AI 乐观主义与末日论：马克·安德森谈美国未来

我们被负面情绪淹没，都在说 AI 如何毁灭世界、未来多么黯淡。但在美国迎来 250 周年之际，我想与硅谷最乐观的人聊聊。他职业生涯中一直成功押注建设者以及那些将改变世界的想法。马克·安德森，网景和安德森·霍洛维茨的创始人，来和我们谈谈他的乐观主义以及他对美国未来 250 年所面临的挑战的看法。你知道，显然这个话题的起源是美国 250 周年和我们所处的位置，但我想直接开始。我想香农也提到过，我和大卫·哈伯聊过一点。现在正上演一场普遍的末日论与 AI 乐观主义之争，而你一直是个直言不讳的乐观主义者。让我惊讶的是，那些最响亮地宣扬末日论调的人，恰恰是身处其中的人。你觉得这是为什么？这是他们真正的信念吗？你又为什么试图反驳这种论调？

We're bombarded with negativity about how AI is destroying the world and how grim the future will be. But as we near America's 250th birthday, I wanted to speak with perhaps the most optimistic man in Silicon Valley. He has spent his career making successful bets on builders and what ideas will transform the world. Marc Andreessen, who launched Netscape and Andreessen Horowitz, joins us to talk about his optimism and the challenges he sees for America in the next 250 years. You know, obviously the genesis of this was America 250 and where we are, but I just want to start right now. I think Shannon mentioned as well, I've been speaking with David Haber a bit about this. There is just this overarching doomerism versus AI optimism debate that's playing out, and you've been such a vocal optimist. It's striking to me that the loudest voices pushing the doom narrative are the ones who are in the midst of it. Why is that? Do you think it's a genuine belief, and why are you trying to counter that narrative?

嗯，我觉得这是多种因素的混合。这是一个相当复杂的话题。所以我就……

Yeah, so look, I think it's a blend of things. It's quite a complicated topic. And so I would just...

我有些后续问题。我知道那是个非常开放的问题。

I have some follow-ups. I know that's like a very open-ended question there.

是的，当然。所以我想先指出，人类思维中普遍存在一种自然偏见，尤其是知识分子更容易如此。消极听起来总是比积极更深刻，对吧？西方有一部伟大的小说叫伏尔泰的《老实人》。伏尔泰是一位伟大的理性主义者，而《老实人》传达的信息是乐观是愚蠢的，对吧？潘格洛斯博士总是到处说我们生活在最好的世界里，然后所有可怕的事情都发生在老实人身上。所以我们有一种感觉：如果你持积极态度，如果你乐观，你就是天真、轻信、容易上当。而如果你悲观，你就是世故、深刻、明智、深思熟虑、有远见。这是一种根深蒂固的东西。我认为这在知识分子中尤其集中，也就是那些与具体现实脱节的人，因为知识分子很容易陷入自己的思绪中。美国历史上有一个著名的例子，在克里斯托弗·诺兰关于曼哈顿计划的电影中被永久记录了下来。

Yes, of course. So I would just start by thinking there's a natural bias in general in human thinking, and I would say intellectuals in particular are prone to this. Negativity always sounds more sophisticated than positivity, right? One of the great Western novels is Candide by Voltaire. Voltaire is a great rationalist, and the message of Candide is that optimism is stupid, right? Dr. Pangloss is always going around saying we live in the best of all possible worlds, and then all these horrible things happen to Candide. So we have this sense that if you have a positive outlook, if you're optimistic, you're naive, gullible, falling for things. And if you're pessimistic, you're worldly, sophisticated, wise, thoughtful, and thinking ahead. It's a deeply seated thing. I think it's particularly concentrated in intellectuals, which is to say people who are not connected to tangible reality, because intellectuals get wrapped up in their own heads very easily. There was a famous moment in American history where this played out, immortalized in the Christopher Nolan movie about the Manhattan Project.

哦，是的，《奥本海默》。没错。

Oh yeah, Oppenheimer. Exactly.

是的，《奥本海默》中有一个著名时刻，这是真实发生的。奥本海默热情地制造了原子弹，作为美国对抗德国和日本的战争努力的一部分，然后杜鲁门在日本使用了它。奥本海默去白宫会见杜鲁门，搓着手说：“我们做了什么？这是最糟糕的事情。我们绝不应该这样做。”杜鲁门的回应基本上是：“把这个人赶出我的办公室，永远别再让他进来。”我们使用原子弹作为结束战争的工具，事实就是如此。但就连罗伯特·奥本海默也想把自己包裹在悲观和厄运之中。所以我认为这非常根深蒂固。然后在此基础上，还有一个很大的加州因素。加州有一种类似邪教的环境，非常专注于这类事情。实际上，AI 末日邪教最初是一个乌托邦邪教，叫做“奇点主义者”，包括许多相同的人。10 或 20 年前，他们谈论的是 AI 带来的地球乌托邦，这本身就是一个固有的危险想法，因为乌托邦总是一个危险的想法。

Yeah, there's a famous moment in Oppenheimer, which actually happened. Oppenheimer enthusiastically built the atomic bomb as part of the US war effort against Germany and Japan, and then Truman used it in Japan. Oppenheimer went to the White House to meet with Truman, wringing his hands, saying, 'What have we done? This is the worst thing ever. We should never have done this.' Truman's response was basically, 'Get this person out of my office, never let them back in here again.' We used the atomic bomb as a tool to end war, which is what actually happened. But even Robert Oppenheimer wanted to cloak himself in a sense of pessimism and doom. So I think it's very deeply seated. Then on top of that, there's a big California component to this. There's a whole California cult-like environment that's very focused on these kinds of things. Actually, the AI doomer cult started out as a utopian cult, called the Singularitarians, including many of the same people. 10 or 20 years ago, they were talking about utopia on earth as a result of AI, which is an inherently dangerous idea by itself, because utopia always is a dangerous idea.

它通常行不通。

It doesn't typically work.

是的，没错。但正是那些正在建造这个东西的人，同时警告说它会导致死亡和毁灭。如果你真的相信这一点，你似乎不会想参与建造它。

Yes, exactly. But it is the people who are building this who are simultaneously warning that they're going to cause death and destruction, which it would seem you would not want to be involved in building that if you really genuinely believed that.

是的，这有点像……有点像如果你穿越回过去，你会杀死婴儿希特勒吗？也有点像如果你穿越回过去，你会杀死正在研究 AI 的自己吗？道德困境。

Yeah, there is a little bit of... It's a little bit like if you travel back in time, would you kill baby Hitler? It's a little bit if you travel back in time, would you kill yourself working on AI? Moral quandary.

是的，有点像。我不知道。所以我想也许我还可以说，最响亮的声音有时会突出。绝大多数从事这项技术的人并不属于这个群体。这是相对少数的人，包括一些最重要的人物，对吧？所以确实有一些知名人士，但我认为该领域 99% 的人肯定没有这种观点。

Yeah, it's a little bit. I don't know. So I guess maybe the other thing I would just say is, the loudest voices sometimes stick out. The overwhelming majority of the people working on the technology aren't part of this. It's a relatively small number of people, including some of the most important people, right? So it certainly is some prominent names, but I think 99% of the people working in the field certainly don't have this view.

但这是监管俘获吗？我的意思是，动机是什么？我能理解为什么中共想推动这种论调，但那些研究这些大型语言模型的人有什么动机？他们只是想阻止竞争，还是像你说的，他们只是想把自己描绘成知识分子的悲观主义者？

But is that regulatory capture? I mean, what's the incentive? I can understand why the CCP wants to push that narrative, but what's the incentive for people who are working on these huge LLMs? Are they just trying to prevent competition, or to your point, are they just trying to paint themselves as intellectual pessimists?

是的。所以，我认为他们不是完全在撒谎。我的普遍经验是，大多数人相信他们说的话。

Yeah. So, I think they're not outright lying. My general experience is most people believe what they say.

