Dario Defends Himself, Datacenter Panic, AI Doomer Trap, Senate Toss-Up

[0:00]All right, everybody, welcome back to the number one podcast in the world, your favorite podcast, the All-In podcast. We got a huge, huge docket with me again, the core four. 好了各位，欢迎回到世界第一的播客、你们最爱的播客——All-In 播客。我们今天的议程超满，和我一起的还是核心四人组。

[0:08]Chamath, Freeberg, Sachs, and your boy J Cal. Topic one. Topic one. Let's get right to it. No small talk, boys. Dario has spoken. Chamath、Freeberg、Sachs，还有你们的 J Cal。第一个话题。第一个话题。直接进入正题，不寒暄了各位。Dario 发话了。

[0:17]Apparently, he listens to the All-In podcast, and he was compelled to respond to Gavin Baker's little discussion last week that many people inside, or some people, allegedly, maybe inside of Anthropic believe, quote, that they will be 显然他是听 All-In 播客的，而且他忍不住要回应上周 Gavin Baker 的那番讨论——据说 Anthropic 内部很多人，或者说某些人，据称也许相信，引述一下，他们将会成为

[0:34]the only private company in the world at some point. Some folks from Anthropic, Sholto Douglas, fired back, completely false. Whoever told Gavin this is lying. 世界上唯一的私营公司。Anthropic 的一些人，比如 Sholto Douglas，直接开怼：完全是假的，跟 Gavin 说这话的人在撒谎。

[0:44]And Dario decided he would come out of hiding, and he entered the chat with an epic two-part essay on regulation. 而 Dario 决定结束隐身状态，亲自进场，发表了一篇关于监管的史诗级两部长文。

[0:53]Sachs, he said, quote, "I know the Silicon Valley shorthand where regulation equals regulatory capture equals concentration of power, but I've always found this to be an overly simplified picture." Sachs，他说，引述：「我知道硅谷的速记法——监管等于监管套利等于权力集中，但我一直觉得这是一个过度简化的图景。」

[1:04]On messaging, he said, quote, "I do not agree my messaging has been disproportionately negative. It has been about equally balanced between risks and benefits." Took a little responsibility, quote, "By far the most accurate criticism of AI companies, including Anthropic, is that we haven't yet 关于宣传口径，他说，引述：「我不同意我的信息传递过于负面。它在风险和收益之间基本是对半平衡的。」也稍微认了点错，引述：「对 AI 公司（包括 Anthropic）最准确的批评是，我们还没有

[1:21]delivered on our big promises to benefit the world. That is totally on us." Most quoted line, "The thing that will work is actually curing cancer, not glitzy marketing." Questions. 兑现造福世界的宏大承诺。这完全怪我们自己。」被引用最多的一句：「真正管用的是治好癌症，而不是浮华的营销。」开始提问。

[1:32]Sachs, uh, you've been talking about Anthropic for a year and a half. He never responded. Sachs，你谈 Anthropic 谈了一年半了，他从来没回应过。

[1:38]Gavin comes out, and he suddenly, uh, is very talkative and responds. What do you take from the fact that he never responded to you, and then he responded to Gavin, and then we'll get into the details. Does he not want to spar with you? Gavin 一出来说话，他突然就变得很健谈、出来回应了。他从不回应你、却回应了 Gavin——你从中读出了什么？然后我们再谈细节。他是不是不想跟你过招？

[1:51]>> he did respond to me. I mean, I I want to center myself, but the reality is he did not respond to what Gavin said. He easily could have said that what Gavin said he said is untrue, like Shalto did. He did not weigh in on that issue. Instead, he >> 他确实回应过我。我是说，我不想往自己脸上贴金，但事实是他没有回应 Gavin 说的内容。他完全可以像 Shalto 那样说 Gavin 转述的话不属实。他没有就那个问题表态。相反，他

[2:07]really weighed in on the regulatory capture issue. So, that's the thing that I think drew him out here is I think he is a little bit irked by the fact that we keep accusing him of regulatory capture. And this is something I've been saying since last year that they were engaged in a sophisticated regulatory 真正回应的是监管套利的问题。所以我认为把他引出山的正是这个——我们不断指控他搞监管套利，这事有点惹恼他了。我从去年就在说，他们在搞一套基于贩卖恐惧的、精密的监管

[2:22]capture scheme based on fearmongering. And those are the points he's really responding to. And I think the thing that he doesn't like about it is that he feels like it impugns his motives, that he's not pure in pursuing all these regulations for the right for the right reasons. And my response to that is 套利方案。他真正回应的就是这些观点。我觉得他不爽的地方在于，他觉得这玷污了他的动机，好像他推进所有这些监管不是为了正当理由、不够纯粹。而我的回应是：

[2:40]look, there's no question that Anthropic has been extremely aggressive about seeking to implement its preferred regulatory frameworks at both the state and federal level. That is regulatory capture. Now, look, I think that Dario may be sincere in his beliefs. I'm not 听着，毫无疑问，Anthropic 在州和联邦两级都在极其激进地推动实施它偏好的监管框架。这就是监管套利。当然，我认为 Dario 的信念可能是真诚的。我不是

[2:55]saying he's doing this for pecuniary reasons. Nonetheless, he is seeking to capture the machinery of the state on behalf of his political agenda. It's just there's not much to debate there, I don't think. 说他为了钱这么做。尽管如此，他就是在试图为代表他政治议程的利益去俘获国家机器。这里真的没什么好争的。

[3:09]There's a few other things to say about this. So, let me go to point number two first. He says that he's not responsible for the negativity around AI, that he's been balanced in his portrayal of AI. Oh, really? 还有几点要说。先说第二点。他说自己对 AI 的负面舆论没有责任，说他对 AI 的描绘是平衡的。哦，是吗？

[3:23]Do you you expect us to have amnesia about the fact that just about a year and a half ago, you said very specifically that within 1 to 5 years 50% of entry-level knowledge workers would lose their jobs. And this was not just some sort of idle speculation on a podcast. This was something that they 你指望我们集体失忆吗？就在大约一年半之前，你非常明确地说过：1 到 5 年内，50% 的入门级知识型工作岗位会消失。这可不是播客上的一句闲聊猜测。这是他们

[3:40]engineered and amplified into a full-fledged media campaign. You may recall that President Obama even retweeted this. And many other Democratic Party officials did the same thing. Moreover, this pattern repeats itself with the whole Anthropic blackmail study. Remember, he's got a 精心策划并放大成了一场完整的媒体攻势。你可能还记得，奥巴马总统都转发了这条，还有很多民主党官员也转了。而且这个模式在 Anthropic 那个「勒索研究」里又重演了一遍。记住，他在公司里养着一个

[3:56]team at the company, they're supposedly studying alignment. They create this highly contrived study where they prompt this model over 200 times until they finally get the headline-grabbing result that they wanted. And again, my point here is that this is not just someone 团队，号称研究对齐。他们炮制了一个高度人为设计的研究，给模型连续提示 200 多次，直到终于得到他们想要的、能上头条的结果。我再说一遍，我的重点是：这不是某个人在播客上

[4:12]dishing on a podcast about the alignment risk of AI. This is them actually creating a study, engineering that study, and then pumping it to create the headlines they wanted. And Dario and his sister, who's the president of the company, and other executives went on 60 Minutes and breathlessly pumped this 随口聊聊 AI 的对齐风险。这是他们真的炮制了一个研究、设计了这个研究、然后猛推它来制造他们想要的头条。Dario 和他妹妹（公司总裁）以及其他高管上《60 分钟》节目，绘声绘色地猛炒这件事。

[4:29]thing. And since that study has come out, a lot of people have looked at it and said, "Look, this thing is highly contrived." Including the UK AI Safety Institute, basically said that this thing was created under highly pressurized conditions, and they engineered the result. So, these guys 自从那个研究发布后，很多人看过之后都说：「这东西高度人为设计。」包括英国 AI 安全研究所也基本上说，这个东西是在高压条件下造出来的，结果是设计出来的。所以这些人

[4:46]have created campaigns to make people fear AI. Why is the public so afraid of AI? They think it's going to take their jobs, they think it's going to become the Terminator. 发起了一场又一场让公众恐惧 AI 的运动。为什么公众这么怕 AI？他们觉得 AI 会抢走工作、会变成终结者。

[4:55]It's exactly the fears that Anthropic has put in the media bloodstream. And no company, and you'd have to say, therefore, no founder CEO has done more to promote these fears and pump these fears and amplify these fears than Dario has. And the fact that he can't see that 这些恐惧恰恰是 Anthropic 注入媒体血脉的。没有哪家公司——因此也得说没有哪个创始人 CEO——比 Dario 做得更多去推广、炒作、放大这些恐惧。而他居然看不到这一点，

[5:12]either makes him I'd say disingenuous or delusional. >> I'm going with the latter. 这要么说明他不真诚，要么说明他在自欺。>> 我押后者。

[5:18]>> So, that that's point number one. And I really think that almost everybody who read that, who's been paying attention, felt the same way. It's like, "Dario, you really have a blind spot with respect to your own behavior." >> 这是第一点。我真的认为，几乎所有一直关注这事、读过那篇长文的人都有同感：「Dario，你对自己言行的盲点也太大了。」

[5:29]>> By the way, can we can we talk about that one for a second or no? >> Yeah, let's do it. Yeah. >> Just to build on what you're saying, the issue that we have right now is actually pretty problematic. If I had to take three different vectors of actions and now put them together, you've had a very disjointed, dis- >> 顺便说一句，我们能就这一点展开聊几句吗？>> 行，聊。>> 在你说的之上再补充一下：我们眼下的问题其实相当棘手。如果把我看到的三条行动向量放在一起看：你面对的是一个涣散、无组织、

[5:44]organized, and frankly, pearl-clutching, disingenuous AI doom narrative. 而且坦白说既圣母又虚伪的 AI 末日叙事。

[5:53]And it's actually seeded and now it's growing. 它已经被种下了，而且正在生长。

[5:58]And as a result, what's happening? I was really shocked when I saw Governor Abbott of Texas put down this executive order around Texas data centers and limiting its growth. Josh Shapiro in Pennsylvania just did the same thing. 结果呢？当我看到德州州长 Abbott 签署行政命令限制德州数据中心增长时，我是真的震惊了。宾州的 Josh Shapiro 随后也做了同样的事。

[6:15]Axios reported that the GOP party sent out a memo to the executives at the various AI companies essentially saying, "Hey, stop rage baiting people because we're about to lose this critical GOP Senate race in Ohio largely around this issue." Axios 报道说，共和党给各家 AI 公司的高管发了一份备忘录，大意是：「嘿，别再煽动群众情绪了，因为我们快要在这个议题上输掉关键的俄亥俄 GOP 参议院席位了。」

[6:33]So, you have doomerism and then you now have mainstream political pushback from both sides of the aisle, which I think is extremely dangerous. 所以，一边是末日论，一边是两党主流政治力量的反弹——我认为这极其危险。

[6:44]And then the third leg of the stool is rising yields. By the way, the Treasury Department just doubled its bond buying. 第三条腿是不断上升的收益率。顺便说一句，财政部刚刚把债券购买规模翻了一倍。

[6:52]If you put all of these things together, as a just a casual observer, what I would say is we have made some really stupid unforced errors in the United States. And the coda to all of this is if you put all of these things together, rising yields constrains the money 把这些放在一起，作为一个普通旁观者我会说：我们在美国犯了一些非常愚蠢的无谓失误。而这一切的尾声是：收益率上升会收紧货币

[7:08]supply in terms of how people want to invest in risk seeking assets because the risk free rate just keeps going up. 供给——就人们投资风险资产的意愿而言，因为无风险利率一直在涨。

[7:16]That causes people to be less likely to invest in all the data centers. Data centers themselves are being shut down. 这让人们更不愿意投资数据中心。而数据中心本身正在被关停。

[7:22]Where does that leave frontier model companies? They're the most at risk. Why? Because they're not investment grade. They're in the worst position in terms of their balance sheets. They're not in a position to then get access to the compute they need to be successful. 那前沿模型公司怎么办？它们风险最大。为什么？因为它们不是投资级。它们的资产负债表状况最差。它们没有条件获得成功所需的算力。

[7:36]So, even as you grow, you're actually making your problem worse in the sense that you need more and more capacity to sustain that growth that you can't fund because all the negativity has turned people against you. So, all roads lead to a big question mark and a and a 所以你越是增长，问题反而越严重——因为你需要越来越多的算力来维持增长，而你又融不到钱，因为所有这些负面情绪已经把大家推到了你的对立面。所以所有路都通向一个大大的问号，

[7:51]head-scratching moment for me where I'm like, "How did we end up in this situation?" But, whether it was accidental or whether it was well-intentioned risk management, whatever it was, we are now in a very precarious situation that we could have frankly avoided. 让我百思不得其解：「我们是怎么走到这一步的？」但无论这是意外还是出于善意的风险管理，我们都已经处于一个非常 precarious 的境地——而这一切本来完全可以避免。

[8:07]>> Totally agree. Freebert, your thoughts on this anti-data center movement or on Dario's going direct and writing the double missive. Do you think he's sincere, delusional, some combination? >> 完全同意。Freeberg，你怎么看这波反数据中心运动？或者怎么看 Dario 直接下场写这两篇檄文？你觉得他是真诚、自欺，还是两者的混合？

[8:22]>> Why don't we steel man it? Yeah. And if we put ourselves in the shoes of a frontier lab leader and the leader is seeing capabilities of models that perhaps we don't see because we're working day-to-day in enterprise settings >> 我们何不替他做个最强辩护（steel man）？如果我们把自己放在前沿实验室领导者的位置上——这位领导者看到的模型能力，可能是我们看不到的，因为我们每天是在企业场景里干活，

[8:38][clears throat] making stuff doing things in a benign, non-nefarious way. 【清嗓子】做东西、做事，都是良性的、无害的用法。

[8:45]And they're sitting there and they're pressure testing models and they figure out, "Hey, there's ways you can jailbreak models to get them to do things that they're not supposed to do. 而他们坐在那里对模型做压力测试，然后发现：「嘿，有办法越狱模型，让它做不该做的事。

[8:53]And then you can use them to do things like design bio weapons, make viruses, make infectious diseases that don't exist today. 然后你可以用它设计生物武器、造病毒、造出今天不存在的传染病。

[9:03]You could use them to design cyber weapons to get them to go and do things online like break into bank accounts and take people's money and all these sorts of things. 你可以用它设计网络武器，让它们在网上干各种事，比如黑进银行账户、偷别人的钱等等。

[9:15]And you can prompt them to do things that could broadly influence society by creating fake media, fake videos, fake images, propagate them through social media, hack the algorithms. I mean, X's algorithm is public, so theoretically an 你还可以提示它做广泛影响社会的事——制造假媒体、假视频、假图片，通过社交媒体扩散、操纵算法。我是说，X 的算法是公开的，所以理论上一个

[9:32]agent could read it, copy it, understand it, and figure out a way to manipulate the broad public. agent 可以读它、复制它、理解它，然后找到操纵大众的方法。

[9:37]And so you realize that as you're building these tools where there's a lot of benign positive world-leading progressive kind of applications, you see all the bad stuff that could happen with it. 于是你意识到：当你在造这些工具时——它们有大量良性的、正面的、世界领先的、进步性的应用——你也看到了它可能带来的所有坏事。

[9:48]And you can keep trying to build systems and keep trying to build ways to stop it from doing that. You can build all your own code that reviews every prompt that comes in, that reviews every API call. 你可以不断构建系统、不断想办法阻止它作恶。你可以写自己的代码审查每一条进来的 prompt、审查每一次 API 调用。

[9:58]And yet somehow a network of agents can be used to do things. And they can start to do things that we don't want them to do. And so let's assume that you're sitting there and you're the executive of the company building this great technology. You believe fully in its capabilities and its benefits to 但不知怎么的，一个 agent 网络还是可以被用来干坏事，开始做我们不希望它们做的事。现在假设你就坐在那里，你是这家建造伟大技术的公司的高管。你完全相信它的能力、它对

[10:13]society, its benefits to the economy. But you can't get a grip on how to keep it under control and how to stop social engineering, cyber weapons, and cyber weapons. What do you do differently? 社会和经济的益处。但你没法完全控制住它、没法阻止社会工程攻击和网络武器。你会做什么不一样的事？

[10:24]>> Okay. Okay. So you're bringing up it's an exceptional going down an exceptional path. I love this. >> 好好好。你这个思路太精彩了，我爱死了。

[10:30]There is one thing that you must do. If you believe that this is true, to further steal my thunder. You cannot obfuscate the thinking tokens of a model then. 如果这是真的，有一件事你必须做——继续偷我的雷。你就不应该对模型的思维 token（thinking tokens）进行遮蔽。

[10:40]What do I mean? If you just compare open-source models versus the closed frontier models. 什么意思？你只要比较一下开源模型和闭源前沿模型。

[10:49]There is no way for us to understand what the thinking tokens are. Meaning you give it a prompt, you get back something, but all of the stuff in the middle is obfuscated. Quad obfuscates it. 我们根本没法理解思维 token 是什么。也就是说：你给它一个 prompt，它返回一个结果，但中间的所有过程都被遮蔽了。Claude 遮蔽了它。

[11:05]Okay? So do the open AI models. And I would say if I was really afraid, the first thing I would do is try to find a way to unobfuscate these thinking tokens so that you could have a third party actually legitimize what you're saying. 对吧？OpenAI 的模型也一样。而我要说，如果我真的害怕，我做的第一件事就是想办法把这些思维 token 解蔽，让第三方能真正为你的说法背书。

[11:20]Because what's interesting to me is I haven't seen the same doomerism from any open-source model company, including Nvidia. And the thing with the open-source models is you can actually see it thinking. So, you can see its misalignment in real time. And 因为让我觉得有意思的是：我没有从任何一家开源模型公司（包括 Nvidia）身上看到同样的末日论。而开源模型的好处是你可以看到它在思考，你可以实时看到它的失配（misalignment）。所以

[11:37]so, instead, we have to basically agree to their interpretation of tokens that we can't see, that only they can see, that they can interpret how they want. And without judging whether it's right or wrong, I would say give them the benefit of the doubt that they're trying to do the right thing. 现在的情况是，我们只能接受他们对那些我们看不见、只有他们看得见、想怎么解释就怎么解释的 token 的解读。先不评判对错，我愿意假设他们是在做正确的事。

[11:53]Make the bloody thing available to other people then. Let somebody else with some amount of judgment as well sit in the room and steel man it with you. 那就把这个该死的东西开放给别人看啊。让其他有判断力的人也坐进房间里，跟你一起做最强论证。

