骗局和歇斯底里的能力的一个明证。数据中心故事里的每一条都是错的。

[20:41] David Sacks

It's a lie.Like you said, they don't use up a community's water.The water use is quite manageable, less than a golf course.The water recirculates.It brings down electricity costs.Assuming that you have the data center stand up its own power generation because they willsell the excess back to the grid and they pay for grid upgrade costs.The real reason why electricity prices are rising in blue states is because of decarbonizationand all the other green type regulations they have that make power generation more difficult.What are the other accusations?

全是谎言。就像你说的，它们不会吃掉一个社区的水。用水量相当可控，比一个高尔夫球场还少，而且水是循环使用的。它会拉低电价——前提是数据中心自建发电，因为他们会把富余的电卖回电网，而且他们会承担电网升级的费用。蓝州电价上涨的真正原因是脱碳，以及各种让发电更困难的绿色监管。还有别的什么指控？

[21:15]

Well, no, no.Those are –They create pollution.Yeah.Noise pollution is legit.So there's nobody in their right mind.They don't have to put it in anyone's backyard.Yeah.Nobody in their right mind is doing that today.We have a lot of land in this country.Yeah.People did do that for a bit.If we just use common sense –Yes, common sense.Exactly.You just use common sense and there's no problem.The idea that you have to have a complete ban on data center construction is just silly.Yes.And the communities that are embracing them are seeing – they have better schools.They have better services because they're making a lot of revenue.Well, and then to your point, Saks, when entrepreneurs get on something, we're all of a sudden seeing solar, batteries, nuclear, and more clean energy being tapped to power these.And that demand will lower the prices of those services over time.Solar panels will get cheaper.Batteries will get cheaper because they're going to be scaled much faster, which is good for humanity.This is basic, folks.It's very basic.The green stuff can't scale in the short term.It's all gas-powered for the new data centers.That's the only way to kind of get the amount of scale that you need in the next year or two.

呃，不不。那些是——说它们制造污染。对。噪音污染是真的。所以没有哪个脑子正常的人会——他们不必把它建在任何人的后院。对。今天没有哪个脑子正常的人会那么干。我们这个国家有大把土地。对。以前确实有人那么干过一阵。只要我们用点常识——对，常识。没错。你只要用点常识，就没有问题。说必须全面禁止数据中心建设，这纯属荒唐。是的。而那些拥抱数据中心的社区已经看到了——他们的学校更好了，公共服务更好了，因为他们拿到了很多财政收入。而且照你说的，Sacks，一旦创业者盯上这件事，我们突然就看到太阳能、电池、核能和更多清洁能源被调动起来给这些东西供电。而这种需求会随时间拉低这些能源的价格。太阳能板会更便宜，电池会更便宜，因为它们会被更快地规模化，这对人类是好事。这是很基本的道理，各位。非常基本。绿色那套在短期内是没法规模化的。新建的数据中心全都是烧天然气的。这是未来一两年里拿到所需规模的唯一办法。

[22:22] Gavin Baker

But in any event, look –You agree with that, Gavin?I think the solar is more like – it'll be viable, but it's obviously not as fast as popping up Nat Gas.Yeah, I largely agree.I will say Elon's building a $10 billion new solar manufacturing plant in Texas.For all reason.My home state.And, you know, I think solar, the world, will eventually run on sunlight.But we're going to run on natural gas for a while.We have a lot of natural gas in America.It's a clean fuel.It's the greenest carbon fuel.And just this media hysteria about data centers is doing such a disservice to low-income Americans, to small-town Americans, often to minority Americans.It's just – it is – it's so good in every way, and the media is doing such a disservice.But no one's taking the other side.And someone needs to take the other side and tell the truth.Yeah, well, it's – I don't know who's relying on the media for their facts now, but I think there's a reason why we were number four in the world last week is because, hey, listen, we actually are part of this build-out.We know about it, and we can explain it to you a little bit more succinctly.Noise pollution, that's an issue.Pollution of the air, that is an issue from NACS.

但不管怎样，你看——Gavin，你同意这个说法吗？我觉得太阳能更像是——它会可行，但显然没有直接架天然气机组那么快。是的，我基本同意。不过我要说一句，Elon 正在德州建一座 100 亿美元的新太阳能制造厂。这是有原因的。那是我的家乡州。而且我认为，太阳能——这个世界最终会靠阳光运转。但我们还得靠天然气跑一阵子。美国的天然气非常多，它是清洁燃料，是最绿色的碳基燃料。而媒体围绕数据中心的这种歇斯底里，正在极大地伤害低收入美国人、小镇美国人，而且往往是少数族裔美国人。它在每一个维度上都是好事，而媒体正在造成巨大的伤害。可是没人站到另一边去。得有人站出来说真话。是啊。我不知道现在还有谁靠媒体获取事实，但我觉得我们上周能排到全球第四是有原因的——因为，听着，我们本身就是这轮建设的一部分。我们了解它，我们能更简洁地把它讲清楚。噪音污染，那是个问题。空气污染，那是天然气机组带来的一个问题。

[23:39] Jason Calacanis

Keep these things far away from people, as Sax just said, and you don't have as much of an issue.There is a minor issue.If you put these at scale in a certain location, you can raise the temperature of that valley or wherever you put it, a couple of degrees.That's something worth monitoring.These are all absolutely manageable issues.And if you really want to get obsessed about water, please, for the love of God, stop eating almonds.Like – and maybe, like, you know, the golf course doesn't have to use a billion gallons of water every year.I will say, on the demand side and on the pricing side, if you pull up this quick note I have here, I think Anthropics' growth rate is going to slow considerably next year.And it's been obviously on a tarot.There's no way they get to a trillion in revenue next year.I think maybe they could, you know, triple, get to $300 million or $400 million.And the reason is, watching startups and then also watching some of the large companies now embrace GLM52 and Zuckerberg, which we'll get to in a moment,releasing really great open source models.You just look at the pricing here of GLM2 versus Claude Opus, and, you know, it is 90% cheaper to use these products.

就像 Sacks 刚说的，把这些东西建得离人远一点，你就没那么多麻烦。有一个次要问题：如果你在某个地点大规模堆放它们，你确实可能把那个山谷、或者你放它的地方的温度抬高个一两度。这值得持续监测。这些都是完全可以管理的问题。而如果你真的要死磕用水，拜托，看在上帝的份上，别吃杏仁了。还有也许，高尔夫球场也不必每年用掉十亿加仑的水。我要说的是，在需求端和定价端——你把我这儿的这条速记调出来——我认为 Anthropic 明年的增速会显著放缓。它当然一直在猛冲。它明年不可能做到 1 万亿美元收入。我觉得也许能翻三倍，做到 3000 亿或 4000 亿美元。原因是，我看到那些创业公司、以及现在一些大公司都在拥抱 GLM-5.2，还有 Zuckerberg——我们等下就会讲到——正在放出非常好的开源模型。你只要看看这里 GLM 和 Claude Opus 的定价对比就知道，用这些产品要便宜 90%。

[24:52]

I'm using them.The founders I invest in are using them.And increasingly, corporate America is going to embrace these solutions.It's just a little bit harder to set them up.And it's coming, folks.And if you think it's not, that corporate America is not going to embrace open source, you have not been paying attention to the last 20 years when they did exactly that.And 20 years ago, there was a big argument.Oh, my God.Corporate America will never trust open source.And corporate America said, we don't want to get rug pulled or have some large tech corporation hold us hostage for their technology, which they're experts at doing.So open source is the safer bet for us.And that's what's happening in corporate America today and inside of startups.This is going to be very real.Gavin, you're jumping out of your seat.Yeah, no, no.This is a larger point.Like, for sure, it's good for corporate America.For sure, open source is good for the world.For sure, it's good for America.But, you know, Dworkesh posted something.Dworkesh, who's a podcaster, a thoughtful guy close to the AI scene, he posted something that really resonated with me.He said, at the end of the day, Anthropics AI and their constitution, it is wired to do what Anthropics thinks is best for humanity.

我自己在用。我投的创始人在用。而且美国的大公司会越来越多地拥抱这些方案，只是配置起来稍微麻烦一点。这波要来了，各位。如果你觉得不会来、觉得美国大公司不会拥抱开源，那你就是完全没关注过去 20 年——他们干的正是这件事。20 年前也有一场大争论：天哪，美国大企业永远不会信任开源。然后美国大企业说：我们不想被人抽地毯，也不想被某家大科技公司拿技术绑架——他们最擅长干这个。所以开源对我们才是更安全的选择。今天在美国大企业内部、在初创公司内部，发生的就是这件事。这会非常真实。Gavin，你都要从椅子上跳起来了。对对，这是一个更大的点。当然，这对美国企业是好事；当然，开源对世界是好事；当然，这对美国是好事。但你知道，Dwarkesh 发过一条东西——Dwarkesh 是个播客主持人，是个离 AI 圈很近的思考者——他发的那条特别打动我。他说：说到底，Anthropic 的 AI 和它的「宪法（constitution）」，是被写死成去做 Anthropic 认为对人类最好的事。

[26:09] Gavin Baker

And let's take Elon's judgment as gold-plated.And he's right.His don't-be-evil detector did not go off.And they're genuinely good people, which I actually believe them to be, who are doing what they think is good for humanity.I actually believe that.The problem is, like, you know, there are always unintended consequences.And the world is chaotic and unpredictable.And I just think what's great about open source is it means that we're going to have a rich variety of AIs.We're going to have a diversity of AIs.And that is a good world for humans in America.We do not want one, two, or three dominant models.And that's why, by the way, you know, not to, you know, don't want to always talk about Grok, but it's great.I think it's Elon will, history will judge Elon kindly for having Grok dedicated towards the pursuit of objective truth.Yeah, talk that book.Yeah, talk that book.You made a book.The whole point, Gavin, is if you made the bet, that means you used your intelligence and your savvy to pick what you think the winner is.That actually is, I think, super accurate.And I think Elon explicitly believes in open source.And he believes that this is a technology that should be shared for the good of it.

而且我们就把 Elon 的判断当成镀金标准好了，他说得对，他的「别作恶探测器」没有报警。他们确实是真心的好人，我也真的相信他们是，他们在做他们认为对人类有益的事。我是真信这一点。问题在于，永远都有意料之外的后果，而这个世界是混沌的、不可预测的。我就觉得，开源之所以了不起，是因为它意味着我们会拥有丰富多样的 AI，我们会有 AI 的多样性。那对美国人、对人类来说是个更好的世界。我们绝不希望只有一个、两个或三个主导模型。顺带一提，这也是为什么——我不想老是讲 Grok，但它确实很棒——我认为历史会善待 Elon，因为他让 Grok 以追求客观真理为使命。是啊，讲你自己的持仓嘛。对对，你讲你的书。这就是你的持仓。Gavin，重点在于，如果你下了这一注，那说明你动用了你的智力和眼力去挑你认为会赢的那个。我认为这其实非常准确。而且我认为 Elon 是明确信仰开源的，他相信这是一项应该被共享、为了它本身的好处而共享的技术。

[27:28] Jason Calacanis

And, in fact, that was his first foray into this, which was backing OpenAI.And OpenAI's manifesto was rewritten this week, topic two, by Mark Zuckerberg.Quite paradoxically or ironically, in a 6,500-word essay.It's titled, The Future is for Everyone, the Path to a Positive AI Future.He laid out an extremely optimistic AI worldview where he would, as now OpenAI 2.0, Facebook, he would release open source models.He would create an agent for everybody.He would make it free.He would provide everybody in the world with their own tutor and unlimited abundance.He has picked up the abundance crown, which Sam Waltman dropped when he went for the private company.And he says superintelligence is invention, not automation.And that AI safety should be based on the balance of power.No singular centralized intelligence.I'm assuming, SACS, you read or scanned or got the coverage of Zuckerberg's manifesto.Do you have any thoughts on his position?

而且事实上，这就是他最早涉足这件事的方式——他资助了 OpenAI。而 OpenAI 的宣言本周被人重写了，这是第二个议题，重写者是 Mark Zuckerberg。相当吊诡、或者说讽刺的是，用了一篇 6500 字的长文，标题叫《未来属于每一个人：通往积极 AI 未来的路径》。他摆出了一个极其乐观的 AI 世界观：作为如今的「OpenAI 2.0」，Facebook 会发布开源模型，会给每个人造一个 agent，会免费提供，会给世界上每个人配一位专属家教和无限的富足。他把 Sam Altman 走向私有公司时丢下的「富足」王冠捡了起来。他说超级智能是发明，不是自动化；说 AI 安全应该建立在权力制衡的基础上，不能有单一的、中心化的智能。Sacks，我猜你读了、或者扫了一遍、或者至少看了关于 Zuckerberg 这篇宣言的报道。你对他的立场有什么看法？

[28:38] David Sacks

Yeah, I mean, I read it and I found myself nodding in agreement with much of it.I really like the point he made pretty early on where he pointed out that a lot of the doomers, and I think he was kind of taking a shot at the effect of altruists and anthropics.Dario.He said, look, why are you guys rushing to create a future that you don't believe in, that is so dystopian?

有，我读了，而且我发现自己对其中很多内容都在点头。我特别喜欢他在文章比较靠前的位置提出的那一点：他指出，很多末日论者——我觉得他其实是在暗指有效利他主义者（effective altruists）和 Anthropic。指 Dario。他说：各位，你们为什么在拼命赶着去创造一个你们自己都不相信的、如此反乌托邦的未来？

[29:01] David Sacks

If you think this is such a bad future where everyone's out of work and AI goes rogue and creates all these safety issues, why are you so excited about creating it?So I think he does point out a fundamental contradiction.Now, there is an answer to that question, which is – actually, there's two answers.One is they could be trying to create a future they don't believe in for, let's call it mercenary reasons.It could be lucrative.Although I don't think that's the real reason.I think most of those guys are missionaries.And what they believe is that that future is dangerous, and only under their enlightened control can humanity be protected.It's right out of the vision of the anointed, which was a book Thomas Sowell wrote 30 years ago, where intellectuals throughout history have thought that if they're maximally empowered, then they can engineer society in a more benevolent direction.And this has always backfired.It's always led to not just empty or broken promises, but also it becomes an excuse for totalitarian schemes and state power.And actually, Zuckerberg goes on to say something that I think gets at this.So right after he kind of made the point about, you know, why are you rushing to build a future that you don't believe in?

如果你认为那是一个所有人失业、AI 失控、制造出各种安全问题的糟糕未来，那你为什么对创造它这么兴奋？所以我觉得他确实点出了一个根本性的矛盾。当然，这个问题是有答案的，其实有两个答案。一是，他们可能是出于——我们姑且叫「雇佣兵式」的理由——去创造一个自己都不相信的未来，因为那很赚钱。不过我不认为那是真正的原因，我觉得他们大多数人是传教士型的。他们真正相信的是：那个未来很危险，而只有在他们开明的掌控之下，人类才能被保护。这完全出自《被拣选者的愿景》（The Vision of the Anointed）——那是 Thomas Sowell 30 年前写的一本书。历史上的知识分子一直认为，只要自己被赋予最大权力，就能把社会朝更仁慈的方向工程化。而这件事从来都是适得其反的。它带来的不只是空头支票和破碎的承诺，它还会变成极权方案和国家权力扩张的借口。而 Zuckerberg 接下来说的一段话，我觉得正好戳中了这一点。就在他讲完「你为什么急着去建一个你自己都不想要的未来」之后，

[30:14] David Sacks

You don't want.He says, the notion that AI is so dangerous that the only safe path is an extreme concentration of power seems inherently problematic.Historically, hoping that an absolute power will benevolently provide for humanity if sufficiently enlightened has not led to safe or positive outcomes.So I think the point he's getting at here is that a major frame on the whole AI debate, it's not just about open or closed.It's also about centralized versus decentralized.And whether you think a centralized outcome is the right outcome or a decentralized outcome.I'm on the side of decentralized.I mean, how could you not be on the side of decentralized?

