布雷特·阿德科克：AI 会比互联网大一百倍

我觉得如果你在谷歌上搜布雷特·阿德科克的净资产，根据《财富》杂志，你有 190 亿美元。这变化可真不小。你感觉如何？

I think if you Google Brett Adcock net worth, according to Fortune, you're worth $19 billion. So that's like a pretty good swing. How does that make you feel?

我根本不在乎这个。完全不在乎。

I don't care about that at all. Give like zero shits about that.

好的。那么，布雷特·阿德科克，简单来说，你在伊利诺伊州的农村长大。你创办了一家叫 Vettery 的公司，以超过 1 亿美元的价格卖掉了。然后你让一家叫 Archer 的公司上市了，那家公司做的是无人驾驶的飞行器，大概是直升机之类的。现在你有一家叫 Figure 的公司，估值我不知道是 400 多亿、300 多亿还是 500 多亿美元。你还有另一个项目叫 Cover，用来阻止或旨在阻止校园枪击案。现在你又搞了个新项目叫 HARK，已经筹集了数十亿美元。你看起来挺疲惫的。

Okay. So you uh Brett Adcock, the the short of it is that you were raised in a rural area of Illinois. You started a company called Vettery, which you sold for over $100 million. Then you took a company public called Archer which is like unmanned uh flying planes I guess helicopters. And then now you have a company called Figure which is worth I don't know how much 40ome 30 something 50 something billion dollars. You have another thing called cover which stops uh or aims to stop school shootings. And then now you have a new thing called HARK which you've raised money at in the billions of dollars. And you seem worn out.

很好。

Great.

你看起来是个大忙人。所以，你这是第三次上节目了。我觉得你过去三年每年都来一次。你之前讲过一个故事，好像是 Figure 刚起步的时候。你基本上说，我当时身价大概几千万美元，我把几乎所有的钱都投进了 Figure 来启动。你当时说，我房子还有按揭，剩下的钱都在 Figure 里，还有一些在 Archer 里，而 Archer 当时表现不太好。从那以后，我觉得如果你在谷歌上搜布雷特·阿德科克的净资产，根据《财富》杂志，你有 190 亿美元。这变化可真不小。你感觉如何？

You look like busy man. So, you've been on this is your third time on. I think you I think you've been on one time each year the last three years. You said uh you were telling a story about how I think it was right when Figure started. You basically said like I had I was worth I don't know how much tens of millions of dollars. I put almost all of it into Figure to get started and at one point you were like I have a mortgage on my house and the rest of my money is in Figure and some of the money is in Archer and that's not doing so great right now. And since then, I think if you Google Brett Adcock net worth, according to Fortune, you're worth $19 billion. So that's like a pretty good swing. How does that make you feel?

我根本不在乎这个。完全不在乎。

I don't care about that at all. Give like zero shits about that.

你是个超级好胜的人。我觉得你说过类似“我就是想赢”的话，而且说过好几次。上次我们见面时，你说“我想赢是因为这些原因。我非常好胜。我想大干一场。”我觉得你肯定得在乎一点这个，而且实际上，我觉得你非常在乎 Figure 成为世界上最大的公司。你说话时确实有种拿破仑式的能量，就是“我想成为最好的，我想征服一切。”

You're a super competitive guy. I think you said something like I just want to You said like win a bunch of times. Last time we hung out, it was like I want to win for these reasons. I I'm I I'm very competitive. I I want to kick ass. I think that like you definitely have to care about this a little bit and you actually have to I think you care a lot about figure being the biggest company in the world. You talk about like you definitely have this like Napoleon energy of like I want to be the best. I want to conquer.

我觉得任何方式来描述，就是我们现在，我的这些公司刚刚到达一个拐点，它们还非常早期，但它们可能会变得非常大。所以如果成功，从这里会增长 100 倍、10 万倍。所以我大部分精力都在想怎么确保这件事成功。这里没有平稳期。要么下降，要么上升，对吧？就像二元的。要么机器人实现规模化，要么不能。所以五年后，要么成为非常庞大的事业，要么非常糟糕。所以我所有的精力都投入到让这件事从我们现在的位置增长一千倍或一百万倍。所以压力就在于真正交付成果。

I I think any way I would characterize is like we're just like we're just now like these companies of mine are just now hitting inlection point and they're really early like they can be like really big. So if it works this will like 100x,000x from here. So most of my energy is like how do I make sure that works? There is no flatline here. It's either like it goes down or goes up, right? Either like it's binary. Either the robots go out of scale or they don't go out of scale. So in like 5 years time, it's either going to be a very big thing or very bad. And so all my energy is going into making this like thousand or a millionx from where we're at here. And so it's it's like the pressure is on to like really just deliver.

你现在处于什么阶段？接下来五年的前景如何？我觉得上次你上节目是三年前，我们当时说，我记得是我说的，你可能会达到 4000 万到 5000 万美元的估值范围，我觉得你现在已经达到了。但就机器人产量而言，你仍然缺乏产出，我们仍然需要看到那部分。五年后你会发展到什么程度？你的预测是什么？

Where where are you now? What's the outlook now for the next 5 years then? I think last time you were on 3 years ago, we said that I think I said it. Uh I was like you'll probably be in the $40 to $50 million valuation range, which I think you are now. But in terms of like you you're still lacking output of robots, like you still need we still need that to come. When where are you going to be in five years? What's your prediction?

我认为从宏观层面看，我们现在看到的 AI 工作将会非常巨大，规模可能是互联网的一百倍。一切都在顺利运转，深度学习确实有效，所有事情的发展速度都超出了我的预期。你知道，我做了 15 年的软件和互联网，那时候没有任何事情在趋势线上发展得这么快，而 AI 领域的发展速度就是这样。

I think at a high level, I think the AI work that we're seeing here now is going to be so much it's going to be like a hundred times bigger than the internet. It's just like everything is just so it's just working so well. like the system is working well like deep learning works and everything's happening faster than I would have think and my like you know having done like 15 years of like software and internet like it was just like nothing was happening faster on a trend line here it's happening like that in AI

你能举个例子，有什么事情让你感到震撼吗？

Can you give an example of something that has happened that's blown you away?

我们从 HARK 开始，大约一年前我成立了一个新的 AI 实验室叫 HARK。我当时对这个想法非常感兴趣，就是在数字世界构建 AI 与人类的共生关系。Figure 将会是 AGI 的最高上限，能够将这种能力释放出来，然后在数字世界也会有对应的版本。人类将拥有这种 AI 配对，最终可能拥有你自己的 AI 权重、你自己的记忆，也许还有你自己的硬件。这非常接近了。这个论点的基础是，你必须弄清楚如何让 AI 通用地使用计算机。你永远不会雇一个不会用电脑的助理。所以你必须能够给它分配任务，让它能做你能做的一切事情：财务模型、订机票、订 DoorDash 外卖，无论你需要什么，它都需要能自主完成。但只有千分之一的网站有 API。所以，在大多数情况下，全球大多数计算机使用都是在互联网和浏览器上。我的主要倾向是，在两三年内，你会有一个系统，你可以对它说“去做这个或那个”，然后它能够上网，像机器人一样熟练地使用互联网，移动鼠标、使用键盘。这就是解决计算机通用性必须做的事情，你不能依赖 API 或 MCP。你必须弄清楚如何像人类一样导航。现在在 HARK，我们上周刚刚发布了我们的第一个模型和研究预览。现在很难找到我们让它去网上做的事情它做不到的。

We started at so harkc I have a new AI lab called hark about a year ago I was like very interested in this idea of like kind of building this AI to human symbiosis digitally it's like Um, figure is going to be like I think figure is going to be like the max ceiling of AGI of like being able to put that out and then there's going to be a version of this in the digital world. It's going to be like a human is going to have this like AI pairing. It's going to have like ultimately maybe your own AI weights, your own memories, maybe your own hardware. It's going like really close. And fundamental to that thesis was like you got to figure out how to get AI to use computers general purpose. You would never hire an assistant that couldn't use a computer. So you got to be able to like give things out to it that can like do everything you can do. financial models, book flights, like order Door Dash, whatever you need to do, you need to be able to do it all autonomously, but only one in a thousand websites have APIs. So, in most, you know, glo like most computer use globally is on the on the internet and browser. My my main inclination within two or three years, you'd have a system that you'd be able to talk to and say, "Go do this or do that." And to be able to like go off go online and like maybe maybe like use the internet really well like almost like a robot would where you can like move the mouse and use the keyboard. That's what you have to do to solve like general purposeness for around a computer is you you can't rely on API or MCP. You have to figure out how to like navigate like a human can. Now at Hark, we've like we just released our first uh kind of model and research preview um last week. It's really hard for us to find now something that we tell it to go do on the internet. It can't do.