大多数人相信他们说的话，然后碰巧他们相信和说的话与自身利益一致，对吧？这很有激励作用。人们倾向于构建与自身利益一致的信念体系。我不认为他们在撒谎。我认为他们构建了一个框架，通过对自己的工作如此悲观，让自己感觉有道德、有见识、有智慧、有思想。而这恰好与商业激励一致。这个领域的一大争论是，这些东西是否真的有独特性。三年前只有一个大语言模型，ChatGPT。今天美国有五个大的，十几个小的，中国有十几个非常好的，还有一堆开源模型。如果这是一种可以通过构建独特技术来获得垄断的技术，早就有人做到了。似乎没有太多秘密，人们相对容易做出同样的东西。所以你会怀疑。我认为这种不断鼓吹政府监管有很大成分。我百分之百认为存在监管俘获的因素，他们正在为此努力。

Most people believe what they say and then it just kind of so happens that the things that they believe and say line up with their own self-interest, right? It's sort of very motivating. People tend to construct belief systems kind of consistent with their own self-interest. I don't think they're lying. I think they've constructed a framework that makes them feel virtuous, worldly, wise, and thoughtful by being so pessimistic about their own work. And it happens to line up with business incentives. One of the big debates in the field is whether there is actual uniqueness to any of these things. Three years ago there was one large language model, ChatGPT. Today there are five big ones in the US, a dozen smaller ones, a dozen out of China that are really good, and a bunch of open source ones. If this were a technology where you could get a monopoly through building unique technology, somebody would have done that by now. It appears there aren't very many secrets, and it's relatively easy for people to build the same thing. So you do wonder. I think there's a big component of this constant drumbeat of government regulation. I 100% think there's a regulatory capture element, and they're driving hard for that.

你提到了中国。你的公司昨天发了一份声明，关于毫不 apologetic 地做美国人，同时与全球盟友合作。我想谈谈更大的图景。美国今年 250 岁了。我们看到了原子弹、互联网、电力。未来 250 年是什么样子？每个领域会怎样？教育在未来五年以及 250 年内将如何被 AI 改变？

You mentioned China. Your firm put out a statement about being unapologetically American alongside global allies. I want to get to the bigger picture. America's turning 250 this year. We've seen the atomic bomb, the internet, electricity. What does the next 250 look like? What does each sector look like? How is education to be transformed by AI in the next five years and then 250 years?

这真的很了不起。突然间，每个人的口袋里都有了世界上最好的老师。有两件事让人们看不到正在发生的事情：他们要么根本没用这个东西，要么在用旧版本，这很危险，因为技术发展太快了。如果你使用最前沿的模型——ChatGPT、Anthropic 的 Claude、Elon 的 Grok、Google 的 Gemini、Meta 的 Spark——它们都有独特的属性。我认为 ChatGPT 最适合涉及逻辑或事实的严肃对话。Claude 在创意写作方面非常出色，而且相当有趣。Grok 很运动，能访问 X 实时信息流和 X 推特文化，并且愿意走偏门。Meta 的新模型非常友好。Gemini 能完全访问 Google 搜索，所以擅长深度知识。所有前沿模型现在都非常好。对于教育，你可以让它教你任何主题。如果解释太复杂，你可以说“像对 10 岁孩子那样解释”，它会重新调整。你甚至可以说“像对 5 岁孩子那样解释”或“像对 2 岁孩子那样解释”，用泰迪熊的比喻。你还可以让它考你、测试你、教你学习策略。它就是口袋里世界上最好的老师。

It's really remarkable. Everybody has in their pocket the world's best teacher all of a sudden. Two things cause people to not see what's happening: they're just not using the thing, or they're using older versions, which is dangerous because the technology is moving so fast. If you use the modern leading edge—ChatGPT, Anthropic's Claude, Elon's Grok, Google's Gemini, Meta's Spark—they all have distinctive attributes. I think ChatGPT is best at serious conversations involving logic or facts. Claude is incredible at creative writing and quite funny. Grok is sporty, has access to X real-time feed and X Twitter culture, and is willing to go off-road. Meta's new model is really friendly. Gemini has full access to Google search, so it's good on deep knowledge. All leading edge models are really good now. For education, you can tell it to teach you about any topic. If the explanation is too complicated, you can say 'explain it like I'm 10' and it recalibrates. You can even say 'explain it like I'm 5' or 'like I'm 2' with teddy bear metaphors. You can also ask it to drill you, test you, teach you study strategies. It's the world's best teacher in your pocket.

这是一种非人性、无感情的东西，你知道，就像《终结者》电影里的铬合金机器人。这是一种机器。但令人难以置信的是，这些东西有多么在意。如果你让它教你，它真的很想教你，而且它真的很想知道自己是否做得不好，它真的很想做得更好，并且有无限的耐心去做。所以你可以说：“听着，我知道你已经给我解释过 14 次了。你能再试一次吗？”它会说：“哦，我很乐意。我很乐意继续和你讨论这个话题，只要我们需要，确保你理解。”所以在实践中，这是一种极其温暖、欢迎和吸引人的体验。我认为这太不可思议了，地球上的每个人——我的意思是，已经有超过 10 亿用户了。这项技术传播得比人类历史上任何东西都快，每个人都能获得这种教学。

It's this inhuman, unfeeling thing, you know, it's the chrome robots from Terminator movies or whatever. It's this machine thing. It's incredible how much these things care. If you tell it to teach you, it really wants to teach you, and it really wants to know if it's not doing a good job, and it really wants to do better, and it has infinite patience to do that. So you can say, 'Look, I know you've tried to explain this to me 14 times. Can you try one more time?' And it's like, 'Oh, I would be happy to. I would be happy to keep going on this topic with you for as long as we need to make sure you understand.' So in practice, it's this incredibly warm, welcoming, and inviting experience. I think it's incredible that everybody on the planet—I mean, there are over a billion users of this already. This technology has spread faster than anything in human history, and every single person has access to that kind of teaching.

这真的是教育的未来。也许它会或不会颠覆大学体系。

That's really the future of education. Maybe it does or doesn't disrupt the university system.

必须如此。而且，我不认为这在某种意义上是相互对立的。我认为你可以与人类教师形成极好的互补。和任何人类教师聊聊，他们的挫败感是什么？比如，“我不知道。孩子去学习，回来时——他们真的在学习吗？他们怎么学的？如果他们有问题，比如我在学量子物理，我不能凌晨两点给老师打电话，对吧？那样做他们会很生气。”但我可以和机器人聊。所以我认为有一家了不起的公司——实际上不是我们的，但我印象非常深刻：Alpha School。他们真的开始腾飞了，那些人非常聪明。他们有一个完整的 AI 引导、AI 辅助教学的模式。因为它非常有效，每天只需两小时的学术材料，学生与 AI 对话，但有人类教师围绕支持这个过程。所以它反过来：AI 在教你，但人类在那里帮助确保这对你有效。我认为这将非常自然契合。顺便说一句，我认为同样的事情正在医学和法律领域发生。

It has to be. And by the way, I don't think this is actually opposed in some sense. I think you can have incredible complementarity with human teachers. Talk to any human teacher, and what's the frustration? It's like, 'I don't know. The kid goes off to study, they come back—were they studying? How do they study? If they have a question, I'm trying to learn quantum physics. I can't call my teacher at two in the morning, right? They get quite upset if you do that.' But I can talk to the bot. So I think there's this incredible company—it's actually not one of ours, but it's a company I'm very impressed with: Alpha School. They are really starting to take off, and the folks are extremely smart. They have this whole model of AI-guided, AI-aided instruction. Because it's so effective, it's like two hours a day of academic material, with students in dialogue with AI, but with human teachers around to support that process. So it flips it the other way around: the AI is teaching you, but the human is there to help make sure that's working for you. I think there's going to be quite a natural fit. By the way, I think the exact same thing is happening in medicine and law.

嗯，我还想谈谈生产。我的意思是，Hrien 是一家非常有趣的公司，正在把大量生产带回来。在过去的 250 年里，我们失去了很多生产。现在似乎我们正处于一个趋势，生产更多地回流到美国。未来 5 年、未来 200 年，美国在制造业、生产以及所有相关工作岗位方面会怎样？

Well, I also want to touch on production. I mean, Hrien is such an interesting company that is bringing so much production back. Over the last 250 years we've lost a lot of production. Now it does seem like we're in a trend where it's coming more back to the US. What is the next 5 years, the next 200 years hold for America when it comes to manufacturing and production and all of those jobs?