你并不想要的未来。他说：认为 AI 危险到唯一安全的路径就是极度集中权力，这个想法本身就有内在的问题。从历史上看，指望绝对权力只要足够开明就会仁慈地供养人类，从来没有带来过安全或积极的结果。所以我觉得他想说的是：整场 AI 辩论有一个大框架，它不只是开源还是闭源的问题，它同时是中心化对去中心化的问题——你认为中心化的结局是对的，还是去中心化的结局是对的。我站在去中心化这边。我是说，怎么可能不站在去中心化这边？

[30:54] David Sacks

I mean, it's like the most obvious argument ever.The EA types are not.And by the way, I experienced this when I got to Washington.And, you know, all the former Biden people had just moved over to Anthropics Government Affairs.I also dealt with think tank people who are part of this, like, click.And their frame on AI was that it was inherently dangerous and that we had to centralize control over it.In the government, there should be, you know, just two or three companies.And they should essentially work together.It should be made into a cartel.Remember, you know, Marc Andreessen had a meeting with those Biden officials.Yeah, they said, don't bother.Don't bother competing.They dispute that characterization.Maybe they didn't use those words exactly.But I do believe that Marc was correct in his interpretation of what he heard because I heard similar things.And they really believe that the way to control what they essentially thought as of being malign or at least very dangerous was, first, you limit it all to the United States.So, you don't export the chips.You don't export the models.It's basically you retain total control over it in the U.S., which, frankly, is a pipe dream because the rest of the world is going to get AI.

这简直是史上最显而易见的论证。可 EA 那帮人就不是。而且顺带一提，我到华盛顿之后亲身领教过这个。当时拜登政府的那些前官员刚刚全体转会到 Anthropic 的政府事务部门。我还打交道过一些属于这个小圈子的智库人士。他们对 AI 的框架是：它天生危险，因此我们必须把对它的控制权集中起来，集中到政府手里，行业里只该有两三家公司，而且这几家本质上应该协同行动——也就是把它做成一个卡特尔。你还记得，Marc Andreessen 和那些拜登官员开过一次会。对，他们说：别费劲了，别去竞争了。他们否认这个说法。也许他们没有原封不动用那几个词。但我确实相信 Marc 对他所听到内容的理解是对的，因为我也听过类似的话。他们真心相信，控制这个他们本质上视为恶意或至少极其危险的东西的办法是：第一，把它全部限制在美国境内。所以芯片不出口，模型不出口。基本上就是把它的控制权完全留在美国手里——坦白说这是白日梦，因为世界上其他地方一定会拿到 AI。

[32:05] David Sacks

Of course.But that was sort of part number one.And then part number two was you form like an atomic energy commission or something like that to cartelize the industry.And then you have it deeply embedded with the government.You effectively have a merger of sort of corporate and state power.So, it was this extremely centralized vision of the future.I think they still believe in that to some degree.Clearly, Zuckerberg listens to the All In podcast.And clearly, I think we have a seat open next week.Mark Zuckerberg, somebody reach out to him.Come on the pod next week.Like, Thursdays, we'll tape at about 10, 11 a.m.Your time, whether you're in Hawaii or in Tahoe, we will get a Starlink to you, and you can do it from your e-foil.Guaranteed.But it would be great to have Mark talk about this.Let me get Gavin involved.Well, just to wrap up, though, I think where Zuckerberg goes with this that I agree is we want maximum empowerment of the individual.That's the way to solve this problem.We want decentralized outcomes.We want open models, like you're saying, J-Cal, we want data sovereignty.Look, there can also be these proprietary frontier models from Anthropic and OpenAI.But we want to create a decentralized, empowering outcome for individuals.

那当然。但那是第一部分。第二部分是，你成立一个类似原子能委员会的东西，把这个行业卡特尔化，然后让它和政府深度绑定。你实际上就完成了企业权力和国家权力的合并。所以这是一个极端中心化的未来愿景。我觉得他们在某种程度上现在还信这一套。很明显，Zuckerberg 听 All-In 播客。而且很明显，我们下周正好空一个位子。Mark Zuckerberg，谁去联系他一下，下周来上节目。周四，我们大概上午 10 点、11 点开录，按你的时区来，不管你在夏威夷还是太浩湖，我们给你架一台 Starlink，你在你的电动冲浪板上录都行。保证没问题。但要是 Mark 能来聊这个就太好了。让我把 Gavin 也拉进来。不过收个尾：我认同 Zuckerberg 的落点，就是我们要的是个体的最大赋能。那才是解决这个问题的方式。我们要的是去中心化的结局。我们要开放模型，就像你说的，J-Cal，我们要数据主权。当然，也可以同时存在 Anthropic 和 OpenAI 这样的闭源前沿模型。但我们要的是为个体创造一个去中心化、赋能式的结局。

[33:18] Gavin Baker

And I think that's the side that Zuckerberg is on, and I very much agree with that.Gavin, I assume you read it, yeah?Or did you just get the hype?You didn't have time for 6,500 words?No, of course I read it.By the way, if he writes nine more, he's got the average nonfiction book, which is right around 60,000, 70,000.So Harper Business, Hollis, go call Zuckerberg and get the other eight parts of this manifesto going.I read it.I agreed with it.And I would just super agree with everything David said.And I would boil it down to, I think Anthropic and people in the effect of altruism movement believe that this technology is too dangerous to distribute.And what Mark Zuckerberg and I think Elon and Jensen believe is that this technology is too dangerous to centralize.And history is spoken.And when given the choice, it is always better to distribute and decentralize.There is, there's really no counter example that I can think of.And just forget as an American, but as a human, I want an AI that looks out for me.Looks out for me.The individual.Be the individual.The sovereign individual was a key point.Yes.Yeah.My conception of what's good for America.I do not want Dario Amadei or Anthropic deciding what's in my best interests.

我认为那就是 Zuckerberg 站的那一边，我非常认同。Gavin，我猜你也读了吧？还是你只看了热闹？没时间读 6500 字？当然读了。顺带一提，他要是再写九篇，就凑够一本非虚构书的平均篇幅了，大概六七万字。所以 Harper Business 的 Hollis，快去打电话给 Zuckerberg，把这份宣言剩下的八部分催出来。我读了，我同意。而且我完全赞同 David 说的每一句话。我会把它浓缩成一句：我认为 Anthropic 和有效利他主义运动里的人相信，这项技术危险到不能分发；而 Mark Zuckerberg、我认为还有 Elon 和 Jensen 相信的是，这项技术危险到不能集中。历史已经给过答案了。当二者必选其一时，分发和去中心化永远是更好的选择。我真的想不出一个反例。而且先别说作为一个美国人，就作为一个人类——我要一个为我着想的 AI。为我着想。为我这个个体。做那个个体。「主权个体」是个关键点。是的。是我心目中对美国好的东西。我不希望由 Dario Amodei 或 Anthropic 来决定什么符合我的最佳利益。

[34:42]

Hmm.I just fundamentally don't want that.You don't trust AI Jesus.I don't.I mean, AI Jesus seems like, he seems like, he seems like a great guy.You know, although it is funny.Somebody said he was cursed with resting, smiling face.And it is a little weird to listen to him talk about this terrible vision for humanity while he's smiling happily.It's hilarious.He's literally got that grin on while he explains to you the Terminator scenario and that only he can save us.And that even he doesn't know that, like, there's a 20% chance that he might not be able to save us all.Gavin, I love that framing about it being too dangerous to centralize.And I think one other closely analogous point is what you think about competition.So I think that the EAs think that competition will create a race to the bottom and will create a lack of safety.Whereas what I believe is that competition brings out the best.And so they think it brings out the worst.Now, look, there are examples of probably both.But fundamentally, they want to centralize control and have less competition.We want to decentralize and have more competition.And I think this –I want in the Constitution, I have the right to bear arms.And I think that's important.

嗯。我就是从根本上不想要那样。你不信任「AI 耶稣」。我不信。我是说，AI 耶稣看着像个挺好的人。不过也挺搞笑的，有人说他被诅咒了一张「面无表情就在笑」的脸。听他一边笑得开心一边描述人类的可怕前景，确实有点怪。太逗了。他脸上真的挂着那种笑容，一边跟你解释终结者场景、一边说只有他能救我们，还说连他自己也不确定——大概有 20% 的概率他救不了我们所有人。Gavin，我特别喜欢你那个说法——「危险到不能集中」。我觉得另一个高度类似的点，是你怎么看竞争。我认为 EA 那帮人觉得竞争会导致向下竞逐（race to the bottom）、会导致安全性缺失；而我相信的是，竞争会激发出最好的东西。他们觉得竞争激发最坏的。当然，两种例子大概都存在。但从根本上说，他们想要集中控制、减少竞争；我们想要去中心化、增加竞争。而我觉得这——我在宪法里有持枪的权利，我认为那很重要。

[35:55] Gavin Baker

I grew up in Texas.I think it is kind of like a lot of defense against tyranny.I love Texas.Gavin, when you're coming home.I'll be back soon, I promise.Okay.But I want the right to have my own AI.Of course.I want the right to my own AI where I – you know, its values are aligned with me.Not someone as good a person as they might be.The right to bear arms is a good analogy because we do allow you to have a gun even though you might misuse it.Guns are used to, like, kill things, right?

我在德州长大。我觉得那在很大程度上是对抗暴政的防线。我爱德州。Gavin，你什么时候回家？我很快就回，我保证。好。但我想要拥有我自己的 AI 的权利。当然。我想要有权拥有我自己的 AI——它的价值观是跟我对齐的，而不是跟别的什么人对齐，哪怕那个人再好。持枪权这个类比是恰当的，因为我们确实允许你拥有枪，尽管你可能滥用它。枪是用来杀东西的，对吧？

[36:25]

Whereas AI has a much more creative use and much more salutary uses.It's beneficial.Fundamentally, it's a consumer technology that everyone's going to want to use.So, you know, if we allow you to have guns, I agree.You should be allowed to have AI.Let me correct you there, Sax, now that you're in the great state of Texas with me.Guns also help you feed your family and defend you against evil forces in the world.So, those are virtuous uses of a gun.Yeah, I get it.They are.That's self-defense.But I'm saying you'd still use the gun to kill the –use the gun to kill or hurt the person in self-defense.I get it.I'm not against guns.I think this is the delusion of Dario.If we think about it with these, you know, effective altruists, you can look it up, folks,but they think that basically they're going to make all the money in the world, create the one company,the money printing machine, and then they're going to save all of us.It's massive delusions of grandeur that, you know, is –The technology is going to diffuse anyway.Of course it is.This idea that you can keep the genie in the bottle is absurd.But by the way –Well, they're treating it –I think, Sax, the way they view it is, they view this literally as super intelligence and the nuclear bomb.

而 AI 的用途要有创造性得多，有益的用途也多得多。它是有益的。从根本上讲，这是一项每个人都会想用的消费级技术。所以，既然我们允许你拥有枪，我同意，那你就应该被允许拥有 AI。Sacks，这里我得纠正你一下，既然你现在跟我一样身在伟大的德克萨斯州。枪也能帮你养家，帮你抵御世上的邪恶力量。所以那些是枪的正当用途。是，我懂，确实是，那是自卫。但我的意思是，你在自卫时用枪，仍然是用枪去杀死或伤害那个人。我懂，我不反对枪。我觉得这就是 Dario 的妄想。如果你把这帮所谓的有效利他主义者想清楚——各位可以自己去查——他们基本上认为自己会赚到世上所有的钱，造出那唯一一家公司、那台印钞机，然后由他们来拯救我们所有人。这是巨大的自大妄想。技术反正是要扩散的。当然会。以为你能把精灵关在瓶子里，这想法太荒唐了。不过顺带一提——他们其实是把它当作——Sacks，我觉得他们看待这件事的方式，就是字面意义上把它当成超级智能加核弹。

[37:30]

And if you convince yourself that you're building the nuclear bomb and you're building the Terminator that you yourself cannot control,and it's super intelligence that will make a billion Terminators and a billion bombs,they literally have deluded themselves while taking P.O. Day or LSD at Burning Man that this is the reality.It's not the reality.You could unplug it at any time.They're suffering from delusions of grandeur.Period.Full stop.This is not the nuclear bomb.Well, you're right about that.When I first came into Washington, that was the dominant analogy that they and their sort of related think tanks,their people in their group had promoted this idea that, A, I was like a nuclear weapon,and we needed an Atomic Energy Commission, a global one, to regulate this.Total pipe dream, never going to happen.But you're right that they were using that analogy.And I think the reason why they were using that analogy is because you do not have a right to basically have a nuclear weapon,to build your own nuclear weapon.You have the right to a gun, but not a nuclear weapon.So they wanted to put AI in that bucket.But as Jensen has said, no individual, no consumer needs a nuclear weapon.Every consumer, every business needs AI.

而如果你说服了自己：你正在造核弹、正在造一个连你自己都控制不住的终结者，而这个超级智能会再造出十亿个终结者和十亿颗炸弹——他们真的在火人节上嗑着佩奥特（peyote）或 LSD 的时候把自己骗进了这个「现实」。这不是现实。你随时可以把它拔电源。他们患的是自大妄想症，句号，完毕。这不是核弹。这点你说得对。我刚到华盛顿的时候，那就是主流类比——他们和他们那一圈相关智库的人一直在推这个想法：第一，AI 就像核武器；第二，我们需要一个全球性的原子能委员会来管它。纯属白日梦，永远不会发生。但你说得对，他们当时用的就是那个类比。而我认为他们用那个类比的原因是：你没有权利拥有一颗核弹、没有权利自己造核弹。你有权拥有枪，但没有权拥有核弹。所以他们想把 AI 塞进核弹那个桶里。但正如 Jensen 说的，没有哪个个人、哪个消费者需要一颗核弹；但每个消费者、每家企业都需要 AI。

[38:42]

It's fundamentally different.This is a consumer technology.It's a consumer technology before it's a military technology.And this idea of that it's like primarily like a nuclear bomb is just kind of absurd.I mean, in fairness.No one's really saying that anymore.In fairness, Gavin, Cyberdyne Industries, they also thought they were building good technology.So science fiction has literally polluted people's brains.If we're going to have a Terminator thing, I want my own Terminators who like me.Okay.Open source the Terminator.Yeah.But like just very simply, just bullying everything else.Like I'm an American.I'm a patriot.If we ban open source, if we let this control be centralized, we will lose AI.We will lose geopolitics.We will lose everything to China.It's just that simple.So if you're an American patriot, you should be for the distribution, democratization, decentralization,individualization of this technology.Hey, J-Cow, man, what are your thoughts?

这是根本性的不同。这是一项消费级技术。它首先是消费级技术，其次才是军事技术。说它主要像核弹，这想法就挺荒唐的。我是说，公平地讲——现在也没什么人还这么说了。公平地讲，Gavin，赛博丁工业（Cyberdyne Industries）当年也觉得自己在造好技术。所以科幻小说是真的把人的脑子污染了。如果我们真要来一场终结者剧情，那我要我自己那些喜欢我的终结者。好吧。把终结者开源。对。但简单粗暴地说，把别的都碾过去：我是美国人，我是爱国者。如果我们禁掉开源，如果我们让这个控制权被集中，我们就会输掉 AI，我们会输掉地缘政治，我们会把一切都输给中国。就这么简单。所以如果你是个美国爱国者，你就应该支持这项技术的分发、民主化、去中心化和个体化。嘿，J-Cal 老兄，你怎么看？

[39:39] Jason Calacanis

What are your thoughts on the essay?Well, like you, Sax, I found myself nodding the whole way because I think he took, hekind of like ran in front of the parade and grabbed the baton and was like, I agree withthe best takes here.And the best take is open source, empowering the individual.And this isn't nuclear technology.I just want to point out two other points that I think got buried because he did try tocover a lot of ground.The first is I want to give him credit for his view of specifically job displacement becausehe says, hey, listen, you know, that is possible job displacement.But he says people fear that automation will outpace individuals' capability growth leadingto job displacement, followed by a difficult period as people learn new jobs.I think we can all agree that that is a concern.But there is no rule that AI must increase automation faster than it increases individuals'capabilities or demand for new skills.Recent statistics suggest it may be more likely that individuals' capability growth could matchor outpace automation.I thought this was like a really good framing and is just another way of saying what Jensensaid.Hey, you're not going to lose your job to AI.You're going to lose it to somebody with AI.