你们和其他人有什么不同？因为每个人都在尝试做计算机使用，对吧？我觉得马斯克有 macro hard，ChatGPT 也有他们的计算机使用功能。大家都在做。你们觉得你们破解了什么。你们做了什么不同的事情？

What did you guys do differently than the other because everyone's trying to do computer use, right? So like I think Elon's got macro hard and chatd had their computer use thing. Everybody's doing it. You guys feel like you've cracked something. What did you guys do differently?

好的，我们做了一些不同的事情。首先，每个人都在尝试使用 API 和 MCP。比如 Open Claw 之所以那么好，是因为它只能……它不能用浏览器。它不能端到端地使用 DoorDash，因为 DoorDash 没有消费者 API。所以我们试图弄清楚如何使用屏幕，其中一点是我们为每个智能体启动一个虚拟计算机。所以他们不需要 MacBook 或其他东西。你可以在沙盒中启动任意数量的这种环境，然后你需要给它能力去看屏幕、移动光标和使用键盘。

Okay, there's a couple things we did a little differently. First is like everybody's tackling this from like using APIs and MCPs. Like the reason Open Claw got so great, it was like it could only it couldn't use the brows. It couldn't like go on and use Door Dash end to end because Door Dash has no consumer API. So we tried to figure out how to use like a how to look at a screen and one is we spin up a virtual computer for every agent. So they don't need like a MacBook or anything. So you can just spin up as many of these environments as you want in the sandboxes and uh and then you need to give it ability to like look at a screen and use like move the move the cursor and use the keyboard.

是的，但我用过 ChatGPT 的计算机使用功能，它也能做到那些。

Yeah, but I used chatp's computer use and it was doing that.

所以这就是我说的。你们做了什么让它工作得很好？是算法上的突破吗？

So that's what I'm saying. What did you guys do to make it work well? Was it like an algorithmic breakthrough?

这在于我们的后训练。我们有一个强化学习过程，我们认为可能世界上没有其他人做过。

It was in our post training. We have a reinforcement learning process that we think is maybe nobody else in the world has done.

好吧，让我们了解一下背景。好的。所以 Figure 非常容易理解。人形机器人，如果成功的话，这个业务将会非常庞大。如果你能破解密码，我想你说过需求是无限的。Hark，我不太明白那是什么。你能像我这么笨的人解释一下吗？

Well, let's get some context behind this. Okay. So Figure is shockingly easy to understand. Humanoid robots and that business is going to be massive if it works. If you can crack the code, I think you said there's unbounded demand. Hark, I don't entirely understand what that is. Can you kind of explain like I'm an idiot?

嗯，我的意思是，那基本上就是我们刚开始时所有的东西。所以，好的，什么是……

Well, I mean that's kind of all we had at the time we started. So, okay, what is...

我认为成功的最佳方式是看看别人是怎么做到的。无论你是要模仿他们，还是仅仅将其作为灵感，因为这样你就知道什么是可能的。所以从 24 岁开始，我坚持不懈地做这件事，而且非常有条理。我创建了一个电子表格，追踪了大约 50 位非常成功的人士，我查看了他们的出生年份、开始学徒生涯的年份、开始做第一件让他们成功的事情的年份，以及最终取得突破的年份。我把这些数据连同他们作为学徒做了什么、最终如何突破的故事汇总在一起，放进了数据库。

I think the best way to become successful is to see how other people did it. Whether you're going to copy them or just use it as inspiration because then now you know what's possible. So starting at the age of 24, I did this relentlessly and I was very methodical about it and I created a spreadsheet where I tracked roughly 50 people who were uber successful and I looked at the year that they were born, the year that they started their apprenticeship, and then the year that they started, the first thing that made them successful, finally the year that they broke through. And I aggregated all this data along with the stories of what they did to be an apprentice and what they did to finally break through. And I put it together in a database.

而 HubSpot 找到了这个东西，坦白说我自己都忘了，但它确实改变了我的生活，他们让它重新浮出水面。他们把它做得更好，并放到了一个你现在可以免费下载的东西里。所以，如果你点击描述中的链接或点击这里的二维码，你可以看到我 24 岁时做的这个数据库，它改变了我的生活。所以，如果你想要成功，或者你已经成功但想要更多灵感，可以去看看。

And HubSpot went and found this thing that I frankly even forgot about, but it did change my life and they resurfaced it. They made it even better and they put it into a thing that you can download for free right now. So, if you click the link in the description or click the QR code right here, you can see this database that I made when I was 24 and it changed my life. And so, if you're looking to become successful or you're already successful and just want some more inspiration, check it out.

我坚信 AI 将朝着两个方向发展，然后在某个时刻，也许甚至会合。第一个方向是，我们将拥有在物理世界中行动的 AI，它会为你做环境中的一切事情，比如洗衣、洗碗、做饭，运行供应链，并参与医疗保健。其载体是人形机器人。它只是一个人形，它会走出去，做你想做的事。你想要一个硬件，你知道，这个硬件能够做所有事情，你把智能 AI 放进去，它就会去完成世界上的一切。这就是 Figure 正在做的事情。

I strongly believe like AI will head in two directions like uh like and then at some point maybe even like maybe like head together like the first is we'll have AI out in the physical world that will like do everything in the in in the environment for you like laundry, dishes, cooking like run the supply chain and be in healthcare. The vessel for that is a humanoid robot. It's just a human form and it will just go out and do like you like want one piece of hardware that can like you know the hardware is capable of doing everything and you put like smart AI into it and it'll go off and do everything in the world. That's what Figure is working on.

除此之外，还会形成一种非常紧密的数字 AI 与人类的共生关系。你将拥有一个非常特别的东西，你可以和它交谈，它随时随地陪伴你，它会知道你的一切，访问你所有的记忆、账户和系统，并能实际为你做事，就像一个超人的助手。它可能最接近的是钢铁侠中的贾维斯，它几乎在各个方面都超乎常人。它会了解你生活的方方面面。你可以在任何需要的时候随时访问它。它会一直在后台帮助你。如果你在航班上，比如长时间中转，或者短时间中转但错过了，它已经为你准备好了备用计划，帮你解决这个问题。它就像是你随身携带的东西。

Separately than that, there's going to be this like really close like digital like AI to human symbiosis that forms. You're going to have like this very special thing that you can like talk to that's with you everywhere you go that will know all your stuff, have access to all your memories, have access to all your accounts and systems and be able to actually go do things for like a superhuman assistant. It'll be like um maybe the closest thing is like Jarvis from Iron Man and it will be able to do like it'll be like super human in almost every way. It'll know everything about your life. You'll be able to access it at any moment whenever you need it. It'll be in the background helping you out at all times. If you're on a like if you're on a flight with like a long layover or flight with like maybe say a short short layover and you miss it, it'll like already have backup plans already help you like figure that out. Like it'll just be something with you everywhere you go.

我们还没有那个。我们有非常好的编码智能体，我们有非常好的聊天机器人，但我们还没有一个能像我的贾维斯那样行动的东西。为了达到那个目标，我们需要在模型方面努力。它必须不仅仅是文本聊天。它必须能够使用计算机，拥有几乎完美的记忆，能够像人类一样与你来回交谈。而且，系统中必须有视觉。你必须能够观察世界并理解你所看到的东西。

We don't have that. We have like really good coding agents. We have really good chat bots, but we don't have like something that can go off and like be my Jarvis. In order to get there, we need to work on the like model side. It's got to be just better than text chat. It's got to be able to use computers, have like basically near perfect memory, be able to talk to you like just like a human would back and forth. And uh we have to have vision in the system. You have to be like look at the world and understand what you're seeing with it.