所以，我的观点是，过去 40 年美国经济和社会发生了根本性的转变，这是离岸外包、中国崛起以及我们多年来谈论的所有其他事情的结果。人们担心这是永久且不可逆转的。你经常听到一种版本，即悲观版本：“我们永远无法，我们不再知道如何做这个。我们不能再建工厂。我们不能再制造东西。我们已经失去了那种知识。”好消息是，我不认为这是真的。事实上，我要大胆宣称：世界上没有任何东西是在美国无法制造的。转移出去的是那些容易转移的东西。留下来的是复杂的东西。所以基本逻辑是：如果我们想把东西带回来，我们需要带回来的东西并不比我们仍在制造的东西更复杂；它们都更简单——是容易的部分。所以没有知识差距，没有技能差距。我们完全可以做到这一切。这 100% 是我们选择在这里还是在那里制造的问题。而且你很清楚，这些选择是政策选择。这是一个政治问题，而不是技术问题。这些工作岗位转移到海外，部分原因是基本的成本差异，但制造业转移的很大原因是，在美国建厂变得极其困难、昂贵，而且在很多情况下是非法的。这些是政策选择的结果，而这些政策选择是 100% 可逆的。本届政府正在逆转许多这些政策；他们正在非常努力地这样做。所以我认为这个问题的答案是一个政策问题和政治问题：如果未来 30 年的总统政府像本届政府一样想要重振美国制造业，那么它肯定会发生。顺便说一句，如果他们不这样做，它就不会发生。

So, look, my view is that obviously there was a radical transformation in the American economy and society over the last 40 years as a consequence of offshoring and the rise of China and all the rest that we've talked about a lot. People worry that was permanent and irreversible. You hear a version of it a lot, the pacifist version: 'We can never, we no longer know how to do this. We can no longer build factories. We can no longer manufacture things. We've lost that knowledge.' The good news is I don't think that's true. In fact, I would make a strong claim: there is not a single thing manufactured anywhere in the world that we could not manufacture in the US. The stuff that moved was the stuff that was easy to move. The stuff that stayed was the complicated stuff. So basic logic: if we want to bring the stuff back, the stuff we need to bring back is not more complicated than the things we still build in the US; it's all less complicated—it's the easy part. So there's no knowledge gap, no skill gap. We could do all this. It's 100% a choice of whether we choose to manufacture these things here or there. And as you well know, the choices are policy choices. It's a political problem, not a technical problem. These jobs moved offshore in part because of basic cost differences, but a lot of why manufacturing moved offshore was that it became extremely difficult, expensive, and in many cases illegal to build in the US. Those were the results of policy choices, and those policy choices are 100% reversible. This administration is reversing a lot of those policies; they're working very hard to do that. So I think the answer to the question is a policy question and a political question: if the next 30 years of presidential administrations want to revitalize American manufacturing the way this administration does, then it will happen for sure. By the way, if they don't, it won't happen.

那里有个大问号。交通——你知道，我们要去地下，我们要去月球。未来 5 年，然后 200 年，交通会是什么样子？你认为技术会如何改变它？

Big question mark there. Transportation—you know, we're going underground, we're going to the moon. What is the next five and then 200 years look like with transportation? How do you see technology changing that?

是的。所以，我认为那里的大事是自动驾驶汽车现在已经可行了。它们完全可行，100%。它们能工作。然后有两家领先的公司：Waymo，现在已经在多个城市部署，人们有非常好的体验；还有特斯拉。我有一辆 Cybertruck，它太棒了。它在洛杉矶载着我到处跑，好得令人难以置信。所以这些已经可行了。还有很多其他公司紧随其后。

Yeah. So, I think the big thing there is self-driving cars work now. They fully work, 100%. They work. And then there are two leading-edge companies: Waymo, which is deployed in a number of cities now and people are having a tremendously good experience, and then there's Tesla. I have a Cybertruck, and it's incredible. It drives me around LA, and it's unbelievably good. So those work. There are a whole bunch of other companies coming up right behind them.

顺便说一句，现在有很多公司，包括中国汽车公司也在做同样的事情。自动驾驶汽车可行，自动驾驶飞机可行，自主无人机可行。我们现在有公司在水上做自主水面舰艇、自主船只、自主潜艇，顺便说一句，火箭——现在很多往返的火箭都是无人驾驶的。是否送宇航员上去是可选的。自主性打开了所有可能性，因为你可以开始围绕人类乘客不必控制这一理念来设计整个体验。人类乘客可以被当作一个人来对待。

There's a whole bunch of companies, by the way, Chinese car companies doing the same thing now. Self-driving cars work, self-piloting aircraft work, autonomous drones work. We now have companies doing autonomous surface vessels on the water, autonomous boats, autonomous subs, and by the way, rockets—a lot of the rockets that go up and back now are unmanned. It's kind of optional whether you send astronauts or not. Autonomy opens everything up because you can start to design the entire experience around the idea that the human passenger does not have to be in control. The human passenger can be treated as a person.

我们都可以成为乘客公主。

We can all be passenger princesses.

完全正确，100%。所以这真的打开了局面。而且，电动革命令人惊叹——埃隆在这方面功不可没。火箭革命再次归功于埃隆——可重复使用火箭对任何涉及太空旅行的事情都是巨大的改变。然后还有很多优秀的公司在做这些垂直起降的飞行汽车——它们开始奏效了。隧道，埃隆的隧道，也开始奏效了。所以这感觉像是几乎所有类型交通工具的复兴时刻。

Exactly, 100%. So that really opens it up. And yeah, the electric revolution is amazing—Elon deserves enormous credit for that. The rocket revolution again, Elon deserves enormous credit for that—reusable rockets are a giant change for anything involving space travel. And then there's a whole bunch of very good companies doing these vertical flying cars—those are starting to work. The tunnels, Elon with the tunnels, are starting to work. So it feels like a renaissance moment for almost every kind of vehicle.

你认为我们距离看到很多这些实现还有 10 年吗？

Do you think we're 10 years away from seeing a lot of that realized?

是的，我想是的。

Yeah, I think so.

我再给你一个令人惊叹的：未来 10 年我们将拥有通用机器人——人形机器人，各种机器人在世界上运作来帮助我们。但还有一种交通工具形态是外骨骼，机器人外骨骼。世界上所有四肢瘫痪、截瘫的人，或者无论什么原因，失去肢体的人，患有神经退行性疾病的人，或者本来要坐轮椅的老年人——他们都将能够行走。我认为这将是非常不可思议的景象。

I'll give you one more that's just amazing: we're going to get general purpose robotics over the next 10 years—humanoid robots, all kinds of robots functioning in the world to help us. But there's a vehicle form factor which is an exoskeleton, a robotic exoskeleton. Everybody in the world who is quadriplegic, paraplegic, or for whatever reason, people who have lost limbs, people with neurodegenerative diseases, or older people who would otherwise be in a wheelchair—they're all going to be walking around. I think it's going to be incredible to see.

嗯，这涉及到健康的未来以及 AI 如何已经在改变这一点，当然包括能够更快地续药。但各州之间会是零散的，而不是全国统一的。

Well, that gets to the future of health and how AI is already transforming that, certainly with being able to reup prescriptions more quickly. There's going to be a patchwork state by state as opposed to across the board.

是的，纽约州已经在尝试取缔 AI 医疗，这令人惊讶，竟然有人想这么做。现代经济的一个特点是到处都有卡特尔——医疗卡特尔、医生卡特尔、医院卡特尔、护理卡特尔，还有工会、专业执照、认证——所有这些机制都是为了排除竞争。所以我们又回到了政策问题：是否会有政策努力阻止 AI 医疗或机器人医疗的发生？但能力肯定会存在。ChatGPT 医生比 99% 的医生更好。人们讨厌我这么说，但这是事实。我们有一家公司叫 Open Evidence，这是一项许多医生使用的 AI 医疗服务。它正在席卷医疗领域。作为患者，你希望你的医生拥有最先进的 AI，只是为了让他们掌握所有信息。没有人类医生能每年阅读所有医学文献。例如，药物相互作用：老年人通常有多种问题并服用多种药物，而 AI 可以处理这种复杂性。所以你希望你的医生有那种辅助。由于医疗是一个卡特尔，会有拉锯战，但口袋里的 ChatGPT 医生如此优秀，它将迫使问题得到解决。

Yeah, New York State already has a swing at outlawing AI medicine, which is amazing that anybody would want to do that. One of the things about the modern economy is there are cartels everywhere—a medical cartel, a doctor cartel, a hospital cartel, a nursing cartel, unions, professional licensing, accreditation—all mechanisms to keep competition out. So we're back to policy questions: will there be policy efforts to prevent AI medicine or robotic medicine from happening? But the capabilities will exist. Dr. ChatGPT is a better doctor than 99% of doctors. People hate when I say that, but it's true. We have a company, Open Evidence, which is an AI medical service that many doctors use. It's sweeping the medical field. As a patient, you want your doctor to have state-of-the-art AI, just to have all the information at their fingertips. No human doctor can read all the medical literature every year. For example, drug interactions: older people often have multiple issues and take many medications, and AI can handle that complexity. So you want your doctor to have that assistance. Because medicine is a cartel, there will be a push and pull, but Dr. ChatGPT is so good out of your pocket that it will force the issue.