[12:01]>> So then, let me make this point. So, Sachs, there was this argument that Demis put out for a FINRA-style regulatory system, where industry participants that understand technology, move quickly, and are aligned with commercialization actively review each other's systems and >> 那我来说一点。Sachs，之前 Demis 提出过一个 FINRA 式监管体系的说法：由懂技术、动作快、与商业化利益一致的行业参与者，通过一个行业联盟来主动审查彼此的系统和

[12:16]each other's models via an industry consortium that hires and assigns scientists to work on this stuff, and they're just doing it, versus having it be owned and managed by the government and create some regulatory capture scheme. But then, this week, you were pretty against the FINRA system in your 彼此的模型——联盟雇佣并指派科学家来做这些事，他们就自己干了——而不是由政府拥有和管理、搞出一套监管套利的机制。但本周你在推文里挺反对 FINRA 体系的。而 Demis 最初

[12:32]tweets. You seemed to be supportive of it the first time around when Demis originally published it. 发布它的时候，你似乎是支持的。

[12:39]It seems to me that the proposal from Demis that Dario signed up to, Sam signed up to, Elon all signed up to, you're now against, but that could be a solution to, if there is genuine concern here, given Chamath's points. That the industry can can can kind of police each 在我看来，Demis 的那个提案，Dario 签了、Sam 签了、Elon 也签了，你现在反对它——但如果这里真有实质担忧，考虑到 Chamath 的观点，它倒可能是一个解决方案：让行业彼此监督。

[12:54]other. >> Let me let me clarify that. I think designing a new regulatory agency for AI that's designed after FINRA, which is a pretty sleepy regulatory agency that writes rules for stock brokers in a fairly mature industry that's not >> 让我澄清一下。我认为照着 FINRA 的样子给 AI 设计一个新监管机构是个可怕的主意。FINRA 是个相当死气沉沉的监管机构，给一个根本不动态的成熟行业里的股票经纪商写规则，

[13:10]dynamic at all, is a horrible idea. And by the way, FINRA is widely perceived as just protecting incumbents. And talk to any startup founders had to deal with them, they view FINRA as existing to protect the big banks from disruption. It's It's 顺便说一句，FINRA 被广泛认为就是在保护既得利益者。去问问任何跟它打过交道的创业公司创始人，他们都认为 FINRA 的存在就是为了保护大银行免受颠覆。这

[13:26]just not a good model to base regulation after. So, now Demis, hold on. >> the NFA, the National Futures Association, which is also one of these self-regulatory organizations in the financial markets, is a better alternative cuz that's a pretty dynamic market. 实在不是个值得效仿的监管模板。所以，Demis，等等。>> 全国期货协会（NFA）也是金融市场里这类自律组织之一，是更好的替代方案，因为那是个相当动态的市场。

[13:40]>> so now what you're saying is very interesting, which is you mentioned the term self-regulatory organization or SRO, and there's many different models. And one of them may work, and that's been my point. The one I think is the most appropriate for this >> 你现在说的很有意思——你提到了「自律组织」即 SRO 这个词，而它有很多不同模式，其中某个可能是行的，这正是我一贯的观点。我认为最适合当前

[13:55]circumstance would be something like the MPAA, which does not report to the government at all, and was formulated by the motion picture industry in response to the Hays Code. And what the government said is, "Get your act together, guys, or we're going to have to do it for you." And and what the MPAA 情形的是类似 MPAA 的东西：完全不向政府汇报，由电影业在《海斯法典》压力下自发成立。当时政府说的是：「自己管好自己，不然我们就替你们管。」而 MPAA

[14:12]did then was promote standards that then forestalled more intrusive, heavy-handed government action. So, that is basically the model that I said I could get behind. But is that the model that Dario supports? No. Dario, and I think Demis, wants a FINRA, which is basically this 随后做的事，就是推广行业标准，从而挡住了政府更具侵入性、更强硬的干预。所以那基本上就是我说我能支持的模型。但 Dario 支持的是这个吗？不是。Dario，我认为还有 Demis，想要的是一个 FINRA——本质上是一个

[14:29]rule-making regulatory agency that would not only promulgate standards, but also do pre-release model testing. 不但制定标准、还要做模型发布前测试的规则制定型监管机构。

[14:37]Okay? And it would then report into the government and then officially be enshrined in law. And the reason why everyone says, "Well, Elon's behind it." 然后它要向政府汇报，并最终正式写进法律。而所有人都说：「Elon 支持这个。」

[14:44]I've talked to Elon about this. Elon does not support a FINRA for AI. What he could get behind is an MPAA for AI. And that's my point is Demis wrote something >> what I've been saying on this podcast, Zach, since the beginning. I have been saying MPAA self-regulate. 我跟 Elon 聊过这件事。Elon 并不支持 AI 版 FINRA。他能接受的是 AI 版 MPAA。我的重点就是：Demis 写的那个东西——>> 这正是我从这个播客第一天起就在说的，Sacks。我一直说 MPAA 式自律。

[15:00]>> I get it, but look, here's my point. Demis wrote something very high-level. And so, when you say something like, "Well, everyone supports Demis's idea." >> 我明白，但听我说。Demis 写的东西非常高层、非常笼统。所以当你说「大家都支持 Demis 的想法」时，

[15:08]There's many versions of Demis's idea. Some of them would be okay, and some of them are bad. And what I'm saying is if it's a FINRA model, that will not work. Demis 的想法有很多个版本，有的还行，有的很糟。我要说的是：如果是 FINRA 模式，那就行不通。

[15:16]>> Having spoken to Demis at length about this, I think that the the general framework is SRO in the right structure. However, FINRA is often used as the example of an SRO simply because it is the most well-known SRO. >> 我跟 Demis 就这个问题深聊过，我认为总体框架是「结构正确的 SRO」。不过 FINRA 之所以总被当作 SRO 的例子，仅仅因为它是最有名的 SRO。

[15:33]And but you are correct. I think that there's a mechanism by which this thing needs to operate that is distinct from any other SRO. But the idea being the industry gets together and to Chamath's point, they get to look at each other's models, they get to police it, they get to have independence, but they don't get to be controlled by government. And that 但你说得对。我认为这个东西需要一套不同于任何现有 SRO 的运作机制。但核心思想是：行业聚到一起，而且像 Chamath 说的，他们可以看彼此的模型、可以监督它、可以保持独立性，但不被政府控制。这样——

[15:49]>> Well, here's what I think is important. And and just to be clear, this FINRA for AI AI idea, Dario's behind the demonstration of that. >> 好，我认为重要的是这点。先说清楚，这个「AI 版 FINRA」的想法，Dario 是支持的。

[15:56]>> an SRO for AI. SRO. >> It's not an SRO because it's not self-regulating. It's a regulatory organization and the S part is a total fig leaf designed to hide the fact that this is a new regulatory agency that reports to the government that does pre-release testing and approval of >> 「AI 的 SRO」。SRO。>> 它不是 SRO，因为它不是自律的。它是个监管组织，那个「S（self 自律）」完全是遮羞布，用来掩盖这是一个向政府汇报、对模型做发布前测试和审批的

[16:11]models. I call it a DMV for AI because I think what's going to happen is all these models are going to get lined up in a queue waiting to get their test done and then released and it's going to slow us down horribly. Have you ever been to the DMV in Arizona? 新监管机构的事实。我管它叫「AI 版车管所（DMV）」，因为我觉得接下来会发生的是：所有模型排成长队等着做测试、等放行，这会把我们拖慢到惨不忍睹。你去过亚利桑那的车管所吗？

[16:25]>> It's a terrible idea and it's just a repackaging of what Dario has always wanted, which is an FDA for AI. Now he's saying it's an FAA for AI. You know how long the FAA and the FDA take to approve new products? Years. Years. >> 这是个可怕的主意，它就是 Dario 一直想要的东西换了个包装——AI 版 FDA。现在他改口叫 AI 版 FAA 了。你知道 FAA 和 FDA 批准一个新产品要多久吗？好几年。好几年。

[16:39]>> Yeah. Yeah, this is China's China's advantage. Yeah, exactly. >> China's going to win if we do this and and it will become pure regulatory capture. >> 是啊。是啊，这就是中国的优势。对，没错。>> 如果我们这么干，中国就会赢，而且这会变成纯粹的监管套利。

[16:46]>> You should look at China's rocket that they landed yesterday. You know it's completely steel and there's a a good analysis that was done on how quickly they were able to do it from the last time they had an explosion of this, then they got full steel done and Elon's still waiting for FAA approval on his >> 你应该去看看中国昨天着陆的那枚火箭。它是全钢的，有很好的分析讲他们从上一次爆炸到这次全钢完成有多快——而 Elon 的全钢 Starship 还在等 FAA 审批。

[17:01]steel Starship. You could actually watch it playing out in China where their regulatory agencies are allowing them to move much more quickly and as a result, they're starting to gain real traction in the space race. And then separately, a lot of biotech has moved offshore from the US for the same reason that you're 你可以眼看着这一切在中国上演：他们的监管机构允许他们快得多地推进，结果是他们在太空竞赛中开始获得真正的牵引力。另外，大量生物科技已经因此流出美国，原因就是你指出的这个：

[17:18]pointing out, which is the FDA approval cycle is so long and the stages slow people down so much. And China's completely revamped their rules for getting to clinicals and getting to phase three and then completing that all the biotech investing in the world is now moving to China. FDA 审批周期太长、各阶段拖得太久。而中国彻底改革了进入临床、推进三期、完成三期的规则——结果全世界的生物科技投资都在流向中国。

[17:33]>> The fact that Dario could want and has called for, including in very recent blogs, a regulatory apparatus modeled after the FAA and that that is the right model for AI in my view is completely discrediting. And yes, we will get >> Dario 想要、并且一直在呼吁（包括在最近的博客里）一个效仿 FAA 的监管机构，还认为那是对 AI 的正确模型——在我看来，这完全是自毁信誉。而且没错，我们会

[17:48]smoked by China because they will not subject all their companies to these constraints. Now, >> Okay, so let's play this out. Let's play this out. Yeah. 被中国打得落花流水，因为他们不会让他们的公司受这些约束。好了——>> 那我们来推演一下，推演一下。

[17:56]>> go back to your steel man for a second. >> Yeah, yeah. >> Just because I believe that we should not have a DMV for AI does not mean that I'm against AI safety. So let's talk about some of the things that can be done. So first of all, I think it's good that these frontier labs take safety >> 回到你的最强论证。>> 好好好。>> 我反对「AI 版车管所」，不代表我反对 AI 安全。所以我们来聊聊哪些事是可行的。首先，我认为前沿实验室重视安全是好事。

[18:12]seriously. Sam just came out with a post, I believe, yesterday saying that actually they're going to slow down the release of some of their frontier models because they felt that their safety stack was not keeping up with the capability. Okay, that's responsible. By the way, do you know why I think they're Sam 昨天好像刚发了个帖子，说他们实际上要放慢一些前沿模型的发布，因为他们觉得安全栈跟不上能力了。行，这是负责任的。顺便说，你知道我为什么觉得他们会这么做吗？

[18:27]doing that? Because they have huge and ample liability to take responsible steps because if they don't, they're going to be subjected to all sorts of product liability lawsuits. Look at what's happening to Meta right now. 因为他们有巨大的法律责任去采取负责任的行动，否则他们会面临各种产品责任诉讼。看看 Meta 现在的遭遇：

[18:39]They're getting sued because supposedly their product led to social media addiction and they're getting sued for over a trillion dollars and this comes on the heel of them losing a $6 million jury trial in Los Angeles where a plaintiff sued them for creating her body image issues. So, we are seeing 他们被起诉，罪名据说是产品导致社交媒体成瘾，索赔超过一万亿美元；此前他们刚在洛杉矶输掉了一场 600 万美元的陪审团审判，原告起诉 Meta 造成她的身体意象问题。所以我们正看到

[18:56]massive product liability lawsuits against big tech companies like Meta and Google and if OpenAI and Anthropic are smart, and I think they are, they do not want to be next. So, these companies have ample reasons for liability protection not to just race out their 针对 Meta、谷歌这样大科技公司的大规模产品责任诉讼。如果 OpenAI 和 Anthropic 聪明——我认为他们聪明——他们不会想成为下一个。所以这些公司有充分的动机出于责任保护而不把产品一股脑抢发出去，

[19:13]product, okay? Safety is part of what they have to get right. Now, how should the industry coordinate with each other? 安全是他们必须做对的一部分。那么，行业之间该怎么协调？

[19:20]There are tried and true ways of doing this. Companies in the industry publish papers, they publish technical papers, they publish blogs. These are good ways of sharing best practices. They attend symposia and conferences, and they present at these things. They do request for comments as part of open standard 有久经考验的方法。行业里的公司发论文、发技术论文、发博客，这些都是分享最佳实践的好方式。他们参加研讨会和行业会议并做演讲。他们搞征求意见（RFC），作为开放标准

[19:37]setting exercises. There are ways of the industry collaborating on safety in very productive ways. But, what is the common denominator of everything I just said? They are all open. They are not in secret, and that makes those judgments contestable to the 制定流程的一部分。行业有很多高效协作安全事务的方式。但我刚才说的一切的共同分母是什么？全都是开放的。不是秘密进行的，这就让那些判断可以被质疑和检验——正呼应

[19:53]point that Jamath made earlier about we need to see the thinking behind this. And what I'm against is these companies meeting in secret, whether it's part of some sort of like FINRA or otherwise, to set standards that are not transparent, that are uncontestable. And you know Chamath 刚才说的：我们需要看到背后的思考。而我反对的是这些公司秘密开会——不管是某种 FINRA 还是别的什么——制定不透明、不可质疑的标准。你知道那会产出什么吗？监管套利和反竞争行为。所以，存在既能保护公众、又开放透明的方法，那才是我们应该走向的方向。

[20:10]what's going to come out of that? Regulatory capture and anti-competitive behavior. So, there are ways of doing this that will protect the public that are open and transparent. That is what we should lead towards. >> 不——你不会看到它发生的。我告诉你为什么。还是那句话，技术细节很重要。

[20:20]>> not You will not see it happen. And I'll tell you why. Just again, the technical details matter. 如果竞争对手 A、B、C 突然能看到你的模型从输入 token 到输出 token 过程中生成的每一条 trace 和 span，你基本上就把自己的配方送出去了大半。

[20:26]If all of a sudden competitor A, B, and C could see every single trace and span that your model generated as an artifact of going from input tokens to output tokens, you're essentially giving away uh a lot of your recipe. 而且我不相信——连我这个 API 客户、Anthropic 都不愿意把我自己公司的思维 token 给我看——他们更不可能把它拿出来跟 OpenAI 和房间里的任何人分享。

[20:43]And I just don't think that if Anthropic isn't willing, even if I'm an API customer, to share my thinking tokens from my company with me, they're sure as not going to share it with OpenAI and anybody else in a room. 行吧。>> 好，我们再做一个最强论证。假设 Dario 成功了。

[21:00]Okay? >> Okay, let's And let's just do one more steel man. So, let's say that Dario's successful. J Cal，这个问题我来问你，因为这一段的主持人已经被我接管了。>> 行，我是说……不，Sacks 显然是这方面的行家，我想听完整的。不不，老实说，这是 Sachs 的主场，我想听你的完整立场。

[21:05]And Jaq, I'll ask you the question cuz I've taken over as moderator of this segment. >> Yeah, no, I mean I No, Sachs has This is Sachs's wheelhouse. I wanted to hear your whole No, no, to be honest, this is Sachs's wheelhouse. I wanted to hear your whole position on it. >> 假设 Dario 成功了。真的建立了一套监管体系，而且正如 Sachs 所说，它确实拖慢了行业、拖慢了模型开发。

[21:17]>> Let's just say that Dario's successful. There's some regulatory scheme that's created that does in fact, to Sachs's point, slow down the industry, slow down model development. 而中国的前沿实验室取得重大进展、开始反超，他们搞出了 RSI——递归自我改进——他们的模型成了世界最好的，而且只会越来越好。先从你开始。

[21:26]And China's frontier labs make significant progress and they start to leap ahead and they get this RSI, this recursive self-improvement going and their models become the best in the world and they only get better and better and better. we'll start with you. >> 好。>> 我们挨个来。>> 好的，世界上最伟大的主持人。>> 在一个中国拥有最好 AI、拥有所有数据中心的世界里，这对美国工人、美国企业和美国消费者意味着什么？我们的世界会发生什么？后果是什么？因为

[21:40]>> Okay. >> we'll go around the horn. >> Okay, world's greatest moderator. >> What is the implication for the American worker, for the American business and for the American consumer in a world where China just has the best AI and they have all the data center? What happens in our world? What is the consequence cuz 大家一直在听「中国会赢、中国会赢」。那又怎样？这到底意味着什么？>> 我不认为中国会赢。我觉得不会有一个赢家通吃的局面，我认为会是渐进式改良。所以我不

[21:56]everyone keeps hearing, "Oh, China will win. China will win." Who cares? What does that actually mean? >> I don't think China is going to win. I think it'll be, you know, there's not going to be one person who who takes the whole prize. I think it's going to be incremental improvement. So, I don't 觉得那有多重要。我最近一直在想数据中心反弹为什么发生——看到 Abbott 和 Shapiro 跳出来真的挺让人震惊的，因为这俩人本来是最支持数据中心的，对吧？如果你往下深挖一层，

[22:11]think that that matters. I you know, I've been thinking about why this backlash is happening actually for the data centers and it was kind of shocking to see Abbott come out and Shapiro come out because those two people were the most pro data centers, right? And if you kind of double click 他们只是想确保你没有接入电网……我们节目里聊过的用水问题其实不算太严重，但我觉得这非常表演性质。我和你在私下频道聊过：他们俩这么做，是因为他们的

[22:27]on it, they just want to make sure you're not tapping into the grid and you know, the water concern which we've talked about here is not too serious, but I think it was very performative and I think you and I were talking about this on the back channel is they both are doing this because their 选民、投票给他们的人，真的很在意这些数据中心，并且想让它们停下。那你就要倒推：为什么选民这么想拦数据中心？

[22:43]constituents, the people who vote for them, really care about these data centers and they want to stop them. So, then you have to work backwards. Why do they want to stop data centers so much? 为什么美国人现在这么反对 AI？当然，公众对 AI 的想象早就被科幻小说预装好了，所以我们本来就需要努力摆脱「天网终结者」的印象。但 Dario 告诉所有人：

[22:52]Why Why are Americans so opposed to AI right now? And the perception of AI was already pre-baked with science fiction, of course. So, we we needed to work to get past the Skynet Terminator vision. But, Dario told everybody, 「我们要失业了。」Elon 说：「你以后不用工作了。」这个视角对世界更正面，但他确实说了。Sam Altman 做了全民基本收入（UBI）研究，然后告诉大家：「你不用工作了，有 UBI。」对那些靠工作谋生、