你对这篇文章怎么看？跟你一样，Sacks，我从头点头到尾，因为我觉得他基本上是跑到游行队伍前面把指挥棒抢过来了，然后说：这里最好的观点我都同意。而最好的观点就是开源、赋能个体，以及这不是核技术。我只想再点出另外两点，我觉得它们被埋掉了，因为他确实想覆盖的面太广。第一点，我要为他关于「岗位替代」的看法给他点个赞。因为他说：听着，岗位替代是可能发生的。但他说，人们担心的是自动化的速度会超过个体能力增长的速度，导致岗位被替代，随后是一段人们重新学习新工作的艰难时期。我想我们都同意那确实是个隐忧。但并没有任何规律规定，AI 提升自动化的速度就一定快于它提升个体能力或对新技能需求的速度。最近的统计数据反而暗示，个体能力的增长更可能追平甚至超过自动化。我觉得这个框架非常好，其实也是换个说法在讲 Jensen 说过的那句：你不会被 AI 抢走工作，你会被一个会用 AI 的人抢走工作。

[40:52] Jason Calacanis

And that's the game on the field right now.You know, every time I use this technology, three new business opportunities open up and Icreate two more job wrecks and people are doing five or 10 times as much work.If they embrace the technology and that that's a big if there and if people will do it.The second thing I thought was pretty interesting was he makes a play on having the right toindex anybody's content and learn from it.The quote he says here is, all AI models are derived from human knowledge.OK, fair enough.I'm just pretty, pretty plain vanilla there.Some have tried to frame distillation as harmful, but I think it's important to protect the principlethat you can learn from anything you can observe.In other words, he's preemptively saying, I want to scan the world's data.Well, I have a message for him, which is Zuckerberg.Please explain how this concept is compatible with the fact that you have sued every startup,many of which I've invested in, that has ever tried to index or use the social graph atInstagram or Facebook.They have hundreds of lawsuits where they have sued people for trying to distill, you know,who are the top influencers on Instagram?

这就是当下场上的实况。我每次用这项技术，就会冒出三个新的商业机会，我就会再开两个招聘岗位，而人们在做五倍到十倍的工作量——前提是他们拥抱这项技术，这个「如果」很大，得看人们愿不愿意做。第二件我觉得挺有意思的，是他在为「有权索引任何人的内容并从中学习」这件事铺路。他的原话是：所有 AI 模型都源自人类知识。行，这话公允，挺平淡无奇的。有些人试图把蒸馏（distillation）描述成有害的，但我认为保护这样一条原则很重要：凡是你能观察到的东西，你就可以从中学习。换句话说，他是在先手声明：我要扫描全世界的数据。那我有句话要带给他：Zuckerberg，请解释一下这个理念，跟你起诉过每一家试图索引或使用 Instagram、Facebook 社交图谱的初创公司（其中很多我都投过）这件事，是怎么兼容的？他们打过几百场官司，去告那些试图蒸馏出「Instagram 上的头部博主是谁」的人。

[42:02]

Who are who's in my social graph?And they lock that whole thing down.So there's a bit of hypocrisy here.That reminded me, Sachs, of our discussion on previous episodes with Claude ripping apartold books and distilling them.Yeah, we've had the distillation debate many times.Let me just go back to this point about there's a huge irony in the fact that Dario did notget his way in Washington for the most part during this administration, because I thinkif he had gotten his way, if he had realized his dream of this highly regulated AI apparatus,I don't think that company would be set up for a hugely successful IPO.Why?

谁在我的社交图谱里？他们把这整套东西全锁死了。所以这里有点双标。Sacks，这让我想起我们前几期讨论过的：Claude 把旧书拆开来蒸馏这件事。是啊，蒸馏这个辩论我们已经打过很多次了。我想回到这个点上：本届政府任内，Dario 在华盛顿基本上没能如愿，这里面有一个巨大的讽刺。因为我觉得，如果他真的如愿了，如果他真的实现了那套高度监管的 AI 体制，那家公司就不会有今天这种冲击超级成功 IPO 的架势了。为什么？

[42:40] David Sacks

Because their whole right to charge this massive premium for tokens, which, Jake, how that'sthe point you were making before, is you're wondering how sustainable that is becauseof all these open models that are coming up and are offering tokens for much less.Their whole pricing power, that premium they're able to charge is because they're six monthsahead of those open models.And that six months would be eaten up in the blink of an eye if they were subject to regulatoryapproval.Dario himself keeps writing these blog posts where he says, we did an FAA for AI.You know how long it takes the FAA to certify and approve a new model, in this case forairframes, minimum of five years.I mean, if you want to change a table in your plane, it takes eight years.You know how long it takes the FDA to approve a new medicine?

因为他们之所以有权为 token 收取如此高的溢价——J-Cal，这正是你之前那个点，你在怀疑这个溢价还能维持多久，因为那些开源模型正在冒出来，token 报价便宜得多——他们全部的定价权、他们能收的这个溢价，来自他们领先那些开放模型六个月。而如果他们要过监管审批，那六个月会在眨眼之间被吃光。Dario 自己不停写博客说，我们需要一个「AI 版 FAA（美国联邦航空局）」。你知道 FAA 认证批准一个新机型要多久吗？最少五年。你要是想换飞机里的一张桌子，得花八年。你知道 FDA 批一款新药要多久吗？

[43:25] David Sacks

Years.Okay.You're going to apply that regulatory framework to AI model releases.Game over.Forget about it.Your six months will be gone like that.Your model will be commoditized.To Gavin's point, the Chinese competitors will catch up.And that's bad for the United States.It's bad for our competitiveness.It's bad for the AI race.I mean, obviously, we don't want China beating us.But aside from just the geopolitical dimension, it'd be horrible for Anthropic as a companybecause, again, all their pricing power depends on them remaining in the lead by six months.If they ever slip up, hold on.If they ever slip up and lose that six months, they're toast.What does that do to revenue?

好几年。好吧，你要把那套监管框架套到 AI 模型发布上？游戏结束，想都别想。你那六个月立刻就没了，你的模型会被商品化。按 Gavin 说的，中国的竞争对手会追上来，那对美国不利，对我们的竞争力不利，对这场 AI 竞赛不利。当然，我们不希望中国赢我们。但撇开地缘政治维度不谈，这对 Anthropic 这家公司本身也是灾难性的，因为，再说一遍，他们全部的定价权都取决于保持六个月的领先。一旦他们失手——等一下——一旦他们失手、丢掉那六个月，他们就完了。那对收入意味着什么？

[44:07] David Sacks

Exactly.So now, look, I actually think that their pricing power may be quite sustainable because I thinkpeople are willing to pay, not the whole market, not the whole market, but a decent subset ofthe market is willing to pay a premium for true frontier intelligence.So as long as...If the top 20%, Sax, if the top 20% is willing to pay 10x, they're golden, right?

正是。不过说回来，我其实认为他们的定价权可能相当可持续，因为我认为人们是愿意付钱的——不是整个市场，不是全部，但市场里有相当一部分人愿意为真正的前沿智能付溢价。所以只要……Sacks，如果最上面那 20% 的人愿意付 10 倍的价钱，他们就稳了，对吧？

[44:29]

Like I said before, it's Apple versus Android, right?Most of the market is willing to go with Android.It's cheaper, but there's a significant subset of the market that's willing to pay a huge premiumfor Apple because it's a better experience.Same thing.If you're a company that's in a competitive market and you want the absolute best forcompetitive or other reasons, you're going to pay a premium for true frontier intelligence.But Anthropic and OpenAI are on that hamster wheel.They have to stay in the lead in order to justify that premium.And by the way, I think they can, and they think they can, and they think, in fact, theirlead is growing because they're on the way to recursive self-improvement.I just want to make another point about the going direct movement.Zuck has had his Twitter account since the beginning when he tried to buy the company atFink, D-F-I-N-K-D, which was his handle in college, I believe.And in the last 30 or 60 days, Nikesh from Palo Alto Network's friend of the pod, he startedtweeting and doing long form.And Jensen opened an X account.He started going long form.Thanks for the follow, Jensen.And then Zuckerberg has activated his account, and he's going direct.And I want to just point out that Zuckerberg previously, I think, was letting other people

就像我之前说的，这是苹果对安卓，对吧？市场上大多数人愿意用安卓，因为便宜；但有相当一部分市场愿意为苹果付很高的溢价，因为体验更好。同理。如果你是一家身处竞争市场的公司，出于竞争或别的原因你想要绝对最好的，那你就会为真正的前沿智能付溢价。但 Anthropic 和 OpenAI 是踩在跑步轮上的——他们必须保持领先，才能撑得住这个溢价。顺带一提，我认为他们做得到，他们自己也认为做得到，而且他们其实认为自己的领先在扩大，因为他们正走在通往递归自我改进的路上。我想再说一个关于「直接对外发声（going direct）」这场运动的点。Zuck 从一开始就有推特账号，那会儿他还试过要收购这家公司，他的账号是 finkd，F-I-N-K-D，我记得那是他大学时的网名。而在过去 30 天或 60 天里，Palo Alto Networks 的 Nikesh——本节目的朋友——开始发推、开始写长文了。Jensen 也开了 X 账号，也开始写长文了。谢谢你的关注，Jensen。然后 Zuckerberg 把他的账号激活了，他开始直接对外发声。我想指出的是，Zuckerberg 以前，我认为，是让别人

[45:46] Jason Calacanis

craft his message and send it to the market.And I think he has matured and decided, I am going to write my own missive.This missive, I believe, was written by him, probably with some help, but I ran it throughan AI detector.It's 100% human-crafted, bespoke human words.I think it's fantastic that he's now getting ahead and saying, these are the potential problemsin the future.And it would have been so great if he had done that with teenage girls and the impactof social media on kids and addiction with Instagram and Facebook earlier on.I give him a lot of credit for deciding, I will go direct.I will write my essay.I will speak my truth.And man, X is getting exceptional at these dialogues and these missives now.It's fantastic.Yeah, so I profoundly agree with everything you said.And I want to come back to a couple of things that Zach said about how open source actuallymakes Anthropic more valuable and computational efficiency of humans is the case for humans.But I think what you said is really important.If you are a public figure, if you are the CEO of a company, and you do not have your ownvoice on X, you are at risk.It's just that simple.Risk of not unpacking.You ultimately, in the same department as the truth shall set you free, the only person

去打磨他的信息、再发给市场。而我认为他成熟了，他决定：我要自己写我的檄文。这篇檄文我相信是他自己写的，可能有人帮忙，但我拿 AI 检测器跑了一遍，是 100% 人类手工写的、定制的人话。我觉得他现在能走到前面、说出「未来可能存在这些问题」，这非常棒。要是他当年在青少年女孩、社交媒体对孩子的影响、Instagram 和 Facebook 的成瘾问题上也这么早这么干，那该多好。我要为他这个决定给他很多分：我要直接对外发声，我要写我自己的文章，我要说我自己的真话。而且，X 现在在这类对话和长文上做得非常出色，太棒了。是啊，你说的每一句我都深深赞同。而且我想回到 Zach 说的两点：开源其实让 Anthropic 更值钱，以及人类的计算效率恰恰是人类存在的理由。但我觉得你说的这点非常重要：如果你是个公众人物，如果你是一家公司的 CEO，而你在 X 上没有自己的声音，你就处在风险里。就这么简单。是那种「没有人替你解读」的风险。归根结底，在「真理使人自由」的同一个部门里，唯一能说出一个人自己的真相的，

[47:09] Gavin Baker

who can tell someone's own individual truth is themselves in their own authentic words.And obviously, Mark Zuckerberg has super voting stock.He's never going to be thrown out of meta.But there's many CEOs who think, oh, my corporate PR department's keeping me safe.They're my voice.Well, guess what?

就是他本人、用他自己真实的话说出来。当然，Mark Zuckerberg 有超级投票权股票，他永远不会被赶出 Meta。但有很多 CEO 会想：哦，我的公司公关部在保护我，他们就是我的声音。那你猜怎么着？

[47:29] Gavin Baker

Maybe you retire.Maybe the next CEO doesn't like you.And you get called in front of Congress.And I do think, you know, at some point, you know, the Democrats, you know, have talked abouthow they're going to have an accountability agenda.And they're going to go after everyone.And I mean, I'm not sure that's like a good message for the Democrats.But be that as it may, if you do not have your own voice in your own brand on X, you are at risk.And having your own voice, it is protection.It's power.It's armor.And I think it's a sigil.Hey, yeah.It is a superpower.And I don't know about you, Sax, but I am loving Zuckerberg's Jeremy Strong era much better than the Jesse Eisenberg era.Are you sure you're doing this MMA stuff?

也许你退休了。也许下一任 CEO 不喜欢你。然后你被叫到国会去。而且我确实认为，某个时点上，民主党人一直在谈他们要搞一套「问责议程」，他们会去追究每一个人。我倒不确定那对民主党是个好口号。但不管怎样，如果你在 X 上没有自己的声音、自己的品牌，你就处在风险中。而拥有自己的声音，它是保护，是权力，是铠甲，我觉得它还是一枚印记。嘿，对。它是一项超能力。Sacks，我不知道你怎么想，但比起 Jesse Eisenberg 版的 Zuckerberg，我要爱死他现在这个 Jeremy Strong 版了。你确定你在练这个综合格斗的东西吗？

[48:14]

There was a video of him on like a barge or something?Yes, on a floating platoon.A floating barge.Platoon.What is that?A barge in Lake Tahoe?Yeah.I would be careful.I'd be careful about saying bad things about him, Jake.No, no, no.Listen, I too have turned over a new leaf.I'm into Zuck 2.0.I hope he likes Jake Al 3.0.Come on the pod, Zuck.I want him on the e-foil with the American flag every 4th of July.And I want it like live streamed.And if he falls, that would be aura maxing for him.If he gets back up and records again, it does it live.Yo, he's got to do that in Kauai.With the bull sharks.Let me see some fins chasing him.He is aura farming times a thousand.Way to go, Zuck.And like we need more of this like authenticity.Face the CEO.Go direct.100%, man.Go direct.Protect yourselves.Yeah.And be yourself, man.And like he's being himself.And like he's a good guy.It's all good, folks.There's two points that I think are really important.Open source makes frontier tokens more valuable.So like let's just say, let's to use an example.Let's say that, you know, Anthropic, OpenAI and Grok, they're at like a 250 IQ or 200 IQ.Well, if you have a 200 or 250 IQ intellect that can orchestrate these open source 150 IQ intellects, like that's actually more.

有个视频是他在一艘驳船之类的东西上？对，在一个浮动排（platoon）上。一艘浮动驳船。排。那是什么？太浩湖上的一艘驳船？对。我要小心点。Jake，说他坏话你可得小心。不不不，听着，我也翻开新的一页了，我入坑 Zuck 2.0 了。我希望他也喜欢 Jake 3.0 版。来上节目吧，Zuck。我想看他每年 7 月 4 日举着美国国旗踩电动冲浪板，而且要直播。如果他摔了，那反而是给他攒气场。如果他爬起来重录，还是直播的。哟，他得去考艾岛干这事，跟牛鲨一起。让我看到鱼鳍在后面追他。他这是千倍速的气场收割。干得漂亮，Zuck。我们就是需要更多这种真实感。露脸的 CEO。直接发声。百分之百，兄弟。直接发声。保护好自己。对。而且做你自己，兄弟。他就是在做自己。他是个好人。都挺好的，各位。有两点我觉得非常重要。开源让前沿 token 更值钱。举个例子：假设 Anthropic、OpenAI 和 Grok 处在 250 或者 200 的智商水平。那么，如果你有一个 200 或 250 智商的智能体，它能去调度一堆 150 智商的开源智能体，那其实是更……

[49:39] Gavin Baker

That makes the 250 IQ intellect more valuable is because it can farm out tasks.And just in the same way, you know, if we're going to use the Manhattan Project analogy, you know, everybody, you know, talks, you know, talks about Richard Feynman.And, you know, who's the guy who ran it all?

……这让那个 250 智商的智能体更值钱，因为它可以把任务外包出去。同样的道理，如果我们要用曼哈顿计划来类比：大家都会讲 Richard Feynman。那统筹这一切的那个人是谁？

[49:55] Gavin Baker

And, you know, there's these 10 really smart physicists.Who's the guy who ran it?You know, who said, I am become death.Who's the guy who ran the Manhattan Project?Come on.Oppenheimer.Yeah.So you had Oppenheimer and you had 10, you know, effectively Nobel laureates, maybe 30 of them.But they needed like the 2000, you know, merely, you know, top 10,000 in the world physicists.So having more, having more affordable intelligence, it actually makes the frontier even more valuable.And I think a very plausible outcome is one in which frontier tokens are 65 to 85% of the economic value.Open source tokens are like 80% of the volume.And that's good for everyone.And everyone wins.Tell me the revenue, Gavin Baker, of Anthropic by the end of 2027.What is their 2027 revenue if 2026 ends at exactly 100 billion?