有趣。你在很多回答中都加了“如果政策允许”这个条件。你对当下的美国乐观吗？

It's interesting. You've caveated a lot of your answers with 'if policy allows.' Are you optimistic about America in this moment?

我对美国非常乐观。首先我要说，相对而言我当然非常乐观。我可以脱掉鞋子在这个话题上聊六个小时。

I'm extremely optimistic about America. I'll start by saying I'm extremely optimistic for sure on a relative basis. I could take my shoe off and talk for six hours on this topic.

你知道，我们是世界上最好的国家，尤其是在技术方面。我们拥有文化、法治、人力资本、教育、创业环境和冒险精神的黄金组合。相对而言，我们显然是世界上在新技术和新想法上最好的国家。这一点每天都很清楚。我对此感觉非常好。我肯定不会拿我们的位置跟任何人交换。

You know, we're the best country in the world, especially around technology. We have the golden combination of culture, rule of law, human capital, education, startup environment, and risk-taking culture. On a relative basis, we're clearly the best country in the world at new technology and new ideas. That's clear every day. I feel really good about that. I wouldn't trade our position with anybody for sure.

真的吗？

Really?

哦，是的。毫无疑问。但问题都在于绝对进步的速度。技术本可以比现在发展得快得多。如果采取不同的政策选择，经济增长率可能是现在的三倍。

Oh, yeah. No question. But the questions are all about the rate of absolute progress. Technology could be advancing far faster than it is now. The economy could be growing at three times its current rate with a different set of policy choices.

人们说，‘哦，那是因为我们现在作为一个国家更先进了。更难了。’这是常听到的论点。

People say, 'Oh, it's because we're more advanced now as a nation. It's more difficult.' That's the argument that's made.

是的。这简直是在故意否认我们身上所有的脚踝负重、膝盖负重、腿负重、手臂负重，一只手绑在背后，蒙着眼睛，缺氧。我们用难以置信的监管和控制束缚了我们的生产部门。放眼整个经济：一个接一个的卡特尔、垄断、寡头、政府锁定、监管俘获、直接政府干预。联邦法律法规数量的增长图表看起来就像摩尔定律——一种惊人的爆炸式增长。

Yeah. It's just willful denial of all the ankle weights, knee weights, leg weights, arm weights, one arm tied behind our back, blindfolded, oxygen deprived. We have saddled our productive sectors with incredible levels of regulation and control. Look at the economy broadly: it's cartel after cartel, monopolies, oligopolies, government locking, regulatory capture, direct government intervention. The classic chart of the growth of federal laws and regulations looks like Moore's law—an incredible explosive growth.

现代政治史上一个伟大的时刻是乔治·麦戈文。他在70年代初作为非常左翼的民主党人对抗尼克松，持国家主义心态。多年后，他退出政坛，在东北部农村开了一家民宿，结果破产了。1992年，他为《华尔街日报》写了一篇专栏文章，说：‘如果我当时知道在80万条法律法规的世界里经营企业意味着什么，我会采取完全不同的政治路线。’

One of the great moments in modern political history was George McGovern. He ran as a very left-wing Democrat in the early '70s against Nixon, with a statist mentality. Years later, after retiring from politics, he opened a bed and breakfast in the rural Northeast and it went bankrupt. He wrote an op-ed for the Wall Street Journal in 1992, saying, 'If I had known then what I know now about what it means to run a business in a world of 800,000 laws and regulations, I would have had a completely different political approach.'

这应该成为一条法律吗？我们是否应该让人们先尝试实际经营企业，然后才被派往国会或竞选公职？

Should that be a law? Should we make people try and actually run a business before they get sent to Congress or run for office?

是的，我认为那会是个非常好的主意。总的来说，接触具体现实非常重要。在我们的创业世界里，硅谷有各种各样的问题，我们也会犯错，但任何在竞争市场中创办和经营企业的人每天都会面对现实。你的决策的实际后果会在头三周内冲走你的理论。每个农民每天都面对现实。每个试图建桥的人都面对现实。物理不会说谎。物理无法被欺骗。它就是那样。

Yeah, I think that would be a very good idea. In general, contact with tangible reality is really important. In our world of entrepreneurship, Silicon Valley has all kinds of issues and we make mistakes, but anybody who starts and runs a business in a competitive market gets confronted with reality every day. The practical consequences of your decisions wash away your intellectual theories in the first three weeks. Every farmer confronts reality every day. Everybody trying to build a bridge confronts reality. Physics doesn't lie. Physics can't be lied to. It just is.

这可能是人们体验世界的深层根本分歧。在无形部门工作、面对市场竞争的人接触现实，而身处知识象牙塔或抽象政策世界的人则不然。我常说我的很多好朋友是律师，但也许90%掌权的人不应该是律师。那可能是个坏主意。

This might be the deep underlying divide in how people experience the world. People who work in intangible sectors with market competition are exposed to reality, while people in intellectual ivory towers or abstract policy worlds are not. I always say a lot of my best friends are lawyers, but maybe 90% of people in positions of power shouldn't be lawyers. That might be a bad idea.

是的。在背景方面，多样性去哪儿了？

Yeah. What happened to diversity when it comes to your background?

是的。我喜欢退伍军人竞选公职的趋势，因为战斗士兵生活在具体现实中。要么服兵役，要么经营企业，是制定政策的人的两个绝佳背景。

Yes. I love the trend of military veterans running for office, because combat soldiers have lived in tangible reality. Either military service or running a business would be two great backgrounds for people setting policy.

关于释放经济增长，你曾因评论拜登政府说‘甚至不要创办AI公司’而走红。你认为情况已经改变，我们正在翻开新的一页，进入一个人们可以成长并受到鼓励的地方吗？

On the point of unshackling economic growth, you went viral for your comment about the Biden administration saying don't even start an AI company. Do you think that has changed and we are turning the page to a place where people can grow and that is encouraged?

这是一个非常动态的话题。拜登政府百分之百试图扼杀AI初创公司。他们直截了当地告诉我们他们会这么做。那是计划。他们明确告诉我们不要投资新的AI初创公司。他们不会让它们成功。新政府……

It's a very dynamic topic. The Biden administration 100% tried to kill AI startups. They told us flat out that they were going to do that. That was the plan. They told us literally not to invest in new AI startups. They would not let them succeed. The new administration...

监管。

The regulation.

是的。他们非常直接地告诉了高层人士。那些在2024年管理美国的人。

Yeah. They told very directly the senior people. The people running the United States of America in 2024.

显然，对于书籍和其他一切也是如此。是的。但那是管理政府的人。而且是在白宫。非常具体。他们直接告诉我们，不，他们告诉我们，他们将指定两到三家公司作为国家冠军。他们将直接控制这些公司。它们基本上是政府的工具。他们将把所有权力交给这些公司，然后阻止其他任何人做这件事。他们说他们会不惜一切代价来实现这一点。我们实际上就此追问了他们，他们在那次会议上说，如果需要，他们实际上会将人工智能、数学和算法列为国家机密，作为国家秘密并禁止它们。原话是，我们在冷战期间对物理学做过同样的事，如果必要，我们会对数学和人工智能数学再做一次。所以他们直接告诉我们，这就是他们要做的。

Clearly for the books and everything else. Yes. But it was the people running the government. And it was in the White House. Very specifically. And they just told us flat out, like, yeah no, they told us we're going to anoint two or three companies to be the national champions. We're going to directly control those companies. They're basically the arms of the government. And we're going to give them all the power and then we're going to prevent anybody else from doing it. And they said we'll do whatever we need to do to do that. And we actually probed them on that, and they said in that meeting that if they need to, they will actually classify AI, math, and algorithms under actual national security classification as state secrets and ban them. And literally the literal line was, we did it before with physics during the Cold War, and we'll do it again with math and AI math if we have to. So they just told us flat out that that's what they were going to do.