其次，我认为你需要修复 AI 的界面。你这边有 AI，那边有人类，中间有一个旧的硬件系统，比如 MacBook 或 iPhone。它们是 20 年前设计的，对 AI 来说完全是垃圾，不是正确的界面。所以，我们走出去，正在设计我们认为 AI 时代 iPhone 之后的东西。这就像一个升级周期。我们在初创公司中经常看到这种情况。你们也看到了，对吧？我们正处于电脑和手机的升级周期中。它们将会消失，会有新的出现。它们都将是 AI 电脑、手机和系统，它们会很棒。它们都将是实时的，你可以随时访问它们，它们总是理解正在发生的事情，总是能够引用正在发生的事情。你将能够抽象掉大多数应用程序。你可能不会有应用商店，或者你可能有一个 AI 操作系统。它对你来说将是完美的。你最终将在自己的设备上拥有自己的权重，随身携带。这将是一个非常好的配对。

And I think secondly, you need to have um you need to fix the the interface to AI. You have like um AI over here and a human and you have like an old hardware system in between like a call like a MacBook or iPhone. They were designed 20 years ago. They're complete rubbish for AI. They're not the right interface. So, we went out and we are out there designing what we think comes like after the iPhone for AI. And it's like an upgrade cycle. We see this all the time in startups. You guys see it, right? Like we're we're in an upgrade cycle with the computers and phones. They're just going to go away. They're going to be a new ones. They're going to be all AI computers and phones and systems and they're going to be great. They're going to be all real time. You can always access them and you want they'll always be like understanding what's happening. They'll always be able to reference things what's going on. You'll be able to abstract away most apps. You'll probably not have an app store. or you probably have an AI operating system. Uh it'll be perfect for you. You'll ultimately have your own weights on your own devices that you'll own and have with you everywhere you go. It'll be like a really great uh pairing.

我们雇佣了一支不可思议的团队。团队现在大概有 80 或 90 人。领导硬件设计的那个家伙，ABS，之前为最近几代 iPhone、MacBook、MacBook Pro 做过设计。他真是个牛人，很棒。所以，我们正在设计我们认为的下一代 AI 设备，它们将取代手机和电脑。然后，我们也在设计下一代 AI 模型。模型需要变得更加多模态，需要更有表现力。文本编码对我们来说不足以真正获得 AGI 的感觉。所以，我们正在努力。

And we hired an incredible team. Teams like you know maybe like 80 or 90 now. Uh the guy that leads uh hardware design ABS previously designed for last several generations of iPhone, MacBook, MacBook Pro. Like he's just like the he's a stud. He's great. So, we're designing what we think are the next generation of AI devices that will kill the phone and computer. And then, uh, we're designing the next generation of AI models. The models need to get a lot more multi multimodal. They need to get a lot more expressive. Like the text encoding is just not enough for us to like really have a like a like a real AGI feeling with AI. So, we're working on that.

我们推出了第一个 AI 预览，也就是上周发布的计算机使用智能体的首个研究预览。我认为我们在全球一些领先的浏览器计算机使用基准测试中名列前茅。而且它会不断进步，每个月都会变得更好、更聪明地使用计算机，也更快。我们内部还在 AI 方面研发其他几种不同类型的技术。大约一个月后，我们将推出在传统浏览器、iPhone 和 Android 上使用 HARK 的功能。所以你将能开始使用它，然后我们还会推出硬件。我们现在正在研发中。我们实验室里已经有硬件在测试使用，简直疯狂，就像科幻电影里的硬件。

We have our first AI preview, our first research preview of a computer-using agent that we came out with last week. I think we were top on some of the leading browser computer use benchmarks in the world. And it'll keep getting better. This will keep getting better and better. Every month it'll be better and smarter at using a computer and faster. We're working on a couple of other different types of technologies internally on the AI side. And then we'll launch the ability to use HARK on traditional browser and iPhone and Android in about a month. So you'll be able to start using it, and then we'll have hardware coming. We're working on it now. We actually have hardware in the lab now that we're using and testing, and it's crazy. The stuff is like a sci-fi movie hardware.

你觉得这些设备会是什么样子？人们一直在猜测，因为 Jony Ive 的工作室被 OpenAI 收购了，你看到过那个小圆盘和耳环的视频。我不知道那是真的还是假的。又有一个超级碗的泄露广告。那是真的还是假的？是怎么回事？你觉得这些设备最终会是什么样子？是手表、眼镜，还是完全不同的东西？

What do you think those devices look like? People have been speculating because Jony Ive's shop got acquired by OpenAI, and you've seen the videos of the puck, and then there's like an earring. I don't know if that's real or fake. There was a leaked commercial for the Super Bowl again. Is that real or fake? What's the story? And what do you think these devices end up looking like? Are these watches, glasses, something else altogether?

我觉得在过去一年左右，我对这个问题的看法改变了很多。但我们内部有一个非常坚定的观点。我们的观点是，中间位置是那些可能在全球达到每年十亿台销量的设备。目前世界上只有电脑和手机能达到这种超级设备级别。然后周围还有一些辅助设备，像 AirPods、手表之类的，它们每年卖不到十亿台，只占苹果收入的 3% 左右，但它们作为平台帮助了整个生态系统。我们在 HARK 关心的是解决中间那个大块头。要解决这个问题，就必须取代电脑和手机，没有别的办法。所以你必须重新打造一台更好的新电脑或新手机，端到端地替代现有系统。然后周围的设备，我们在 HARK 也会有一系列设备，虽然不是每年十亿台，但对生态系统很重要。

I think I've really changed my mind on this a lot in the last year or so. But we have a really strong opinion here internally. Our opinion is that what sits in the middle are devices that could possibly reach a billion units a year in the world. The only kinds of things we have like that in the world right now are computers and phones that kind of meet that mega-device category. And then you have things on the ancillary around it, like orbiting this big thing, like AirPods and a watch or things like this, that don't sell a billion units a year. They're like 3% of Apple's revenue, and they help the ecosystem as a platform. What we care about at HARK is trying to solve what's in the big middle piece. To solve that, you have to take down the computer and the phone. There's no way around that. So you have to rebuild a new computer or new phone that's better and replaces your existing systems end to end. And then what's around there are things that we will even have at HARK that are a family of devices that are not a billion units a year but important for the ecosystem.

我的理解是，你是在说下一个设备可能就是一个 AI 原生的手机，对吧？你不会试图改变外形。

My understanding is you're kind of saying the next device might be like a phone that is just an AI-native first phone, right? You're not going to try to change the form factor.

不，我根本不是这个意思。你需要彻底重新思考一切。我们现在实验室里的第一版硬件是我这辈子从未见过的。

No, I'm not saying that at all. You're going to want to radically rethink everything. The first version of hardware we have now in our lab is unlike anything I've ever seen in my whole life.

好的。

Okay.

外围的东西是眼镜、吊坠、可穿戴设备之类的。它们不是主角。事实上，Meta 眼镜可能是我买过的最差的产品之一。它们太糟糕了。我甚至不知道怎么用。它没有自己的网络，依赖 iPhone 的网络，这意味着你的手机上的应用必须打开。配对时间很长，不好用。我想不出任何理由需要把这东西戴在头上 14 个小时。这是错误的设备。最终形态是脑机接口，在那之前我们会有 10 年的 AI 语言设备。这才是路径，不是眼镜。我甚至不知道眼镜能不能进我们设备的前十名。

What lives outside of here on the edge are glasses and pendants and wearables and things. They're not the main show. In fact, the Meta glasses are probably one of the worst products I've ever bought. They're horrible. I can't even figure out how to use it. It doesn't have its own network. It piggybacks on the iPhone network, which means your app needs to be open on your phone. The pairing is long. It doesn't work well. I can't think of any reason why I would need this thing strapped to my head for 14 hours a day. It's the wrong device. The end state is BCI in the brain, and we're going to have AI language devices for the next 10 years before that. That's the path, and it's not glasses. I don't even know if glasses will make our top 10 list of devices.

当你和你的团队头脑风暴时，你们有没有一个框架来跳出既有规范思考？因为我完全无法想象你在说什么。

When you and your team are brainstorming, do you have a framework on how to think outside of pre-existing norms? Because I literally can't imagine at all what you're talking about.