[23:08]"We're going to lose our jobs." Elon said, "You're not going to have to work." It was a more world positive view, but he did say that. Sam Altman did, you know, studies on um universal basic income, and he told everybody, "You're not going to have to work in a UBI." Saying that kind of stuff to people who have to work for a living 没有任何积蓄、眼看着一届又一届政府——不管是民主党还是共和党——都压不住通胀、看着最低工资几乎不动而股市

[23:25]and who have no savings and who are watching, you know, in administration after administration not be able to tame inflation, whether they're Democrats or Republicans, and then watching the minimum wage just stay the way it is and barely move while the stock market hits 屡创新高、然后我们这帮科技高管全发了横财的人说这种话——你还奇怪为什么大家现在拥抱社会主义、为什么 Abbott 和 Shapiro 会说「你知道吗」，以及 Ro Khanna——最过分——说「我必须反资本主义。

[23:40]records, and then all of us all of us being tech executives get super rich, and you wonder why people now are embracing socialism and why you have Abbott and Shapiro saying, "You know what? I And and and Ro Khanna, worst of all, saying, 'I have to be anti-capitalistic. 我必须做 decel（减速派），我必须踩刹车」？因为他们本该看到一点喘息，却什么都没等到。那自觉毫无权力的人会做什么？他们会去寻找权力。看看现在发生的事：他们在磨断头台的铡刀。我

[23:57]I have to be a decel. I have to decelerate.'" It's because, you know, they were supposed to see some relief, and they haven't. And then, what do people do who do not feel they have any power? They look to find power. And if you look at what's happened, they are sharpening the guillotines. And I you 觉得这个比喻该让人惊醒了——无论是联合健康 CEO 被杀，还是有人纵火烧 Sam Altman 的房子。

[24:14]know, that metaphor should wake people up, whether it's, you know, the United Healthcare CEO getting killed or people firebombing Sam Altman's house. 我们需要醒来，看看人们为什么这么愤怒。因为他们活不下去。他们在负担能力和工资上挣扎。他们能做两件事：投票给社会主义者，以及真的拦下数据中心。这就是本质。没有——没有任何一个

[24:23]We need to wake up and look at why are people so upset. It's because they can't survive. They're having a hard time with the affordability, with their wages. And what two things can they do? They can vote for socialists, and they can actually stop data centers. And that's what this is about. Nobody No No regular 普通美国人在乎 AI 安全。那不是他们的议题。他们的议题是：「为什么我没变富，而其他所有人都富了？」

[24:42]rank and file American cares about AI safety. That's not the discussion for them. The discussion for them is, "Why am I not getting rich and everybody else is?" >> 我把你的论调微调一度。我真的觉得你抓住了要害，但这一度的偏差是：我不认为他们完全那么想。我其实认为他们是这么想的：想想过去十年谁赢了——这一代赢了的科技寡头

[24:50]>> tune your argument by 1 degree. I I really think you're onto something, but the 1 degree shift is I don't think they think about it entirely that way. I actually think they think about it as if I think about who has won over the last decade this generation of tech oligarchs who 相当令人反感。他们不是励志偶像，不是让人仰望的人，无论如何都不是 Steve Jobs。而现在他们看着下一代说：等等，我们要从十亿富豪、百亿富豪、千亿富豪，走到万亿富豪，

[25:07]have won are pretty displeasing. They're not aspirational and they're not the people that one would look up to. They're not Steve Jobs by any stretch of the imagination. And now they look at this other next generation and they're like, wait, we're going to go from billionaire, decabillionaire, centibillionaire to now trillionaire and 然后是那帮家伙？那帮人比上一茬还烂。这正在发生的就是这么回事。所以我不认为像你说得那么具体，Jason。我认为是「氛围（vibes）」，而氛围糟透了。他们环顾四周想：如果我们能让这事慢下来，也许

[25:24]it's those guys? Those guys are even worse than the last crop. That's exactly what's happening and so I don't think it's as specific as what you're saying Jason. I think it's vibes and the vibes are terrible and they're looking around they're like if we can just slow this down maybe 会再冒出一批人，他们真的打算做好事，至少能让我感觉好受点，我就不会这么愤懑地恨你即将赚到一个万亿。就是这种感受。而站在台前的这帮人并不

[25:39]another batch of people will show up and they will actually mean to do the right thing and will make me at least feel better and I won't be so begrudgingly hating the fact that you are about to make a trillion dollars. It is precisely that sensation. And the folks that are at the front of the show on this are not 值得信任。就这么回事。>> Sachs，让我问你的问题，我真的需要有人回答这个问题。

[25:56]trustworthy. That's what it is. >> Zack, let me ask you my question and I just I need I need to hear someone answer this question. >> 好，我会回答你的。我想——>> 好，我的问题再重复一遍：如果美国的 AI 开发因监管而放慢，而中国全速前进，这对美国企业、美国消费者意味着什么？

[26:03]>> Okay, but I'll answer your question. I'd like to >> Yeah, so my question again, if there's a regulatory slowdown with AI development in the US and China races ahead what are the implications for American business for American consumers? >> 他们会赢。这到底怎么影响——他们会赢。>> 听着，影响很大——什么叫他们会赢？什么意思？帮大家理解一下。比如我坐在家里——>> 好，想象一下。我给你一个反事实。想象中国赢下了互联网。想象现在所有主要互联网公司都是中国的。

[26:15]>> They will win. How does this actually affect They will win. >> Look, there are big What does that mean? What does that mean? Help people understand. So like if I'm sitting at home >> Okay, imagine this. Okay, let me give you a like a counterfactual. Imagine if China had won the internet. Imagine if all the major internet companies were Chinese right now. >> 我需要你——我需要你想象给我看。告诉我这意味着什么，给我一个场景。>> 想象一下，我知道大家不喜欢大科技，但想象 Google、微软、Meta 这些公司实际上换成的是

[26:31]>> I need you to I need you to imagine for me. Tell me what what So give me a scenario of what this means. >> imagine that instead of I I I know people don't like big tech, but imagine if Google and Microsoft and Meta and all these companies were actually it was 百度、阿里、腾讯等等。那对美国显然糟糕得多。我们的经济不会这么有活力，不会这么前沿，而所有那些财富都会流入中国。还有军事或者说

[26:46]Baidu and Alibaba and Tencent and all the rest of them instead. I mean, obviously that would be way worse the United States. Our economy would not be as vibrant. It would not be as cutting edge and all that wealth would have gone into China. They're also military or you 军民两用：这项技术尤其是 AI，算法驱动着各种武器、无人机之类。所以我们会落在他们后面。听着，如果美国输掉 AI，它就不会再是世界上经济和军事的第一名，

[27:03]know, dual use applications of this technology, especially with AI, the algorithms power all sorts of weapons, drones, things like that. So we'll be behind them. I mean, look, America is not going to be the number one country in the world economically and militarily 就这么简单。>> Chamath，你同意吗？在一个中国「赢了」AI 的世界里，你如何看待世界、美国、企业、消费者、个人？

[27:18]if we lose AI. It's just that simple. >> Chamath, do you agree? Like what's your view of the world, the US, business, consumer, individual, in a world where China has won quote in AI? >> 我不知道你说的「赢」是什么意思。什么叫赢？>> 我也不知道它是什么意思。

[27:30]>> I don't know what you mean by won. What do you mean by won? >> I don't know what it means. >> Okay, so I'm what I'm trying to get to is the implication of the implications of regulatory slowdown. >> 好，我想抵达的是监管减速的后果。

[27:40]So if we do slow down AI, what are the implications for people? How do the people understand tangibly what that means? 所以如果我们真的让 AI 减速，对人们意味着什么？人们怎么具体地理解这件事？

[27:46]>> Yeah, unless we actually go and ban, and I don't even know how one would implement a ban, so I think it's just a worthless exercise. >> 是这样：除非我们真的去搞禁令——而我甚至不知道禁令要怎么执行——否则我认为那纯属无用功。

[27:54]Ban open source, I think we're going to be just fine. I think the consumer is going to win. I think American companies will win. I think the American economy will win. 去禁开源，不然的话我们会好好的。我认为消费者会赢，美国企业会赢，美国经济会赢。

[28:04]And the reason why it will win is that what we are finding, I'll give you an example of a really interesting insight. 它会赢的原因是我们在不断发现的东西。我给你举一个特别有意思的洞察的例子。

[28:11]And this is about closed versus open, okay? Three years ago, all there ever were were models, yeah? And you would look at some benchmark and there would be a score free bird and you would say, "Wow, the one with the higher score is better." 这是关于闭源对开源的。三年前，世界上只有模型本身。你去看某个 benchmark，上面有个分数，Freeberg，你就说：「哇，分数高的那个更好。」

[28:26]And then in the last few years, what's emerged is what's called a harness, right? And a harness just wraps the model, right? Like Codex or Claude Code or Hermes or OpenClaw, these are harnesses. 然后在最近几年，出现了一个叫 harness（工具壳）的东西。harness 就是包在模型外面的一层，比如 Codex、Claude Code、Hermes、OpenClaw，这些都是 harness。

[28:42]You know what the interesting thing about open source versus closed source models? When you take a closed source model, it does really well on a benchmark. When you wrap it in its own harness, it decays in capability. When you take an open-source model and you use any other open-source model, it improves in capability. 你知道开源和闭源模型最有意思的地方是什么吗？你拿一个闭源模型，它在 benchmark 上表现很好；你把它包进它自己的 harness 里，它的能力反而衰减。而你拿一个开源模型，配任何一个别的开源组件，它的能力反而提升。

[28:58]What is that trying to tell you? What I think we are finding out, and this data is just going to be a tsunami over the next year, is there are all kinds of ways of using open-source technologies that are meaningfully more performant and dramatically cheaper than the 这想告诉你什么？我认为我们正在发现、而且这波数据在明年会像海啸一样涌来的是：有各种各样使用开源技术的方式，比闭源替代方案

[29:14]closed-source alternative. Unless America stops us, us being consumers, business owners, from using the cheapest, fastest, best thing, unless they stop us from doing that, we will win. 性能明显更强、成本还低得多。除非美国拦住我们——我们指的是消费者和企业主——不让我们用最便宜、最快、最好的东西，否则我们就会赢。

[29:28]What we are debating at this point is which private American company will win or lose. I go back to the previous argument. I don't think Americans care that much whether any of these closed-source companies win or lose. 我们此刻争论的只是哪家美国私营公司赢或输。回到刚才的论点：我认为美国人并不那么在乎这些闭源公司谁输谁赢，

[29:40]And I think you're seeing that in the policy. So, Friedberg, my explicit answer to you is as long as we don't do some draconian thing like trying to stop open-source, the American consumer will win. 我认为你在政策上已经看到了这一点。所以 Freeberg，我明确的回答是：只要我们不做那种 draconian 的事、比如试图掐死开源，美国消费者就会赢。

[29:51]>> got bad news for you, Chamath. An open-source ban is coming. They're not going to call it that. You know what they're going to say? They're going to say that we simply have to apply the same standards to open models that we apply to closed. Okay, here here's how they do it step by step. Let me explain how regulatory capture actually works. >> Chamath，我有个坏消息要告诉你：开源禁令正在路上。他们不会这么叫它。你知道他们会怎么说吗？他们会说：我们只是要对开放模型适用与闭源模型相同的标准。好，我一步步给你演示。让我讲讲监管套利到底是怎么运作的。

[30:06]So, first of all, you got to get this regulatory apparatus, okay? Dario wants an FDA for AI, but he doesn't have enough political support for that, so instead, they do this Trojan horse of a Finra for AI. They call it self-regulating. It's not really, but anyway, that gets them off the ground. 所以，首先，你得先建立这个监管机构，对吧？Dario 想要一个 AI 版 FDA，但他没有足够的政治支持，所以他们换了个 Trojan 木马——AI 版 FINRA。他们管它叫自律组织。其实根本不是，但不管怎样，这就让他们起步了。

[30:22]Now they've created the standard-setting organization. Now they've got pre-release model testing. Then the pressure grows to codify that in law, so that happens next. And then what they do is they say, "Look, all these standards need to apply equally to all models." 现在他们有了标准制定组织，有了模型发布前测试。然后立法化的压力越来越大，这一步随之而来。接下来他们会说：「看，所有这些标准都需要平等地适用于所有模型。」

[30:35]But here's the problem with that. Open models and closed models, as you know, are technologically different. Go back to Dario's Senate testimony in I think it was 2023 where he said that open models, if sufficiently advanced, are dangerous. And why are 但问题在于：开放模型和闭源模型，如你所知，在技术上是不同的。回想一下 Dario 2023 年（我记得是）的参议院证词，他说开放模型如果足够先进，就是危险的。为什么

[30:51]they dangerous? Because they cannot be centrally monitored and controlled and rolled back and withdrawn hold on like like a closed service like the one he hosts. Okay. 危险？因为它们无法被集中监控、控制、回滚和撤回——等等——像他托管的那种闭源服务那样。好。

[31:01]>> I understand David, but you know >> hold on let me just finish the point. Please let me just finish the point then you can >> Yeah. >> 我理解，David，但是——>> 等等，让我把这点说完。请让我说完这点，然后你——>> 好。

[31:06]>> you can respond. Dario basically says that open models are dangerous because of these characteristics they have and those characteristics are immutable. >> 你再回应。Dario 的意思基本上是：开放模型因为具有这些特性所以危险，而这些特性是不可改变的。

[31:17]They're technologically they can't be changed. Once you release an open model into the world, you can't roll it back and you can't monitor exactly how people are using it because they run it on their own hardware. Dario says this is what makes open models dangerous. So what they're going to do is they're going to have the standard setting body say, well, we have to set the standards 在技术上无法改变。一旦你把开放模型发布到世界上，你无法回滚它，也无法精确监控人们怎么用它，因为他们跑在自己的硬件上。Dario 说这正是开放模型危险的原因。所以他们接下来会做的，就是让标准制定机构说：好吧，我们必须制定标准——

[31:33]for AI safety. By the way, Dario and OpenAI, they're going to fund the whole thing. They're going to contribute all the compute. They're going to be behind it. They're going to be the ones coordinating with the government officials because frankly people in government have no idea how to monitor and control and set standards for AI AI 安全的标准。顺便说，Dario 和 OpenAI 会资助整个机构，贡献全部算力，他们会站在幕后，会是跟政府官员对接的人——因为坦白说，政府里的人根本不知道怎么监控、控制和制定 AI

[31:49]safety. Technologically this is way beyond them. So they're going to go to these companies and say, tell us how to do it. 安全标准。这在技术上远超他们的能力。所以他们会去找这些公司说：告诉我们该怎么做。

[31:56]And so what will happen is the standards will get set and then it'll be a very simple matter of fairness to say that the standards need to apply to open as well as closed models. The open models cannot comply in the same way and gradually they will be shut out of the market. 于是标准就会被制定出来，然后只需要一句「公平起见」，标准就必须同时适用于开放模型和闭源模型。开放模型没法以同样的方式合规，于是逐渐被挤出市场。

[32:11]>> I understand but the American economy would overnight. Why and how? >> 我理解，但美国经济会一夜之间——为什么？怎么会？

[32:18]>> You're saying regulators have never done things that are crippling. >> Okay, let your mouth let your mouth answer. So let let's take it again. I'll use the example of Coca-Cola. Coca-Cola is is an extremely profitable, cash generative, well-run business. It has existed for I don't know 100 plus years. At this point >> 你是说监管机构从来没干过瘫痪一个行业的事？>> 好，让你的嘴自己回答。我们重新捋一遍。我用可口可乐举例。可口可乐是一家极其赚钱、现金流充沛、经营出色

[32:34]it will exist for many hundreds more if left to their own devices. It is a global business, right? Take Pfizer, same situation. Well-run business, global business. 的企业。它已经存在了不知道 100 多年，如果不受干预，它还会再存在几百年。它是个全球性企业，对吧？辉瑞也一样：经营出色、全球性企业。

[32:45]And now what you will tell them is, unless you tell them, "Hey, anywhere you operate, you need to follow American rules." Okay, that's one thing. 而现在你要告诉它们的是——要么「嘿，无论你在哪里经营，都要遵守美国规则」，这是一种。

[32:53]Or it'll say, "Inside of America, you need to follow our rules, and other governments will have different rules." 要么「在美国境内要遵守我们的规则，其他政府另有各家的规则」。

[32:59]Now, what is Coca-Cola to do? Run the slower, more costly, less effective, less valuable model in the United States? Sure, they'll do They will have a better flourishing 那可口可乐该怎么办？在美国跑那个更慢、更贵、更差、价值更低的模型？行，它们会照办——然后它们在海外会有一个更欣欣向荣的

[33:14]business abroad. They will invest abroad, and what you will see is FDI, foreign direct investment in the United States, go off of a cliff. And what you will see is direct investments into other countries of the world rise. 业务。它们会把投资投向海外，你会看到 FDI——流入美国的外国直接投资——断崖式下跌，而流向世界其他国家的直接投资上升。

[33:28]You've seen this, by the way, in China. When China made all these moves to the Chinese tech companies, I think this is the best analogy. 顺便说，这一幕你在中国见过。当中国对中国科技公司出手的时候——我认为这是最好的类比。

[33:36]There was a moment where China was at a fork in the road, where the large oligopolists there, the large tech oligarchs, were trying to seize power, and the tip of that spear was when Jack Ma gave that very famous speech on the eve of the IPO of Ant Financial. What 中国曾站在岔路口：那边的大寡头们、大科技寡头们在试图夺权，矛头就是马云在蚂蚁金服 IPO 前夜那场著名的演讲。然后

[33:52]happened? Xi Jinping cracked down. They took complete control. They said, "We are now operating together. Everybody that is worth anything will have a representative from the Chinese government on their board. We have the golden vote. We're going to work together." 发生了什么？习近平出手镇压。他们完全接管。他们说：「现在我们是一起运作的。任何有价值的企业，董事会里都要有中国政府派驻的代表。我们拥有黄金投票权。我们要合作。」

[34:06]When you look at what happened to foreign direct investment, and you look at what happened their economy, it fell off of a cliff. 你看看那之后的外国直接投资，再看看他们的经济——双双断崖下跌。

[34:13]That will happen here if this happens. So, they can try, and everybody will say, "Okay." And what what the Coke CEO will do is redirect all of his capital allocation to outside of America. So will Pfizer. So will I. So will anybody." And then what are you going to 如果我们这么干，同样的事就会在这里上演。他们可以试，所有人都会说「行」。然后可乐的 CEO 会把所有资本配置转向美国以外。辉瑞也会。我也会。任何人都会。那你接下来

[34:29]do? Take away our passports so you can't travel? So, I think it will be a very visible way of just destroying the United States. 打算怎么办？没收我们的护照不让我们出行？所以我认为，这会是一种非常直观的、摧毁美国的方式。