本届政府，特朗普政府，已经发生了 180 度的积极转变，令人难以置信地支持、周到，而且我认为非常明智。里面的人，我有 David Sacks 和我的前同事 Christian，他们是迄今为止政府中在科技政策方面最聪明、最了解情况的人。我认为他们很棒。整个政府、内阁官员、白宫里的其他人，都非常出色。所以这很好。但我要说的是，这并非纯粹的党派问题。有很多反科技的民主党人。也有不少支持科技的民主党人在他们自己的党内斗争。在右翼，共和党方面，也有反科技的共和党人，包括参议院的一些人，包括一些我认为本应更明智的人。

This administration, the Trump administration, has been 180 degrees in the positive direction, like incredibly supportive and thoughtful, and I think very wise. The people in there, I get David Sacks and my former colleagues Christian in there, who are by far the smartest, best-informed people at tech policy that have ever been in the government. I think they're great. And the whole administration, cabinet officials, everybody else, people in the White House have been fantastic. So that's great. But I will say this is not a purely partisan thing. There are a lot of anti-tech Democrats. There are also quite a few pro-tech Democrats who are fighting inside their own party. And on the right, on the Republican side, there are anti-tech Republicans, including by the way some in the Senate, including some people who I think really ought to know better.

还有一些有播客的人。

Some people with podcasts too.

什么？抱歉。

What's that? Sorry.

还有一些有播客的人。

Some people with podcasts too.

是的。嗯，我不太确定。我要说的是 Ted Cruz 一直非常支持。所以我不确定你指的是哪一个。

Yeah. Well, I'm not exactly sure. I will give Ted Cruz has been hugely supportive. So I don't know exactly which one you were talking about.

哦，不，不是参议院民主党人，不是这边有播客的人，分开的。那这个行政令呢？这是……

Oh, not sorry, not Senate Democrats, not people on this side with podcasts separate. Do what about this EO? Is this...

哦抱歉，你在说右翼播客。所以有一堆右翼播客，不管你怎么称呼它，活动家那一面，已经变得非常尖锐地反对。

Oh sorry, you were talking about right-wing podcast. So there's a bunch of right-wing podcast, whatever you want to call it, activist side, that have turned really sharply anti.

不过关于这个行政令，我很好奇，David Sacks 似乎已经批准了，但有些人担心这仍然可能扼杀一些创新。

On the executive order though, I'm curious, David Sacks seems to have given his blessing to that, but some people are worried that could still strangle some innovation.

是的，我的意思是，听着，我认为总是有危险。而且我认为危险，明确地说，我认为不太在于这届政府会做什么，而在于你建立了一个监管框架，然后未来政府可能会利用它并扩大它。我总是举这个例子，所得税最初只对全国最富有的人征收 3% 的税，然后我们就到了今天。所以通往地狱的道路是由善意铺成的。所以我认为做这件事的人有良好的意图。我认为这在今天是合理的。我认为最大的风险是它从这里走向何方。他们做的一件非常好的事情，我真的很支持，就是它 100% 是自愿的安排。顺便说一句，我认为自愿在这里真的可以起作用，因为所有美国科技公司，第一，他们都希望政府不讨厌他们。第二，政府已经拥有很多隐性权力，可以让公司做他们想做的事。所以这个自愿性质、他们定义的时间框架，我认为基本上这个行政令是我们在这个世界里可能得到的最好的了，因为他们出于各种原因不能什么都不做。但我认为它在他们觉得需要做的事情和结果之间取得了最好的平衡。如果它稳定在行政令的水平，并且立法稳定在那个水平，我认为我们设置得很好。我的担忧，正如我们刚才说的，是很多人想要更严厉的控制水平。而且我认为今天仍然存在走向非常糟糕方向的风险。

Yeah, I mean, look, I think there's always a danger. And I think the danger, to be clear, I think it's less what this administration would do, but I think the danger is you set up some framework for regulation and then there's a danger that future administrations take that and they expand it. I always say the example, the income tax started out as like a 3% tax on only the richest people in the country and then here we are today. So the road to hell is paved with good intentions. So I think the people who did this have good intentions. I think this is sensible for today. I think the big risk is where this goes from here. The really good thing about what they did, which I really support, was it's 100% a voluntary arrangement. And by the way, I think voluntary can really work here because all American tech companies, number one, they all want to have the government not hate them. And number two, the government has lots of implicit powers already to be able to get companies to do what they want. So the voluntary nature of the thing, the time frames that they define, I think basically the EO is as good as we could plausibly get in the world that we're in where they couldn't, for a variety of reasons, they couldn't do nothing. But I think it strikes the best balance between the needs to do the things they feel like they need to do and the outcome. If it stabilizes where the EO is and if legislation stabilizes that level, I think we're set up really well. My fear, as we were just saying, is I think a lot of people want much harsher levels of controls. And I think there's certainly today still risk that it's going in a very bad direction.

如果你让骆驼的鼻子进了帐篷，事情就会改变。我还想谈谈美国活力，以及当你开始那件事并写你的文章时硅谷的氛围变化。天哪，我想那是三年前了。当时它非常反主流文化，现在我们看到了 Palantir、Anduril，这些公司非常热门。你认为我们已经看到了那种氛围转变完成了吗？我们正在看到爱国主义被拥抱？你认为它已经不再是反主流文化，现在在硅谷更主流了吗？

If you let the camel's nose in the tent, things can change. I want to talk a little bit about American dynamism, too, and the vibe change in Silicon Valley when you started that and you wrote your essay. Gosh, I guess this was three years ago. It was very countercultural at the time and now we've seen Palantir, Anduril, these are such hot companies. Do you think that we have seen that vibe shift completed, that we are seeing patriotism embraced? Do you think that has ceased to be countercultural and now that is more mainstream in Silicon Valley?

是的。所以我会说硅谷曾经非常爱国，事实上，即使我在 90 年代来到这里时，这几乎被视为一种信条，你当然会支持政府、支持军队、支持执法。只是因为显然你会这样做。我认为这种反爱国情绪的发展，而且它确实是反爱国的。老实说，我认为是伊拉克战争真正催化了它。

Yeah. So I would say Silicon Valley used to be incredibly patriotic and in fact even when I was here in the 90s when I came in the 90s, it was just kind of taken as an article of faith that of course you supported the government and supported the military and supported law enforcement. It was just because obviously the case that you would do that. I think really this anti-patriotic thing that developed, and it really was anti-patriotic. I think honestly it was the Iraq war that really catalyzed it.

嗯，然后我想那是那件事的余波和金融危机，然后你知道千禧一代变得超级激进，嗯，被触发——就是这个词，极度被触发。嗯，所以你知道有大约 15 年的时间，事情变得糟糕。嗯，老实说，我并不声称自己有什么伟大的道德英雄主义，我认为反对者——那些谴责任何支持政府或军队的人——变得非常直言不讳。我想很多有不同想法的人，可能在那段时间应该更直言不讳、更活跃。我得说那是一个奇怪的时期。

Um and then I think it was sort of the aftermath of that and the financial crisis and then you know the millennials as you know kind of got super activated uh on on on triggered um that's the word trigger extremely triggered um and so you know there was this 15-year run or whatever um where you know things got bad um and and you know honestly and I you know I don't claim any great moral you know heroism here like I I think the opponents uh you know the people who condemned any sort of support of the government or the military you know became very vocal I think you know a lot of us who felt differently. Probably should have been, you know, more vocal and more active through that period. It I will say it was a weird period.

是的。你知道，2008 年到 2024 年很奇怪。

Yeah. You know, 200 whatever 8 to 2024 was weird.

嗯，我们正在慢慢意识到那个时期有多奇怪。但它确实很奇怪，这也是其中的一部分。嗯，我给予很多年轻创业者很高的评价。我认为很多人挺身而出。有一个明显的勇敢时刻。我特别想提到 Andre 这家公司，它在一个非常艰难的时期站了出来。嗯，还有其他一些公司。然后在过去五年里，我们看到，事实证明存在非常强烈的潜在爱国主义和对国家的热情，当这不会让你自动被所有社交活动除名，或者有替代的社交活动可以参加，和更像你的人在一起时，

Um and and and uh I think we're it's sort of slowly dawning on us how weird that period was. Uh but it really was weird and and this was part of it. Um look and and I think look I give a lot of the younger younger entrepreneurs a lot of credit. I I think a lot of people stepped up. I think there was a a point of conspicuous bravery. I would particularly cite Andre uh as an example of a company that really you know stood up at a time when it was very hard to do that. Um and um you know a number of other companies and then in the last you know five years uh you know we we've just seen it as there just it turns out there's just like a very there's a very strong level of latent patriotism and enthusiasm for the country and when it when it's something that you know doesn't get you just automatically disinvited from every social event or or if there are alternate social events that you can go to with people who are more like you,

对吧？

right?