让我们深入到底层。首先，变化在于我们有了一种新型计算机。我认为 AI 是一种新型计算机，一种新的自动化。这种自动化能做几件事。当我们设计时，我们希望围绕能带来 10 倍提升的关键原则来设计。如果只比你的手机或电脑好一两倍，你不会用它。它必须是 10 倍好。深度学习现在带来了哪些 10 倍好的东西？有几个。一是 AI 现在基本上可以像人一样为你思考和操作计算机和系统。它可以和你对话。它有视觉理解能力。它有实时的语音到语音转换。它可以像人一样快，甚至接近人速地为你操作计算机和系统。随着时间的推移，它会和人类一样好，甚至在成功率上更快。所以你有一个能力几乎像人一样的系统。它还可以有记忆，意味着你可以把记忆放进去。它不会忘记任何东西，长期来看近乎完美。所以你有一个几乎像装在盒子里的真人一样的系统，拥有真人所有的功能。这几乎就像你走到哪里都能带一个小人，肩上扛着电脑，那太疯狂了。就像只有 Sam 能拥有，只有 Sam 能看到，只有 Sam 能和它说话，它只帮助 Sam。那就是你的全部生活。而且它会越来越聪明，有完美的记忆，能操作电脑，能和你说话，能看见。你会想，天哪，那东西能像你一样在电脑上做任何事。

Let's get down to the substrate level. First order, what has changed is we have a new type of computer. I think of AI as a new type of computer, a new type of automation that's here. That automation can do a few things. When we're designing this, we want to design around key principles that could be 10x better. If it's one or two times better than your phone or computer, you're not going to use it. It's literally 10x better. What are things now that deep learning brings that are 10x better? There are a few. One is AI can basically now think and use computers and systems for you just like a human can. It can talk to you. It has visual understanding. It has real-time speech-to-speech. It can use computers and systems for you as fast as a human can, or close to it. Over time it'll be just as good as a human and faster in terms of success rate. So you have a system that's almost humanlike in capabilities. It also can have memory, meaning you can put memory into it. It won't forget anything, near perfect over time. So you have a system that's almost like a human in a box that has all the same affordances a human has. It's almost like if you could bring a little human around with a computer on your shoulder everywhere you went, that'd be insane. It would be like it was only for Sam. Only Sam could see it. Only Sam could talk to it, and it was only there to help Sam. And that was your whole life. And it's going to be smarter and better along the way, had perfect memory, could use computers, talk to you, and see. You'd be like, damn, that thing would be able to do anything you do on a computer.

好的，所以你和团队的第一步就是，让我们摆脱所有约束。如果我们肩上有个小东西，是 AI，无所不知，能看见和听到我们看到和听到的一切，然后给我们建议，那会是最酷最神奇的事情吗？

Okay, so your first step with your team is like, let's just get rid of any constraint ever. What would be the coolest magical thing if we had a little guy on our shoulder that was AI, all-knowing, and could see and hear everything we see and hear, and then give advice to us?

比如，什么会从根本上重塑这一切？

Like, what is the thing that's going to fundamentally reshape all this?

好的。

Okay.

然后从那里出发，我们必须围绕这个系统来设计。这里的竞争优势在于它能力像人，而且有近乎完美的记忆，可以随时间回溯和参考。我的手机没有这个。

And then from there, we've got to design around that system. The competitive advantages here are that it is humanlike in capabilities and it has almost near-perfect memory, can go back and reference over time. My phone doesn't have that.

比如上周我在手机上存一个联系人，当时忙着输号码，结果一天后有人问我：“你给那个人打电话了吗？”我说：“我连他名字都不知道，忘了，连问手机都问不了。”这系统真是蠢透了。然后我每天像个猴子一样在手机上点 DoorDash。现在用 Harkc，我什么都不用做了，它端到端全给我搞定。上班路上，我只要说“给我点杯咖啡”，它就办好了，全程后台自动完成，我啥都不用碰，全被抽象掉了。就像你身边随时跟着一个小助理，你只要说——它甚至能预判：“Brett，你今天想喝咖啡吧？”然后你说：“嗯，对，来一杯，不过今天换成双份浓缩，送到 Hark 办公室而不是 Figure 那边。”我就说：“搞定，交给我。”而不是像猴子一样在手机上折腾三分钟去 DoorDash 结账。

Like I put a contact on my phone like last week and I was busy when I was putting the phone number in, and like a day later somebody's like, 'Hey, did you call that person?' I'm like, 'I don't even know the name. I forgot. I can't even ask my phone.' It's just so stupid. The whole system is. And then I go in there like order DoorDash like a monkey every day now. I'm pushing things. I don't do any of that now with Harkc; it does it end to end for me. On my drive to work, I just say, 'Order me coffee,' and it's just done. It does it all for me in the background. I don't have to touch anything. It's all abstracted away. It's like if you had that little human with you everywhere you go, you would just say—you would even predict, probably, 'Brett, you want coffee today?' and be like, 'Ah yeah, I do. Let's order. But you know what, make it a double shot today,' and route it to the Hark office instead of Figure. I would just say, 'Done, I got it, let me take care of it.' Instead of thinking like a monkey on my phone for the next three minutes trying to do checkout on DoorDash.

这几乎就像手机是个工具——就像一把锤子，对吧？你想让锤子干点实事，就得自己拿起来挥。而下一代产品基本上就像你身边随时有个杂工。你只要跟他说：“嘿，能修下那扇窗户吗？”他就去修了。你根本不用拿起锤子去琢磨怎么用。

It's almost like the phone is a tool—it's like a hammer, right? If you want the hammer to do anything functional, you have to pick it up and start swinging it. Whereas the next generation is basically like having a handyman next to you at all times. So you just tell them, 'Hey, can you fix that window?' Just go fix the window. You don't have to pick up the hammer and start figuring out how to use it.

从那儿开始，然后就得快速原型迭代。所以你过来的时候，我们已经把能想到的设计全做出来了，都 3D 打印好了。

Start there. And then from there, you got to rapidly prototype. So when you come over, we've designed everything you could possibly think of. We 3D printed it.

有哪些设计没成功但还挺酷的？

What were the designs that didn't work but were kind of cool?

问题是，我们现在在造很多不同的设备，覆盖了相当大的范围。我们设计了一些相当疯狂的东西。所以不是说你一看就觉得“这像那个，那个在这儿好用”，没那么容易画等号，相当激进。我们全都快速原型化了。我们有个制造车间专门干这个，还有个完整的设计工作室。接下来几周、几个月，我会随身带着、戴着，或者怎么着都行。最后我们做筛选。之前有位全球最大的电信公司 CEO 来过，他当年跟史蒂夫·乔布斯一起做过 iPhone 1，两周前刚来过。他刚从苹果那边见完蒂姆·库克过来——你知道库克快从苹果退了——他看了我们的东西，就说：“天哪，这是我第一次见到有人真有可能干掉那些巨头。”

The thing is, we're building many different devices now that cover a pretty wide area of this. We have some pretty crazy stuff that we were designing. So it's not like you look at it and think, 'That looks like this and it does well over here.' It's not as easy as drawing those parallels. It's pretty radical. We rapidly prototype all this. We have a fabrication facility that does this stuff. We have a whole design studio where we work on this. Over the coming weeks and months, I'll either carry it around with me, wear it, whatever. We end up doing it and we'll do a down selection. We had one of the biggest telecom CEOs in the world here—he actually helped with the work with Steve Jobs on iPhone 1—and he was here two weeks ago. He just came from meeting Tim Cook. You know, Tim Cook's on his way out at Apple, but he was over there at Apple and came over here. He saw our stuff and he's just like, 'Holy man, this is the first time I've ever seen anybody that could possibly take out the big guys.'

那么，是不是可以说，对于 Archer、Figure 和 Harkc 来说，难题似乎是“我能大规模量产吗？”

Well, is it true to say that with like Archer, Figure, and Harkc, the hard problem seems like, 'Can I just mass-produce this?'