[34:37]And you can do that if you want to, but you will see it play out almost instantaneously in the equity markets and in the capital allocation of companies. 你当然可以这么干，但你几乎会在股权市场和企业的资本配置上立刻看到后果。

[34:44]>> I agree with you about all the negative consequences if we ban open models, but that doesn't mean it can't happen. Look at all the negative consequences of banning data centers. Data centers are going to be 3% of GDP. It's basically creating all the gains in the stock market. >> 我同意你说的禁掉开放模型的一切负面后果，但那不代表它不会发生。看看禁掉数据中心的负面后果吧。数据中心将占到 GDP 的 3%，股市的涨幅基本全是它创造的。

[34:59]>> But they're not voting against that. They're voting against two people that they deeply dislike becoming trillionaires. That's what they're voting against. >> 但人们投票反对的不是这个。他们投票反对的是两个他们深恶痛绝的人成为万亿富豪。他们投的是这个。

[35:06]>> That's a big part of it. Jacob, what do you think? >> Well, I was just going to add to this. We already know what would happen with AGI, Freiberg, cuz your original question of like trying to explain to Americans what happens when China wins. >> 这确实是很大一部分。Jason，你怎么看？>> 我正想补充。Freeberg，关于你最初那个问题——怎么向美国人解释中国赢了会怎样——其实我们已经知道 AGI 时代会发生什么了。

[35:16]We already know what happens when they win. Products get cheaper. And and uh you get more access to them. 我们已经知道他们赢了会发生什么：产品变得更便宜，你能更多地用到它们。

[35:22]If they solve cancer, what are they going to do with that solution? They're going to just keep it for themselves? No, they're going to sell it to Americans just like they sell phones cheaper, just like they sell cars cheaper, and we try to block them. And then what Americans are going to do? 如果他们攻克了癌症，他们会怎么处理那个解决方案？藏着掖着自己用吗？不，他们会卖给美国人，就像他们卖更便宜的手机、更便宜的汽车一样——然后我们试图封锁他们。那美国人会怎样？

[35:33]We're going to lose our jobs to them. And so that would be the way I I think to explain it to them. But if we want to turn this around, this really is just about convincing Americans that they're going to have jobs and that this is going to benefit them. The American people are scared of losing their 我们会输掉我们的工作，输给他们。所以我觉得这就是向他们解释的方式。但如果想扭转局面，归根结底就是让美国人相信：他们会有工作，这件事会让他们受益。美国人民害怕失去

[35:49]livelihoods. That's the beginning and end of it. And you know, the the place where I have a some expertise is in having invested in Zipline and Uber and a bunch of self-driving companies. And that's where you're >> You're saying being poor. 生计。这就是全部。而我有一点发言权的地方是我投过 Zipline、Uber 和一批自动驾驶公司。这正是——>> 你这是在说贫穷。

[36:03]>> The money's not the issue here, but the thing that I think is, you know, where this is going to play out is really on the autonomy and and those jobs. Zipline just did a partnership with Uber. I'm investors in both of those companies, uh obviously, and they are going to do a million deliveries a day. Then you look at just massive >> 钱不是这里的问题。我认为真正会引爆的点在自动驾驶和那些工作上。Zipline 刚和 Uber 达成了合作——我两家都投了，显然——他们要做到每天 100 万单配送。然后你看 Waymo 的巨大

[36:20]progress by Waymo. They're doubling the number of cars on the road like every 6 months or so. Tesla's about to go cybercab. And what people are perceiving is in their community that you're not going to have any money staying in their community. Cuz those 进步：大约每 6 个月路上的车就翻一倍。Tesla 马上要上 cybercab。而人们的社区里正在感知到的是：不会再有钱留在你的社区里了。因为那些

[36:35]drivers, you know, as as much as hand-wringing has occurred, that money at least stayed in the community. Oh, they weren't making enough. Oh, they're only making $18, $16 an hour, $20, $21 an hour, whatever it is. But now we're taking that away from them. And then where does that money go? 司机——不管之前有多少唉声叹气，那笔钱至少留在了社区里。哦，他们赚得不够多，哦，他们一小时才挣 18 美元、16 美元、20、21 美元，随便多少。但现在我们连这个也要拿走。那这笔钱去哪了？

[36:49]Where does the money for a robo-taxi, Waymo, Ride, Zipline, where does it all go? It goes up into the cloud and then comes back down into Silicon Valley, polarizing more wealth. And that's where you're going to see, I will predict it now, they're going to ban Waymos in New Robo-taxi、Waymo、Ride、Zipline 的钱都去哪了？它升上云端，然后落回硅谷，让财富更加极化。所以我现在就可以预言：他们会在纽约、

[37:05]York, Boston, DC, and they're going to limit them with licenses. We must figure out how to explain to Americans that we're going to raise the minimum wage, something that when I say we should raise the minimum wage and focus on that, people are like, "That's crazy." 波士顿、华盛顿特区禁 Waymo，或者用牌照限制它们。我们必须想清楚怎么向美国人解释：我们要提高最低工资——每当我说我们应该提高最低工资、聚焦这个的时候，人们就说「你疯了」。

[37:20]But when we talk about an experimental tax cut, people are like, "Oh yeah, we do an experimental tax cut. We'll change the cost basis of capital gains. Great, let's try and see what happens." But nobody wants to try to raise the minimum wage. Nobody wants to give people an explanation of where their job will be 但一说到实验性减税，人们就说：「哦好啊，来个实验性减税。改改资本利得的成本基础。太好了，试试看会怎样。」但没有人愿意试着提高最低工资。没有人愿意给人们一个解释：他们未来的工作

[37:36]in the future. And that's where, you know, whoever is making the point, I think it was You Tur Off about like who are the spokespeople here. We need spokespeople who are saying this technology will empower you to get a great job. Paradoxically, the guy who did the best job explaining this, or the two guys, were Jensen, obviously, who 在哪里。所以——刚才谁在说代言人那个点，好像是说这里谁是代言人。我们需要的是这样的代言人：这项技术会让你有能力得到一份好工作。吊诡的是，把这事解释得最好的——一个或两个人——是 Jensen，显然，他

[37:52]said, "You're not going to get replaced by AI. It's going to be somebody who works with AI." And then Zuckerberg who said, "I'm going to give you a bunch of free tools." In his manifesto going direct, he was like, "I'm going to give you free tools. I'm going to empower you. You're going to get agents. Your life's going to get better." 说：「AI 不会取代你。取代你的是会用 AI 的人。」然后是 Zuckerberg，他说：「我会给你一大堆免费工具。」在他直接下场的宣言里，他说：「我给你免费工具，我给你赋能，你会有 agent，你的生活会更好。」

[38:05]Dario and Sam have not done that and have done the exact >> great thing. Zuck did a really great thing, which is he created this new program or it's almost like a training school >> Yes. Dario 和 Sam 没有这么做，而且做得恰恰相反——>> Zuck 做了一件很棒的事，他搞了一个新项目，几乎像是一所培训学校——>> 是的。

[38:16]>> for it's like a trade school for people to go and learn how to get these various kinds of jobs, laying fiber, doing these new blue-collar construction jobs, >> Yeah. which we have a huge shortage of. >> 像一所职业学校，让人去学怎么得到这些各种各样的工作：铺光纤、做这些新型蓝领建筑工作——>> 对，这些岗位我们缺口巨大。

[38:27]And that's why it actually it's like hard to build these data centers. >> or something. It's a great program. >> Yeah. >> Yeah, I mean that's that we need more of. And so, how is Zuckerberg all of a sudden the best spokesperson for our industry? Like >> It's a very low bar. I mean >> I guess that's [laughter] it. 而这也正是为什么建数据中心那么难。>> 什么的。是个很棒的项目。>> 是。>> 是，我的意思是我们需要更多这样的东西。所以，Zuckerberg 怎么突然就成了我们行业最好的代言人了？>> 门槛实在太低了。>> 我猜也是——【笑】就这么回事。

[38:40]>> I mean this all goes back to the fact that Dario, not exclusively, but again, he was the one who put the most specific numbers on it and therefore made it the most tantalizing and salacious for the media. 50% of entry-level knowledge workers going to lose their job, not in >> 我是说，这一切都回到那个事实：Dario——不 exclusively 是他，但再次强调，他是给出最具体数字的人，因此对媒体来说最撩人、最劲爆。50% 的入门级知识型工作者将失业，而且不是

[38:55]the distant future, as soon as 1 year. We're now past that 1 year mark, by the way. Has he updated? Has he said I was wrong? Not really. Sam did, by the way, to his credit. 在遥远的未来，最快 1 年内。顺便说，我们现在已经过了那个 1 年之期。他更新过观点吗？他说过「我错了」吗？没有。倒是 Sam 更新了，给他记一功。

[39:05]>> Sam Altman and Dario should start talking about jobs. >> No, but Sam actually came out and said I was too negative. I was wrong. >> Sam Altman 和 Dario 真该开始谈谈就业。>> 不过 Sam 确实公开说过「我太负面了，我错了」。

[39:12]Dario has doubled down, as far as I know. He's never recanted. He's an ideologue. So, but I mean to your whole point, yes, I agree with you that the animus and anger towards data center flows out of the fears about job, as well as other fears about AI, but it's 而据我所知，Dario 是加倍下注。他从不认错。他是个 ideologue（教条主义者）。但回到你的核心观点——是的，我同意：针对数据中心的敌意和愤怒源于对失业的恐惧，以及对 AI 的其他恐惧，但

[39:28]the job loss that is the main concern and there is no evidence of that. We're now years into the AI boom and there's no evidence of that. We only see net job creation. 失业才是主要担忧，而完全没有证据。我们进入 AI 热潮已经好几年了，没有证据。我们只看到净就业创造。

[39:38]>> Exactly. It's always been the opposite. >> I mean, listen, I was I had I had massive concerns about it and I had business opportunities so great I got seven job openings. So, I keep saying >> Hold on, hold on. This is important. >> 没错，一直恰恰相反。>> 我是说，听着，我本来也对这事有巨大的担忧，结果我的商业机会好到我开了 7 个招聘岗位。所以我一直在说——>> 等等等等，这很重要。

[39:50]Hold on, J Kaskas has got an important announcement here. >> No, I I listen, what what I've discovered is the if you can convince your team to embrace the tools, so many business opportunities emerge that you wind up hiring people to work for the people in your organization 停下，J Cal 有重要宣布。>> 不不，听着，我发现的是：如果你能说服团队拥抱这些工具，会涌现出那么多商业机会，最后你会雇人来给团队里那些

[40:05]who embrace the tools. So, we have become so fast at building the syndicate, building our programs that I'm like, you know, we need to add a couple people. So, I thought we were going to automate them away and we're adding >> say I'll just say another general point. 拥抱工具的人打工。所以我们建 syndicate、建项目的速度快到我说：我们需要再加几个人。所以，我原以为会把人自动化掉，结果我们在加人——>> 我再说一个一般性的观点。

[40:19]Silicon Valley has completely lost what made it great. It was a place for weirdos, for people at the fringes, for people that really were idealistic. 硅谷已经完全丢掉了让它伟大的东西。它曾是一个属于怪人、边缘人和真正理想主义者的地方。

[40:32]And it has become something that's far from that now. It is the equivalent of what Harvard used to be. It's a credentialing place. 而现在它变成了离那很远的东西。它成了当年哈佛的等价物——一个发证书镀金的地方。

[40:40]It's a place you come to to say, "Oh, I came, you know, to Silicon Valley. I worked at a startup. You know, I got into YC." You know, when we came, we used to aspire to meet the Bill Campbells of the world, Mike Moritzes, John Doerrs. Those folks are gone. We 一个你来了可以宣称的地方：「哦，我来过硅谷，我在创业公司干过，我进过 YC。」我们刚来的时候，向往的是结识这个时代的 Bill Campbell、Mike Moritz、John Doerr。那些人都不在了。我们

[40:57]used to aspire to work for Steve Jobs. He's not here. And in that breach, all we have done is focused on making money in every single way that we can, hand over fist. And I think again, there's nothing wrong with that, except 曾向往为 Steve Jobs 工作。他不在了。而在这个真空里，我们唯一做的事就是想尽一切办法拼命赚钱。我再说一遍，这本身没什么错，

[41:13]now is a point where we need to do a better job of acknowledging it and trying to do a little bit better. And every time we don't, and we're given a chance to do be better and we're not, the hatred builds. The animus, as you said, David, builds. And that's what we're seeing now. 只是现在已经到了一个节点：我们得更坦诚地承认这一点，并努力做得好一点。而每当我们没有做到——每当我们被给予变得更好的机会却错过——仇恨就在累积。你说的那种敌意，David，就在累积。我们现在看到的就是这个。

[41:29]>> Yeah, there's no question about that. But can I defend Arie? Let me just say something in his >> my my headphone went [laughter] out. You said you WANTED TO DEFEND ARIE? Breaking news. >> 是，这毫无疑问。但我能为 Dario 辩护一下吗？让我说两句——>> 我耳机刚才没声了。【笑】你说你要为 DARIO 辩护？突发新闻。

[41:39]>> think I actually think he is very idealistic. I don't think money is his main motivation. It's just what are his ideals? He wants to bring forth this new super intelligence, machines of loving grace that are going to apportion all the resources of >> 我认为他其实非常有理想。我不认为钱是他的主要动机。只是：他的理想是什么？他想迎来那个新的超级智能——「恩典机器」——由它来分配社会的全部资源，

[41:56]society while humans twiddle their thumbs. It's not a vision that the average person believes in or is attracted to. In fact, it's dystopian. 而人类在旁边无所事事地掰手指。这不是普通人相信或向往的愿景。事实上，它是反乌托邦的。

[42:05]And that's why a lot of their messaging is dystopian. And because of that vision, no matter how idealistic his motives, it is why he does not inspire trust. I mean, I actually I I don't think he's lying to us and what he's really trying to do here is capture all the wealth for 这也为什么他们的很多宣传是反乌托邦的。因为那个愿景，无论他的动机多么理想主义，他无法激发信任。我的意思是，我真的不认为他在对我们撒谎、

[42:20]himself. I actually think he's he's trying to bring forth this future, but it's not a future that I think the American public wants. And by the way, he should have to defend that. I mean, if you think you are bringing about the singularity, and let's face it, he probably is in the lead to do that, to 真正图的是把财富都抓到自己手里。我真的认为他在努力迎来那个未来——但那不是美国公众想要的未来。而且说真的，他应该为那个未来辩护。如果你自认为在带来奇点——说白了，他很可能确实是实现

[42:36]bring about recursive self-improvement, he should defend this future. I see all these people simping on X, so like, "Oh my god, Dario, thank you for coming forth. Oh, you know, thank you for coming down from Mount Olympus and slumming it with the rest of us on X and finally engaging on social media, you know?" And it's like, no, he 递归自我改进的领跑者——那他就该为这个未来辩护。我看到 X 上一堆人跪舔：「天哪 Dario，谢谢你出来说话，谢谢你从奥林匹斯山下凡，屈尊跟我们在 X 上同框，终于肯上社交媒体了。」不，他

[42:53]should have to do that. You know, if you're going to put yourself in the lead to bring about this crazy future, you should tell us why you're doing that. 本来就该这么做。你要是把自己放在引领这个疯狂未来的位置上，你就应该告诉我们你为什么这么干。

[43:01]>> Well, and he should speak more. I think this is a key point. And to to your point about it's not about money for him, the seven co-founders of This is my understanding. It's been documented. >> 而且，他应该多说话。我觉得这是关键。回应你说的「钱对他不重要」——据我了解，七位联合创始人，这是有报道的，

[43:12]Um I think in a bunch of press reports, but they never talk about it. They donated 80% of their shares. So, like 10% of Anthropic's in some sort of trust, which at two or three trillion, you know, 10% of that's going to be the largest >> Well, it's a >> It's going to be the largest Democratic PAC. It's going to be the largest PAC 很多媒体写过，但他们从不谈：他们捐掉了 80% 的股份。比如 Anthropic 的 10% 放在某种信托里，按两三万亿估值算，那 10% 会成为最大的——>> 呃，那是——>> 会成为最大的民主党 PAC。最大的 PAC。

[43:28]>> However they use it, but why aren't they talking about how they're >> It's going to be the largest DSA >> dollars to EA causes. What could go wrong? >> 不管他们怎么用，可他们为什么不说清楚钱会——>> 会是最大的 DSA——>> 捐给 EA 事业的美元。能出什么岔子呢？

[43:34]>> Well, I mean, let's hope they put it towards something that actually is uh humanity positive. But they they And he only they all they all own only like 1.8% each. >> 呃，希望他们把它用在真正对人类有益的事情上吧。但他们——还有，他们每个人只持有大约 1.8%。

[43:44]>> That's a term of art. >> Yeah. Well, and then I do think going direct is something you just I think we can move on after this. >> 这是个术语了都。>> 是。然后我确实认为「直接下场」这件事——我觉得说完这点我们就可以翻篇了。

[43:52]is critically important. >> What have you got to learn? You got an AI to to talk. >> No, no, no. You got a The playbook is very simple. You got a number one, you got to write it all yourself, which he clearly did. And, you know, Jensen's been doing >> about that? 是至关重要的。>> 你有什么资格说？你自己让 AI 替你说话。>> 不不不。剧本非常简单：第一，你得全部自己写，他显然是自己写的。Jensen 一直在这么——>> 何以见得？

[44:06]>> I think so. It felt like to me. It felt very Claude-like to me. >> You think >> [laughter] >> you know the Claude voice? I mean, he is the father of Claude. >> 我觉得是。我的感觉是，非常有 Claude 的味道。>> 你觉得——【笑】>> 你懂的，Claude 的腔调？毕竟他是 Claude 之父。

[44:13]>> That's true, actually. Claude could sound like him rather than vice versa. >> be the other way. Causation correlation. But, what he did is he did the exact playbook, which is you have to exhaust the audience. You have to go thousands of words like Zuck did, and then you have to get ahead of problems. That's >> 这倒是真的。可能是 Claude 听起来像他，而不是他像 Claude。>> 因果倒置。但他做的是标准剧本：你必须耗尽受众的耐心，必须像 Zuck 那样写上几千字，然后你必须抢在问题前面。这正是

[44:29]where the leadership is breaking down. You have to talk about what's coming next, and then never ever apologize or give an inch. Just attack whoever's attacking you, fake news. 领导力正在失灵的地方。你必须谈接下来会发生什么，然后永不道歉、寸步不让。谁攻击你你就攻击谁，假新闻。

[44:40]>> right now, I don't think he can win the argument by being transparent because what he's selling is fear in order to motivate the government to give him control to bring about this dystopian future that he believes in. At least that's that's my distillation of >> 而且我认为，眼下他不可能靠透明赢得这场辩论，因为他卖的是恐惧，为的是推动政府把控制权交给他，去实现他信奉的那个反乌托邦未来。至少这是我对