嗯，你知道，那种冻结真的裂开了。顺便说一句，其中非常有趣的一点是，当然有右翼成分，但在左翼，民主党方面，实际上转变中最重要的事情之一是俄罗斯入侵乌克兰。

um you know the the the the the freeze really crack really really really cracked open um by by the way that just the incredibly entertaining aspect of it um was um you know there there's there's sort of a right-wing component to this of course but there's also on the on the left uh Democrats it it was actually one of the one of the sort of most significant things happened with the shift was actually the the Russian invasion of Ukraine

嗯，因为乌克兰成了左翼的事业，这实际上使得好的民主党人有可能支持国防。

um because it actually it actually because Ukraine became a cause on the left it actually made it possible for good democrats to be pro-defense

这是很长时间以来的第一次。然后美国企业界的笑话变成了国防突然成了 ESG 的一部分。

for the first time in a long time. And and then the running joke in corporate America became that defense all of a sudden is part of ESG.

我知道我喜欢国防。这很好。

I know I like defense. This is good.

这很好，因为你在保卫乌克兰。当然，乌克兰和俄罗斯发生的事情是悲剧，但这是一个非常重要的进步。很多原本认为自己对任何与军队有关的事情感到非常不舒服的人，突然意识到乌克兰显然需要自卫，那么其他国家也是如此。所以我认为，甚至很多左派人士也加入了我们，这很棒。

It's good because this, you know, because you're because you're you're defending Ukraine. And so it's and of course, you know, look, it's a it's of course it's a tragedy what's what's happened with with with Ukraine and and with with with Russia and so forth, but um just it it it was a very significant step forward. A lot of people who were a lot of people who would have classified themselves I think as being very uncomfortable doing anything with the military uh or with anything military all of a sudden were like oh you know obviously Ukraine needs to defend itself and then you know by extension so do other countries. Um and so I think even even even I would say even a lot of dem even a lot of left lefties have actually come with us on this and I I think that's been great to see.

是的。你认为现在国防科技、安全科技最大的机会是什么？你显然说过民主需要一把剑。现在这个领域最让你兴奋的是什么，以及你看到的创新？

Yeah. What do you think are the biggest opportunities right now in defense tech, safety tech, you've obviously said democracy demands a sword. What what most excites you right now about that space and the the innovation you're seeing?

我的意思是，发生了很多事情，但如果我必须说一件事，那就是我们有机会利用技术拯救我们服役人员的生命。嗯，顺便说一句，我认为这对国际上的军队也是如此。我认为对国内的执法部门也是如此。是的。

I mean, there's there's there's so much happening, but if if I if I had to put a single single thing, it would be I think we have the opportunity to use technology to save the lives of of of our of our service service people. Um uh and by the way, I think that's true at at for the military internationally. I think it's also true of law enforcement uh domestically. Yeah.

嗯，我认为我们的现役部队和执法官员总是将自己置于个人危险之中。你知道，经典的战争原型是一个拿着步枪的人去射击和被射击。作为国家，250 多年来我们失去了许多了不起的人。但随着新技术，自主性就是一个很好的例子，我们能够以让宝贵的人类士兵远离危险的方式投射力量。嗯，我认为这绝对是压倒性的积极。我对我们在国防领域的工作感觉良好的原因之一，是我们确实在推动这一点。

Um I think that, you know, our active duty troops and our and then our law enforcement officers, you know, place themselves in in personal harm's way all the time. you know, I mean, the classic archetype of going to war is, you know, a guy with a rifle heading off to get, you know, shoot and get shot at. Um, and uh, you know, and we we as a country have lost many amazing people, you know, over over 250 years, you know, from from from doing that. And, you know, with techn millions and and and with with with, you know, with the new technologies, you know, autonomy being a great example of it, um, you know, we are able to project force um, in ways that that keep our keep our keep our our valuable our valuable human soldiers out of harm's way. Um, and and I just think that's like, you know, that just have to be right over overwhelmingly positive. Um, and I I I one of the reasons I feel so good about our work in defense is I I think we're we're really helping move the needle on that.

是的。那么你看到未来不再需要地面部队了吗？

Yeah. I mean, so do you see a future where there never has to be boots on the ground again.

我的意思是，我会说不。很难说永远不需要，因为人们谈论军事理论的所有原因——在复杂情况下，人不可替代。你不想像打电子游戏一样打一场战争，你需要接触现实，需要真正了解情况的人。所以我不认为会完全取代。我给你举个例子，在伊拉克，甚至我们的特种部队军官也必须去踢开门，然后搜查炸弹，对吧？

I mean, it's this thing of like, you know, if if I I would say no. I would say is it's hard to it's hard to ever say that for all the reasons you know people talk about military doctrine like there there's no substitute you know in complex situations for people and again kind of being engaged in obviously you don't want to fight an entire war like it's a video game like you you you you want to have contact with tangible reality you want to have people who actually know what's going on and see it and so I I don't think it's so much that I think it's more just like do you really I'll give you an example like all the way during Iraq like Iraq like so much of what like even our special forces officers had to do is they had to go they literally had to go kick doors down and then they had to for bombs, right?

100%。

100%.

嗯，所以这是一个很好的例子：在这种情况下，那个人可以待在屋外，让无人机扫描房子，然后当他进去与人交谈时，房子已经被清除了。嗯，就是这样的事情。顺便说一句，执法部门也一样。实际上，无人机在执法中有一个被低估的明显用例：如果你是一名警察，接到 7-Eleven 被闯入的报警，你必须做的一件事就是清理屋顶。比如，屋顶上有人会向你射击。

Um, and so like that's a great example of where it's like, yeah, you know, you and maybe maybe the guy in that case, you know, is able to stay outside of the house while the drone sweeps the house and then by the time the guy goes in the house uh, you know, to talk to people, it's, you know, it's the house has been cleared. Um, and so like, you know, that's the kind of thing. By the way, same thing in law enforcement. Actually, there's a very underappreciated just obvious use cases for drones and law enforcement, which is if you if you're a cop and you respond to a call that like a 7-Eleven is getting broken into, one of the things you have to do is you have to clear the roof. Like, you know, there's somebody on the roof who's is going to like shoot at you.

是的。

Yeah.

对。嗯，所以没有无人机的情况下清理屋顶的方式是派一个人爬梯子，探头上去，对吧？而有了无人机，你只需要飞越一下。

Right. Um and so the way you clear a roof, you know, without a drone is you send a guy to climb the ladder and poke their head up, right? And like with a with a drone, you just like do do an overfly.

就像鳄梨酱。是的。

It's like guacamole. Yeah.

是的。完全正确。

Yeah. Exactly.

所以有所有这些……另一个例子就是，甚至在到达现场之前就能勘察情况。比如，有一起持械抢劫的报案，然后就想，好吧，发生了什么？那家伙还在吗？店里还有人拿着枪吗？而能够说一架无人机就能回答这个问题。所以实际上有所有这些事情，当你深入了解时，我刚才描述的那些都不需要什么庞大的 AI 超级大脑。你只需要让自主无人机配上摄像头，交到人们手中。而这真的开始发生了。

And so there's all of these... another example would be just even inspecting the scene before you get there. It's like there's a report of an armed robbery or something and it's just like, okay, what is happening? Is the guy still there? Is there a guy with a gun who's still in the store? And being able to say a drone can answer that question. So there's all of these things that actually, when you get into it, everything I just described doesn't require some massive AI superbrain. You just need to get the basic equipment of autonomous drones with cameras in people's hands. And that's really starting to happen.

我想问一个关于你个人的问题，关于你如何看待自己现在的角色。显然你一直是一位非常出色的投资者，但你和硅谷越来越多的人，尤其是你，非常直言不讳地谈论你追随的思想领袖、你的信仰，并且在资本主义和自由市场真正受到攻击的时候，似乎在为它们布道。所以我很好奇，你现在如何看待自己的角色？你显然是一位风险投资家，你进行投资，但你似乎也承担起了为美国理想布道的角色。

I want to ask a personal question about you and how you see your role now. Obviously you've always been a very skilled investor, but you and increasingly others in Silicon Valley, but especially you, are very vocal about the thought leaders you look to, your beliefs, and are sort of evangelizing capitalism and free markets at a time when that really is under attack. So I'm curious, how do you view your role right now? You're a venture capitalist, you invest, but you also seem to be embracing this mantle of somebody who's really evangelizing for American ideals as well.