那里面有十来个非常聪明的物理学家。统筹的那个人是谁？说出「我成了死神」那句话的人。统筹曼哈顿计划的是谁？说啊。奥本海默。对。所以你有奥本海默，还有十位、实际上大概三十位诺贝尔奖级别的人物。但他们还需要那两千名——姑且说是全世界排名前一万的——物理学家。所以拥有更多、更便宜的智能，反而让前沿变得更有价值。我认为一个非常合理的结局是：前沿 token 占掉经济价值的 65% 到 85%，而开源 token 占掉大约 80% 的用量。这对所有人都好，所有人都赢。Gavin Baker，告诉我 Anthropic 到 2027 年底的收入是多少。如果 2026 年正好收在 1000 亿美元，他们 2027 年的收入是多少？

[50:53] Jason Calacanis

What is their 2027 revenue, Gavin Baker?I will take the over.Over what?A trillion?A lot of estimates.I'm not going to take the over on a trillion.Okay, I'm going to put 500 billion over or under Gavin Baker.How about this?

Gavin Baker，他们 2027 年的收入是多少？我押「大于」。大于多少？1 万亿？很多估计。我不会在 1 万亿上押「大于」。好，那我给你划 5000 亿，Gavin Baker，大于还是小于？这样如何？

[51:08] Gavin Baker

I'm going to take the over.Oh, 500 billion.No, I'm going to take the over.Whatever the Wall Street consensus is for 27, which I don't know.I'll take a 25 to 30% over on that.Assuming that this regulation campaign doesn't succeed.So you're basically going out of a limb here saying that the underwriters are going to standback revenue.Okay, congratulations, Gavin.Yes.Give me my answer.No, no, no.Over 500 or under?

我押「大于」。哦，5000 亿。不，我押「大于」。不管华尔街对 2027 年的一致预期是多少——我不知道那个数——我押比它再高 25% 到 30%。前提是这轮监管运动不成功。所以你基本上是在冒险断言：承销商会把收入预期压低。好，恭喜你，Gavin。是的。给我个答案。不不不，5000 亿以上还是以下？

[51:34]

I think four to five.I think four to 500 billion.You're under buying.Is within the zone of realistic and achievable.Can I just say one other thing?Hold on, Saks.Over under 500 billion, Saks.I actually was going to say four to 500 billion as exit ARR for next year.Seems very comfortable.So right under.Okay, great.Two unders.Yeah.And the only reason it's not higher is because I do think that somehow you get into physicallimitations at that point.Yes.But maybe not, you know, maybe there'll be enough compute.I mean, let that sink in, though, that we all believe that they're going to hit 400billion next year and they're going to 4X.4X, 100 billion.Like, just let it sink in.They'll be at 120.So I guess that, you know, implies, you know, 4X would be more like 500.Yeah, 4X.Yeah.That's kind of where we're coming at.I'd say the over under really is, you know, I think the over under really is 500 billionexit ARR for next year.And I think they think they'll do the over.We were talking about them at 10 billion 10 months ago.10 billion to 400 billion is 40X, gentlemen.Are they going to thank me for preventing Dario from shooting himself in the foot?

我觉得是四到五。我觉得是 4000 亿到 5000 亿美元。你是在押小。这在现实且可达成的区间之内。我能再说一件事吗？等一下，Sacks。5000 亿的大小盘，Sacks。我本来也要说 4000 亿到 5000 亿，作为明年的退出年化收入（exit ARR）。这个看起来非常稳。所以是刚好押小。好，太好了，两个「小于」。对。之所以不更高，唯一的原因是我确实认为到那个量级你会撞上物理极限。是的。但也许不会，也许算力会够。不过，各位品一品：我们三个人都相信他们明年会做到 4000 亿，也就是翻 4 倍——在 1000 亿的基础上翻 4 倍。就让这句话沉一沉。他们会做到 1200 亿。所以照这么算，4 倍其实更接近 5000 亿。对，4 倍。对。我们大致就落在这里。我觉得真正的大小盘线，是明年退出 ARR 5000 亿美元。而且我觉得他们自己认为会打出「大于」。十个月前我们还在讨论他们 100 亿的时候。从 100 亿到 4000 亿是 40 倍，先生们。他们会不会感谢我阻止了 Dario 自己朝自己脚上开枪？

[52:47]

I think they should.That's what I'd like to know.He literally was like, he put two guns on two ankles and he just emptied both clips andyou had taken the bullets out.They should give you 1%.I actually don't think the bullets were taken out.I think he had them pointed at his temples.Oh, you're saying diddy Mao?

我觉得他们应该感谢。这正是我想知道的。他当时简直是两只脚踝上各绑一把枪，然后把两个弹匣全打空了，而是你把子弹取了出来。他们应该给你 1%。我其实不认为子弹被取出来了，我觉得他是把枪顶在自己太阳穴上。哦，你是说「快走」那段？

[53:04] David Sacks

He did deer hunter?And Saks pulled him down and like he just shot himself in the foot.This is where their hostility to open models comes from, I think, is that if they're goingto slow down, they have to slow down the open models as well.Otherwise, they do get commoditized and their business goes away.However, the United States does not have jurisdiction over what happens in China and China was notgoing to slow down.That was always, I think, a huge flaw in this whole conception.These EA types thought they're going to institute global compute governance.Somehow they're going to get some global agreement with China.Total pipe dream, never going to happen.If we slow down, China is just going to race ahead.It was, I think, always a naive vision.I thought he was going to create the Paris Accords 2.0 and that didn't work.So, slow your roll, kid.Can I just say one thing that I think is also important about open source and just like whyI think AI is going to be good for humans.It's going to be good for Americans.It's that like if we are short compute, we are talking about compute slowing down anthropicsgrowth because literally we're just, you know, we're running out of energy.We can't bring on...

他演的是《猎鹿人》？然后 Sacks 把他的手按了下来，结果他只是打中了自己的脚。我认为他们对开放模型的敌意就是从这儿来的：如果他们要放慢速度，那他们必须让开放模型也一起慢下来。否则他们就真的会被商品化，生意就没了。可是，美国对中国境内发生的事没有管辖权，而中国是不会放慢的。我一直觉得这是整套构想里的一个巨大缺陷。这些 EA 类型的人以为他们能建立起全球算力治理，以为他们能跟中国达成某种全球协议。纯属白日梦，永远不会发生。如果我们慢下来，中国只会一路冲到前面去。我一直认为那是个天真的愿景。我以为他要搞一个《巴黎协定》2.0，而那玩意儿并没有奏效。所以，稳一稳吧，孩子。关于开源，我能不能再说一件我觉得同样重要的事——就是为什么我认为 AI 对人类会是好事、对美国人会是好事。是这样：如果我们算力短缺，我们讨论的是算力会拖慢 Anthropic 的增长，因为我们真的、我们就是能源不够了，我们没法上……

[54:09] Gavin Baker

We're running into physics and like, you know, planetary scale limitations.And I think this is like, this is a reality.Just think about how energy efficient we are.All of our brains, to approximate our brains, you need to have a data center that consumesthe power of a million American homes running for probably six to nine months.And then you get our brains.And then every time you want to answer a question, you need the power of like a thousand Americanhomes.Okay.To operate at the level of our minds.You know, we're smart.We're smart.But like, and all those figures are going to get better.But we're so energy efficient.Capitalism.There's going to be a use for the most computationally efficient intelligences on planet Earth.And those are human intelligences.It's so well said, because if you look at what Cerebris and Grok and NVIDIA are up to,they're making inference 10 times, five times every year.And you're the chip master here, Gavin.But they'll be five to 10 times more efficient every 12 to 18 months.Then you look at the models.And by the way, congratulations.You got Zuckerberg on your ass.Go ask MySpace.Go ask Snapchat what it's like to have Zuckerberg in the review mirror.That's literally the T-Rex coming at the frontier models.

我们撞上的是物理学，是行星尺度的限制。我觉得这是现实。想想我们人类有多节能：要近似模拟我们的大脑，你需要一个消耗一百万户美国家庭用电、连续跑大约六到九个月的数据中心，然后你才得到我们的大脑。而之后每次你想回答一个问题，你还需要大约一千户美国家庭的电力。就为了达到我们大脑的水平。我们是挺聪明的，我们很聪明。但——所有这些数字都会改善——可我们实在是太节能了。资本主义。这个星球上计算效率最高的智能，一定会有它的用武之地，而那就是人类的智能。说得太好了，因为你看 Cerebras、Grok 和英伟达在干什么，他们每年都在把推理效率提升 10 倍、5 倍。你才是这里的芯片大师，Gavin。但他们每 12 到 18 个月就会效率提升 5 到 10 倍。然后你再看模型。顺带一提，恭喜你，你现在被 Zuckerberg 盯上了。去问问 MySpace，去问问 Snapchat，后视镜里出现 Zuckerberg 是什么滋味。那简直就是霸王龙朝前沿模型冲过来了。

[55:35] Jason Calacanis

He is a beast at copying, mimicking, and competing.He's going to be the open source hero from America.He is going to have the best open source model in the next year.I guarantee it here on the audience.I might take the under.Oh, what?

他在抄袭、模仿和竞争上是头野兽。他会成为来自美国的开源英雄。他会在未来一年里拥有最好的开源模型。我在这儿向听众打包票。这个我可能要押「小于」。哦，什么？

[55:52] Jason Calacanis

There's another American hero.Okay.Jensen Wang.Okay, Nemotron.You might see NVIDIA at the open source frontier.He doesn't like us talking about that.I got the back channel.He's like, we're not in the open source business.I mean, we do have Nemotron and we can't give you a full stack.But one of the hyperscalers, by the way, that we know, Gavin,one of the hyperscalers told me they just did a nine figure double Lindy reversewhere they went from doing a frontier model on $100 billion in infrastructure.And they're like, bing, it's going to be open source, not frontier.So people are looking at this saying it's time to bing,maybe move over to some open source models.All right, listen.All in Summit is coming September 13th to 15th in Los Angeles.We told you about the giants last week, Jensen, Satya, Jared, Gwen.I mean, this is all first name club.But here we go.Steve Hilton running for governor of California is going to come.Spencer Pratt going to talk about his failed LA run,but successful in terms of capturing people's imagination.Abigail Schreier, author of Bad Therapy, a fantastic book,is going to be with us with a very special surprise guest.And hey, Bestie Gavin, our guy, Bestie Gavin might even make an appearance.

还有另一位美国英雄。好吧。黄仁勋（Jensen Huang）。好，Nemotron。你可能会在开源前沿看到英伟达。他不喜欢我们谈这个。我收到过内部消息，他说：我们不做开源生意。我们确实有 Nemotron，但我们没法给你一整套全栈。不过，Gavin，我们都认识的一家超大规模云厂商（hyperscaler）跟我说，他们刚刚做了一个九位数级别的空中大回环转体：本来他们是要拿 1000 亿美元的基础设施去做一个前沿模型，结果一个转身——砰，改成开源，不做前沿了。所以有人正看着这局面在想：是时候转身、也许挪到一些开源模型上去了。好，各位听着。All-In 峰会将于 9 月 13 日到 15 日在洛杉矶举行。上周我们已经公布了几位巨头：Jensen、Satya、Jared、Gwynne，全是「只叫名字」俱乐部的成员。但现在还有：正在竞选加州州长的 Steve Hilton 会来；Spencer Pratt 会来聊他失败的洛杉矶竞选——虽然败了，但在抓住大众想象力这件事上很成功；《Bad Therapy（坏疗法）》的作者 Abigail Shrier 也会和我们在一起，还带一位非常特别的神秘嘉宾。还有，我们的「铁哥们」Gavin，我们的自己人，说不定也会露个面。

[57:19] Jason Calacanis

And my guy, Blake Scholl, CEO of Boom Supersonic,who I mentioned earlier in the program.And this year, we're going to turn the outside of the shrineinto a festival-like experience from you, for you,a frontier tech hub featuring AI and robotics demosand a gifting suite, the gifting suite, like you're a celebrity.Merch.com has made an incredible all-in swag suite for you,wellness suite.With red light therapy facials and IVs,if you, hey, you stay out a little too late,saxy poo, that sometimes happens,don't miss these important conversations.And by the way, we're going to announce all the DJs and parties,which is the reason Freeberg does this, going to blow your mind,theallinsummit.com to apply today.Go ahead and apply.Hey, speaking of our guy, Jensen, sixth bestie here on the pod,NVIDIA is partnering with Goldman, BlackRock,and others to raise $500 billion in AI compute.Jensen laid out his vision in an X article.As we mentioned, Jensen is now on X,titled, NVIDIA AI Factory Compute is becoming an investable asset class.The firms will establish independent financing platforms,raising third-party capital.The idea is to treat GPUs like a financeable income-producing asset,like, hey, maybe mortgage-backed securities,

还有我的哥们 Blake Scholl，Boom Supersonic 的 CEO，我前面在节目里提到过他。而且今年我们要把 Shrine 会场外面变成一个节日般的体验，为你打造一个前沿科技中心，有 AI 和机器人演示，还有礼品间——就是那种明星待遇的礼品间。Merch.com 给你们做了一整套超棒的 All-In 周边套装，还有健康养护间，有红光疗法面部护理和输液——如果你，嘿，前一晚玩得太晚了，Sacks 宝贝，这种事有时候是会发生的——别错过这些重要的对话。顺带一提，我们还会公布所有的 DJ 和派对阵容，那才是 Friedberg 干这事的真正原因，绝对让你惊掉下巴。上 theallinsummit.com 立刻申请，去申请吧。说到我们的哥们 Jensen，本节目的第六位铁哥们——英伟达正在和高盛、贝莱德等机构合作，为 AI 算力募集 5000 亿美元。Jensen 在一篇 X 长文里摆出了他的构想，我们前面说过 Jensen 现在也上 X 了，标题是《英伟达 AI 工厂算力正在成为一类可投资资产》。这些机构将设立独立的融资平台，募集第三方资本。核心思路是把 GPU 当成一种可融资、能产生收益的资产来对待——比如说，也许像抵押贷款支持证券（MBS），

[58:48] Jason Calacanis

or hopefully not, with the same outcome.Here's how it works at a basic level.Instead of a company paying billions to buy chips up front,they borrow money.They buy NVIDIA-based systems and use them to start generating revenue,i.e. renting compute.And then they can use the cash flows to pay back the loan.NVIDIA's role is matchmaker,connecting customers who need compute with lenders.Obviously, GM, Ford have their own finance divisions.Gavin, what do you think of this announcementvis-a-vis round-trippingand the hand-wringing concerns we talked abouton this very podcast a year or two ago?

但但愿别是同一个结局。基本的运作方式是这样：一家公司不用先掏几十亿美元把芯片买下来，而是借钱，买入基于英伟达的系统，然后用它们开始产生收入，也就是出租算力；再用这些现金流去还贷款。英伟达的角色是牵线人，把需要算力的客户和放贷方对接起来。当然，通用汽车、福特也都有自己的金融部门。Gavin，你怎么看这个公告——特别是对照一两年前我们在这档播客上聊过的「资金空转（round-tripping）」和那些忧心忡忡的担心？

[59:27] Gavin Baker

Well, I do think what's important here iswhat's essentially happenedis Blackstone, KKR, these firms, Goldman Sachs,they would not have done this.They would not have done that CNBC episode with Jensenif they did not believe that NVIDIA GPU Computewas a financeable asset.And so what's really smart hereis you're getting some of the smartest asset managersin the world who charge $2.20 for many productsto essentially validate this market.And NVIDIA is being a matchmakerand what they're basically saying is,hey, our compute, because it's so flexible,is going to have a long enough lifethat you can finance it at lower ratesthan other kinds of compute.And I think that's smart, that's good for everyone.And it is interesting, if you look at the underlying architecturesof the three, what I call big Chinese open source models,and maybe even throw a few or four,if we have Quinn, if we have Kimmy,if we have DeepSeq, and then we have GLM,they're actually evolving in very different ways.And that architectural variation works in NVIDIA's favorand in GPU's favor.Because it means that you do need this more flexible computeto be able to finance it,to be able to think that you can have a long life.But what NVIDIA is doing that I think is really smart

我觉得这里重要的是：实际上发生的事情是，黑石（Blackstone）、KKR 这些机构，还有高盛，如果他们不相信英伟达的 GPU 算力是一种可融资资产，他们是不会做这件事的，也不会去上 CNBC 跟 Jensen 一起录那期节目。所以真正聪明的地方在于，你把世界上最聪明的一批资产管理人——那些在很多产品上收「2% 管理费 + 20% 业绩分成（two and twenty）」的人——拉来给这个市场背书。而英伟达在做牵线人，他们基本上是在说：嘿，我们的算力因为足够灵活，会有足够长的生命周期，所以你能用比其他类型算力更低的利率去融资。我觉得这很聪明，对所有人都好。而且有意思的是，如果你去看那三个——我称之为「中国三大开源模型」的底层架构，甚至可以说四个，如果算上 Qwen、Kimi、DeepSeek，再加上 GLM——它们的演进方向其实很不一样。而这种架构上的分化对英伟达有利、对 GPU 有利。因为这意味着你确实需要这种更灵活的算力，才能给它融资，才敢相信它有很长的寿命。但英伟达在做的、我觉得非常聪明的一件事是，

[1:00:56] Gavin Baker

is they're saying, hey,we are going to help create this market,grow this market.But they're going to provide two very important functions.They're going to say, hey, Blackstone, Goldman Sachs,KKR, Apollo, BlackRock,I'm leaving someone out and I'm so sorry.There are six of them.Come to us.If you have a deal,you can bring it to usand we will give a residual value guarantee.And I think the way that that residual,which will further lower the cost of financing these,and that residual value guaranteecan then be incorporated into those companies underwriting.And what that basically means is they're sayingthat after three years, four years,we guarantee that these GPUs can be rented at a certain rate.And then the risk that they are bearingis the risk between that valueand wherever the market rate is.And right now, everything's going straight up.NVIDIA has better telemetry than almost anyoneinto the supply and demand of compute.So they can put that at a very, very smart place,further make it financeable.You're risking, you know,they're only bearing 25% of the risk.The world's, some of the world's smartest underwriters,you know, in terms of reputation.What could go wrong, Gavin?