[44:55]of what I think >> did. Create a program for tube bell generation. Come up with ideas like that. Do more of that. Freeberg, last you get the last last last last. Go. 这件事的提炼。>> 搞一个面向普通人的培训项目吧。多想出这样的点子，多做点这种事。Freeberg，最后你收尾。说吧。

[45:03]>> I think that it's worth having a view on whether or not you believe in recursive self-improvement. >> 我认为，你对「是否相信递归自我改进（RSI）」持有一个立场，是值得的。

[45:11]And recursive self-improvement is this theory that at some point an AI model can spin up a whole bunch of agents to do a bunch of work, and they can work together to make a new AI model, and the new AI model is better than the 递归自我改进是这样一个理论：某个时刻，一个 AI 模型可以生成一大堆 agent 去干一堆活，它们协作造出一个新的 AI 模型，而新模型比

[45:26]last AI model. And at some point this all becomes effectively fully automated and self-evolutionary. 上一个更强。到了某个时刻，这一切变得完全自动化、自我进化。

[45:35]The AI is now improving itself continuously. And that is this idea that you have the singularity as some people would call it, or you have the ASI, the artificial superintelligence that arises from this capacity. AI 在持续地自我改进。这就是有些人所说的奇点，或者说从中诞生的 ASI——人工超级智能。

[45:50]But, the capacity is recursive self-improvement, RSA. Anthropic has stated they're not sure if it's going to happen, but it could, and they're working towards it, and they expect that there's a path to get there. 但这个能力本身叫递归自我改进，RSI。Anthropic 已表态：他们不确定它会不会发生，但它有可能，他们在朝它努力，他们预期存在一条通向它的路径。

[46:02]I happen to believe having looked at this and thought about it, I think that it is possible. I think it is likely to happen soon. People I've talked to in industry do think it's real. 我恰好研究过、思考过这个问题——我认为它是可能的，而且很可能很快发生。我跟业界的人聊过，他们确实认为这是真的。

[46:12]And there is this kind of philosophical debate because companies have a different view on the path forward where it really needs to be human constructed, human designed. 还有一个哲学层面的分歧：各家公司对前进路径看法不同——有的认为必须由人类构造、人类设计，

[46:21]And there isn't this like let it all go, let humans detach from the AI and let the AI evolve. And this is really like out of the movie Her where the thing just takes off and it evolves. 而不是撒手不管、让人类脱离 AI、让 AI 自行演化。后者就很像电影《Her》：那东西直接起飞、自行演化。

[46:31]But if it does happen and if RSI is a real construct, then you're almost fighting it's like a fool's errand. And it's a fool's errand in the sense that like all this idea that you've got human intervention in the stages of model 但如果它真的发生、如果 RSI 是真实存在的构造，那你几乎是在——这就像徒劳无功的傻事。之所以是傻事，是因为「人类在模型开发各阶段介入」这一整套想法

[46:46]development. >> Uh you're scaring people again. >> Well, no. I I actually No, this is my point. If RSI is a thing and and by the way, RSI doesn't need lead to some dystopian outcome. It just means that the models improve faster and faster. 都会失效。>> 你又吓到人了。>> 不，听我说。这正是我的观点：如果 RSI 成立——顺便说，RSI 并不必然导向反乌托邦结局，它只是意味着模型改进得越来越快。

[46:59]>> Without humans. >> Without humans necessarily constructing, architecting, and designing the next evolution. And then you don't necessarily have space for oh, we just did one big model run every 6 months. >> 不需要人类。>> 不一定需要人类去构造、架构和设计下一代。然后你就不再有空间说「哦，我们每 6 个月跑一次大模型训练，

[47:11]We're going to go to the government regulator or we're going to go somewhere and get it approved. There's just this continuous policing and monitoring service that needs to take place at some point. And the truth is that all that you need to pull this off is you don't need governments in the US to say yes. 然后去找政府监管机构报批」。到那时需要的是某种持续运行的巡查与监控服务。而真相是：要做成这件事，你根本不需要美国政府点头，

[47:25]Or governments in China to say yes. Or governments anywhere to say yes. >> you're going to say. You need satellites in space. 不需要中国政府点头，也不需要任何地方的政府点头。>> 你要说的是——你需要天上的卫星。

[47:32]>> You need some So you need some sovereignty, some space where you can. And I I don't know if it's in space. >> 你需要的是某种主权、某个可以落脚的空间。我不知道是不是在太空。

[47:37]I don't know if it's with Balaji and the deal he just did in Kazakhstan and you've got some unique geography. 不知道是不是 Balaji 刚在哈萨克斯坦做的那个交易、那片独特的地理。

[47:42]>> Nice. >> It could be in Iceland. It could be in the ocean. >> People are floating it. >> It could be in the ocean. But the The point The point is if RSI is a thing, you don't have a human process where you're stepping in to run this thing. So you don't necessarily have a cohort of >> 不错。>> 可能在冰岛，可能在海上。>> 有人正在推动。>> 可能在海上。但重点是：如果 RSI 成立，就不存在「人类介入运行这个过程」这回事。也就不一定需要一队

[47:58]frontier lab scientists running it. Theoretically, an individual that's smart enough or a group of individuals that's smart enough can go spin up an RSI factory anywhere in the world or anywhere in space. And And your attempts to ban the data centers or the attempts to ban or to create a 前沿实验室科学家来运行它。理论上，一个足够聪明的个人或一群足够聪明的人，可以在世界任何角落、甚至太空任何角落建起一座 RSI 工厂。而你们禁数据中心、

[48:14]regulatory body, that's why I call it a fool's errand. Because the thing with the open internet and the thing with compute is all that it needs is the chips and the power and a connection. 设监管机构的企图，所以我说那是徒劳的傻事。因为开放互联网和算力这个东西，它需要的只是芯片、电力和一条连接。

[48:23]And if it's got the chips and the power and a communication connection, it can run. And I'm I'm not painting a dystopian view. I'm just saying practically, pragmatically the the labs are all talking about RSI. 有了芯片、电力和通信连接，它就能跑。我不是在描绘反乌托邦图景。我只是说，务实地讲，各家实验室都在谈论 RSI。

[48:34]And so if RSI is where this goes, then the labs can go and set up data centers in Iceland, in China, in wherever. 所以如果 RSI 是归宿，实验室们就可以去冰岛、去中国、去任何地方建数据中心。

[48:42]>> Whoever will give them the freedom. It's a great point. >> better off? Wouldn't we be better off having systems where we actually have those labs based where we have those data centers based in the US, where we have jurisdiction over them, where we have capacity to monitor and control and police them, rather than just have them >> 谁给他们自由，他们就去哪。说得好。>> 我们难道不是更好过吗？让这些实验室、这些数据中心留在美国——我们对它们有司法管辖权、有能力监控、控制和监管——而不是放任它们

[48:57]sprung up in some random country or some random territory? So, I guess my point is we should be attracting and retaining the frontier labs, not trying to >> Yes. 在某个随机国家、随机领土上冒出来？所以我的观点是：我们应该吸引并留住前沿实验室，而不是试图——>> 是。

[49:07]>> you know, chase them off. >> And then here's the famous wait, but why image of a human progress going on a nice clip, and then recursive self-improvement and it just goes out of control and >> I'll tell you why I think RSI >> Yeah. >> 把它们赶走。>> 然后就是那张著名的《Wait But Why》图：人类进步原本平稳前行，然后递归自我改进一来，直接失控——>> 我告诉你们我为什么认为 RSI——>> 是。

[49:20]>> is the legitimate concept. >> to it? And that's I think what people would want to know. How close do people >> You know, honestly, time there there's a massive problem with time dilation in AI model development. And And what I mean by that is we would typically consider some sort of thing that takes a linear amount of time as a certain amount of compute to get somewhere. There are >> 是个站得住的概念。>> 离我们多近？我想这才是大家想知道的。大家离它有多近——>> 老实说，AI 模型开发存在巨大的「时间膨胀」问题。我的意思是：我们通常认为某件事需要线性的时间和一定算力才能抵达某处。但

[49:36]changes that happen that are non-predictable, that you don't actually know what the next phase, what the next step change takes in terms of time. The time models change a lot in terms of AI model development. And that's what's so crazy to make it so hard to predict. 实际发生的变化是不可预测的，你根本不知道下一个阶段、下一次阶跃变化需要多少时间。在 AI 模型开发中，时间模型变化非常大。这就是为什么预测如此疯狂、如此困难。

[49:50]Which is why no one's really good at predicting when is AGI When's the next breakthrough happening? And then all of a sudden everyone's like, "Oh my god, we're in the singularity." They just solved eight unproven math theorems in a week off of some model that's fairly cutting edge. It's now going to be the second cutting edge model. So, it's moving very quickly. 这也解释了为什么没人擅长预测 AGI 何时到来、下一次突破何时发生。然后突然之间所有人都说：「天哪，我们进奇点了。」某个相当前沿的模型一周之内证明了 8 个未证明的数学定理，而它马上就要变成第二前沿的模型了。所以，它跑得非常快。

[50:06]But, a lot of the the models are this mixture of experts concept where you've got these kind of smaller models that call each other to do things. And then you wrap it up and you call it a big model. 而且，很多模型是这种混合专家（MoE）的概念：一堆较小的模型互相调用来干活，然后你把它打包，称之为一个大模型。

[50:18]And those smaller models, it seems over time look a lot more like you see in biology where you have genes. A gene makes a protein and proteins interact in really complicated ways that we don't fully grok. Networks of proteins do things. 而这些小模型，随着时间推移，看起来越来越像生物学里的基因：一个基因制造蛋白质，蛋白质以我们无法完全理解的复杂方式相互作用，蛋白质网络完成

[50:31]And then they complete tasks. And I think the same evolution is likely to happen with the models where they break down into smaller models and that makes them so much more efficient and that makes them so much more performant. 任务。我认为模型会发生同样的演化：它们分解成更小的模型，这让它们效率大大提高、性能大大增强。

[50:41]And we're seeing a lot of this in in how a lot of the models are evolving. We call it a model, but it's actually many models underneath it that work together. 我们在很多模型的演化中已经看到大量这种迹象。我们称之为「一个模型」，但它底下其实是许多协同工作的模型。

[50:48]And so, if the models actually do evolve this way, then RSI is the perfect system for figuring out the right path, which is how many sub models are there? What does each sub model's functions look like? How do they call each other? That evolution is likely going to happen without a human architect behind it. 所以，如果模型真的沿这个方向演化，那 RSI 就是找出正确路径的完美系统：有多少个子模型？每个子模型的功能是什么？它们如何互调？这个演化过程很可能不需要背后有人类架构师。

[51:05]It's going to be developed through a test and deploy system and it's going to be fully automated. And that's why this approach to actually solving that problem ends up becoming the future of how model development takes off, which is why I think RSI is probably the future. 它会通过一套测试与部署的系统发展，完全自动化。这就是为什么这种解决问题的路径最终会成为模型开发的未来形态——所以我认为 RSI 很可能就是未来。

[51:20]And when that happens, again, it's just wherever the data center and the power is. >> I remember a few years ago everyone was like, "Oh, AGI is imminent. AGI is imminent." And then where is it? I mean >> I don't think we're soaking in it. 而到那时，还是那句话：数据中心和电力在哪，它就在哪。>> 我记得几年前人人都说「AGI 迫在眉睫、AGI 迫在眉睫」。然后呢？AGI 在哪？>> 我不觉得我们正泡在 AGI 里。

[51:32]>> Okay, fine, but did it turn out to be the Terminator? >> Right. No, that would be super intelligence. I mean, artificial general intelligence is like as smart as any human. Super intelligence is like you're getting towards what >> I get it. >> 好，就算这样，但它变成终结者了吗？>> 不，那是超级智能。人工通用智能是跟任何人类一样聪明，超级智能才是接近我们

[51:44]>> we're talking about. That's far off. >> I Do Do I think RSI could possibly happen? 在谈的那种。那还远着呢。>> 那你觉得 RSI 有可能发生吗？

[51:50]Yes. I think there's lots of black swans that could happen. It's very hard to assess the probability of those things. 有可能。我认为存在大量可能发生的黑天鹅，很难评估它们的概率。

[51:56]What I do know is like you're saying, I don't think some new government regulatory agency will stop it if that's the thing cuz it won't be technically sophisticated enough. All it will do actually is concentrate power around that thing. It'll limit the options. You know, when it comes to Let me talk about 我确定的是，像你说的：如果那事真来了，某个新的政府监管机构拦不住它，因为其在技术上不够精进。它实际会做的只是把权力集中到那个东西周围、限制选项。说到这，让我谈谈

[52:12]AGI >> want to have that technology in this country is my point. And I just want to go back to the to the point I was asking earlier. Why do we not want China to win? Because we want the best labs. We want the first breakthrough system to be here. Not to be there. AGI——>> 我们要的就是让这项技术留在本国，这是我的观点。回到我之前的问题：为什么我们不想让中国赢？因为我们想要最好的实验室，想让第一个突破性系统诞生在这里，而不是那里。

[52:26]>> the argument Dario actually made this argument in the context of trying to ban export of chips, which is one of his hobby horses is Anthropic has really been behind a lot of the political activity to restrict Nvidia's sales internationally. And the argument was >> 这个论证 Dario 其实用过——在推动禁售芯片的语境下。这是他的执念之一：Anthropic 一直是限制 Nvidia 国际销售的大量政治活动背后的推手。当时的论证是，

[52:42]that we want to have a data center full of geniuses in the US. And that eventually, you know, what AI's going to be used for is you're going to be able to create all these agents that are basically Nobel Prize winning level scientists. And if we have thousands or millions of those and China has only 我们要让美国拥有一座「天才数据中心」。最终 AI 的用途是：你能创造出成千上万个 essentially 诺贝尔奖水准科学家的 agent。如果我们有几千个、几百万个，而中国只有

[52:59]dozens or a hundred, whatever, then we're going to win everything. And that was the whole argument. But now we're basically banning the data centers. We're not going to have those data centers full of geniuses. And he's done his best to restrict sale of chips to our allies as well. So, I guess our 几十个、一百个，那我们将赢得一切。整套论证就是这样的。但现在我们在禁数据中心。我们不会有那些天才数据中心了。而他还竭力限制把芯片卖给盟友。所以我们的

[53:16]allies aren't going to have the data centers either. So, we are totally shooting ourselves in the foot. And probably it would be a good thing if Dario really does believe in the data centers full of geniuses uh idea that he was arguing for last year, he should come forward and say we're making a big mistake by 盟友也不会有数据中心。我们完全是在搬起石头砸自己的脚。如果 Dario 真心相信他去年主张的「天才数据中心」理念，他其实应该站出来说：我们在过度监管数据中心上犯了

[53:31]over-regulating data centers. Here's how we should do it. I haven't heard him step up in that way. And it makes you kind of question, do you only fight for the doomer side of the argument or are there versions of the positive side that you're willing to fight for politically? 大错，正确做法是这样的。我没听到他这样站出来。这让人不禁怀疑：你是不是只为末日论那一面战斗，还是说正面版本的事你也愿意在政治上为其奋斗？

[53:47]Now, Friedberg, you're just one last point about RSI. I think this is really important is look, again, black swan, very hard to say if it's going to happen. I you know, I do think that we're headed for AGI or ready in AGI or whatever, but the most important thing about AGI is that we have choice. Like 最后，Freeberg，关于 RSI 我再说一点。我认为这真的很重要：还是那句，黑天鹅，很难说会不会发生。我确实认为我们正在走向 AGI、或已身处 AGI 之类，但 AGI 最重要的一点是：我们要有选择。

[54:03]I'd much rather have multiple AGIs than just one, right? So for example, Elon says that the way to keep AI or AGI aligned is to make it maximally truth-seeking. 我宁可要有多个 AGI，也不想要只有一个。比如 Elon 说，让 AI 或 AGI 保持对齐的方法是让它最大程度地求真。

[54:16]I kind of like that, right? That's a way to prevent it from becoming woke. What Anthropic says is the way to keep it aligned is to have this clawed constitution that is full of liberal values and you train it on that and they've kind of got this EA ethicist or whatever who's writing these things. So 我挺喜欢这个说法的，对吧？这是防止它变 woke 的办法。而 Anthropic 的说法是，让它对齐的方法是用那部充满自由派价值观的 Claude 宪法来训练它，他们还养着个 EA 伦理学家什么的在写这些东西。

[54:31]there's different approaches. I kind of want all them to produce sophisticated AI to keep checks and balances to keep these things in check. What I don't want is just a Skynet or you know, like a woke Skynet is sort of the the worst possible outcome. And you know, if we're going to 所以存在不同路线。我有点想要它们全都去产出成熟的 AI，形成制衡，让这些东西互相约束。我不想要的只是一个天网——一个 woke 天网大概是最糟的结局。而且，如果要

[54:48]go like sci-fi for a second, you know, there was like a Terminator TV show. >> Yes, that was pretty good. They they stopped >> good. It got canceled. This is like years and years ago. But anyway, what they figured out in the show, like the plot was that the way they finally defeated Skynet was they created a good 科幻一下——以前有部《终结者》电视剧。>> 对，还挺好看的。他们停——>> 剧被砍了。好多年前的事了。不管，那部剧的设定是：他们最终打败天网的方式，是造出了一个善良的

[55:03]AGI to take on Skynet. They realized like humans just couldn't do it by themselves. So they created a choice. AGI 来对抗天网。他们意识到光靠人类自己做不到，所以他们创造了选择。

[55:08]>> Maximally good. >> Yeah, they trained it to be more benevolent you know, benevolent towards humans. So in any event, and if you're going to get like really sci-fi about it, I look, I think choice has to be our hallmark. Competition has to be our hallmark. We want multiple winners here. >> 至善。>> 是，他们把它训练得更仁慈、对人类更仁慈。总之，如果要彻底科幻的话——我认为「选择」必须是我们的标志，「竞争」必须是我们的标志。我们要多个赢家。

[55:22]We don't want a small number of elites guardrailing this thing. 我们不要一小撮精英给这个东西装护栏。

[55:28]And that that is what you will get if you head in this FAA Finra direction. >> Breaking Friedberg, new nickname club just just dropped. There it is. 而那正是你们走 FAA/FINRA 方向会得到的。>> Freeberg 昵称俱乐部又添新成员了。就在刚才。

[55:37]>> [laughter] >> It's It's This is what you sounded like for a bird. Doc Brown, Marty. Marty, your future's not predetermined. You can make your own future. You can use RSI. >> 【笑】>> 这就是你刚才的既视感，鸟。Doc Brown：「Marty，你的未来不是注定的。你可以自己创造未来。你可以用 RSI。」