难题不是那个。我们现在认为，真正要解决的最重要约束，是给世界造一个真正智能的机器人系统。现在市面上有一堆机器人可以买。你可以从中国买一些，但它们完全是垃圾，啥也干不了。你只能拿摇杆推着它走，按个按钮它挥挥手。它没有手，只有小凸起。你会想：“我拿这东西干嘛？就是个玩具。”就像我几年前买了个大疆无人机，玩了一会儿，第二天就想：“我拿这东西干嘛？”它难设置，不好用，我撞了好几棵树，根本不行。我就想：“我拿这东西干嘛？”现在的机器人就是这样。我们可以量产一大堆，但如果它们不够聪明，其实没什么用。我们想攻克的是 Figure 真正的人类级智能。我们真的想解决怎么让它能进任何家庭，能干我想让它干的每一件活。这就是我们在做的。我们认为这是我们要解决的最大差距。除此之外，人们有时会把消费电子制造和汽车制造搞混。世界上没有哪家大公司会说：“如果需求这么大，我害怕高速量产这种消费电子——这完全可行。”我的意思是，全世界能造十亿部手机，几乎半手工加一些自动化。但汽车是另一回事，造汽车你会死。我见过宝马是我们的商业客户，我进过宝马和其他一些工厂，早期阶段非常棘手。汽车难的原因是你没法把零件拿在手里。手机你总能拿在手里，换、移动、拿着都行。汽车你物理上做不到。所以你需要机器人真的把零件传给其他机器人，把它们装到底盘上。如果其中任何一个在几千或 800 个机器人里坏了，你就完了，整条线就停了。所以这就像你在造一个巨大的机器人，它在造汽车。而 Figure 呢，你任何零件都能拿在手里。所以我觉得我们介于汽车和消费电子之间——我们更靠近手机这边，大概 40% 的水平。我们上周或上上周刚造出第 1000 台 Figure 3 的 EVT 机器人。

The hard problem is not that. We believe now the most important constraint to really solve is building a really intelligent robot system for the world. There's a bunch of robots you can go buy now. You can buy some from China and they're complete crap. They can't do anything. You can joy-stick it around, that's all. You hit a button and it waves. It's got no hands, it's got nubs. And you're like, 'What do I do with this thing? It's a toy.' It's like early when I bought a DJI drone years ago and I was playing around with it. Then a day later, I was like, 'What do I do with this thing?' It was hard to set up, didn't really work well. I flew it into a bunch of trees. It just didn't work. I was like, 'What am I doing with this thing?' Robots are like that now. We can go manufacture a ton of them, but if they're not really smart, it's not really going to be that helpful. We're trying to crack true human-level intelligence for Figure. We really want to tackle how to make it so I can put it into any home. It can do every job I'd want it to do. That's what we're working on. We think that's the largest gap in the schedule of what we need to solve. Beyond that, people sometimes confuse consumer electronics manufacturing with car manufacturing. No big company in the world would say, 'I'm scared of manufacturing this consumer electronics at high rate if there's so much demand—it's just possible to go do.' I mean, you can make a billion phones almost pseudo by hand in the world with some automation. But cars are a different story. You will die trying to manufacture cars. Having seen BMW as a commercial customer of ours—I've been in BMW and a few other groups—it's gnarly early. The reason why cars are so hard is that you can't hold the part in your hand. Phones you can always hold in your hand and go change or move and hold. Cars you physically can't. So you need robots that literally pass parts to other robots that put things on the chassis. And if any of those break across thousands or 800 robots, you're dead. The whole line's done. So it's like this huge giant robot you're building that's building the car. With Figure, you can hold any part in your hand. So I think we're between cars and consumer electronics—we're over here closer to the 40% level by cell phones. We just made our thousandth EVT robot for Figure 3 last week or the week before.

你说造了一千台——这些是给宝马这样的客户的，还是内部做的原型？这是什么意思？

When you say you made a thousand—are those a thousand that go to customers like BMW, or are you making prototypes internally? What does that mean?

我们有两个大客户。一个是我们自己，作为工程和 AI 研究组织，需要机器人——每个工程师都需要一台，每个实验室都需要，我们要做大量测试。内部有很多工作要做，我们称之为工程车队。第二个是给客户的。所以我们现在两边都在供货。

We have two big customers. We have us as an engineering and AI research org that needs robots—like every engineer needs a robot, every lab needs robots, we need to do tons of testing. There's a lot of work we need to do internally. We call that maybe the engineering fleet. The second one is going to customers. So we're going to both right now.

呃，我们这周实际上已经把机器人发货给了第三个客户。

Uh, we've actually shipped out robots to our third customer this week.

它们到客户那里后做什么？机器人能做什么？目前可能还不能做什么？

When they go to customers, what do they do? What can the robot do? What maybe can't it do at this point?

我们现在做很多物流方面的工作，比如包裹。过去我们也做过制造业的其他事情，主要是制造物流。但我们也在和其他行业的人洽谈。

We do a lot of logistics stuff right now, like packages. We've done other stuff in manufacturing, mostly manufacturing logistics, in the past. But we're also talking to folks about other industries.

那么目前，当它到客户那里，做的是，我不知道，比如包装工作还是什么？是分拣还是撕裂还是做什么？

And at this point, when it goes to a customer and it's doing, I don't know, like packaging work or what? Is that like sorting or tearing or what is it doing?

他们刚做了一个 YouTube 直播，有几十万甚至上百万的观看量，人们看着这个机器人从传送带上分拣包裹。

They just did a live YouTube video, and they had hundreds of thousands, maybe millions of views, of people watching this robot sort packages off of a conveyor belt.

是的，我看到了。那么这是那种工作吗？这能给我一个工作的例子吗？

Yeah, I saw that. So is that the type of job? Would that give me an example of one of the jobs?

这是我们做的工作的一个例子，非常接近。

That's an example of one of the works we do, like a very close one.

那个客户是觉得‘哦，这太棒了，因为我找不到劳动力来做这个，付钱给人类太贵了，这便宜多了’？还是只是觉得‘嘿，看，今天它不一定更快、更便宜或更好，但这是对未来的投资，两年后成本曲线会起作用，它会更快、更便宜，你知道的，随便什么’？

Is that customer like, 'Oh, this is awesome because I can't find the labor to do this, it's too expensive to pay humans, this is way cheaper'? Or is it just like, 'Hey, look, today it's not faster, cheaper, or better necessarily, but it's an investment in the future where two years from now that cost curve is going to work and it will be faster, cheaper, you know, whatever.'

不，不，不。情况是这样的，他们来找我们，说‘我们被劳动力问题折磨得要死。我们的人员流动率非常高。有些地区年流动率超过 100%。找人才非常贵。我们有大量的人才缺口。人才非常昂贵。工资在上涨，我们没有解决办法。我们想不出怎么自动化所有这些工作，我们需要你们来帮我们。’我们有能力在合同中赚很多钱，客户也能获得非常好的投资回报。你得想想，一个机器人可以每天多班次工作，一周工作七天。我们可以有很高的运行时间。你在我们直播的物流线上看到的任务，实际上是某个客户的真实用例。那个任务需要每 3 秒处理一个包裹，最初需要每天做 5 小时，我想是每周 5 天。我们以每包 2.9 秒的速度连续做了 200 小时。所以我们已经达到了人类的速度。我们现在就在这里做这件事。他们已经有了投资回报，我们现在正处于把这些产品交付给客户并扩大规模的早期阶段。随着时间的推移，这将为这些群体带来数十亿的收益。

No, no, no. It's like the pitch is they come to us and they're saying, 'We're dying with labor. We have really high turnover. Some areas have over 100% turnover per year. It's really expensive to find talent. We have a large talent shortfall. The talent's really expensive. Wages are going up, and we don't have a solve for this. We can't figure out how to automate all this work, and we need you to come in and help us.' We have an ability to make a lot of good money in our contracts, and the customers make really good ROI on this. You got to think a robot can do multiple shifts per day, work seven days a week. We can have a lot of uptime. The task you saw on the logistics line that we live-streamed was actually a real use case for one of our customers. That needs to be done at 3 seconds a package, and it initially needs to be done five hours a day, I think it's like 5 days a week. We did that 200 hours straight at 2.9 seconds a package. So we're already at human speeds. We're already doing this here now. They're already having ROI, and we're now in the early stages of getting these out to these customers and scaling it up. Over time, it will just put billions out to these groups.

你能帮我区分一下事实和虚构吗？因为奇怪的一点是，作为一个爱好者或外行人，对这个未来感到兴奋，你实际上没法测试它，对吧？测试它非常贵或很难。所以我会看到一个中国机器人，如果我想买这个机器人，要 2 万美元。我真的不知道它能做什么。我看到埃隆会出去说，‘我们明年要造一百万个这种东西。我们要把它们发出去。’然后你看到 1X，他们展示他们的手，他们说，‘看看我们的手。这是你见过的最好的手。’然后旧金山有一种服务，他们会派一个机器人去打扫你的公寓，他们说，‘是的，今天就能用。’所以你能帮我区分事实和虚构吗？这似乎真的很难，相比大多数类别，我可以快速在线试用产品，或者买下来测试它们。

Can you help me with the kind of truth versus fiction? Because one of the weird things is, as an enthusiast or a lay person who's excited about this future, you can't really test it, right? It's really expensive or hard to test. So I'll see a Chinese robot and it's 20 grand if I want to buy this robot. I have no idea really what it can do. I see Elon will go out there and say, 'We're going to build a million of these things in the next year. We're going to ship them.' Then you get 1X and they're showing their hand and they're like, 'Look at our hand. This is the best hand you've ever seen.' And then there's this service in San Francisco where they'll send a robot in to clean your apartment and they're like, 'Yeah, that works today.' So can you help me separate fact from fiction? It seems really hard compared to most categories where I can just try the products quickly online or buy them and test them out.