他们在说：嘿，我们会帮忙把这个市场创造出来、做大。而他们要提供两个非常重要的功能。他们会说：嘿，黑石、高盛、KKR、阿波罗（Apollo）、贝莱德——我肯定漏了谁，实在抱歉，一共有六家——你们来找我们。如果你有一笔交易，可以拿到我们这儿来，我们给一个残值担保（residual value guarantee）。我认为这个残值担保会进一步压低这些资产的融资成本，而且可以被并进那些机构的承销模型里。这基本上意味着他们在说：三年、四年之后，我们保证这些 GPU 还能按某个价格租出去。而他们承担的风险，就是那个保证价格和届时市场价之间的差额。而现在一切都在往上走。英伟达对算力供需的遥测数据（telemetry）几乎比任何人都好，所以他们可以把那个价格定在一个非常非常聪明的位置，进一步让它变得可融资。你承担的风险——他们只承担 25% 的风险，而承担另一部分的是世界上声誉最好的一批承销人。那能出什么问题呢，Gavin？

[1:02:28] Jason Calacanis

I mentioned mortgage-backed securities, et cetera.And obviously, the lifespan is the key issue here.The different hyperscalers were saying,hey, four, five, six years,Amazon and the big debate happened.But what we've seen is,to your point with the open source models,is they can be tweakedto run on the oldest hardware.So maybe we get years six, seven, eight out of these.That's the key to this?

我刚才提到了抵押贷款支持证券之类的。而且显然，使用寿命是这里的关键问题。不同的超大规模云厂商说法是四年、五年、六年，亚马逊那边还发生过一场大辩论。但我们看到的是，按你说的，因为有开源模型，它们可以被调优到能跑在最老的硬件上。所以也许我们能从这些卡上榨出第六年、第七年、第八年。这就是关键？

[1:02:53] Gavin Baker

More than that, Corweave saidthat they are renting Amperesat economically profitable ratestoday for 2029.So an Ampere that came out in 2020is going to have a nine-year life.And I think that's what NVIDIA sees, you know?

还不止。CoreWeave 说，他们今天就已经能把 Ampere（A100 那一代）以经济上有利可图的价格租到 2029 年。也就是说，一块 2020 年出的 Ampere 会有九年的寿命。我觉得这就是英伟达看到的东西，你懂吗？

[1:03:12] David Sacks

It's like old American cars,they get sent overseas.Like, just the idea that these...Where could it go wrong?Here, I'll tell you where it could go wrong,in theory, okay?In theory, just so we're steel manning it.The biggest risk, to me,is not on the demand side.The biggest risk is that you get a glut of computeand you get an overbuild.And the same way that we had dark fiberafter the dot-com crash,if you had dark GPUs,that'd be a disaster for everyone,especially if you built outyour compute infrastructureexpecting a spot price of $30 to $50 a watt,as Elon said that they were expecting, right?

就像美国的老车，最后被运到海外去。就是说这些东西……那哪里可能出问题？我来告诉你理论上哪里会出问题，好吧？理论上，我们来给它做个「钢人论证」。对我来说，最大的风险不在需求端。最大的风险是你出现算力过剩、出现过度建设。就像互联网泡沫破裂后我们有一堆「暗光纤（dark fiber）」一样，如果你有一堆「暗 GPU」，那对所有人都是灾难——尤其是如果你当初建算力基础设施时，是按每瓦 30 到 50 美元的现货价（spot price）来预期的，就像 Elon 说他们当时预期的那样，对吧？

[1:03:50] David Sacks

So if all of a sudden there are too many peopleracing to provide this computeand now there's an oversupplyand the market crashes,that'd be the risk factor.In a weird way,all the political headwinds,I think, ensure against that outcome.Because it is so hard to build data centersfor all the reason we said.There's a whole moral panicslash hysteria slash hoax going onthat actually it's those political headwinds,I think, will almost guaranteethat there's not an oversupplyrelative to the exponentially growing demand.So in a weird way,you're protected against that.But Jake, can I shift gears for a second?

所以如果突然之间有太多人抢着提供这种算力，出现供给过剩，市场崩了，那就是风险因子。但以一种很怪的方式，所有这些政治阻力，我觉得反而给这个结局上了保险。因为出于我们刚说的所有原因，建数据中心实在太难了。现在正在发生一场道德恐慌／歇斯底里／骗局，而恰恰是这些政治阻力，我认为几乎能保证：相对于指数级增长的需求，不会出现供给过剩。所以以一种奇怪的方式，你被保护住了。不过 Jason，我能不能岔个话题？

[1:04:26]

Absolutely.Tell you what's so brilliantabout what Jensen did here.There's one other part.In return for that,they get a revenue shareabove that floor.Who gets the revenue share above the floor?In India.Yeah.So I think the wayI think the deals will workis, hey, we're willing to guarantee offtakeat nowhere near $30 or $50 a watt.We're willing to guarantee offtakeat a very low risk levelthat with our really high resolution viewof productive capacityand future model trendswe're comfortable with.And in return,we get some revenue shareabove probably a higher level.And Morgan Stanley wrote a great note,I think two days ago,NVIDIA could become,and these are effectively royalties,these revenue shares,they could very quickly becomea very large cloudwith a capital-like business.Let me tell you why I thinkwhat Jensen did was so brilliant.It's because the numbers are getting so bigthat the TAM is getting constrainedby the ability to finance this build-out.And what he's doingis alleviating that finance constraintso that he can growas big as the TAM actually is, right?

当然可以。我来讲讲 Jensen 这一手到底妙在哪儿。还有另外一部分。作为对残值担保的交换，他们能拿到那个下限之上的收入分成。谁拿到下限之上的收入分成？印度。对。所以我认为这些交易的做法会是：嘿，我们愿意承诺包销（offtake），价格远远低于每瓦 30 或 50 美元；我们愿意在一个我们凭借对产能和未来模型趋势的高分辨率认知、感到很放心的低风险水平上承诺包销。作为交换，我们在一个大概更高的水平之上拿一部分收入分成。摩根士丹利大概两天前写过一篇很好的研报，说英伟达可能会变成——而这些收入分成实际上就是特许权使用费（royalty）——它们可能很快就变成一门规模巨大、但资本占用很轻的类云业务。我来说说我为什么觉得 Jensen 这一手如此高明。因为这些数字正在变得太大，以至于总可寻址市场（TAM）反而被「有没有能力为这轮建设融到资」给卡住了。而他做的事，就是把这个融资约束解开，好让他能长到 TAM 真正该有的那么大，对吧？

[1:05:37] David Sacks

It's removing that constraint.So just to take one example,Elon wants to add somewherearound six to eight gigawatts next year.We know that that would costthree to $400 billion of CapEx.The company just raised $100 billionin its equity and debt offering.So obviously,they would have to go outand finance that somehow.And as we talked abouton our previous episode,the simplest way to finance itwould be to get seller financingfrom NVIDIA,especially given that the payback periodcould be as quick as one year.So now,Jensen is creating the,you could say,the line of creditusing these big banks,using these big private equity shops,and he's making that available.And that's going to now benefitall of these downstream purchasers.I don't think it's circular.This is not a circular situationlike Gurley's worried about.This is,I think,Wall Street banksand private equity firmsproviding financingbased onthe expected cash flowsthat will be deliveredfrom these GPUs.In some ways,NVIDIA is kind of becomingthe central bank of AI.The Federal Reserve of AI.That is like a wayto kind of conceptualizewhat they are doing.And in the same way,the Federal Reserve,like a lot of monetary policy,it works through private banks.

他是在移除那个约束。举个例子：Elon 明年想加上大约六到八吉瓦（GW）。我们知道那大概要花三到四千亿美元的资本开支（capex）。这家公司刚刚通过股权和债券发行募了 1000 亿美元。所以显然，他们必须出去以某种方式融到剩下的钱。而正如我们上一期聊过的，最简单的融资方式就是从英伟达拿卖方融资（seller financing），尤其考虑到回本周期可能短到一年。所以现在，Jensen 是在用这些大银行、这些大私募股权机构，造出一条——你可以说——授信额度，并把它开放出来。这会立刻惠及所有下游采购方。我不认为这是循环的。这不是 Bill Gurley 担心的那种循环交易。我认为这是华尔街的银行和私募股权公司，基于这些 GPU 预期能产生的现金流来提供融资。从某种意义上说，英伟达正在变成 AI 的中央银行——AI 的美联储。这是理解他们在做什么的一种方式。而且同理，美联储的很多货币政策也是通过私人银行来发挥作用的。

[1:06:54] David Sacks

Here you're going to have,and probably more bankswill opt into this,you have private marketswho are charging fees,underwriting these deals,and NVIDIA is helpingto facilitate that.So I don't thinkit's really circular financing.Yeah, I watched that CNBC special as well.One point that really stuck with mewas when the private equity guyssaid thatthis was a little bitlike what they dofor plane financing,where when an airlinebuys airplanes,they're able to finance itnot just based on the creditworthinessof the airline,but rather becausethe airplanes themselveshave value.And that makes itmuch easier to financebecause you know thateven if the airlinegets in trouble,you're protected, right?

在这里你会看到——而且大概会有更多银行选择加入——你有私募市场在收费、在承销这些交易，而英伟达在帮忙撮合。所以我不认为这真的是循环融资。是啊，那期 CNBC 特别节目我也看了。有一点让我印象特别深：那位私募股权的人说，这有点像他们做飞机融资的方式——航空公司买飞机的时候，融资依据不只是航空公司本身的信用，更是因为飞机本身有价值。而这让融资容易得多，因为你知道，即便航空公司出了事，你也是被保护的，对吧？

[1:07:39]

You have asset-backed financing.And that's what they'rebasically doing herewith GPUs.Now, in order for that to happen,NVIDIA has to standardize,they have to createessentially reference designsand so forth and so onso that it can allbe standardized enoughfor these Wall Street guysto now package it upand securitize it,turn it into securities.And that's basicallywhat they're doinglike they did with mortgages.Yeah.I'll say there's two,there's a reason whyI'm not worriedabout this build-out.There's two,there's a governor,like a throttle,Gavin, if you will.There's a governor on this,which is the physicsof the real worldthat we talked aboutin building data centers.There's also a dead man switchon this,which is, you know,the switch that you haveto hold downto run the train.But if you have a heart attack,your hand comes off the switch,which is Anthropicand OpenAI.These are the two biggest customers.One of them has scale back,OpenAI,their ambitions in this regardof their build-outfrom like 1.4 trillionto like,but 600 billion.And then there's Anthropic,which we just said,okay,they're going to go to 400 billionin revenue next yearor something like that,or they'll end the yearat that run rate.If they don't need the compute

你拿到的是资产支持融资（asset-backed financing）。他们现在用 GPU 做的基本就是这件事。当然，要做到这一点，英伟达必须做标准化，他们必须造出一套参考设计等等，让这一切标准化到足以让华尔街那帮人能把它打包、证券化，变成证券。这基本上就是他们在做的事，跟当年对房贷做的一样。对。我要说，有两点——我之所以不担心这轮建设，有两个原因。有一个调速器，一个节流阀，Gavin，如果你愿意这么叫的话。第一个调速器是我们前面说的、建数据中心时现实世界的物理约束。还有一个是「死人开关（dead man switch）」——就是那种你必须一直按住才能让火车开的开关，一旦你心脏病发、手松开了，车就停了——而这个开关就是 Anthropic 和 OpenAI。这是两个最大的客户。其中一个已经收缩了，OpenAI 把这方面的建设野心从大约 1.4 万亿降到了大概 6000 亿。然后是 Anthropic，我们刚才说了，好，他们明年收入会做到 4000 亿，或者年底会达到那个运营收入水平。如果他们不需要那么多算力，

[1:08:52]

or compute gets so efficientor open source models,create a headwind,whatever it is,they're going to take their handoff the buy button.And that dead man switchwill then slow everything downso we don't geta mortgage-backed security.Well,that could lead to a car crash.But actually,this is why I thinkit's the Anthropic IPOis really importantfor the market,is getting those quarterly earningsand be able to see their numbersevery quarter.I think it'll becomeprobably the most important signalthat the entire industry has.Why?

或者算力效率变得极高、或者开源模型形成逆风，不管是什么原因，他们就会把手从「买入」按钮上拿开。而那个死人开关就会让整个系统慢下来，这样我们就不会搞出一个抵押贷款支持证券式的结局。嗯，那也可能导致连环追尾。不过实际上，这正是我认为 Anthropic 的 IPO 对市场特别重要的原因——你能拿到季度财报，能每个季度看到他们的数字。我觉得那会成为整个行业最重要的信号，可能没有之一。为什么？

[1:09:22] David Sacks

Because you've gotall these people out theresaying that AI is a bubble,this CapEx is not justified,there won't be an ROI,and everyone's going to be lookingto Anthropic's quarterly earningsas the pace car.Not just them,it'll be themand OpenAI and SpaceX,but you'd have to saythat Anthropic right nowis the pace car.And they're going to be lookingto them to seeif the demand signal is there.And if there's a hiccupor a wobble,you know,the industry...It'll show up there first,yes.If Anthropic slams on the brakes,there'll be a pileupof companies behind it.Because Anthropic,they're making$100 billion per gigawatt,okay, of compute.That's why they're ableto pay SpaceX,say, $50 billion a gigawatt,you know,spot price for compute.That's why SpaceXis then able to pay$30 billionto NVIDIA for chipsfor the build-out.That's why NVIDIAis able to pay TSMCand Micronand SK Hynixand all the way down.It's the entire food chain.So if there is likea wobble...If somebody slamson the brakes...If the pace car slows down,the horse is going to feel it.But it's still betterfor those numbersto come outon a quarterly basisin a vetted,gap-compliant way.So Gavin,as we're sort of saying here,there's a kill switch,there's a brake,

因为现在外面有一大堆人在说 AI 是泡沫、这些资本开支不合理、不会有投资回报，而所有人都会盯着 Anthropic 的季度财报当作领跑车（pace car）。不只是他们，还有 OpenAI 和 SpaceX，但你不得不承认现在的领跑车就是 Anthropic。大家会看他们，来判断需求信号在不在。而一旦出现打嗝或者晃动，整个行业……会最先在那里显形，对。如果 Anthropic 一脚急刹，后面就会追尾一整排公司。因为 Anthropic 每吉瓦算力能做出 1000 亿美元收入。这就是为什么他们付得起 SpaceX，比如说每吉瓦 500 亿美元的算力现货价。这也是为什么 SpaceX 接下来付得起 300 亿美元给英伟达买芯片来搞建设。这也是为什么英伟达付得起台积电、美光、SK 海力士，一路往下。这是一整条食物链。所以如果真出现晃动……如果有人一脚急刹……如果领跑车慢下来，后面的马会感觉到。但这些数字以季度为周期、以经过审计、符合会计准则（GAAP）的方式披露出来，仍然是更好的事。所以 Gavin，就像我们这儿说的，有一个终止开关，有一个刹车，

[1:10:39] Jason Calacanis

whatever you want to call it.Is there enough space?Is there enough distancebetween the carsto not have a pileup?In other words,if Dario does slamon the brakes,if there are headwinds,does everybodygo slamming into them?