[55:48]>> [laughter] >> Doc Brown next >> I'm not trying to be an I'm not trying to be an advocate for RSI. I'm trying to be pragmatic about it. >> 【笑】>> 下一个 Doc Brown——>> 我不是要当 RSI 的鼓吹者，我只是务实。

[55:54]>> future. You're back. >> If If RSI is the way things are going to go, you definitely don't want to be trying to regulate AI to stall out because someone else will win it. And then you're literally handing the keys to RSI to them because they're suddenly going to be able to get there faster. >> 未来。你回来了。>> 如果 RSI 就是趋势，你绝对不想试图靠监管把 AI 拖停，因为别人会赢下它。那你就等于亲手把 RSI 的钥匙交了出去，因为他们会突然跑得更快。

[56:09]And you don't want to be in the data centers because then they're going to be offsite. >> All right, there was a peculiar story this week. Andreessen Horowitz is allegedly being investigated. This is all according to a Bloomberg report that they're being investigated by the Department of Justice for over a year. 而且你也不会想失去数据中心，否则它们会跑到境外去。>> 好，本周还有个蹊跷的故事：Andreessen Horowitz 据称正被调查。据 Bloomberg 报道，他们被司法部（DOJ）调查已超过一年。

[56:24]This is all according to Bloomberg. Under something called Section 8 of the Clayton Antitrust Act. This is a 1914 law. It's like over 100 years old prohibiting interlocking directorates, which I guess is like old-timey English for directors, like board of directors. 一切都出自 Bloomberg。依据的是《克莱顿反垄断法》第 8 条。这是部 1914 年的法律，一百多年了，禁止「interlocking directorates」（连锁董事）——我猜这是老派英语，指董事，就是董事会成员。

[56:40]And the conflict seems to stem from Ben Horowitz being on Data Bricks board while another partner from Andreessen Horowitz, Martin Casado, sits on Five Trains board. 冲突的起因似乎是 Ben Horowitz 在 Databricks 董事会任职，而 a16z 的另一位合伙人 Martin Casado 在 Five Trains 董事会。

[56:51]I don't know what if you guys have heard about this story. It was like one of the top stories this week. Uh obviously Andreessen and uh Ben Horowitz have been all in with uh Trump and big big donators. Uh put that into the mix as you will. Why 不知道你们听没听说这事，是本周头条之一。显然 Andreessen 和 Ben Horowitz 一直全力押注 Trump、是大金主。这层背景你们自行掂量。为什么

[57:08]is this happening now? >> I can tell you. >> Enemy of Andreessen Horowitz jealous of their returns >> What are we talking about? Do they not understand how venture works? Like is it like you invest in one company, you can't invest in another? I don't understand. 现在爆出来？>> 我可以告诉你。>> a16z 的仇家，嫉妒他们的回报？>> 我们到底在谈什么？他们不懂风险投资怎么运作吗？投了一家公司就不能投另一家？我不理解。

[57:23]>> What did Bloomberg say? >> They said they've been under investigation for this conflict as described between these two companies. >> Bloomberg 怎么说的？>> 他们说 a16z 因这两家公司之间的上述利益冲突被调查。

[57:29]And >> Which means Which means Which means that enemy reported them to the DOJ and then said to Bloomberg they're under investigation. 然后——>> 这意味着——这意味着有仇家向 DOJ 举报了他们，然后又对 Bloomberg 说「他们在被调查」。

[57:36]>> Well, that's why I bring it up is like it seems so odd. >> Which to me which to me likely seems I will make a perspeculative guess. >> 呃，这也是我提这事的原因，太反常了。>> 在我看来——我做一个投机性的猜测：

[57:45]The CEO of one of these companies is pissed off that the other partners on the other board. And reported it to the DOJ and reported it to Bloomberg at the same time. 其中一家公司的 CEO 对另一边的合伙人进对手董事会感到恼火，于是同时举报给了 DOJ 和 Bloomberg。

[57:53]>> Interesting. All right, well there you have it folks. Moving on in the docket and I think we made quick work of that one. Let me let me let me just say you know you've been seeing like the distribution of like cursor and then space x. >> 有意思。好了，各位，真相如此。继续议程，这块我们算是速战速决。我只说一句：你们都看到了 Cursor 的份额分配，然后是 SpaceX。

[58:06]And Andreessen Horowitz is on fire. They are crushing it. So. a16z 现在火得不行，他们杀疯了。所以。

[58:12]>> They've had two wins in the last few weeks. >> I mean my god. Congrats to them. They are on fire. Keep going. >> 他们最近几周赢了两把。>> 我的天。恭喜他们。火得不行，继续保持。

[58:19]>> Well, it's cursor and didn't they make a lot of money off of open router? And space x >> And open router just went out million. >> 呃，是 Cursor，还有他们不是在 OpenRouter 上赚了一大笔？还有 SpaceX——>> OpenRouter 刚出了个 million（级回报）。

[58:26]Open router was more like an amuse-bouche. It was like 6 or 700 million. >> [laughter] >> The cursor the cursor was decent. Yeah, it was like an appetizer but space x was the main course. That's like 15 17 billion. OpenRouter 更像餐前小点，大概六七亿美元吧。【笑】Cursor 还不错，算开胃菜，SpaceX 才是主菜——那可是 150 到 170 亿美元。

[58:37]>> Yeah. >> Just just by the way even if you want to say that VC firms should not invest in competitors, the reality is that startups exist a very dynamic environment. They frequently pivot and change their business models and two companies that start off in totally different spaces can end up competing >> 是。>> 顺便说，就算你坚持认为 VC 不该投相互竞争的公司，现实是创业公司活在非常动态的环境里，频繁转型、改商业模式，两家起点完全不同的公司最后可能

[58:54]with each other. So now these are different board members and they can probably set up a you know a firewall between them. 打起来。何况这还是两个不同的董事，他们之间大概可以设一道防火墙。

[59:00]>> Yes. Yeah, trust me sorry these are not these are not filing companies. What confidential information did any of these private businesses have? >> 是。说真的，这些又不是上市公司。这些私人公司能有什么机密信息？

[59:09]>> And and by the way. Andreessen Horowitz could also just put an independent director in there and just be done with it. So like if this was some ticky-tacky very important thing I don't know why the DOJ is on this. >> 而且顺便说，a16z 完全可以派一个独立董事进去，一了百了。如果这真是多要紧的事，我不明白 DOJ 为什么盯这个。

[59:20]>> I think Freeburg is right now because when you when you make a claim to the DOJ they technically if they open a file you are technically under investigation. >> 我觉得 Freeberg 说得对：你向 DOJ 提出一项指控，他们技术上只要立了案，你就「在被调查中」。

[59:29]So then you can you can report that to a buffoon reporter who doesn't know from Shinola and they'll just repeat it. 然后你就可以把这透露给一个分不清鞋油的傻冒记者，对方照单全收。

[59:35]It means nothing. All right. And then the DOJ is doesn't tell you that that the you're under investigation and then they also don't tell you when it's closed. 这说明不了什么。而且 DOJ 不会告诉你被调查了，也不会告诉你案子什么时候了结。

[59:43]>> Same as the SEC. >> PR ratchet for an enemy to exact some sort of chaos. >> SEC 也一个样。>> 这是仇家制造混乱的 PR 绞索。

[59:49]A H is crushing it. God bless them. Godspeed to them. >> It is interesting that there's even this rule, but we had some fun with that. So let's keep moving. I want you, Freeberg, to join me and Sachs invest each of them up at the All-In a16z 杀疯了。上帝保佑他们，一路顺风。>> 有意思的是居然还有这么一条法规。不过这事我们乐过了，继续。我需要你们，Freeberg、Sachs——

[60:04]Summit happening September 13th, 14th, and 15th in Los Angeles. All-Star Weekend has been announced and hey, guess who's headlining the party? The same guys who sell out Madison Square Garden. 峰会将于 9 月 13、14、15 日在洛杉矶举行。「全明星周末」已官宣，猜猜谁在派对压轴？就是那支能唱爆麦迪逊广场花园的组合。

[60:16]I know that Freeberg has a personal relationship. Go ahead, tell everybody who's going to be on the music and smokers. >> which is going to be awesome. So we'll have them at the closing party. It's going to be great. And then we're also doing another wow full takeover at Universal Studios. 我知道 Freeberg 有私交。来，告诉大家谁会上台——音乐和「Chainsmokers」。>> 肯定炸裂。他们会出现在闭幕派对上，会非常棒。另外我们还会在环球影城再搞一次全场包场。

[60:30]We got a whole party at Universal, which was awesome last year. Uh where Sachs and I were with Diplo on stage. Jake, I you were there, too. It was great. 去年在环球的整场派对太棒了，我和 Sachs 跟 Diplo 一起在台上。Jason，你当时也在。特别嗨。

[60:39]>> Yeah, I was on stage dancing. >> Who's Diplo? >> Diplo's [laughter] great. The DJ. >> Just say Diplo. >> 是，我还上台跳舞了。>> Diplo 是谁？>> Diplo 很【笑】棒啊，那个 DJ。>> 直接说 Diplo 就行了。

[60:44]>> You're getting Diplo and Chainsmokers mixed >> They blew our minds. >> Come to that concert. We have other special unannounced guests that we're not going to announce. So this is it. >> 你把 Diplo 和 Chainsmokers 搞混了——>> 他们把我们震到了。>> 来那场演唱会吧。我们还有其他未官宣的神秘嘉宾，就不公布了。就这样。

[60:52]We're done making announcements. >> All right, we know the Democrats have been favored to take the House in the midterms, which are just around the corner. Polymarket has them at 88% favorites, but Polymarket now has the Dems at 52%. Basically a toss-up. 该宣布的都宣布完了。>> 好，我们都知道中期选举近在眼前，民主党一直被看好拿下众议院。Polymarket 一度给到 88%，但现在 Polymarket 显示民主党只有 52%，基本是五五开。

[61:07]Current balance of power, GOP 53 seats, Dems 47. Which means I guess the Dems have to flip four of those without losing any. 当前席位格局：共和党 53 席，民主党 47 席。也就是说民主党要翻 4 席且一席不能丢。

[61:16]35 seats are up in 2026. Uh there are some special elections going on in Florida and Ohio. 23 of those 35 seats are held by Republicans. 2026 年有 35 个席位改选。佛罗里达和俄亥俄还有一些特殊选举。35 席里 23 席由共和党持有。

[61:26]Big races to watch, Alaska, Maine, North Carolina and Ohio. On February 25th, the day before the Iran war started, GOP was favored 60 to 41. It's been as high as 70%. Here's your Polymarket. 重点观战选区：阿拉斯加、缅因、北卡、俄亥俄。2 月 25 日，伊朗战争开打前一天，共和党还被看好 60 比 41。曾经高到 70%。这是你的 Polymarket。

[61:40]Sachs, what do you take from this? Why are the Republicans set to get trounced in the midterms? Sachs，你怎么看？为什么共和党会在中期选举里被痛击？

[61:49]>> Well, first of all, polls are notoriously wrong in the summer before an election and there's some data on this. So, we should just If you're going to cite polls, let's just actually talk about the topic of whether political polls are even accurate at this point. >> 首先，大选前夏季的民调出了名的不准，这有数据支撑。所以，既然要引用民调，我们就干脆聊聊「政治民调到底还准不准」这个话题。

[62:02]>> These aren't polls. This is Polymarket, just to be clear. The polls are slightly different, yeah. These are people betting. >> 这些不是民调，是 Polymarket，说清楚。民调还稍有不同。这是真金白银在下注。

[62:08]>> Yeah, but the polls are shaping public opinion right now. So, political scientist Patrick Ruffini analyzed over 3,000 polls over the last four election cycles, every 2 years since 2018. And what he saw is that there is a huge >> 对，但现在左右舆论的正是民调。政治学家 Patrick Ruffini 分析了 2018 年以来四个选举周期、3000 多份民调，发现存在巨大的

[62:23]bias. The preponderance of polls in all four cycles favored the Democrats and the average polling error across all the polls was D plus 3.7 relative to the election outcome. And I think this polling problem has become even worse recently and you see it where there 偏差。四个周期里绝大多数民调都偏向民主党，所有民调相对选举结果的平均误差是 D+3.7。而且我认为这个民调问题最近更严重了，你能看到

[62:41]seems to be an over sampling of young progressives in these polls because it's not just the fact that the polls favor Democrats over Republicans, they also favor the DSA Democrat over the more moderate or centrist Democrats and you've seen this in a few elections 这些民调明显过度抽样了年轻进步派——不只是民调偏向民主党那么简单，它们还偏向 DSA 民主党人而非温和派或中间派民主党人，这一点你在好几场选举里都见过：

[62:56]where the DSA candidate was supposed to win by a lot and then they basically eked it out or actually lost like in Wisconsin. So, in any event, I just think that the the polls are very unreliable at this time of the year and a lot of the polls have They're actually created as internals DSA 候选人本被看好大胜，结果只是险胜，甚至干脆输了，比如威斯康星。总之，我只是认为这个时节的民调非常不可靠，而且很多民调其实是竞选团队的内部民调，

[63:14]by campaigns to motivate their donors. >> So, basically, people are upset, then they throw people out of office. 用来给自己的金主打气的。>> 所以 basically 就是：老百姓不爽，然后把人投下台。

[63:21]That's just the trend. And that's what happened over and over again, yeah? >> No, I just think that the polls somehow are over sampling Democrats and progressives and have a bias. That's what the data is showing. So, I just wouldn't read too much into this right now. I do think that the Democrats, I 这就是趋势，而且一次又一次重演，对吧？>> 不，我只是认为民调不知怎么过度抽样了民主党人和进步派，存在系统性偏差。数据就是这么显示的。所以现在别太当回事。我确实认为民主党——

[63:38]think it is legitimate to say that they are favored to win the House. I think the Senate right now is a toss-up and I would say that the Republicans will ultimately keep it if I had to predict right now. And I just think that when you get back from Labor Day, that's when the campaign for the midterms really starts and I think that 说他们被看好赢下众议院是合理的。参议院目前是五五开，如果非要我现在预测，我会说共和党最终守住参议院。而且我认为，劳动节假期回来之后，中期选举的竞选才算真正开始，我认为

[63:56]Republicans need to focus on the right messages in this campaign and I think there are plenty of good ones. They just have to remind them >> What are they? 共和党需要聚焦正确的竞选信息，而且好素材多得是，他们只需要提醒选民——>> 比如哪些？

[64:03]>> things. >> What What are they What would you say the top three they should focus on are? >> Well, look, everyone's been over indexing on the war and that's been the focus of the news and hopefully the kinetic phase of that is over and doesn't restart. And assuming that's the case, I think the Republicans can remind >> 哪些事。>> 你觉得前三条该聚焦什么？>> 听着，所有人的注意力都过度压在那场战争上，那也是新闻的焦点，希望其动能阶段已经结束并且不再重启。假设如此，我认为共和党可以提醒

[64:18]the electorate of everything that's happened over the last couple years. I think there is a good story to tell. So, first of all, remind people that the border has been sealed for the last year and a half. We haven't just slowed Biden's invasion, we've stopped it completely. Contacts at the border have fallen off a cliff to the lowest level 选民过去两年发生的一切。这是有故事可讲的。首先，提醒大家：边境已经被封了一年半。我们不只是放缓了拜登的入侵，而是彻底拦停了。边境接触量断崖式下跌到 50 年来

[64:34]in 50 years. Illegals aren't even really trying to cross the border anymore. It's been so successful. Second, sealing the border and deporting hundreds of thousands of the worst criminals has actually made the country much safer. 最低。非法移民甚至都不怎么尝试越境了。就是这么成功。第二，封锁边境并遣返数十万最恶劣的罪犯，确实让这个国家安全多了。

[64:45]Murders have dropped around 20% year over year and the murder rate is the lowest in 70 years. So, that's number two. Third, deaths from drug overdoses have dropped 20%. Again, thanks to tougher border enforcement. Fourth, Americans saw a 17% increase in tax 凶杀案同比下降约 20%，凶杀率处于 70 年来最低。这是第二条。第三，药物过量死亡下降了 20%，同样归功于更严的边境执法。第四，美国人的退税

[65:00]refunds thanks to the president's no tax on tips and no tax on overtime and no tax on social security. So, these are real increases in take-home pay as a result of those tax cuts. Also, the tax provisions in big beautiful bill incentivize new equipment buying. We 增加了 17%，得益于总统的小费免税、加班费免税和社保免税。这些是减税带来的实实在在的到手收入增长。另外，大美丽法案的税收条款激励购置新设备。我们

[65:16]talked about the accelerated depreciation and perhaps that's why small business optimism just reached its highest level in nearly a year and more small businesses are saying they plan to hire in the months ahead. So, I think we do have a strong economy. Prescription 聊过加速折旧——也许正因为此，小企业乐观指数刚创下近一年新高，更多小企业表示未来几个月计划招人。所以我们确实有一个强劲的经济。处方药

[65:31]drugs and auto insurance costs both fell last year after increasing a lot during the Biden administration. I think prescription drugs has had their steepest annual decline since 1963. 和汽车保险费用去年双双下降——拜登时期可是大涨过。处方药出现了 1963 年以来最大的年度降幅。

[65:44]So, there are positive things to talk about there. Everyone made a big deal of the surge in egg prices at the start of the administration, but few have bothered to note that there's now been a 39% decrease year-over-year in July. So, look, there are many good things I think for Republicans to talk about. They just 所以有的是正面素材可讲。所有人都揪着本届政府初期蛋价大涨做文章，却没人提 7 月蛋价同比已经下降了 39%。总之，共和党有的是好东西可讲，

[66:00]have to kind of remember everything that's happened in the last two years. >> Okay. Yeah, Freeberg, do you have thoughts on what is going to happen this midterm and what people are making their decisions based on? 只需要记起过去两年发生的这些事。>> 好。Freeberg，你对这次中期选举的走向、以及人们基于什么做决定，有什么想法？

[66:14]>> Things are expensive. People can't afford housing, health care, or education. >> 东西太贵。人们负担不起住房、医疗和教育。

[66:19]And are leaning towards this DSA concept. There was a frightening statistic on young conservatives embracing socialism. 并且正在倒向 DSA 那套主张。有个数据很吓人：年轻保守派在拥抱社会主义。

[66:27]>> Really? Wait, the the conservatives like MAGA folks you're saying are embracing socialist tenants. >> 真的假的？你是说 MAGA 那样的保守派在拥抱社会主义信条？

[66:37]>> How are they conservatives then? >> Well, it's the name. I guess I'm sensibly they're flipping or they're changing the composure of the issues they care about in some way. >> 那他们还叫什么保守派？>> 呃，名头还是那个名头。大概在实质上，他们在翻转，或者说在改变自己关心的议题构成。

[66:46]>> 53% of self-identified conservatives under 40 support government-run grocery stores. >> 40 岁以下自我认同为保守派的人里，53% 支持政府开办国营食品杂货店。