一是市场中的噪音相对于信号来说，就像你提到的，完全失控了。市场上东西太多了。真的很难分辨到底发生了什么。所以让我总结一下我认为最重要的，然后倒推。

One is the amount of noise in the market for a signal is just out of control, like you mentioned. There's just so much out there in the market. It's really hard to tell what the hell's going on. So let me summarize what I think is the most important and work backwards.

好的。

Okay.

我认为最重要的事情是能够在有用的工作环境中大规模自主部署机器人，比如它们能做晚餐、洗碗、铺床、端到端运行供应链、在医疗保健领域工作、建造建筑、做物流，诸如此类。这些事情从根本上需要机载 AI，你可以运行它来做自主工作。你不能用代码解决它；你需要自主地做。你需要长时间地做，而且你可能需要移动，用手拿东西，并在世界中移动东西。你明白我的意思吗？它真的需要经济地做事，比如移动电子。所以我认为在高层次上，我们关心的不是做后空翻、跑最快英里、跳舞、游行或在树林里跑步的最好的机器人。我们不关心那些。

What I think the most important thing to do is to be able to ship robots autonomously at scale in useful work environments, like they can cook your dinner, clean your dishes, make your bed, run the supply chain end to end, work in healthcare, build a building, do logistics, that sort of stuff. That stuff requires fundamentally onboard AI you can run so you can do autonomous work. You can't solve it with code; you need to do it autonomously. You need to do it over long periods of time, and you probably need to move around and use something in your hands and move stuff through the world. You know what I mean? It's really got to do stuff economically, like move electrons around. So I think at a high level, what we care about is not the best robot that's doing backflips and running the fastest mile or dancing or in a parade or running outside in the woods. We don't care about that stuff.

老兄，我等不及看到 Figure 机器人在宝马工厂抽烟休息的样子。就像，我高中时本可以很出色，但我搞砸了。现在我在宝马工厂工作。

Dude, I can't wait till I see a Figure like on a smoke break at the BMW factory. Like, I could have been a great back in high school, but I blew it. Now I'm working at BMW factory.

我以前跟你开过玩笑，我说，你从 Veter 开始，那只是个招聘的事情。现在你在做这些改变世界的事情，你说，‘嗯，Veter 实际上是改变世界的，原因如下。’然后你做了这个推销。非常好。你很擅长推销。你很擅长筹集资金。你很擅长有魅力并说服别人。当你精心设计一个推销，来招募并说服人们改变他们的生活，连根拔起他们的生活，信任你，来建立一家公司，你是怎么设计那个推销的？你的一些公司的推销是什么？

I've made jokes with you before where I was like, you started with Veter, which was just like a job recruitment thing. Now you're on these world-changing things, and you were like, 'Well, Veter actually is world-changing, and here's why.' And you gave this pitch. It was very good. You're very good at pitching. You're very good at raising money. You're very good at being charismatic and convincing people of stuff. When you're crafting a pitch to recruit and convince people to change their lives, to uproot their lives, and to trust in you and to come and build a company, how do you craft that pitch? And what was that pitch for some of your companies?

我的意思是，最重要的是，这些都在网上。Figure 的总体规划在网站上有。Archer 的也挂了很长时间。我发过帖子。我认为内心深处我真的很想找到真正关心和痴迷的人。我认为我的大部分时间不是——我知道你想知道推销——我的大部分时间是在试图找到那些人。我发现即使在湾区，那里可能是世界上 AI 和工程人才最丰富的地方，这里 90% 的人都不擅长他们的工作。

I mean, most of all, these are online. The Figure master plan is on the internet on the site. Archer's was up for a long time. I posted about it. I think deep down I really want to find folks that really care and are obsessed. I think most of my time is not—I know you want to know about the pitch—most of my time is trying to find those folks. I found that even in the Bay Area, where it's probably the richest AI and engineering folks in the world, 90% of everybody out here is not good at their jobs.

你怎么分辨谁好谁不好？

How do you tell who's good and who's not?

我技术性地评估他们。

I technically assess them.

所有人。

All of them.

是的。

Yeah.

要做到这一点，是不是意味着你需要和他们一样好或更好，才能评估一个人？

To do that, does that mean you need to be as good or better than them technically to be able to assess somebody?

我需要知道某些指导原则。例如，我需要知道是你做了工作，还是你看着别人做工作。

I need to know certain guiding principles. For instance, I need to know if you did the work or if you watch somebody do the work.

如果你真正做过这件事，它就像你身上的一道伤疤，刻在你身上。你了解所有细节，可以自如地谈论，不需要思考。你会懂得如何逆向工程你做过的一切并讨论它。没做过的人做不到。他们只触及一层就会立刻崩溃，无法谈论，也不知道为什么。

If you've done the work, it's like a scar you carry with you. It's dug into you. You know all the details. You can talk about it freely. You don't need to think. You'll understand how to reverse engineer everything you've done and discuss it. The folks that haven't done it can't do that. They get one layer and they instantly blow up. They can't talk about it. They don't know why.

在一百个听起来不错、简历好看、招聘人员也认为不错的候选人中，你觉得有多少真正能达到这个标准？

Out of a hundred candidates who sound good, whose resumes look good, and the recruiter thinks they're good, how many would you say actually hit that bar?

我给你举个例子。在 Figure，要成为一名机械工程师，我们要经历一个非常有挑战性的流程。你必须能够从零开始构建执行器。系统里有轴承、电机、齿轮箱和其他传感器。它非常紧凑，非常难做，要求也很高。我们六个月来每周做 10 个案例研究，但一个人都没招到。

I'll give you an example. We have a really challenging process to become a mechanical engineer here at Figure. You have to be able to build actuators from scratch. There are bearings, motors, a gearbox, and other sensors inside the system. It's very compact. It's a very difficult thing to do, with really hard requirements. We've been doing 10 case studies a week for six months and have not hired anybody.

这太疯狂了。

That's insane.

确实疯狂。

It's insane.

但当你确实找到合格的人选时，他们的竞争性 offer 来自比你更大或更流动的公司，而且这些 offer 我觉得每年有数千万美元，对吧？AI 领域肯定如此。这主要是 Meta 推动的。在 Hark，我从未见过——我一年前以为 Meta 在付这些人钱，而且这种情况会消失。但他们从未停止。

But when you do get someone qualified, their competing offers are from companies that are larger or more liquid than you, and the offers are, I think, tens of millions of dollars a year, right? The AI side is certainly like that. It's mostly driven by Meta. At Hark, I've never seen—I thought maybe Meta was paying these people like a year ago and it would go away. They've not stopped.

那你听过什么疯狂的故事吗？

So what's a crazy story you've heard?

我想我们给一个非常资深的人发了 offer，他来自 XDI。XDI 完全崩了——大约 6 个月前所有人都离开了。就像宏观硬件被完全解散，发生了很多事情。我们面试了一个 AI 基础设施方面相当资深的人。面试很顺利。我给了他一份很好的 Hark A 轮股票 package。

I think we gave an offer to somebody really senior who was coming from XDI. XDI completely blew up—everybody just left about 6 months ago. It was like macro hard got fully disbanded, a bunch of stuff happened. We interviewed a pretty senior guy on the AI infra side. It was great. I gave him a really good package of Series A stock at Hark.

那是，我不知道，1500 万到 2000 万美元的股票？

And it was, I don't know, 15, 20 million dollars of stock?

分四年。

Over four years.

分四年。

Over four years.