随你怎么叫它。车距够不够？车与车之间有没有足够的距离，让大家不至于追尾？换句话说，如果 Dario 真的一脚急刹，如果真出现逆风，是不是所有人都会撞上去？

[1:10:53] Gavin Baker

Or is there enoughfor the entire industryto say,okay, yeah,we're going to go20% slower,we're going to go30% slower?It's very simple.If he's slammingon the brakesbecause there's not demand,there will be a pileup.Okay, there's...Oh, there will be a pileup.There's not enough roombetween the cars.Yeah, if it'sbecause of demand.Now, if Dario'sgetting passedby SpaceXor OpenAIare an AmericanOpen Source champion,there won't be a pileup.But I do thinkthe Anthropic S1is going to bereally important.And I think it's goingto break a lotof people's brains.And the reason is,is I think there area lot of macroand value investorswho are very confidentlymaking theseprognosticationsabout AIwith the assumptionthat tokensare subsidized.Anthropicis generatingcash.And like,they kind of,this fact is,I would say,kind of known in the world.And I've actuallynever really spokento Anthropic,but it just,you know,it feels likethere are enough signalsthat if you'reto read the tea leavesof what has beensaid in public,they're generatingcash.They're profitable.Open Source tokensare profitable.Anthropic is profitable.OpenAI,if they're notgenerating cash,they will be imminently.SpaceX,SpaceX,you know,same thing.And this,

还是说整个行业有足够的余地说：好吧，我们慢 20%，我们慢 30%？很简单。如果他踩刹车是因为没有需求，那一定会追尾。哦，一定会追尾。车距不够。对，如果原因是需求。但如果 Dario 是被 SpaceX、或 OpenAI、或某个美国开源冠军超过去了，那就不会追尾。不过我确实认为 Anthropic 的招股书（S-1）会非常重要，而且我觉得它会把很多人的脑子搞崩。原因是，我认为有很多宏观投资者和价值投资者，正非常自信地对 AI 做出各种预言，而他们的假设是 token 是被补贴的。可 Anthropic 是在产生现金的。这件事在世界上算是半公开的常识。我其实从没真正跟 Anthropic 打过交道，但就是，如果你去解读公开场合说过的那些话里的茶叶渣，信号已经足够多了：他们在产生现金，他们是盈利的。开源 token 是盈利的，Anthropic 是盈利的。OpenAI 如果现在还没产生现金，很快也会。SpaceX，SpaceX 也一样。而这个，

[1:12:14] Gavin Baker

when I hear thesemacro and value investorswho are ignorant,they're very smart,but they're ignorantof the factsand they're makingan assumptionthat tokensare subsidizedand this is allgoing to collapsein some circularfinancing bonfire,they're just wrong.The overwhelmingmajorityof tokensare profitablefor everyonein the chain.everyone.And I thinkthe Anthropic S1is kind of goingto make that clear,but when I listento these macroand value investorstalk about AI,it's a little bitabout me saying,you know,I'm very bearishon the world economybecause oil'sat $500 a barrel.And you know what?

当我听到这些宏观投资者和价值投资者——他们很聪明，但他们对事实无知——他们假设 token 是被补贴的、这一切最终会在一场循环融资的大火里崩塌，他们就是错的。链条上绝大多数 token 对每一个环节都是盈利的，每一个环节。我认为 Anthropic 的招股书会把这一点讲清楚。但当我听这些宏观和价值投资者谈 AI 的时候，感觉有点像我说：我对全球经济极度看空，因为油价在每桶 500 美元。你猜怎么着？

[1:12:58]

If oil was at$500 a barrel,that'd be a goodreason to be bearish.It's just not.Like,they're just wrong.Like,the simple factis wrong.They're paidin some waysto be doomscrollersand merchantsof doom.I can tell you,you know,one framing of thisthat I've been working on,which is the numberof people employedin the United States,like $150 million,$160 million,what their collectivesalaries are,$10 trillion,$12 trillion,whatever it winds up being.There is no worldin which I don't seecorporationsand people in the economyspending 5% or 10%of the salariesof their employeeson the equivalentin tokens.In other words,somebody makes $80,000a year on average,they're going to,you will spend $8,000,10% of their salaryor a minimum of $4,000to make them more efficient.We have an analogy for this.It's SaaS softwareand computerson their desktop,mobile phones,et cetera,just their general useof technology.If you put those numbers together,you're looking at$8 trillion in AI spendjust in the United States.It's obviouslygoing to happen.If I'm wrongand it's halfand it's $500 billion,well, that's kind ofwhere these companiesare trendingand that's not even talkingabout the restof the planet.I think we're goingto be fine, folks.

如果油价真是每桶 500 美元，那确实是个看空的好理由。可它不是。所以他们就是错的，最基本的事实就是错的。他们在某种程度上是被雇来当末日刷屏者和贩卖末日的商人的。我可以告诉你一个我一直在打磨的框架：美国的就业人数，大概 1.5 亿、1.6 亿人，他们的工资总额是 10 万亿、12 万亿美元，最后不管落在哪。我完全想象不出有哪个世界，是企业和经济体里的人不会把员工工资的 5% 或 10% 花在等价的 token 上的。换句话说，某人平均年薪 8 万美元，那你就会为他花 8000 美元——工资的 10%，或者至少 4000 美元——来让他更高效。这件事我们有类比：SaaS 软件、桌上的电脑、手机等等，也就是他们对技术的一般使用。把这些数字加起来，光美国就是 8 万亿美元的 AI 支出。这显然会发生。就算我错了，只有一半，是 5000 亿——那也差不多正是这些公司现在的走势，而且这还没算地球上其他地方。各位，我觉得我们会没事的。

[1:14:21] Jason Calacanis

All right.Let's wrap up here.I've got two more storiesI want to get to.We'll go lightning round here.Very interesting storycoming out ofNew Jerseyand New Yorkaround Amazon.New Jersey's AGhas sued Amazon last week.They're arguingthe company is cheatingbasicallyby hiring driversthrough subcontractors.This has been well-known.It's called DSP.This conceptof delivery service partnersis a very clever,perhaps too clever,of a wayfor Amazonto shield themselvesaccordingto these lawsuitsthat are startingto pile upbecauseif an Amazon drivercrashes,there is liabilityand it gets youout of somethings you mighthave to pay forif they wereAmazon employeeson Monday.New York CityMayor in frontof the pods,the ZoranMondami,jumped into the fray.He is throwingsupport and he didone of hisbig grinsocialist,democratic socialistvideosthat he is nowbecoming famousfor,tellingand demandingthrough municipallaws thatwill requireAmazon to hirethe driversthemselves.Quote,corporations likeAmazonbuild billion-dollarbusiness modelsby insulatingthemselves fromaccountabilitythrough a systemof exploitativesubcontracting.This has been anissue for a longtime.There have beentragically a numberof deathsbecause thesesubcontractors

好，我们收个尾。我还有两条新闻想聊，我们走闪电轮。有一条很有意思的新闻来自新泽西和纽约，关于亚马逊。新泽西州总检察长上周起诉了亚马逊，他们的论点是这家公司在作弊——基本上是通过分包商来雇司机。这事早就人尽皆知，叫 DSP，配送服务合作商（delivery service partner）这个概念。这是一个非常聪明、也许聪明过头的做法，按这些开始越堆越多的诉讼的说法，它让亚马逊得以给自己挡枪：如果一个亚马逊司机出了车祸，是有责任的，而这套安排能让你规避掉一些「如果他们是周一上岗的亚马逊正式员工」你就得赔的东西。上过本节目的纽约市市长佐赫兰·曼达尼（Zohran Mamdani）也跳进了战局。他表态支持，还录了一条他现在越来越出名的那种「咧嘴大笑的社会主义者、民主社会主义者」视频，要求通过市政立法强制亚马逊自己直接雇佣这些司机。原话是：像亚马逊这样的企业，通过一套剥削性的分包体系把自己与问责隔离开来，从而建起了十亿美元级的商业模式。这个问题存在很久了。已经有一些人因此丧生的悲剧，因为这些分包商

[1:15:45] Jason Calacanis

are put undermassive pressureand people getkilled.tragically,when they runinto peoplebecause ofthem operatingin the realworld.And becausethe driversare technicallyemployed bythe DSP,Amazon claims,you know,hey, we're notthe employer,so we're noton the hook.And we don'thave to getinto unionbargaining orlabor obligations.So, Saks,I'm curiousyour take onthis in termsof the overallsurge inpeople believingin socialismnow andbelieving thatcorporations arecheating on themargins.Is Amazoncheating here?

被压上了巨大的压力，然后有人因此丧命——很惨，因为他们在真实世界里作业时撞到了人。而由于司机在法律上是由 DSP 雇佣的，亚马逊就可以说：嘿，我们不是雇主，所以我们不担责，我们也不需要进入工会集体谈判或承担劳动义务。所以，Sacks，我很想听听你的看法——放在当下越来越多人相信社会主义、相信企业在边边角角上占便宜的大背景下。亚马逊在这件事上是在作弊吗？

[1:16:24] David Sacks

Is this tooclever?And maybe theyneed to rollthis back.I'm curious youropinion becauseit's gettingmomentum andit's going tobe one of thebig issues inthe midterms andobviously in2028.Well, I mean,this is apractice that'swidespread.It's not focusedon New York.I think Amazon'sbeen using thesenetworks ofindependentcontractors asdrivers for awhile.And whatthey say isthat if youdeny them theability to usethese independentcontractors, itwill raisedelivery costs.And there was astudy thatshowed that theaverage householdcould pay hundredsof dollars extraif you changethis.It would slowdown service.It would putthousands of jobsat risk atsmall contractorbusinesses.And Amazonsays that itcould lead themto shift operationsoutside thecity.Also, onething that thisDSP model doesis it enablesrapid scalingfor peaks.So holidaysurges, denseurban routing,lower costs.It's a much moreflexible arrangementthan if everyonewas just anAmazon employee.So this is howthe market hasevolved is Ithink maybe whenthey started,Amazon probablywas doingeverything themselvesand they startedto decentralize itto these networksof small businesses,these contractors.It actually endsup creating morejobs.It creates moreentrepreneurship.

这是不是聪明过头了？他们是不是该往回收一收？我很想听你的意见，因为这个势头在起来，而且它会成为中期选举、以及 2028 年大选的重要议题之一。我是说，这是一个非常普遍的做法，并不只发生在纽约。我觉得亚马逊用这种独立承包商网络当司机已经有一阵子了。而他们的说法是：如果你剥夺他们使用这些独立承包商的能力，配送成本会上升。有一项研究显示，如果改掉这个模式，平均每个家庭每年可能要多付几百美元。服务会变慢，还会让小承包商企业里成千上万的岗位面临风险。亚马逊说这可能会促使他们把运营迁出这座城市。另外，DSP 模式的一个作用是能针对高峰快速扩容：节假日的爆单、密集的城市配送路径、更低的成本。这比让所有人都当亚马逊正式员工要灵活得多。所以市场就是这么演化过来的：我猜刚开始的时候，亚马逊多半什么都自己干，后来他们开始把它分散到这些小企业、这些承包商网络里。这实际上创造了更多岗位，创造了更多创业机会。

[1:17:46] David Sacks

I think theseDSPs employthousands ofpeople in NewYork City.Those jobswould bethreatened ifyou basicallymade themillegal.So I thinkthis is acase wherethey shouldjust let...Where do youstand, Sachs?Where do youstand on it?

我认为这些 DSP 在纽约市雇了成千上万的人。如果你基本上把它们判成非法，那些岗位就会受到威胁。所以我觉得这属于那种应该就让它……Sacks，你的立场是什么？你在这件事上站哪边？

[1:17:59] David Sacks

I would justlet freedomof contractand volunteerexchange prevailhere.And look,this marketstructure evolvedfor a reason.The workersliked it.The DSPowners liked it.Amazon liked it.These are mutuallybeneficial arrangements.I don't see thereason whyMamdani has toinsert the cityin the middleof this.Gavin, do youhave a take here?

我会让契约自由和自愿交易在这里说了算。而且你看，这个市场结构是有原因才演化成这样的：工人喜欢它，DSP 的老板喜欢它，亚马逊喜欢它。这些都是互利的安排。我看不出曼达尼有什么理由要把市政府插进这中间。Gavin，你有什么看法吗？

[1:18:19] Gavin Baker

What I would doif I was Amazon,and I'm surethey will notdo this,I would justput asurcharge.Hey, this isthe Mamdani,this is theZoran Mamdanifee that isbeing imposedon you.Loads of peopleuse Amazon,not just wealthypeople in NewYork.Lots of peopleuse Amazon.Amazon.And this isthe Zoran Mamdanifee as a resultof this ill-advisedpolicy.I would have itas a line itemin every bill.Because that's whatit's going to resultin.When you compareAmazon service toanything governmentprovides,it's insane howefficient Amazonis.You order yourthing and it canbe there the sameday, or at worse,it gets delivered toyou the next day.It's unbelievablehow amazingAmazon service is.It's optimized forlow prices forconsumers.And we're going tobasically get in themiddle of that andhave the governmentget in the middle ofAmazon's drivernetwork.What city servicedoes Mamdaniprovide thatremotely is asefficient as whatAmazon does thathe now is going totell them how torun their business?

如果我是亚马逊，我会怎么做——我相信他们不会这么干——我会直接加一笔附加费。嘿，这是「曼达尼费」，这是佐赫兰·曼达尼强加到你头上的费用。用亚马逊的人非常多，不只是纽约的有钱人，很多普通人也在用亚马逊。这就是这项不明智政策带来的「佐赫兰·曼达尼费」。我会把它当成一个单独的行项目列在每一张账单上。因为最终的结果就是这个。当你把亚马逊的服务和政府提供的任何东西对比一下，你会发现亚马逊的效率高到离谱。你下个单，东西当天就能到，最差第二天送到。亚马逊的服务好到不可思议，它是围绕「给消费者低价」优化出来的。而我们现在要横插一杠，让政府插手亚马逊的司机网络。曼达尼提供的哪一项市政服务，效率能跟亚马逊沾上一点边，让他现在有资格去教人家怎么做生意？

[1:19:28] Jason Calacanis

It's insane.If you look atwhat Starbucks didwhen Schultz wasrunning it, hegave people ajob worth havingand he tookpride in takingthe lowest paidpeople in theorganization andgiving them aliving wage.And that livingwage is somewherearound $20, $22in these cities.These DSP driversare actually paidquite well, $18,$19, $20.But this was notcreated in thefree market asyou're sort ofsaying, Sax, andit was just whateverybody wanted.No, this wascreated explicitlyto benefit Amazonand to take theburden of theseemployees and putit onto theAmerican taxpayer.And they aregetting called outon it, rightfullyso, by New Jerseyand New York.And I believe, youknow, and I'm along-term Amazoncustomer.I've been, youknow, a primesubscriber since thebeginning.It would be ade minimis cost forthem to make theseemployees, full-timeemployees, and youcould still havesome DSPs to scaleup and down, sure.But this is anopportunity for themto make thosefrontline workers partof the Amazon family.It would cost but25 cents perdelivery in NewYork to move themfrom this $18, $19an hour to givethem a slightlyhigher pay raise,maybe $21, $22,and give thembenefits.This is an easything to do.And what ourindustry, what ourindustry has a crazy

太荒唐了。你看看舒尔茨（Howard Schultz）还在管星巴克时干了什么：他给了人们一份值得拥有的工作，而且他以「把组织里工资最低的那批人给到能过日子的工资」为荣。在这些城市里，那个生活工资大概是每小时 20、22 美元。这些 DSP 司机其实拿得也不错，18、19、20 美元。但 Sacks，这套东西并不像你说的那样是自由市场自然长出来的、是大家都想要的。不，这套东西的设立是明确为了让亚马逊获益，把这些员工的负担转嫁给美国纳税人。而现在他们被新泽西和纽约点名了，而且点得有道理。我是亚马逊的长期客户，我从一开始就是 Prime 会员。让这些人变成全职员工，对亚马逊来说成本微不足道；而且你完全可以保留一部分 DSP 来做弹性扩缩，当然可以。但这对他们来说是个机会——把那些一线工人纳入亚马逊这个大家庭。在纽约，把他们从 18、19 美元时薪提到稍高一点，比如 21、22 美元，再给上福利，每单配送的成本只会增加 25 美分。这是件很容易做的事。而我们这个行业，我们这个行业有一个很疯狂的

[1:20:53] Jason Calacanis

blind spot for isthat they look forefficiencies, andAmazon is the mostefficient company inthe world, to thepoint of it beingabusive to the mostvulnerable members ofour society, and forbeing against theesprit de corps ofcapitalism.And this is a levelof capitalism thatneeds to be self-corrected for.And I think thegovernment's going tohave very validarguments here.I think it would beamazing for Amazon toget ahead of this,because this will bewhat the democraticsocialists, alsoknown as socialistcommunists in mymind, is going tothis is a veryvalid argument thatwill sell withAmericans whobelieve that thisgroup of people isbeing taken advantageof, and that we caneasily afford to pay25 cents more perpackage delivery.Yeah, Sachs, you'reright.It might add $100.Okay, well, why don'twe have Uber do thattoo, then, J. Cal?