[66:54]Not Democrats, conservatives. Take that in. First line of this Wall Street Journal article. July 2026 Fox News poll found that 61% of Republicans view capitalism favorably from 72% in 2019. 不是民主党人，是保守派。品品这句话。这是《华尔街日报》这篇文章的第一行。2026 年 7 月福克斯新闻民调显示，61% 的共和党人对资本主义持正面看法，而 2019 年是 72%。

[67:10]While the share who feel strongly favorable has fallen from 54 to 41%. 其中「强烈正面」的比例从 54% 跌到 41%。

[67:16]The share who say the system is rigged towards the wealthy has climbed from 30% in 2018 to 42% amongst Republicans. 认为「体制偏向富人」的共和党人比例从 2018 年的 30% 升到 42%。

[67:24]So, I don't think that this is a principled line in the sand anymore. 所以我认为，这已经不是一条有原则的底线了。

[67:31]I think that the the problem with unaffordability, which I would argue continuously and I've argued it since the first day of the show, is a function of government spending. 我认为负担能力问题的根源——我从节目第一天就一直在论证——是政府开支的函数。

[67:43]The more government spends on education, health care, and housing, the more expensive those things get, rather than the opposite. They're supposed to be able to provide access to those things, but when the government intervenes in those markets, those things get more expensive. And no one wants to kind of 政府在教育、医疗、住房上花得越多，这些东西就越贵，而不是相反。这些本该是政府提供可及性的领域，但政府一干预市场，东西就变贵。而两党都没人愿意

[67:58]acknowledge or admit that in either party. And I've been to DC enough now to know that no one actually acknowledges or cares about this point. That they all think that their job is to get more money out of the government for their constituents. Every single person in every party. 承认这一点。我去华盛顿去得够多了，我知道没人在乎这个观点。所有人都认为自己的工作是为选民从政府手里多抠钱出来。每一党、每一个人都是。

[68:14]There's very few principled conservatives. Many of them are conservatives by title, by party affiliation, but they're not conservatives by principle in the traditional sense. Small government, free markets, et cetera. 有原则的保守派凤毛麟角。很多人只是名义上、党派归属上的保守派，但不是传统意义上小政府、自由市场那种原则上的保守派。

[68:28]And I think that this is the motivating factor, I I would argue, if you take the the look at the macro cycle. 而我认为这就是驱动因素——如果你看宏观周期的话。

[68:36]The macro cycle being we printed so much money, we've got so much debt. The 30-year is now at 5.3% this week. It continues to hit new 20-year records every day. We I think are going to continue to face a problem of unaffordability. 宏观周期是：我们印了太多钱、背了太多债。30 年期国债本周到了 5.3%，天天刷新 20 年纪录。我认为我们将持续面临负担能力问题。

[68:52]And ultimately there is going to be some sort of big movement towards more socialist policies that we're going to see play out between now and 2028. It's why I continue to believe I think someone like an AOC becomes president. 而最终，2028 年之前会出现某种向更社会主义政策的大转向。这也是我一直相信 AOC 这号人会当总统的原因。

[69:07]I think there's going to be a wave this year and then there's going to be a bigger push in 2028 because I don't see the arithmetic or the fiscal situation changing to actually alter the affordability landscape. I just don't. I I don't know what you could do from a policy perspective and I don't hear either 我认为今年会有一波，然后 2028 年会有更大的浪潮，因为我看不到财政状况的算术有什么变化能真正扭转负担能力困局。我就是看不到。我不知道政策上还能怎么办，而两党

[69:22]party talking about actual solutions to solve this affordability problem between now and 2028. And then the other kind of big problem that I've talked about is assets. Middle-class people, people that worked in private companies, people that don't depend on social security or public pension funds, 也都没在谈 2028 年前解决负担能力问题的真正方案。另一个我聊过的大问题是资产。中产阶级、在私企工作的人、不靠社保或公共养老金的人，

[69:39]they've all had 401(k)s and owned homes. So, they have seen their net worth climb with asset inflation surge. So, the top 50% of Americans have 178 trillion dollars of net worth. 都有 401(k)、都持有房产。所以他们的净资产随着资产通胀的狂潮水涨船高。美国财富前 50% 的人拥有 178 万亿美元净资产。

[69:55]And the bottom 50% of Americans have 6 trillion of net worth. It's not the billionaires, it's the the whole middle class. The middle class has all the wealth in the United States. And that's because they've owned assets. They've owned houses and they've owned stocks. 而后 50% 的美国人只有 6 万亿。关键不是那几个十亿富豪，而是整个中产阶级——中产阶级拥有美国的绝大部分财富。因为他们持有资产：房子和股票。

[70:09]Houses and stocks have surged in value. And so, if you don't own a house and you don't own stocks, you were left behind. 房子和股票价值狂涨。所以，如果你没有房、没有股票，你就被甩下了。

[70:16]And so, I don't see how from a uh public policy perspective we don't find ourselves walking down this very nasty winding dark path of socialism because the alternative policies that could be put in place to solve these problems are not being put forth by either party 所以从公共政策视角，我看不出我们怎么可能不走上一条肮脏、曲折、黑暗的社会主义之路，因为能解决这些问题的替代性政策，两党现在

[70:31]right now. So, I do think that there's going to be a big socialist embrace. I've said this obviously for a number of years and I think we're going to see this um play out in this midterm and that's going to create the the momentum going into 28. And look, I mean this this polling data is very worrying. It's not just the Democrats, it's not just a 一个都没提出来。所以我确实认为会出现一次对社会主义的大拥抱。这我说了好几年了，我认为这次中期选举就会上演，并为 2028 年积蓄动能。而且，这组民调数据非常令人担忧——不只是民主党，不只是

[70:48]party-based system, it's the horseshoe. It's the 63% of Americans that are living paycheck to paycheck that can't afford That's the Americans that are going to vote for this system and it doesn't matter what party it is. They're just going to vote for it. 政党问题，这是「马蹄铁」现象。是那 63% 月光族、负担不起生活的美国人——是这批人会把票投给这套制度，跟哪个党无关。他们就是会投。

[71:01]>> Chubbs, 20 cents on the dollar going to our interest approximately on the federal side. Inflation. >> Chubbs（指 Freeberg），联邦层面大约每 1 美元就有 20 美分拿去付利息。通胀呢？

[71:09]We thought we would get there. It's been three or four percent for the whole year. Haven't been able to get down to two. 我们原以为会降下来。结果一整年都在 3% 到 4%，始终下不到 2%。

[71:15]And the GDP concept which we talked about here. Getting to growth again and if we can get it, we can grow our way out of it that is hasn't happened as well. 还有我们节目里聊过的 GDP 增长逻辑：重回增长、靠增长走出困局——这也没有发生。

[71:25]Obviously the war big disappointment for many folks in both parties who were promised no war and here we are six months into it. What do you think is going to happen over the next through the midterms and then into 2028 显然，那场战争让两党很多人大失所望——当初承诺的是不打仗，结果打了半年了。你觉得从现在到中期选举、再到 2028 年会怎么走？

[71:40]and to freebird's point any way to get out of this? Is this a debt spiral or is it something we can course correct? 还有，顺着 Freeberg 的观点：有出路吗？这是债务死亡螺旋，还是我们能纠偏的事？

[71:46]Out of in your mind. >> I think that we can course correct. I just think it takes will. 你怎么看。>> 我认为我们能纠偏，只是需要意志。

[71:51]And I think that we need corporate America to step up and do their part of the job and we need governments mostly to get out of the way and not have any hairbrained schemes on new inventive legislation that they 而且我认为我们需要企业美国站出来做好自己那份，需要政府基本上让开路，别搞什么自以为能解决问题的新发明立法。

[72:08]think can solve the problem. I don't trust any polls just to be clear. I don't believe this polling data. 另外说清楚：我不信任何民调。我不信这组数据。

[72:15]And I think that we've gotten into a very dangerous point which we already got into with the media. 我认为我们已经到了一个非常危险的节点——媒体那边我们早就是这样了：

[72:21]Where the media lies. We all now know that the media lies and the media can't be trusted. 媒体撒谎。我们现在都知道媒体撒谎、媒体不可信。

[72:29]I would like to now go out on a limb. I don't think it's much of a limb though. Pollsters lie. And I don't think you can trust them either. 现在我再往前探一步——虽然也不算多冒进：民调机构也撒谎。我认为它们同样不可信。

[72:36]And I think that the media acts in concert with pollsters to find the data that allows them to write the headline that hopefully allows them to get the clicks. 而且我认为媒体和民调机构是合谋的：一起找到能支撑标题的数据，指望靠标题拿点击。

[72:50]So I would just start there. Pollsters, trash. 所以我就从这里说起：民调机构，垃圾。

[72:55]Media, trash. So, what is really happening on the ground? I don't know, but I don't trust what this data tells me. Now, what do I know is helping and hurting the incumbent sitting president? 媒体，垃圾。那地面上真实发生着什么？我不知道，反正我不信这些数据告诉我的东西。那我知道什么在帮、什么在伤这位在任总统？

[73:08]I think that if we as in corporate America get ourselves in a position to actually present the better face of capitalism, we allowed the system as it stands to 我认为，如果我们企业美国把自己摆到能真正展示资本主义更好一面的位置上，我们就能让现有体系

[73:23]continue to work. And I think all of these things, if we don't wake up, are telling us that corporate America is not trustworthy anymore. 继续运转。而这一切，如果我们还不醒，是在告诉我们：企业美国不再可信了。

[73:35]Corporate America is not aspirational. Corporate America does not do the things that engender some amount of trust and respect. 企业美国不再是励志的存在。企业美国不再做那些赢得信任与尊重的事。

[73:46]And specifically in Silicon Valley, we've completely lost the script because we've gotten to the far, far, far end of that. 具体到硅谷，我们已经彻底丢了剧本，因为我们在这条路上走到了远远远远的极端。

[73:53]And everybody is in it for themselves to such a degree that it's either pearl clutching on the one hand or getting your hand caught in the cookie jar on the other. You know, I mean, I I posted the cookie jar thing that Anthropic did on the healthcare side. 人人各怀私心到这种程度：一边是圣母式表演，一边是手伸进饼干罐被抓个正着。我是说，我发过 Anthropic 在医疗保健领域干的那档子「饼干罐」事。

[74:08]>> Cookie jar. >> Why are we Why are we in this? I mean, it's just so avoidable. Those are headwinds for the incumbent president, even if he has nothing to do with it. >> 饼干罐。>> 我们为什么要掺和这种事？完全可以避免的。这些对在任总统是逆风，哪怕跟他毫无关系。

[74:17]What is a tailwind? I think getting out of the Iran war can be a massive tailwind. 那什么是顺风？我认为走出伊朗战争可以是巨大的顺风。

[74:24]I think structurally solving energy on behalf of American citizens can be a tailwind. 我认为为美国公民从结构上解决能源问题可以是顺风。

[74:35]I think making sure that all of the long-term and immediate-term investments of GDP can actually come and bear fruit can be a tailwind. All of those things, there's still a lot of levers to be pulled between now and November. And so those can create pretty 我认为确保所有长期和短期的 GDP 投资真正落地结果可以是顺风。这些杠杆，从现在到 11 月还有很多可拉，能创造出相当

[74:53]meaningful tailwinds. But the headwinds, unfortunately, are pretty strong and it requires some real coordination that the White House is going to have to do. To get a bunch of these corporate CEOs in the room, to get a bunch of these Silicon Valley CEOs in the room, and to tell them, "Guys, enough of this bull Play nice in the sandbox cuz you're 有分量的顺风。但逆风 unfortunately 也很强，这需要白宫真的去做协调：把企业 CEO 们请进房间、把硅谷 CEO 们请进房间，告诉他们：「各位，这套胡闹够了，在沙坑里好好玩，因为你们马上

[75:09]about to lose the standard operating procedure that allowed you to exist in the first place." 要失去那套让你们得以存在的基本运作规则了。」

[75:14]>> Yeah. >> And they need to get this message clearly. >> Yeah, and I I can tell you the exact reason democratic socialism is going to win, continue to win, um and these far-left democratic candidates are going to dominate. And it's, you know, it's the Republicans' fault. They had all three >> 是。>> 他们必须清清楚楚听到这个讯息。>> 是。我可以精确告诉你民主社会主义为什么会赢、会继续赢，这些极左民主党候选人为什么会主导——这得怪共和党。他们可是把三权

[75:31]branches of government here really tightly in their control. And if you look at what the podcast bros and the pro podcast referencing stack is responding to, and I listen to them, you know, not every episode, but I I I I I dip in and out of Tucker, who's now 都攥得很紧。你去看看播客兄弟们、那个亲播客信息流在回应什么。我也听他们，不是每期都听，但我断断续续听 Tucker——他现在

[75:47]making his own political party, Joe Rogan, Tim Dillon, you know, all these folks who were all in for Trump, they cite Epstein investigation, which, okay, whatever. I don't think it's particularly material, but it is to them. The Iran war is obviously top of 在组建自己的政党——还有 Joe Rogan、Tim Dillon，这批当初全力押 Trump 的人。他们提爱泼斯坦调查——行吧，随便，我觉得不重要，但对他们重要。伊朗战争显然是他们眼中

[76:04]what they believe is the problem here cuz they were promised that wouldn't happen. And then, it's to your point, Freeberg, this inflation thing has not been solved and it just keeps going and we're going to be sitting here, I think, in another 6 months or 12 months talking 问题的首位，因为当初承诺不会开战。然后就是你说到点子上的，Freeberg：通胀没解决，还在持续。我觉得再过 6 个月、12 个月，我们会坐在这里谈：

[76:20]about, "My lord, every investigation possible into the Trump administration, and we need to start looking at what what what would be things, to your point, Chamath, and I thought you summarized some of the 「天哪，对 Trump 政府的每一种调查都安排上」——而且我们需要开始看，顺着你说的，Chamath，我觉得你对资本主义

[76:36]the broken parts of capitalism, I think, well. But we got to get granular about it. And I brought up last week just the edge behaviors in capitalism where we over optimize. I felt the delivery providers that and Sachs and I had a pretty hearty debate about it last 溃烂处的概括很到位。但我们必须落到颗粒度。我上周提过资本主义里的「边界行为」——我们过度优化。我感觉那些配送服务商——上周我和 Sachs 为此大吵了一架——

[76:51]week for just being too clever, squeezing too much out of the the last mile, quite literally, on Amazon's part. How we treat the people who are at the bottom of the employment stack is what's causing this. 就是太精了，字面意义上把「最后一公里」榨得太干，亚马逊就是个中典型。我们怎么对待就业链条最底端的人，正是这一切的根源。

[77:04]And not being able to solve the cost of homes and not making that a focus, not being able to solve the cost of child care, not being able to move the minimum wage, I think we have to look at all of these things and say, that's or Trump should 解决不了住房成本、不把这当重点，解决不了育儿成本，推不动最低工资——我们必须正视所有这些事，然后说：该动手了。或者说 Trump 应该

[77:21]look at all of those things and he should build the second half of his term just around those things. No more crypto stuff, no more wars, no more attacking our allies, all of this performative stuff and and just mistakes, they got to go away and we got to focus on what he sold people on. 正视所有这些，把任期的下半场完全围绕这些来搭。别再搞加密货币那些事，别再打仗，别再攻击盟友，所有这些表演性的东西、这些失误，都得靠边，回到他卖给选民的东西上。

[77:37]People feel they were sold that this would be a populist presidency that took care of the working man and woman in this country and it has been the opposite. Sachs, I'm sure you have some thoughts and we'll give you the final word. 人们觉得自己买到手的承诺是：这会是一个照顾本国工薪男女的民粹主义总统任期，而实际恰恰相反。Sachs，你肯定有想法，最后的话给你。

[77:50]>> on the inflation thing cuz you talked about it and Freeberg talked about it. I mean, basically you're saying the cost of living's going to drive everyone to socialism. I agree there's some risk of that, but let's break it down into a little bit more nuance here. >> 谈谈通胀，因为你也提了，Freeberg 也提了。基本上你是说生活成本会把所有人推向社会主义。我同意存在这个风险，但咱们拆得更细一点。

[78:02]If you look at where the most inflation has occurred, it's always in those sectors of the economy that the government's most involved. So, healthcare and colleges where and housing where the government has massive amounts of subsidies and regulation, it only seems 你看通胀最严重的地方，永远是政府介入最深的经济部门。医疗、大学、住房——政府补贴和管制最重的领域，结果只是

[78:19]to make the problem worse. Kind of makes sense that if the government's paying for something, then the providers have a huge incentive to charge as much as possible. And in fact, a lot of them engage in fraud as well. So, we know these areas are rife with abuse. But 让问题更糟。道理不难懂：政府买单的东西，供给方就有巨大动机往死里收费，甚至大量造假。我们知道这些领域舞弊横行。而

[78:34]those are the areas that are most inefficient and where you see the price going through the roof. I mean hospital services man, that's like just exponential just about. And then if you look at the areas where we've had not just no inflation but actually prices have come down 这些恰恰是最没效率、价格飞天的地方。医院服务那曲线简直指数级。再看不但没通胀、价格反而下跌的领域：

[78:49]it's in all the tech areas. It's basically TVs or computer software cell phones things like that the prices have come down a lot. By the way this chart is showing all the way from 2000 to 2026. So it's a 26 year time 全在科技那边。基本就是电视、电脑软件、手机这类，价格大跌。顺便说，这张图覆盖 2000 到 2026 年，26 年的

[79:06]frame. And during that time overall inflation went up 96%. So the price level almost doubled in that 26 year period. Fortunately >> on average >> on average. That's right. But like I said with certain things like health care becoming 时间跨度。期间整体通胀涨了 96%，也就是 26 年里物价几乎翻倍。幸运的是——>> 平均而言——>> 平均而言，对。但就像我说的，某些东西比如医疗变得

[79:22]you know 300% almost more expensive whereas computer software went down by 67% in real dollars. 贵了近 300%，而电脑软件按实际美元算跌了 67%。

[79:30]Meanwhile the average hourly wage grew 135% so at least the average hourly wage grew by about 40% more than the inflation level but obviously it still bites that you have all this inflation. 同期平均时薪涨了 135%，所以至少平均时薪比通胀多涨了约 40%，但 Obviously，身处这么多通胀里还是很扎心。

[79:44]I guess my point is just how do you solve it? We know that you bring prices down through vigorous competition not through more government interference in these markets and unfortunately what happens is that socialism causes the problem and then what happens is that 我想说的是：怎么解决？我们知道要靠激烈竞争来降价，而不是靠更多政府干预。而不幸的是现实是：社会主义制造了问题，然后