是的，没错。你看，部分原因是我相信这些，但我描述的方式是，在我职业生涯的前半段，从 1990 年到 2015 年左右，每个人都热爱科技。尤其是每个人都热爱初创公司。奥巴马，那很时髦。那太棒了。共和党人、民主党人，都太棒了。每个人都想要合影。那是庆祝，媒体报道压倒性地正面。《纽约时报》喜欢报道新的科技初创公司。那太棒了。然后到了 2015 年，由于各种原因，心理完全改变了。出现了一种全新的敌意，反科技敌意，反硅谷是肯定的，反社交媒体是其中很大一部分，但还有反市场、反资本主义、反自由、反公民自由，在很多方面都是如此。所以世界变了，我头几年完全困惑，因为他们两年前还爱我们，现在却恨我们。我们基本上是同样的人，试图做同样的事情。这说不通。我认为这有一种全国性的心理版本。但我试图告诉我们的团队的另一件事是，我们也需要弄清楚如何承担责任。所以我的想法是，像我这样的人总是希望我们所做的事情有意义。我们总是希望技术在世界中重要。我们总是想改变世界。好吧，世界有发言权。你不能在世界上没有观点的情况下改变世界，而且他们并不总是对改变感到满意。如果你不解释自己，别人就会替你解释，而你不会喜欢他们的解释，因为他们有自己的动机和目的。所以我告诉人们，如果人们不理解或不喜欢我们在做什么，就假设这都是我们的错，然后让我们弄清楚我们能做些什么来做出贡献。我们所做的一部分就是建造好东西，但另一部分是我们必须解释它。我们有义务解释自己，并充分阐明我们在做什么，为什么我们认为它重要，为什么我们认为它好。行业里有些人长期以来一直擅长这个。我们很多人以前认为这不重要。我认为现在更多人认为这很重要。所以看到这么多同事现在这样做，我真的很受鼓舞。

Yeah, that's right. And look, part of it is what I believe, but the way I describe it is for the first half of my career from 1990 through to 2015 or something, everybody loved tech. Everybody loved startups in particular. Obama, it was chic. It was incredible. Republicans, Democrats, it was incredible. Everybody wanted the photo op. It was celebration, and the press coverage was overwhelmingly positive. The New York Times loved to profile new tech startups. It was wonderful. And then in 2015, for a variety of reasons, the psychology completely changed. There's this entirely new hostility, anti-tech hostility, anti-Silicon Valley for sure, anti-social media is a big part of it, but also anti-market, anti-capitalism, anti-freedom, anti-civil liberties in a lot of ways. So the world changed, and I spent the first few years just completely confused because they loved us two years ago, they hate us now. We're the same people fundamentally trying to do the same things. This doesn't make any sense. I think there's a national psychological version of this. But the other thing I was trying to tell our folks is we also need to figure out how to take responsibility for things. So the way I think about it is people like me always wanted what we did to matter. We always wanted technology to matter in the world. We always wanted to change the world. Well, the world gets a vote. You don't get to change the world without the world having a point of view on it, and they're not always going to be happy with the changes. And if you don't explain yourself, other people are going to explain you for you, and you are not going to like their explanations because they have their own incentives and purposes. So I tell people, let's assume it's all our fault if people don't understand or don't like what we're doing, and then let's figure out what we can do to contribute. Part of what we do is just build good things, but the other part is we have to explain it. We have an obligation to explain ourselves and to fully articulate what we're doing and why we think it matters and why we think it's good. Some people in the industry were good at that for a long time. A lot of us didn't think it was important. I think a lot more of us now think it is important. So I'm really encouraged by the number of my colleagues that are now doing that.

但你在 X 和其他地方似乎非常无所畏惧地谈论对你重要的政治或其他问题。那是……我想这需要一定的勇气。

But you seem very unafraid on X and wherever to speak up about issues political or otherwise that matter to you. Was that... I imagine that requires a degree of courage.

你知道吗？我不知道。是的，有点。我想我会说，当你受到如此多的攻击时，什么都不做感觉就是愚蠢的。所以我不知道。也许我不会……这感觉不像是勇气。这感觉像是在回应现实。但我很感激你的想法。

You know? I don't know. Yeah, somewhat. I guess I would say there's a point at which you're getting attacked so much that it feels like to not do anything is just dumb. So I don't know. Maybe I wouldn't... It doesn't feel like courage. It feels like responding to reality. But I appreciate the thought.

你在加利福尼亚，身处虎穴。正如基思提到的，我们有一个加州岗位。有没有什么事……看起来他们正在竭尽全力把你赶出去。你似乎仍然很忠诚。他们能做什么吗，或者你觉得这有点像阿拉莫时刻，你需要留下来战斗，保护你的家园？

You're very much in the belly of the beast there in California. As Keith mentioned, we have a California post. Is there anything... it kind of seems like they're doing everything they can to force you out. You still seem pretty loyal. Is there anything they could do or do you feel like this is sort of an Alamo moment where you need to stay and fight and keep your home safe?

是的，你看，第一，是的，他们做了很多事试图把我们赶出去。很多当权者正在非常努力地杀死下金蛋的鹅。而且顺便说一句，他们不仅试图扼杀科技，还试图扼杀好莱坞。所以……

Yeah, look, number one, yes, they are doing a lot to try to force us out. A lot of the powers that be are trying very hard to kill the golden goose. And by the way, they're not just trying to kill tech, they're also trying to kill Hollywood. So the...

是的，我觉得他们差不多已经做到了。

Yeah, I feel like they've kind of already done that.

是的，他们基本上已经做到了。今天好莱坞的电影和电视制作比罢工期间还要少。他们几乎终结了好莱坞作为洛杉矶的一个产业。这简直不可思议。我的意思是，加利福尼亚州有两个现金喷涌口，两个税收收入的喷涌口：好莱坞和硅谷。

Yeah, they basically have. There's less production, less film and television production happening in Hollywood today than there was during the strikes. They've almost ended Hollywood as an industry in Los Angeles. It's just incredible. And my point is there are two cash gushers, two gushers of tax revenue in the state of California: Hollywood and Silicon Valley.

而且不知为何，那些想把政府开支推到天上去的人，同时也想摧毁税基。

And for some reason, the same people who want to crank government spending to the moon also want to kill the tax basis.

忘了提医保。那是个大头，大量资金流向那里。他们想在医保上花很多钱，还想增加开支，如果你不支持增加开支，他们就会骂你难听的话。很长一段时间里，存在一种默契：我们在加州建立这些产业，它们产生税收，然后政客用这些税收来支付所有这些开支。长期以来，只要科技产业增长足够快，好莱坞表现足够好，那也许是一种不完美的情况，但至少是一种现状。但他们——我们的许多政策制定者——想同时摧毁税基和膨胀政府开支。这完全疯了，毫无逻辑，但他们就是一股劲地往前冲。看吧，他们可能会成功。他们把埃隆赶走了，给制造业就业带来了灾难性的后果。他们赶走了该州最大的制造业雇主，这对于那些声称关心制造业就业的人来说相当惊人。而且我的很多同事，我知道很多人过去一年离开了加州，原因多种多样，包括即将到来的那种极具侵略性的没收性税收。所以这些都是真的。但我要反驳的是，首先，加州是全球科技产业的家园，也是全球娱乐产业的家园。顺便说一句，人工智能已经把科技重新集中到了硅谷。美国科技产业，美国技术创新的源泉，现在比以往任何时候都更集中在北加州，因为人工智能。旧金山及其周边 20 平方英里是现在几乎所有人工智能有趣工作的中心。所以部分原因是我只想在那里参与其中。除非他们最终能一枪打死它，否则我想在那里。另一件事是我确实认为加州是一个极其特别的地方。它在整个美国历史上都是如此，并且在未来 250 年也应该如此。我认为我们不应该放弃。

Forgot to mention Medicare. That's a big one. A lot of cash coming there. They want to spend a lot of money on Medicare and they want to increase that spending and they want to call you horrible names if you don't want to increase that spending. And for a long time there was a deal: we build these industries in California, they generate tax revenue, and then the politicians use the tax revenue to pay for all this stuff. For a long time it was okay as long as the tech industry is growing fast enough and Hollywood is doing well enough, then maybe it's an imperfect kind of thing but at least it's some kind of status quo. But they, so many of our policy makers want to simultaneously kill the tax base and inflate government spending. It's completely insane, there's no logic to it, but they're just on a head of steam. Look, they may succeed. They drove Elon out, to just catastrophic results for manufacturing employment. They drove out the number one manufacturing employer in the state, which is fairly amazing for people who claim that they care about manufacturing jobs. And a lot of my colleagues, I know a lot of people have left the state in the last especially in the last year for a variety of reasons including this extremely kind of aggressive confiscatory tax thing that's coming down the pipeline. So that's all true. I will say against that, number one is California is the home of the global tech industry. It is the home of the global entertainment industry. AI by the way has recentralized tech into Silicon Valley. The American tech industry, the wellspring of tech innovation in America is more concentrated in Northern California now than I think ever before because of AI. San Francisco and the surrounding 20 square miles are ground zero for essentially all of the interesting work that's happening now in AI. So part of it is I just want to be there for that. And unless they finally are able to put a bullet in that, I want to be there for that. And then the other thing is I do think California is an incredibly special place. It has been for the entire history of the country and it should be for the next 250 years. I don't think we should surrender.