盲区：他们只盯着效率。亚马逊是世界上效率最高的公司，高到对社会中最脆弱的成员构成了压榨，也违背了资本主义应有的那种集体荣誉感（esprit de corps）。这是资本主义里需要自我纠偏的一个层面。我认为政府在这件事上会有非常站得住脚的论点。我觉得亚马逊如果能主动走到前面去做，会非常漂亮。因为这将是民主社会主义者——在我心里也可以叫社会主义共产党人——手里一个非常有效的论点，会在美国人当中很有市场：他们相信这群人正在被占便宜，而我们完全负担得起每个包裹多付 25 美分。是啊，Sacks，你说得对，可能一年下来会多花 100 美元。好啊，那 J-Cal，我们干脆让 Uber 也这么干？

[1:21:46] Jason Calacanis

Absolutely, Uber has.Because Uber wasbasically castigated forthe exact same thing,using independentcontractors, instead ofmaking them full-fledgedemployees.Sachs, you make a greatpoint about Uber, andthis would be theequivalent of if Uberhad 3,000 or 4,000people representingtheir drivers, andthey had no exposure.Uber actually hasmassive insuranceexposure, massivelawsuits that theyhave to contend withfor what thosedrivers do.So it's distinctlydifferent, and they'vegone through this.I'm not a proponentof socialism, but Ido think that there'san opportunity forAmazon, which is goingto be automating alongthe way, to justtake a little bitof a haircuton their amazingbusiness, but thenbe on the right sideof history, and nottake their employees'benefits, unemployment,food stamps, andhealth insurance, andgive that to thepublic to do.And these socialistsare going to pointthis out in the nexttwo election cyclesto the point ofabsurdity.And it's going toplay.Socialism isplaying.Socialism is makinginroads, and it'sinfecting people'sbrains, and it's upto capitalists to say,you know what?

当然，Uber 已经这么干了。因为 Uber 当年正是因为完全一样的事被口诛笔伐：用独立承包商，而不是把他们变成正式员工。Sacks，你拿 Uber 举例很好，但这件事的等价物是：如果 Uber 找了三四千家公司来代表它的司机，而 Uber 自己一点风险敞口都没有。可 Uber 实际上有巨额的保险敞口，为司机的行为背了巨额诉讼。所以这是有本质区别的，而且他们已经趟过这一遭了。我不是社会主义的拥护者，但我确实认为，亚马逊——一家反正一路上都在自动化的公司——有机会在自己极好的生意上让出一点点利，然后站到历史正确的一边，而不是把员工的福利、失业保险、食品券和医保甩给公众来兜。而这些社会主义者会在接下来两个选举周期里把这件事翻来覆去说到荒谬的程度。而且这招是奏效的。社会主义正在奏效。社会主义正在攻城略地，正在感染人们的大脑，所以该轮到资本家站出来说：你猜怎么着？

[1:22:59] Jason Calacanis

We don't need to be ascutthroat and don'tneed to be as clever.So that's my message.As a long-termshareholder in Amazon,I have a very bigposition in Amazon, andI keep increasing itevery year.They should do theright thing here andget ahead of this sothat they can say,hey, we are anincredible employer, asopposed to, hey, wereally are trying tohide and move theseexpenses around.That's my only pointon it.Not a socialist, nota cuck, notinterested in, youknow, messing withcapitalism.I think they got alittle too clever here.That's my only point.And I think theyshould do the rightthing here andtreat those, themost vulnerableemployees and thelowest paid ones.I know why you putthis on the docket.What's that?

我们不必这么心狠手辣，不必这么精于算计。这就是我要传达的意思。作为亚马逊的长期股东——我在亚马逊上有很重的仓位，而且每年都在加仓——他们应该在这件事上做正确的事、主动走到前面去，这样他们才能说：嘿，我们是一家了不起的雇主；而不是：嘿，我们其实是在到处藏匿和挪移这些开支。这就是我在这件事上唯一的观点。我不是社会主义者，不是软蛋，也没兴趣去动资本主义。我只是觉得他们在这件事上算计得有点过头了。这就是我唯一的观点。我认为他们应该做正确的事，善待那些最脆弱、工资最低的员工。我知道你为什么把这条放进议题单。什么？

[1:23:41]

I know why you putthis on the docket,so you can get it onyour soapbox.Well, no, no.I put it on the docketbecause it's one of thetop news stories, and Iput it on the docketbecause we talk aboutsocialism everyweek, and we talkabout all the virtuesof capitalism, andthere are downsides tocapitalism where we canget too clever and thentry to take expenses.I think we spent toomuch time on this, butlet me ask you aquestion.You said that you wantAmazon to employ allthese drivers.By the way, they don't.I think they should employa bunch of them.I know.You said you wanted abunch to still remainwith the DSPs.I think the DSPs canhappen.So how do you decidewhich is which?

我知道你为什么把这条放进议题单——好让你站上你的道德讲台。不不不。我把它放进议题单是因为它是最热门的新闻之一；也因为我们每周都在聊社会主义、每周都在聊资本主义的种种优点，而资本主义确实有一些糟糕面——我们会算计过头，然后试图把该自己承担的开支甩出去。我觉得我们在这个话题上花了太多时间，不过让我问你一个问题：你说你希望亚马逊把这些司机全都雇下来。顺带一提，他们现在不是。我觉得他们应该雇一部分。我知道。可你自己也说了，你希望一部分继续留在 DSP。我觉得 DSP 可以存在。那你怎么决定哪些归哪边？

[1:24:12]

I think they couldproactively do that.It's a great question.Andy Jassy couldproactively decide, justlike Schultz did.Well, Manon, he wantsto get rid of the DSPs.He wants to make themall employees.And I think there's acompromise here, whichis there are DSPs toscale up when there'sdemand during theholidays.Seems like a greatidea, but I think heshould get ahead of it.As soon as youacknowledge that theDSPs are better becausethere are more flexibleworking arrangements thatpeople want, why are youmeddling in thisarrangement to begin with?

我觉得他们可以主动去划。这是个好问题。Andy Jassy 完全可以主动决定，就像舒尔茨当年那样。可曼达尼想要的是把 DSP 全废掉，他要把他们全变成正式员工。我觉得这里存在一个折中：保留 DSP，用来在节假日需求高峰时扩容。听着像个好主意，但我觉得他应该主动走到前面。可你一旦承认 DSP 更好、承认它提供的是人们想要的更灵活的工作安排，那你一开始又何必去插手这个安排呢？

[1:24:39]

That's actually nottrue.The people who work inthe DSPs, the DSPsthemselves, the peoplewho own those fragmented,very small companies thatgo out of business thesecond they get in anaccident to shieldAmazon, those employeeswould very much like tohave Amazon equity.They would love to getRSUs and they wouldlove to get healthcareand benefits.Some of those drivers,they work for Uber alot of the time.They toggle back andforth between being adriver for Amazon, beinga driver for Uber.They like the flexibleworking arrangements.You're actuallymisinformed here.You're incorrect.These are people who areworking full shifts.They're working 8 to 12hour shifts.I think it's actually 10to 12 hour shifts.I don't think you canspeak for all theseworkers.You can speak for themif you, they can speakfor themselves and theyare organizing and Amazonhas been trying to breakthe organization effortswhen they try tounionize.And this is a tool forbreaking that byforcing them to be partof the DSP.That's the game on thefield, Zach.Well, you already

这其实不对。在 DSP 里工作的人、DSP 本身、那些拥有这些高度分散的小公司的人——这些公司一出事故就立刻倒闭，好给亚马逊挡枪——那些员工是非常想拿到亚马逊股权的。他们会很想拿到限制性股票单位（RSU），很想拿到医保和福利。那些司机里有不少人同时也给 Uber 开车，他们在「给亚马逊开」和「给 Uber 开」之间来回切换，他们喜欢这种灵活的工作安排。你在这一点上其实是被误导了，你说错了。这些人是在上整班的，他们上的是 8 到 12 小时的班，我觉得实际上是 10 到 12 小时的班。我觉得你没资格代表所有这些工人发言。他们可以自己为自己发言，而且他们正在组织起来，而亚马逊一直在试图破坏他们想成立工会的努力。把他们锁死在 DSP 里，正是破坏工会的一个工具。这就是场上的实况，Sacks。可你已经

[1:25:36] Jason Calacanis

acknowledged that youthink the DSP shouldstick around.I think a portion ofthem should do that.And I think Andy Jassyand the team at Amazonshould get ahead of thisbecause it's a bad look.It's not in the bestinterest of capitalism tobe this cutthroat.I think he should do itproactively.I don't think he should beforced to do it.That's my personal belief.The same way Schultz saidthere's an opportunity hereto make baristas have aliving wage and havebenefits.And it was one of thebest things Starbucks everdid.And it's really whatcapitalism should startthinking about.What is our role?

承认了你觉得 DSP 应该继续存在。我觉得其中一部分应该保留。而我认为 Andy Jassy 和亚马逊的团队应该主动走到前面去，因为这个形象很难看。这么心狠手辣并不符合资本主义的最大利益。我觉得他应该主动做，我不觉得他应该被逼着做，这是我的个人观点。就像舒尔茨当年说的：这里有个机会，让咖啡师拿到能过日子的工资、拿到福利。那是星巴克做过的最好的事情之一，也正是资本主义该开始思考的东西：我们的角色是什么？

[1:26:05] Jason Calacanis

What is our role indriving the move tosocialism amongst youngpeople?And this is an example ofone of the reasons peopleare starting to believe insocialism is because theybelieve that companies likeAmazon are cheating alittle bit on the marginsand maybe trying to takethe expenses that theyshould incur and put themon the American taxpayer.And you know what?

我们在推动年轻人转向社会主义这件事里扮演了什么角色？这就是人们开始信社会主义的原因之一：他们相信像亚马逊这样的公司在边边角角上占了点便宜，把本该自己承担的开支推给了美国纳税人。而你猜怎么着？

[1:26:27] David Sacks

Mondami and New Jersey aregoing to win these lawsuits.I predict it now.Well, look, J. Cal, thatall sounds really nice andI want to save capitalismas much as you do.And if Amazon doing thatwould really savecapitalism, then I guessI'm game for it.But I'm not sure it will.What we do know it willdo is raise the cost ofpackages by about $5.20per package and it willcost families $664 moreper year.And it puts something like5,000 jobs at risk.So there are real costsin terms of doing thingsthe way that you're saying.When people are wearingtheir hat as voters, yeah,they fall for the socialismstuff.But then when they'reconsumers, when they'rewearing their consumer hat,they just want to pay lessfor packages.They will hate the fact thatpackages cost $5 more.And like Gavin said, theyshould put the Mamdanisurcharge on their orders.And it's not 25 cents.It's $5.$5.20 per package.$664 per household per yearin New York.That's crazy.That's a real cost.Talk about cost of livingadjustment.That's a huge cost ofliving adjustment.JCal has had internetissues and we're justmoving on.Gavin, last issue.What did you think ofGrokBot?

曼达尼和新泽西会赢下这些官司。我现在就预测。行吧，J-Cal，这些话听起来都很好，我跟你一样想拯救资本主义。如果亚马逊那么做真能拯救资本主义，那我也认了。但我不确定它能。我们确实知道它会带来什么：每个包裹的成本上升大约 5.20 美元，每个家庭一年多花 664 美元，还会让大约 5000 个岗位面临风险。所以按你说的方式做，是有真实代价的。人们戴着选民帽子的时候，是啊，他们会吃社会主义那一套；可当他们戴上消费者帽子的时候，他们只想为包裹少付点钱。他们会恨透「包裹贵了 5 美元」这件事。就像 Gavin 说的，他们应该在订单上加一笔「曼达尼附加费」。而且那不是 25 美分，是 5 美元。5 美元。每个包裹 5.20 美元，纽约每户每年 664 美元。太夸张了。这是实打实的成本。说到生活成本调整——这可是个巨大的生活成本调整。J-Cal 网络出问题了，我们直接往下走。Gavin，最后一个议题：你怎么看 GrokBot？

[1:27:35] Jason Calacanis

What did you think of thereleases that XAI madethis week?I guess it was GrokBot andthen also it was Grok4.6, right?And people were saying thatit is close to the frontieror at the frontier but alot cheaper.So I've been maintainingas of even last week thatwe were down to a frontierlab duopoly withAnthropic and OpenAI.Was I wrong about that?

你怎么看 xAI 本周的这几个发布？我想是 GrokBot，还有 Grok 4.6，对吧？有人说它已经接近前沿、甚至就在前沿，但便宜得多。而我到上周为止一直坚持认为，我们已经收敛成 Anthropic 和 OpenAI 的前沿实验室双寡头了。我这个判断错了吗？

[1:28:01] Gavin Baker

Is this number three?Is it more flexible andfluid than I thought?Yeah, I think it is.I mean, I think it's can weshow up, show that graph ofthe Pareto frontier?This is pretty wild.Now, on the left, what thisshows is kind of likequality or intelligence onthe y-axis and cost on thex-axis.So you want to be in theupper right quadrant.And you can see here, youjust want to be on theoutside of this.And you can just see likehow disruptive Grok 4.6 ispricing.Right.Sorry, sorry.As you move right on thex-axis, it's gettingcheaper.It's getting cheaper.And the y-axis is thecapability and it's gettingmore and more capable asyou move up.Exactly.They invert the x-axis sothat as you move to thetop right, it's better,right?

这是第三家吗？局面是不是比我想的更有弹性、更流动？对，我觉得是。我是说——能不能把那张帕累托前沿（Pareto frontier）的图放出来？这个挺野的。左边这张图，纵轴是质量或者说智能水平，横轴是成本。所以你想待在右上角象限。你可以看到，你想待在这条曲线的外侧。你能直接看出来 Grok 4.6 的定价有多具破坏性。对。抱歉抱歉，横轴往右走是越来越便宜。是越来越便宜。纵轴是能力，往上走能力越来越强。正是。他们把横轴反转了，所以越靠右上越好，对吧？

[1:28:54] Gavin Baker

Exactly.And I mean, this is now,and this is on CursorBench and XAI, you know,SpaceX is, you know, almostcertain, you know, they havea right to acquire Cursor,they're working withCursor.So maybe this is gradingyour own homework.Well, right here on theright, this is fromDatabricks, which israised money at $190billion.This is a very realcompany that is verysophisticated.And you can see, like,even on this, you're wellahead of Fable 5, which,you know, I think mostpeople would kind of agreeis the, you know, the goldstandard.Like, this is your higherquality at a slightly lowerprice.So I think Grok 4.6 is,you know, the numbers arethe numbers, you know, onX, I think we put it down.People, you know, we do haveto wait and see the vibes,but I think it's prettyinteresting.DHH, who I think createdRuby on Rails.David Hanmeyer-Hanson.Yeah.He posted something reallypositive.Basecamp, Basecamp.The DHAs.David Hanmeyer-Hanson.37 signals.Great, great.37 signals.Basecamp, yes.He posted something reallypositive.You saw the Databricks, theevaluation.You saw our Merkurevaluation.And you saw, I think theearly vibes are really goodin these benchmark scoresare bearing out in the realworld.Elon has caught up, in other

正是。而且这一张现在是 Cursor Bench 上的数据，xAI——SpaceX 几乎可以肯定拥有收购 Cursor 的权利，他们在跟 Cursor 合作——所以这也许算是自己给自己批卷子。不过右边这张来自 Databricks，Databricks 是按 1900 亿美元估值融的资，这是一家非常真实、非常成熟的公司。你可以看到，即便在这张图上，你也明显领先 Fable 5——我想大多数人会同意 Fable 5 是当下的黄金标准。而这里是质量更高、价格还略低。所以我觉得 Grok 4.6，数字就是数字，我们在 X 上也放出来了。当然还得等一等看实际口碑（vibes），但我觉得挺有意思的。DHH，我记得是他创造了 Ruby on Rails。David Heinemeier Hansson。对。他发了条非常正面的评价。Basecamp，Basecamp。DHH 他们。David Heinemeier Hansson。37signals。对对，37signals。Basecamp，是的。他发了条非常正面的评价。你也看到了 Databricks 那份评测，看到了我们那份 Mercor 评测。而且我觉得早期口碑非常好，这些基准分数正在真实世界里被兑现。换句话说，Elon 追上来了。

[1:30:23] Jason Calacanis

words.And if you were toattribute why he's caughtup, it would really be tworeasons.Number one, he boughtCursor.He's got that team.They're a dialed-in team.Number two, he reshuffledthe management of XAI andbrought in his killers fromSpaceX.Yeah, Gavin, those are thetwo reasons.He brought in, you know, hebrought in, you know, theSpaceX Aces along with theCursor Aces and some newpeople.And he did that in sixshort months, Gavin.Never count Elon out.He's frequently late.He's never wrong.And he tends to over-deliverwhen it does come.And that's to his ownadmission.He's like, I'm always late,but we get it righteventually.Yeah?