[80:01]groups like the DSA argue that the way to solve the problem is through more socialism. So it's a disease that claims to be a cure and then people take it and they take more of it and the disease gets worse and worse. And that that is the spiral that I think we risk being in. DSA 这类团体跳出来说，解决问题的办法是更多社会主义。这是一种自称是解药的病，人们吃下它、越吃越多，病就越来越重。我认为我们面临的就是这个螺旋。

[80:16]>> We are in a socialist death spiral now. All this spending now is we're not in a death spiral. Take it easy. >> 我们现在已经处在社会主义死亡螺旋里了。现在这些开支——>> 我们没在死亡螺旋里。淡定。

[80:24]Um >> We're not in a death >> I know I mean it's great it's great it's great for you to say cuz it sounds like really like you got some real hot take we're not there. 呃——>> 我们没在死亡——>> 我知道，我的意思是，你这么说显得你像有什么惊天暴论似的，其实我们还没到那一步。

[80:34]>> Our mayor is doing fine. We are cranking. The economy is great. There are things that need to be fixed. >> 我们市长干得挺好。我们火力全开。经济很好。该修的东西是有的。

[80:41]I would say start with corporate America. Let's get our act together and let's start doing things that are honorable, interesting, and aspirational. 我建议从企业美国开始。让我们把自己整顿好，开始做体面的、有趣的、令人向往的事。

[80:49]>> Yeah, I think you totally missed the point. I think the real point here is if you hadn't started the war, we wouldn't have the same inflation issues. We wouldn't have the same expense issues and this would be a totally different picture. And it was his job sacks when he promised the American people he would focus on affordability and he had control of both >> 是，我觉得你完全没抓住重点。真正的重点是：如果没开那场仗，我们就不会有这些通胀问题，不会有这些开销问题，局面会完全不同。而且 Sachs，这是他的职责所在——他向美国人民承诺过聚焦负担能力，而且他当时手握

[81:06]houses and the presidency and he's willing to do any executive order on crypto, on wars, on anything, but he didn't do any executive orders on affordability. And that's the failure of this presidency. 参众两院和总统职权，加密货币的行政令他愿意签、打仗他愿意干、什么都愿意干，唯独负担能力上一个行政令都没出。这就是这届总统任期的失败。

[81:16]>> missed the point where Jamal said it would be a tailwind to get out of this war. I think the kinetic part's already over by the way. So, it's just a >> Well, let's hope. I mean, it just should never been started um and I think >> But in any but in any event, look, the president passes economic program last year, a big beautiful bill and I think it's working. We're seeing a huge >> 你是没听到 Chamath 刚才说「走出战争是顺风」。顺便说，我认为战争动能阶段其实已经结束了。所以——>> 希望如此。我是说，这仗根本就不该打，而且——>> 但不管怎样，听着，总统去年通过了经济计划，大美丽法案，我认为它在起效。我们看到巨大的

[81:32]investment boom. We're seeing small business confidence rise to its highest level in a long time. They are saying they're going to create more jobs over the next year. I think the economy is doing pretty well. I mean, we do have these issues. I'm not saying we don't and spending is a long-standing problem. 投资热潮。小企业信心升到很久未见的水平，他们说明年会创造更多就业。我认为经济相当不错。我们确实有这些问题，我不否认，开支是长期顽疾，

[81:47]Can't disagree with you there. However, I don't think Democrats have the solution to the problem and the Democratic Socialists definitely don't have the solution to the problem because they want even more spending. 这点没法反驳你。但是，我不认为民主党有解，民主社会主义者更没解，因为他们要的是更多开支。

[81:58]>> we we are now incentivizing young people in the Republican Party to embrace socialism. Moderates are embracing socialism. They're all looking at it saying, "Well, that sounds better. If Trump won't be a populist, maybe these people are the populist." And and that's what I mean by the death spiral because they're going to just keep spending >> 我们现在是在激励共和党里的年轻人拥抱社会主义。温和派在拥抱社会主义。他们都看着说：「嗯，听起来更好。如果 Trump 不当民粹派，也许这帮人是民粹派。」这就是我说的死亡螺旋，因为接下来只会是没完没了地花钱。

[82:13]money. They're going to increase the velocity of spending the way it was in >> You may be right, which is scary. 他们会加大花钱的速度，就像当年——>> 你可能是对的，这很吓人。

[82:19]>> By the way, that's that's what I I don't think we're like in it now. We're kind of we're risking going to >> Just just one last just one last point. >> 顺便说，这就是我想说的——我不认为我们现在已经在螺旋里，我们是「有风险滑进去」——>> 最后就说最后一点。

[82:25]They did a cost estimate of the policy platform proposed by the DSA. And we've talked about that before. And the estimate is that it would project new federal spending of between 71 trillion and 212 trillion >> By the way, 有人给 DSA 提出的政策纲领做了成本估算，我们之前聊过。估算显示它将新增 7.1 万亿到 21.2 万亿美元的联邦支出——>> 顺便说，

[82:46]total >> There's no prison. >> of the entirety of the American population, 182 trillion. 全部——>> 没得跑——>> 美国全部人口的总财富是 18.2 万亿。

[82:52]182 trillion. Yeah. Okay, but I'm just saying that's how much they >> the DSA program would cost. If you guys believe that the root of our problems is overspending and basically too much government, unfortunately, >> This is not the >> the DSA is going to >> No, but here's what's going to happen, 18.2 万亿。对。好吧，我只是说这就是 DSA 计划的造价。如果你们认为问题的根源是过度开支、本质是政府管太多，那很不幸——>> 这不是——>> DSA 将会——>> 不，接下来会发生的是：

[83:07]and I'll just I'll let me just walk through the sequence of events. Sequence one, they they enact a policy starting in 2028 with uh comrade AOC in charge and the Politburo uh Ro Khanna, Secretary of Housing and Urban Development, and the rest of the Politburo organizing themselves to 我把事件序列走一遍。序列一：他们从 2028 年开始立法，AOC 同志主政，政治局——Ro Khanna 任住房与城市发展部部长——以及其余政治局成员组织

[83:24]launch these new programs. Those new programs get launched, then costs balloon. The bond market sells off 30 years. 上线这些新项目。项目一上线，成本立刻气球式膨胀。债市抛售 30 年期国债。

[83:32]And so now the 30-year yield spikes 7%, 8%, maybe more. 于是 30 年期收益率飙到 7%、8%，甚至更高。

[83:37]Things become more unaffordable. So then the question is, okay, well, we can't bridge it from a fiscal perspective. We can't just get the Fed to start buying bonds. 东西变得更负担不起。于是问题来了：财政上我们桥不过去，也不能光靠美联储下场买债——

[83:45]There's not enough buying power there to bring rates down, to bring inflation down. So, we need more income. 那边的购买力不够把利率压下来、把通胀压下来。所以我们需要更多收入。

[83:52]So now we need to pass a national you know, asset seizure tax. And there's not enough money in the billionaire class or the 50-millionaire class. The total net worth of all people were that are worth over 50 million dollars is only 23 trillion. 于是现在我们要通过一项全国性的资产没收税。而十亿富豪阶层、五千万富豪阶层的钱根本不够——净资产 5000 万美元以上人群的总净资产只有 23 万亿。

[84:07]You could take 10% of their net worth, and it would only cover 3 months of fiscal spending, today's fiscal spending. 拿走他们净资产的 10%，只够覆盖 3 个月的财政开支——按今天的财政开支算。

[84:13]So at the end of the day, the only way to get there is to start doing massive taxation. Seizures. And so, seizures. 所以归根结底，唯一的出路就是开始大规模征税。没收。没收。

[84:20]Seize. >> So this just came out this week >> Cato >> by >> Cato Institute >> by the Cato Institute. And again, it shows the low end of the DSA spending is 71 trillion. The high end is 211 trillion. 没收。>> 这份报告本周刚出——>> Cato——>> 出自——>> Cato 研究所——>> Cato 研究所。它再次显示 DSA 开支的低端是 7.1 万亿，高端是 21.1 万亿。

[84:34]Now, you compare that to things like corporate profits or private sector wealth. Corporate profits are only 35 trillion. The Forbes 400, all the wealth of the Forbes 400 is only 6.6 trillion. 拿它对比企业利润或私营部门财富：企业利润只有 35 万亿。福布斯 400 的全部财富只有 6.6 万亿。

[84:47]So, it's less than 10% of the low end of DSA spending. >> I I keep telling you, this is not about mathematical realism. This is about anger towards a class of people that are increasingly odious and misperceived. 连 DSA 开支低端的一成都不到。>> 我一直跟你说，这不是数学现实主义的事。这是对一个日益令人反感、又常被误读的阶层的愤怒。

[85:02]>> Yeah. But, people often can't afford to buy a house. People are 35 years old with a roommate. And people and people go to the grocery store and they're shocked at the fact that they have to spend an entire paycheck or a third of a paycheck to buy groceries. >> 是。但是，人们常常买不起房。人们 35 岁了还在跟人合租。人们走进超市，被震惊到：买趟菜要花掉整整一张工资单、或三分之张工资单。

[85:15]>> It's not It's not that It's not that people want to buy a yacht. It's that the yacht class are just disproportionately >> no. I mean, look. In every socialist revolution, every revolution we've seen historically around this, there is always an elite branch of society and they're typically classified >> 不是——不是说人们想买游艇。而是游艇阶级实在太——>> 不。听着，历史上每一场社会主义革命、每一场这样的革命，社会中总有一个精英分支，被以各种方式

[85:33]different ways. It's It's wealth. It's Jews. It's some race. It's immigrants. There's some There's some organizational assignment, some identity assignment that says, "This is the cause of unaffordability." At the end of the day, the dysfunction of the relationship between government and all the other 归类贴标。是财富，是犹太人，是某个种族，是移民。总有某种组织性标签、身份标签，写着「这就是负担不起的根源」。归根到底，是政府与社会中其他所有

[85:49]organizations that make up that society. Those organizations can be states. They can be private companies. They can be the NGOs. It is the dysfunction of the government's relationship with the other organizations that fundamentally causes things to become more expensive and 组织之间关系的功能失调——那些组织可以是州、私企、NGO。正是政府与其他组织关系的功能失调，才从根本上让东西变得又贵又

[86:05]dysfunctional. And the actual market for getting access to things for the individual fails. And at that point, you have to identify. And you don't identify the government cuz the government is the aggregate of all the people. 失灵。个人获取生活所需的真实市场失效了。到了那个地步，就必须找出一个对象。而你不能指认政府，因为政府是全体人民的集合。

[86:16]>> No, of course. >> So, So, the the gov- And then And then And remember, the government is system, the means of action. So, for the people to organize, they have to organize the government to go and then they have to have an alternative system to go and destroy. >> 那当然。>> 所以——记住，政府是系统、是行动的手段。所以人民要组织起来，就必须组织政府出手，然后还必须有一个可供摧毁的替代系统。

[86:30]>> You have to point the finger and you have to point the finger at somebody and if you're going to scapegoat somebody in 2026, the best class of people to scapegoat are the billionaires and frankly even better than them, the trillionaires. >> 必须有人被指着鼻子，必须在 2026 年找一只替罪羊的话，最好的替罪羊阶层就是十亿富豪，而且说实话，比他们更好的是万亿富豪。

[86:41]>> At the end of the day, it's only tech, right? The data center is the temple where the billionaires and trillionaires pray. It's where they go to make their sacrifices. It is their gathering spot. >> 归根结底，矛头只会指向科技，对吧？数据中心就是十亿富豪和万亿富豪祈祷的神庙，是他们献祭的地方，是他们的聚会场所。

[86:50]So, I think the data center is this like, you know, temple on the mount. And so, that's why I think the data center gets tied up in all of this. There's no rational discourse about data centers. 所以我认为数据中心就是那座山上的神庙。这就是为什么数据中心会被卷进这一切。关于数据中心已经不存在理性讨论了。

[86:59]It's like >> No, no, no. Look, you're if you take it through the Rene Girard lens, everything maps. It's like the billionaires will be scapegoated and the trillionaires will be I mean, even worse. And because that will allow the catharsis that will allow society to reset and realize, wait a 就像——>> 不不不。你要是用 René Girard 那副透镜看，一切都对上了：十亿富豪会被献祭替罪，万亿富豪更是——下场更惨。而那样带来的宣泄会让社会重置，然后才会意识到：等等，

[87:15]minute. Oh my gosh. >> Yeah, there's a generation ahead of that though. There's a generation ahead. >> And only then it is a long process. It is a long >> Here's an idea. Next president or Trump in his second half, just putting it out there. 我的天。>> 是，不过在那之前还有一代人的距离。还有一代人。>> 而且只有到那之后——那是个漫长的过程。很漫长——>> 我有个主意。下一任总统，或者 Trump 的后半场，我就这么一说：

[87:27]Do an executive order that you're going to build like a million homes, 2 million homes. If you can have executive orders, if you can start wars without >> You know that's not how it works, right? 签个行政令，建它个一两百万套住房。既然行政令能签、仗都能不经国会就开打——>> 你知道事情不是那么运作的吧？

[87:36]Go. >> What's that? >> You know that that's not how it works, right? There are local laws, there are state laws. 说吧。>> 什么？>> 你知道事情不是那么运作的吧？还有地方法律、州法律。

[87:43]>> Of course, but he pulled out all the stops for getting data centers greenlit. >> 当然，但为了给数据中心开绿灯，他可是把吃奶的劲儿都使出来了。

[87:48]We know that. And he worked with the state governments. I'm talking about him coming on saying, you know what, we did a great job for the data centers, we did a great job for the crypto industry and now we're going to do two great things. 我们都知道。而且他跟州政府协作了。我说的是他站出来说：你看，我们为数据中心干了件漂亮事，我们为加密行业干了件漂亮事，现在我们要干两件漂亮事。

[87:58]We're going to do two great things. We're going to do two great things. We're going to make trade shows, we're going to make trade schools and we're going to make affordable homes. I'm just going to make that the next 24 months. 我们要干两件漂亮事。两件漂亮事：建职业学校，建负担得起的住房。就把这个当成接下来 24 个月。

[88:07]We're going to figure out a way to mobilize the private sector, mobilize the public sector. We're going to make more homes and we're going to get you a great education that turns into a job. 我们要想办法动员私营部门、动员公共部门，多建房子，给你一个能变成好工作的好教育。

[88:16]That would be a great thing for Trump to do. >> federal The federal government does control the largest amount of land in the United States. So, it could do something on federal land without the need for a bunch of other state and local gobbledygook. So, there is >> Fannie Mae, Freddie Mac. 这对 Trump 来说会是件了不起的事。>> 联邦政府确实掌握着美国最大量的土地。所以它可以在联邦土地上做文章，不需要一堆州和地方的繁文缛节。所以是有的——>> 还有房利美、房地美。

[88:29]>> Now, there it may not be located in places that actually make sense for commerce and for people to live or that are already living there cuz I don't think it really abuts you know, uh most cities. So, unfortunately, it's not in his control. >> 不过那些土地可能不在适合商业、适合人居住、或已经有人住的位置，因为我不觉得它们挨着大多数城市。所以很不幸，这不在他的掌控之内。

[88:44]Even if he wanted to just kind of say, "Let's make this happen." >> I think if the president can go spend a couple hundred billion on this war, he can go spend a hundred couple billion on homes that are affordable. Uh and he maybe can't put them in Manhattan, but he can put them other places and he should fight for the regulations to get 就算他想说「就把这事办成」，也办不成。>> 我认为，总统既然能为这场战争花几千亿，他就能为负担得起的住房花上千几百亿。也许塞不进曼哈顿，但可以放在别的地方，而且他应该为法规去打这场仗，

[89:01]Just like he fought for the regulations on data centers and energy. He made an a point on this very podcast at the AI summit, "Hey, we're going to clear all the pathways to nuclear, to nat gas, we're going to >> They have. 就像他为数据中心和能源打法规之战一样。他在这个播客的 AI 峰会上说过：「我们要扫清通往核能、天然气的所有障碍，我们要——>> 他们做到了。

[89:13]>> beautiful clean coal which is really And I give him credit for that. Now, [clears throat] do it. Now, do it for homes. >> 还有美丽清洁的煤，这个真的——这个我给他记功。现在【清嗓子】，接着干。现在为住房干一次。

[89:19]Let me see you do it for homes, Trump. Use that same energy. That's what people want to see. >> They just actually just passed a housing bill. 让我看看你为住房干一次，Trump。拿出同样的劲头。这才是人民想看到的。>> 他们其实刚刚通过了一个住房法案。

[89:26]>> When's the last time we heard >> Unfortunately, it was written by us with Warren, so I don't know how good it is, but >> Yeah. >> 我们上一次听到——>> 不幸的是，那是跟我们和 Warren 一起写的，所以我不知道质量如何，但是——>> 是。

[89:32]Anyway, I'm just giving him the solution to save his legacy. All right, everybody. >> As always, you're looking out for everybody. 总之，我只是在给他留拯救政治遗产的药方。好了各位。>> 一如既往，你在替所有人操心。

[89:37]>> You're looking out for everybody. You're looking out for the audience. for the audience. >> Nostradamus, thank you for watching. >> Nostradamus is also looking out for you. >> 你替所有人操心，替听众操心。>> 诺查丹玛斯，感谢观看。>> 诺查丹玛斯也在替你操心。

[89:44]>> Okay, rap. I got to go. Come on. >> All right, everybody. It's been another amazing episode >> boys. >> of the All-In podcast. >> 好了，收工。我得走了。来吧。>> 好了各位，这又是精彩的一集——>> 兄弟们——>> All-In 播客。

[89:50]>> Good to have the quartet back. Let's go. >> [music] >> [music] >> You let your winners ride. >> 核心四人组回归真好。走起。>>【音乐】>>【音乐】>> 让你的赢家继续跑。

[90:03]>> And I said we [music] open sourced it to the fans and they've just gone crazy with it. >> Love you as I am. >> Queen of quinoa. 「我说过我们【音乐】把它开源给了粉丝，然后他们彻底玩疯了。」>> 就爱这样的我。>> 藜麦女王。

[90:09]>> I'm going all in. >> Besties are back. >> 我全押了。>> 兄弟们回来了。

[90:16]>> I'm going all in. >> That's my dog taking a piss in your >> We should all just get a room and just have a one big huge orgy because they're all just useless. It's like this like sexual [music] tension that they just >> 我全押了。>> 那是我的狗在你家——撒尿。>> 咱们干脆开个房，来一场大乱斗算了，因为这帮人全都废物。就像这种性【音乐】张力，总得

[90:32]need to release somehow. >> What you say? >> What about your feet? >> What you say? >> What about your feet? 找个方式释放。>> 你说什么？>> 你的脚怎么了？>> 你说什么？>> 你的脚怎么了？

[90:38]>> What? >> We need to get merch. >> Besties are back. >> 什么？>> 得搞点周边。>> 兄弟们回来了。