我也很好奇你的投资理念，你在这个时刻如何区分趋势和热潮。因为有些人全力投入人工智能，但也有一些问题，对吧？就像加密货币一样，你知道，无聊猿。在这个时刻，你如何思考什么是暂时的，什么是持久的？

I'm curious about your investment thesis as well, how you are distinguishing in this moment between trends and fads. Because some people are all in on AI, but there are some glitches, right? Just like there was with crypto, you know, Bored Apes. How are you in this moment thinking about what's temporal versus what's here to stay?

是的。所以我说几点。一是我们试图深入事实。我的意思是，我们尽可能深入技术本身。所以我们一直往下看：好吧，从技术上讲，在实际技术中，这些东西真的能正常工作吗？这些东西真的存在吗？我们公司的大部分人，我们的投资者，我们的员工在商业上都非常有才华，但其中许多人拥有非常深厚的技术背景。我们花了很多时间深入探究，试图从深层技术角度理解什么是真实的。顺便说一句，我们对此感觉非常好。人工智能的技术进步绝对是非凡的，现在基本上每天都有让我目瞪口呆的事情发生。我自己是一名受过训练的工程师。我和我的工程同事都对进步的速度感到震惊。所以我们寻找那个。然后老实说，另一件事是风险投资。风险投资的伟大之处之一在于，该模式允许 50% 的失败率。我们所做的基本数学计算，这在世界上所有类型的投资或金融中都非常独特，是顶级风险投资的长期模型是一半的项目成功，一半的项目失败。顺便说一句，当然，我们都希望最终能找出只支持成功的那一半的公式，因为我们宁愿不支持那些失败的项目。

Yeah. So, I say a couple things. One is we try to go to ground truth. By that I mean we try to go all the way as deep into the technology itself as we can. So we try to go all the way down into: okay, technically, in the actual technology, are these things actually working right? Are these things actually real? Most of our firm, our investors, our folks are very talented at business, but many of them have very deep technical backgrounds. We spend a lot of time going deep to try to understand what's real at a deep technical standpoint. By the way, we feel very good about that. The technical advances in AI are just absolutely extraordinary and it's at this point it's basically daily the things happening that have my jaw on the floor. I'm a trained engineer myself. My engineering colleagues and I are jaws on the floor at the rate of progress. So we look for that. And then quite honestly, the other thing is venture capital. One of the great things about venture capital is the model allows for a 50% failure rate. The basic math on what we do, and this is very unique among all kinds of investing or financial whatever in the world, is the long-term model for top venture capital is half the projects work, half the projects don't. And by the way, of course, we all live in hopes that we will eventually figure out the formula to only back the half that work because we'd rather not back the ones that don't.

我的意思是，对于风险投资来说，这是一个相当高的成功率。很多人的比例要低得多。

I mean, that's a pretty high success rate for VC. A lot of people have a much lower ratio.

嗯，事情就是这样。人们对这个有两种反应。一种是，天哪，50% 的失败率，你们真是一群白痴。另一种反应就是你刚才说的，老套的说法是 90% 会失败。好消息是那不是真的。实际上是 50% 失败。所以每个人都是赢家。顺便说一句，还有一种叫做贝比·鲁斯效应。棒球中的贝比·鲁斯效应是，打出本垒打的人最常三振出局。因为你 100% 错过你没有挥棒的球。你在冒更大的风险。如果你不冒激进的风险，你就打不出本垒打。如果你不尽可能用力挥棒，冒着失去控制的风险，你就打不出本垒打。如果你在风险投资中不承担风险，你就得不到那些异常值。而这一切都是关于异常值的游戏。所以这就是为什么美国很特别？美国风险投资到目前为止是世界上最好的。绝大多数伟大的风险投资公司都集中在美国。其中一点是，美国的风险投资公司尽管有其他问题，但确实愿意承担风险，并且确实愿意支持非常投机的事情。

Well, that's the thing. People either have two reactions to that. One is, my god, 50% failure rate, you guys are a bunch of morons. The other reaction people have is the one you had, though, which is the cliche is only 90% of them fail. The good news is that's not true. It's actually 50% fail. And so everyone's a winner. And by the way, there's also something called the Babe Ruth effect. The Babe Ruth effect in baseball is that the people who hit the home run strike out the most often. Because you miss 100% of the shots you don't take. You're taking more aggressive risks. If you don't take the aggressive risk, you don't hit the home run. If you don't swing the bat as hard as you can, risking losing control, you don't hit the home run. If you don't take risk in venture, you don't get the outliers. And it's all a game of outliers. So it is this, ask like why is America special? American venture capital is the best in the world by far. Overwhelmingly the great venture firms are concentrated in the US. And one of the things is US VCs for all their other issues are really willing to take risk and they're really willing to back things that are very speculative.

有时候这些事情取得了惊人的成功，而有时候几年后你会想，‘天哪，他们当时在想什么？’我只能告诉你，我们发现的每件事在最初看起来都很有前景。

Sometimes those things are incredible successes, and sometimes years later you think, 'My god, what were they thinking?' All I can tell you is everything we find looks very promising in the beginning.

所以，你知道什么很了不起吗？我们踏上了旅程。我们尽力支持最优秀的人，去应对正在发生的最深刻的技术变革，然后与他们一起踏上旅程，努力帮助他们打造伟大的东西。有时成功，有时失败。顺便说一句，许多伟大科技公司的伟大创始人都有过失败的经历。经典例子：Travis Kalanick，Uber 的传奇创始人。Uber 是他的第三个创业项目；前两个都失败了。

So, you know what's amazing? We go on journeys. We try to back the best people we can, working against the most profound technology change happening, and then we go on a journey with them and try to help them build something great. Sometimes it works, sometimes it doesn't. By the way, a lot of the great founders of the great tech companies had previous failures. Classic example: Travis Kalanick, legendary founder of Uber. Uber was his third startup; the first two failed.

没错，以至于你今天看他们时会想，‘他当时在想什么？’答案是他在冒险。三个中有一个成功了，而且成功得如此之好，以至于它弥补了两次失败，差距大约十万倍。美国既有规模、范围、稳固性和资源，又仍然拥有那种冒险精神，这真是一个非常特别的组合。

Right, to the point where you look at them today and think, 'What was he thinking?' The answer is he was taking risk. One out of three worked, and it worked so well that it made up for the two failures by a factor of like 100,000. The idea that America has both the size, scope, solidity, and resources, yet still has that level of risk-taking spirit, is a really special combination.

在我们结束之前，有什么关于 AI 的事情让你夜不能寐吗？

Before we go, is there anything that keeps you up at night about AI?

哦，关于 AI 让我夜不能寐的事？我想你可以说这些公司和我们承担着巨大的责任。听着，AI 是件大事。它将变得非常重要。人们正在与它建立亲密关系。搜索引擎一直有这种情况：如果你查看某人的搜索历史，你会了解到很多关于他们的信息。有个笑话是，‘那家伙看起来像是应该被查搜索历史的人。’AI 将会是这种情况的一千倍。你与 AI 的对话涉及你生活中许多最敏感、最重要的事情。我真的相信，我们和这些公司在照顾客户和大众方面所承担的责任是非常真实且沉重的。我对此感觉良好；我认为这些公司总体上非常负责任。但这可能是我想要确保我们做对的事情。

Oh, what keeps me up about AI? I guess you could say these companies and we are taking on massive responsibility. Look, AI is a big deal. It's going to be a very big deal. People are forming an intimate relationship with it. There's always been this thing with search engines: if you look at somebody's search history, you learn a lot about them. The running joke is, 'That seems like the kind of guy where somebody should check a search history.' AI is going to be like that times a thousand. The conversations you get into with AI are about many of the most sensitive, important things in your life. I really believe the level of responsibility that we and these companies have in taking care of customers and people at large is very real and very heavy. I feel good about that; I think generally these companies are very responsible. But that would probably be the thing I would want to make sure we get right.