我们早期公司是分五年，后来才过渡到四年。我们现在还是五年。我说，我觉得我们很快就能让 Hark 涨 10 倍。所以我告诉他，你有 1500 万到 2000 万。涨 10 倍，你就有几亿。再涨 10 倍，你就有几十亿。我觉得我们能行。显然这很难，但我觉得我们能行。他收到了 Meta 的 offer，四年 3600 万美元的 RSU。他说这基本上是保证的现金。他必须权衡，在 Hark 可能拿到 2 亿美元，或者在 Meta 确定拿到 3600 万。他离开了，去了 Meta。他们对我们接触的每个候选人都是这样——开出一些荒谬的 offer。他们从未停止。他们已经这样做了大约一年，购买人才。他们一直在用钱买进 AI 竞赛。

We do five for my companies in the early days, and we transition to four a little bit later. We're still at five. I was like, I think we can 10x Hark pretty quick. So I told him, you have 15 to 20 million. 10x that, you have a few hundred million. 10x one more time, you have a few billion dollars. And I think we can do it. Obviously it's going to be hard, but I think we can do it. He got an offer to go to Meta for 36 million over four years of RSUs. And he said it's kind of guaranteed cash. He had to weigh maybe $200 million at Hark or $36 million for sure at Meta. He left and went to Meta. They've been doing that with every candidate we speak to—making some absurd offer. They just haven't stopped. They've been at it for like a year, buying talent. They've been buying their way into the AI race.

你怎么看这个策略？即使你有点讨厌它，你尊重它吗？还是你觉得这是徒劳？你怎么看？

What do you think of that strategy? Even if you kind of hate it, do you respect it? Or do you think it's a fool's errand? What do you think?

我真的很喜欢这个策略。在 AI 领域，我发现真正理解如何做语言预训练、中期训练和后训练——尤其是预训练——以及超算、数据、评估等基础设施，并知道如何正确设置这些的人，以及真正理解 Transformer 在哪些配方下表现良好的人，真的很难找到。实际上很难找到真正懂行的人。我粗略估算，加州大概有 20 到 30 个人知道如何构建真正好的 AI 模型。

I really like it. In the AI space, I've found that the folks who really understand how to do language pre-training, mid-training, and post-training—especially pre-training—and the infra around supercomputing, data, evals, and all the right stuff you need to put in place to do that right, and the amount of folks who really understand the right kind of recipes that Transformers do well in, is really hard to find. It's actually really hard to find the actual folks who know what they're doing. My rough back-of-the-envelope calculus is probably 20 to 30 people in California know how to build really good AI models.

等等，那这种情况会向下传导吗？你说有一个资深的人，但那些不那么资深的，二十多岁、三十出头的人，他们还能拿到八位数的年薪吗？

Wait, so is that trickling down? You said there was a senior guy, but are even some of the less senior, the 20-somethings, the young 30-somethings, still getting eight figures a year?

不，二十多岁或二十多岁的初级员工总共能赚几百万。他们的基本工资是 20 万到 25 万，另外每年还有大约 100 万的 RSU。所以他们每年支付的薪酬范围在 75 万到 200 万之间。

No, the junior guys in their 20s or late 20s are making a few million total. They're making 200 to 250 in base, and another million or so in RSUs every year. So they're paying in the range of 750,000 to 2 million per year.

这是由 Meta 推动的，但其他所有实验室都跟随了薪酬。

And that's been driven up by Meta, but then all the other labs have followed comp.

当我问你对此怎么看时，你说你喜欢。你是在讽刺，还是说实际上这很聪明，考虑到获得这种人才有多难？

When I asked you what you think of that, you said you like it. Were you being sarcastic, or are you saying no, actually that is smart given how hard it is to get this talent?

我认为这非常聪明，如果我是马克，我也会做同样的事情。我会用钱买进竞赛，我认为他现在正在这样做。我不认为我会那样做。我想理解它，我想第一性地找到真正关心这件事的合适人才，而不是雇佣雇佣兵。

I think it was really smart, and I would have done the same thing if I were Mark. I would have bought my way into the race, and I think he's doing that now. I don't think I would have done that. I want to understand it, and I want to first-order find the right folks who really care deeply about this, not hire mercenaries.

所以他雇佣了一群雇佣兵。他们纯粹是金钱驱动的。没有人想去 Meta；他们去那里是因为他们坐在那里就能得到有保证的 RSU 包。发生的事情是，你不需要一千人或 500 或 300 人来设计 AI 模型。你需要一个真正优秀的 20、30 或 40 人团队。而且你不需要这样做就能达到那个目标。那些人可能会更深入地关心使命和你的处境，比纯粹用钱砸问题更投入。但我认为如果我是他，这是一个非常好的策略，而且正在奏效。致敬。他们在招聘工作和如何构建这些东西方面执行得非常好。这对他们来说正在得到回报。他们能否真正推出真正的产品，还没有定论。我认为我对这些群体的问题是，他们传统上无法做新的事情。

And so he hired a bunch of mercenaries. They're purely money-driven. Nobody wants to go to Meta; they're going there because they're getting paid a guaranteed RSU package by sitting around. What's happening is you don't need a thousand people or 500 or 300 to design AI models. You need a really good team of 20 or 30 or 40 people. And you can get there without doing this. And those people probably would care more deeply about the mission and where you're at and be more committed than if you purely throw money at the problem. But I think if I were him, it was a really good strategy and it's working. Hats off. Really good execution in their recruiting efforts and how they're structuring this stuff. It's paying off for them. The jury's still out if they can actually ship real products. I think the problem I have with those groups is they've traditionally not been able to do things new.

嗯，我是说我觉得 Facebook，可能 Meta 会作为史上最伟大的收购者之一被铭记，比如收购 Instagram、WhatsApp，他们用各种方式买进了这些领域。但是，你看 Ray-Ban 眼镜和他们做的其他东西，就是不太行。所以我觉得真正的问题是，怎么在这里做出真正伟大的工作？我觉得就像我们现在聊的，我们正在用 heart 系统，它太好了，比我今天用的任何东西都好。

Well, I mean I think Facebook is probably Meta is going to go down like one of the greatest acquirers in all time with like you know Instagram and WhatsApp and different way they've like bought their way into those spaces. Uh but like you know if you look at like the Ray-Bans and everything they're doing is just like it's it's not great work and so I think the question really is how do you really do great work here? I think like we're even talking like we're using like the heart system right now and it's so good. It's so much better than anything I use today.

你得把它发给我们。

You got to send it to us.

是啊，我们能用它吗？

Yeah. Can we use it?

嗯，当然。我们让你们提前用。对，肯定的。

Well, yeah. We get you guys early on. Yeah, for sure.

这就像研究预览。有大概 500 个博士，然后还有我和 Sam。

It's like research preview. There's like 500 PhDs and then me and Sam.

对，没错。不，我们喜欢，就像其他平台一样……

Yeah, exactly. No, we like like every other platform is

Park。外面天气怎么样？

Park. What's the weather outside?

我可以回答那个。对，没问题。我想说的是，每周都有大概 5 到 10 个垃圾 AI 初创公司或者类似的东西冒出来，它们就是不太好。整个领域已经到了一个地步，就是没什么好东西出来。我觉得现在编程方面的东西可能真的很棒，但除此之外的东西就有点不太行。

I can answer that. Yeah, no problem. Like I I think what I'm trying to say is like every week there's like five or 10 like junk AI slop startups or like things that are coming out. They're just like not very good. Like this whole space has gotten to a point where like there's just not great things coming out the door. I think the stuff in coding is probably really excellent right now, but everything beyond that is like just kind of like not great.

今年 1 月 1 日，你为今年做了四个预测。我想检查一下，看看你觉得它们进展如何。第一个，嗯，第一，人形机器人将在从未见过的家庭中执行无人监督的多日任务，完全由神经网络驱动。长时间跨度，直接从像素到扭矩。这个我们做得怎么样了？在轨道上、偏离轨道，还是已经完成了？

On January 1st of this year, you've made four predictions for the year. I want to check in and see how how you think they're going. First one, uh, number one, humanoid robots will perform unsupervised multi-day tasks in homes they've never seen before, driven entirely by neural networks. Long time horizons going straight from pixels to torques. How are we doing on that one? On track, off track, or done?

在轨道上。

On track.

在轨道上。你还有四个月。

On track. You got four months.