而如果要归因他为什么追上来了，其实就两个原因。第一，他买下了 Cursor，他拿到了那个团队，那是一支状态极佳的团队。第二，他重排了 xAI 的管理层，把 SpaceX 的杀手们调了进来。对，Gavin，就是这两个原因。他把 SpaceX 的王牌和 Cursor 的王牌，再加上一些新人调到了一起。而且他是在短短六个月里做到的，Gavin。永远别把 Elon 排除在外。他经常迟到，但他从不做错。而且一旦交付，他往往是超额交付。这是他自己承认的，他说：我总是晚，但我们最终会做对。对吧？

[1:31:08] Gavin Baker

Absolutely.And this is still arelatively small model.It's 1.5 trillionparameter models.A larger model, Grok 4.7,that should be significantlymore capable, is comingout, you know, in a fewshort weeks.GrokBot is also seemsjust as important.It feels kind of likeanother open claw momentfor AI, for this kind ofpersonalization of AI, ofmaking it, democratizingAI, making it easy for-The same thing thatZuckerberg said in hismanifesto.He wants to create bots foreverybody.Yes.And that's what Claudeco-work and Claude tag bothdo exceptionally well, aswell as Hermes, PerplexityComputer, and OpenClaw.I'm just going to admit,for me, some of those arelike, and I try to be verytechnical, but I'm like,RockBot feels-Too hard to use.Yeah.Not too hard to use, but itfeels much easier to use forNo, no, RockBot is mucheasier to use.RockBot is much easier to use.OpenClaw and Hermes, youneed to be-Hermes?

完全同意。而且这还是一个相对小的模型，是 1.5 万亿参数的模型。更大的模型 Grok 4.7 会显著更强，再过几周就出来。GrokBot 看起来同样重要。它感觉有点像 AI 的又一个「OpenClaw 时刻」，是 AI 的个人化、把 AI 民主化、让它变得容易用——就是 Zuckerberg 在他的宣言里说的那件事。他想给每个人造一个 bot。是的。而这正是 Claude Cowork 和 Claude Tag 做得极其出色的地方，Hermes、Perplexity Computer 和 OpenClaw 也是。我得承认，对我来说其中有些是——我一直努力让自己很技术流，但我觉得 GrokBot 感觉——太难用了？不，不是太难用，而是它感觉容易用得多——不不，GrokBot 是容易用得多。GrokBot 容易用得多。OpenClaw 和 Hermes，你得——Hermes？

[1:32:06] Jason Calacanis

Hermes.Are you buying a scarf?Yeah, that's what I call them.Hermes, they're just likethe freedom.That's very funny.If you've ever had thatchannel, it's like that,Saks.Like, our wives order somechannel once in a while,and they break-worked us.It is pretty amazing that insix months he caught up, butSaks, this creates atension.Yeah.This does create a littletension.He's hosting Anthropic.He's catching up toAnthropic.What do you think about thattension here, that he'sthere, Amazon WebServices, Elon WebServices, at the same timeis competing with themincreasingly?

爱马仕（Hermes）。你是要买围巾吗？对，我就这么叫它们。爱马仕，它们就像那种自由。太逗了。你要是有过那张卡就知道了，Sacks，就跟那个一样。就像我们的太太们时不时下单买点香奈儿，然后账单就把我们打垮了。他在六个月里追上来，确实相当惊人。但 Sacks，这就制造出一种张力。对。这确实制造了一点张力。他在给 Anthropic 提供算力托管，他又在追赶 Anthropic。你怎么看这里的张力？他既是「亚马逊云服务」、哦不，「Elon 云服务」，同时又在越来越直接地跟他们竞争。

[1:32:36] David Sacks

What I love about it isthat he's creating thesehuge compute clusters,right?And that gives him theability to train a bigmodel, a big potentiallyfrontier model.So, you can think of thatas like the call option isto catch up and become atrue frontier lab.But if he fails at that,he's got the put option ofjust selling the computeat pretty attractive pricesto the frontier labs.And he has the right, Saks,to take that compute backfrom Anthropic.He put that clause,according to reports,into the Anthropic dealso he could take backcompute if he wants to.Right.Both sides have a 90-daycancellation, which meansthat if the spot price ofcompute goes up, hebenefits.If it goes down, by theway, it could hurt himbecause either side couldrenegotiate within 90 days,right?

我喜欢这件事的地方在于，他在建这些巨型算力集群，对吧？而这给了他训练一个大模型、一个潜在的前沿模型的能力。所以你可以把它理解成一个看涨期权（call option）：追上来，成为真正的前沿实验室。而如果他在那件事上失败了，他还有一个看跌期权（put option）：以相当有吸引力的价格把算力卖给前沿实验室。而且 Sacks，他有权把那部分算力从 Anthropic 手里收回来。据报道他把那个条款写进了和 Anthropic 的协议里，所以只要他想，就可以把算力收回去。对。双方都有 90 天的解约权，这意味着如果算力的现货价上涨，他获益；顺带一提，如果现货价下跌，也可能伤到他，因为任何一方都可以在 90 天内重新谈判，对吧？

[1:33:22] David Sacks

So, basically, Elon is ableto charge spot minus 90.So, that's the risk.But it all goes back towhat we were talking aboutbefore, which is, do youthink that compute will besupply constrained goinginto next year and theyear after?

所以基本上，Elon 能收的是「现货价减 90（天）」。那就是风险所在。但这一切又回到我们之前聊的那个问题：你认为明年和后年算力会不会是供给受限的？

[1:33:38] David Sacks

And if so, that's a veryadvantageous dynamic forElon because, again, he'svery close to chargingtrue spot for hiscompute.But he can alsoessentially reroute it tohis own service if he'sable to take the lead.Yeah.And just...Can I say what'sinteresting?

如果是，那对 Elon 就是极其有利的格局，因为，再说一次，他基本上是按真实现货价在卖他的算力。而如果他能取得领先，他也可以直接把算力调回自家服务上。对。而且——我能说个有意思的点吗？

[1:33:53] Gavin Baker

Sorry to interrupt.No, no.I talk to a lot ofinvestors about SpaceX,okay?Yeah.And what's just amazing tome is everyone will have anuanced view on Starlink.Maybe they're a skeptic onStarlink direct-to-sell.I think the plan to usekind of, you know, thesedistributed Starlink,effectively almost, youknow, we'll call themsmall cells, femto cells,is interesting.Everybody will have anuanced view on orbitalcompute.Whether they're bullish orbearish, they will have aview.They'll have a view onterrestrial compute.Very few of theseinvestors talk aboutGrok at all.And it is Paretodominant on a lot ofmeasures.And we have anexistence proof thatAnthropic went fromwhatever it is, abillion to 50 billionreally fast.And if Anthropic isworth, you know, withtwo trillion per theFT, you know, three orfour trillion, you know,according to yourselfon a previous episode,settle out.I said I thought that wasa market clearing price.I didn't say that'swhere I thought it wasworth.You know, maybe peopleshould start to considerGrok when they thinkabout SpaceX.SpaceX has so many waysto win.And I think that's thehigher bit here is, andit's, and we're obviouslyall talking our bookhere.The folks on thispodcast are amongst the

抱歉打断一下。没事没事。我跟很多投资人聊过 SpaceX，好吧？对。让我特别惊讶的是，每个人对 Starlink 都有一套细致的看法。也许他们对 Starlink 的直连手机（direct-to-cell）持怀疑态度——我觉得那个用分布式 Starlink、实际上相当于小型基站、飞蜂窝（femtocell）的方案挺有意思的。每个人对轨道算力（orbital compute）都有一套细致的看法，不管看多还是看空，他们都有观点。他们对地面算力也有观点。可这些投资人里几乎没有人谈 Grok。而 Grok 在很多指标上都是帕累托占优（Pareto dominant）的。而且我们手上有一个现成的存在性证明：Anthropic 从大概 10 亿美元干到 500 亿美元，快得不得了。如果按《金融时报》说的 Anthropic 值 2 万亿，或者按你上一期自己说的三四万亿，最后落定的话——我说的是我认为那是个市场出清价，我没说我认为它值那么多。那也许人们在想 SpaceX 的时候，该开始把 Grok 也算进去了。SpaceX 有太多条赢的路径。我觉得这才是这里更高维的那一位，而且我们显然都在讲自己的持仓。这档播客上的人，是

[1:35:12] Gavin Baker

earliest investors inSpaceX.I was part of leading, asmall part of leading thesecond round to 10million.Hey, there is one thingthat broke during theshow that we shouldtalk about.And that's at Silver Lakemaybe buying Workday.Workday is up 17%.And if this is true,this is, I think, asignificant moment forthe stock market.You know, software hasbeen in a death spiralthat really kind of hasstarted, you know,that probably, youknow, troughed maybethree, four months ago,but it was, it was in atough spot for 12, 15,18 months.And the return of aprivate equity bid forsoftware really changesthe investing landscape.And I don't know whatthe potential buyers ofWorkday are thinking,but I do know that therise of open source andthe increasingcompetitiveness of opensource is a godsend forthe American softwareindustry.It is an absolutegodsend.Yeah, if you want tothrow up a chart ofWorkday, you know, likea five-year chart, youcan just see it.But a world where youhave one, two, or threedominant frontier models,I mean, you can justsee the stock went from,yeah, we'll call it 300straight down to 100, youknow, in a really shortperiod of time.And even with this bid,you've only recovered alittle bit.But a world where

SpaceX 最早期的一批投资人。第二轮 1000 万美元的融资我也参与领投了一小部分。嘿，节目进行中还有一条新闻爆出来，我们得聊一下：银湖（Silver Lake）可能要收购 Workday。Workday 涨了 17%。如果这是真的，我认为这对股市是个重要时刻。软件板块一直在死亡螺旋里，那大概是从——可能三四个月前触底的，但它有 12、15、18 个月都非常难受。而私募股权对软件重新出价，真的会改变整个投资格局。我不知道 Workday 的潜在买家在想什么，但我确实知道，开源的崛起和开源竞争力的提升，对美国软件产业是天赐之福，是绝对的天赐之福。对，如果你把 Workday 的股价图、比如五年图放出来，你能一眼看出来。但一个只有一两三家主导前沿模型的世界，我是说，你能看到这只股票从——姑且说 300 美元一路直下到 100 美元，而且是在很短的时间里。即便有了这次出价，也只收复了一点点。但一个

[1:36:33] Gavin Baker

Anthropic, OpenAI, and,you know, maybe evenSpaceX are kind of theonly model providers thatdominate the modellayer, that is a hardworld for software.If I were to guess whythe private equity bid forsoftware might be comingback, it's because theysee what's happening withopen source.Pretty amazing.And that is thebackstop, isn't it,Saks?

由 Anthropic、OpenAI、可能还有 SpaceX 垄断模型层、成为仅有的几家模型供应商的世界，对软件公司来说是个很难过的世界。如果让我猜为什么私募股权对软件的出价可能回来了，那是因为他们看到了开源正在发生的事。挺惊人的。Sacks，这就是那个托底，对吧？

[1:36:56] Jason Calacanis

We've talked about this.If nobody in the publicmarkets wants to ownthese companies, andprivate equity sees away,we talked about bendingspoons, I think, lastweek, a whole bunch.There is now, I think,I don't want to call thembottom feeders, but whatwould you call them,Gavin?

我们聊过这个。如果公开市场没人想持有这些公司，而私募股权看到了一条路——我们上周还聊了 Bending Spoons 一大堆。现在出现了一批，我不想叫他们「食腐者」，那你会怎么称呼他们，Gavin？

[1:37:11]

Like, you know, sharpacquirers of moneyprinting businesses thatwant to extract even morerevenue from it.Yeah, Saks?This is a good turn ofevents, that there is abuyer for these companies.Absolutely.I mean, what Silver Lakeis saying, they've beenoversold.They've been oversold,yeah.And so here they go.There must be some valuein that.And if you use AI to runthese, I think that wasthe thing Bending Spoons isdoing really well.They get rid of 80% of theemployees and then put the20% who are left as AIfirst employees, and theyfigure out how to get allof this incredible revenue.There's your breakingnews, Gavin Baker.Amazing.Tied right now, just soyou know, I looked at themetrics before we got ontoday.You're now tied with BradGerstner in terms ofratings.You and Brad Gerstner areneck and neck and tied forfifth bestie, which meansone of you will drop tosixth bestie if thiscontinues.And we'll see.Brad Gerstner may be onnext week, so we'll seewhich bestie takes thefifth slot.David Sachs, great to seeyou.Hey, maybe we get ayear round later.Maybe we can get asteak.Want to get a steaklater?

叫他们「精明的印钞机收购者」吧——买下这些印钞生意，然后想办法从里面榨出更多收入。是吧，Sacks？这是个好转折，这些公司终于有买家了。绝对是。我是说，银湖的意思就是：它们被错杀了。它们被错杀了，对。所以他们就上了。那里面一定有价值。而且如果你用 AI 来经营这些公司——我觉得 Bending Spoons 这件事就做得非常好：他们裁掉 80% 的员工，然后把剩下的 20% 变成 AI 优先的员工，再想办法把这些惊人的收入全部拿出来。这就是你的突发新闻，Gavin Baker。太棒了。顺带一提，我今天上节目前看了下数据：你现在和 Brad Gerstner 打平了，在评分上你和 Brad Gerstner 并驾齐驱，并列第五铁哥们。这意味着你们俩里会有一个掉到第六。我们走着瞧。Brad Gerstner 下周可能会来，所以我们看看第五这个位子最后归谁。David Sacks，很高兴见到你。嘿，也许我们晚点见一面？也许可以吃个牛排。晚点想去吃个牛排吗？

[1:38:17] Jason Calacanis

I got Thursday night offhere.I got a hall passtonight.Oh, okay.All right.Maybe we can go get asteak.This has been an amazingepisode of the All InPodcast for August 14th.We'll see you next time.Bye-bye.We'll let your winnersride.Rain Man David Sachs.And it said we open

我今天周四晚上有空，今晚我拿到「通行证」了。哦，好。行。那我们也许可以去吃个牛排。这是一期非常精彩的 All-In 播客，8 月 14 日这期。下次见。拜拜。让你的赢家继续跑。雨人 David Sacks。我们说我们把它

[1:38:39]

source it to the fans andthey've just gone crazywith it.Love you, Westies.I'm queen of kid wine.I'm going all in.What your winnersride.What your winnersride.Besties are gone.Go 13th.That's my dog takingnotice in your driveway.Sex.We're not a bitch.Oh, man.My habbitasher will meetme at what it's like.We should all just get aroom and just have one bighuge orgy because they'reall just useless.It's like this like sexualtension that they just needto release somehow.Wet yourbaby.Wet yourfeet.Wet your feet.We need to getmerch.These are back.I'm going all in.I'm going all in.

开源给了粉丝，然后他们就玩疯了。爱你们，铁哥们。（片尾曲及历年金句混剪，ASR 识别混乱，以下为大致内容）我是「酸雨女王」。我要全押（going all in）。让你的赢家继续跑。让你的赢家继续跑。铁哥们都走了。第 13 号。那是我家狗在你车道上撒欢。Sacks。我们不是孬种。哦，天啊。我的男装师傅会告诉我那是什么感觉。我们干脆都开一间房，来一场大杂烩算了，因为他们全都没用。就像那种性张力，他们就是得找个地方释放出来。全押。全押。我们得做周边。这些又回来了。我要全押。我要全押。