是的。我每天都能看到我们在做什么。我们在轨道上。难点在于，嗯，我们已经实现了从像素到扭矩。这意味着我们接收摄像头输入，然后输出电机应该放在哪里，比如告诉电机怎么做才能让手或关节到达正确位置。所以我们准备好了。嗯，进入一个从未见过的新房子去工作，这才是这个问题的难点。嗯，我们正在努力解决。我每天都在解决这个问题。我每天花大约 3 到 4 个小时，一周七天，都在这个问题上。

Yeah. I see every day like what we're doing. Like we're on track. The hard part here is um we already do pixels to torques. It just means like we're taking camera feeds and we output like where to put the motor like put the like you know we want to put a we want to like tell the motor like what to what to do to get to the hand in the right spot or the joints. So we're ready to do that. Um getting into a new house never seeing to do work. That's the hard part of this problem. Um we're working on that. I'm working on that every day. This where I spend about 3 four hours a day every single day seven days a week on this problem.

那么如果一个 Figure 机器人出现在我家里，它会怎么样？现在的瓶颈是什么？比如它不知道要做什么，不知道要去哪里，它无法微调，处理我的盘子。对我来说，它会在哪里表现糟糕？

So if a figure robot showed up in my house what would it what's the bottleneck right now? Like it wouldn't know what to do. It wouldn't know where to go. it wouldn't be able to, you know, fine-tune, handle my dishes. Where would it suck for me?

我们可以叠衣服，举个例子，但然后去一个新地方，在不同的位置叠，不同的光照，可能不同的桌子高度，不同类型的衣物，以及它从未见过的不同类型场景。这就像模型分布外了。它不知道该怎么办。就像如果你从 LLM 的预训练中移除所有金字塔数据，它就不会知道怎么谈论金字塔。

We can fold laundry, like as an example, but then going to a new place where we're folding in different location with different lighting and maybe different like table height and different types of laundry and different like types of scenarios it's never seen before. It's like the model is like out of distribution. It doesn't know what to do. It's like if you removed all the pyramid data from the pre-training of LLM, it wouldn't know how to talk about pyramids.

而且我们就是没有足够的数据。这些数据不在互联网上。所以你必须出去收集。所以我们需要知道的是，我们必须去世界上采样多少数据，才能训练模型，让它能进入你的房子，比如说叠衣服就是一个好例子。

And we just like we don't have enough of that data out there. It's not on the internet. So you have to go out and collect it. So what we need to know is like how much of that data we have to go sample in the world to be able to train the model to be able to go into your house and say fold clothes is a good example.

嘿，愚蠢的问题。为什么所有机器人公司都关心叠衣服和洗衣服？商业化来说，直接说“嘿，我们要打造最好的仓库工人”不是更好吗？因为世界上已经有 2000 万个这样的工人，这代表了多少十亿美元。当然，这能为我们赢得时间，让机器人叠衣服，你知道，完成机器人。但为什么你今天要关心这个呢？为什么不直接做工业工作，那些人们不想做、公司需要完成、他们愿意付钱的工作，而不是像我家那样有各种其他敏感问题。你们为什么现在关心这个？

Hey, stupid question. Why do all the robot companies care about folding clothes and and doing laundry? Wouldn't it be commercial better just to say, "Hey, we're going to build like the best warehouse worker cuz there's already 20 million of those in the world and that represents this much billies." And of course that buys us the runway to like get the robot folding, you know, robot done. But like why do you care about that at all today? Why not just industrial work that people don't want to do, companies need done, they're ready to pay, and it's not like my home where there's all these other sensitivities. Why do you guys care about that right now?

嗯，我们过去并不关心这个。我们刚推出时，我们说基本上要做商业方面来长期支付家庭方面。嗯，那是策略。这很有道理。比如我们在商业市场可以收取更高的费用。这容易得多。变化性更低。我们在一个小工作场所，24/7 工作。简单多了。

Uh we didn't care about it in the past. We like when we first launched, we're like, we're going to basically do the commercial side to pay for the home long term. And uh that was the strategy. It made a lot of sense. Like there's like we can charge a lot more in the commercial market. It's like much easier to do. It's like lower varitability. We're in like a little work site and just work 24/7. Just so much simpler.

嗯，我现在学到的是，家庭问题今天超级可解决。

Uh what I've learned now is that the home is super solvable today.

所以我们可以不去解决那个问题，就坐在这里在仓库里工作，但我或我的团队没有人想解决那个问题。我们想解决一个机器人，它可以通过语言进入任何环境并工作。我们想成为第一个做到这一点的。你可能可以用 100 个机器人和 50 人的团队做到这一点。所以那家公司一夜之间就会达到万亿美元市值。

So like we can like not go work on that problem and just like sit here and work in a warehouse, but me or none of my guys want to solve that problem. We want to solve a robot that can go into any environment just through language and do work. We want to be the first to do that. You can probably do that with a 100 robots and a 50 person team. So that that company overnight would be a trillion dollar market cap.

听起来不错。去做吧。

That sounds good. Do that.

我们正在做。这就是我们在做的。我们要解决这个问题。我认为我们会是第一个。我们称之为解决通用机器人。

We're doing that. That's what we're doing. Like we're going to solve that. I think we'll be the first. We call it like solving general robotics.

还有《我，机器人》。他们不是攻击人类吗？我不太记得这部电影了。只是……

And I Robot. Don't they attack the humans? I don't remember this movie very well. Just

是的。别担心那个。

Yeah. Don't worry about that.

好吧。不是那部分。谁能赢？现在打架谁能赢？人类还能赢吗？

Okay. Not that part of who could win. Who could win in a Who can win in a fight right now? Can a Can a human still win?

是的，人类还能赢。

Yeah, human can still win.

好的。嗯，其他预测是什么？

Okay. Uh what are the other predictions?

另一个预测，嗯，你这里有一个。日常 AI 使用将发生转变。人们将超越文本，走向高度多模态。具有持久记忆的语音智能体将变得普遍，这将推动 AI 更接近我们在科幻小说中想象的合成人类智能。

Other prediction um one you had on here. Daily AI usage will shift. People will move beyond text to highly multimodal. Voice agents with permiss uh persistent memory will become common which will push AI closer to the synthetic human intelligence we've imagined in sci-fi.

我们正在 Harkc 做这个。我们将在一个月内发布第一个版本。它会越来越好。嗯，我认为我们在这方面在轨道上。

We're doing that at Harkc. We'll ship that in a month and our first version of it. It'll get better and better. Uh I think we're on track for that.

实验室有没有，比如 ChatGPT 或 Claude，他们有没有发布过这方面的数据？比如我用很多语音功能。Sam，你经常用语音功能吗？

Have the labs ever like has Chad GPD or Claude, have they ever released the data on this? Like I use a ton of the voice thing. Sam, do you use the voice stuff a lot?

是的，我几乎不打字。

Yeah, I don't type really at all.

是的。我想这可能已经占了很大比例，已经到了办公室需要改变的地步。比如这些在初创公司流行的开放式办公室，现在有点糟糕，因为我想私下交谈。

Yeah. I wonder it's probably already a huge percentage of gotten to the point where like offices need to change. Like these open air offices that are like popular in startups, they're kind of whack right now because like I want to talk in private.

是的。

Yeah.

是的。很多工程师现在都有麦克风，他们低声细语，用压低的声音对着电脑说话。

Yeah. A lot of engineers have microphones now where they're whispering and they're just like in hushed tones whispering to their computers.

是的。就像我昨晚在说话，我说：“Claude，为什么我这么优柔寡断？”然后我妻子就说，她说：“天哪，老兄。她现在能听到我和 Claude 说的所有话了。”是的。

Yeah. Like I didn't I was like talking last night and I was like, "Claude, why am I so indecisive?" And then my wife was like like she was like like damn, dude. She can hear everything I'm talking to Claude about now. Yeah.

不，我一直在说，但这很尴尬。比如语音还是不行，还是不太好。你得去那里，打开它，而且它不太记得你刚才跟它聊了什么。它不能很好地做工具调用和电脑使用。它就在这些方面受限。你得在特定会话里使用它。我不知道今年能不能实现，但到 2027 年，肯定能通过完整的语音图灵测试。你手机上的 AI 系统能接电话，能骗过你们。我可以让一个人给你打电话，再让一个机器人给你打电话，我觉得你们分不出来。这是 2027 年的事，我很有把握。

No, I talk all the time, but it's embarrassing. Like even speech still sucks. It's still not great. It's like you have to go there, you have to turn it on, and it doesn't really remember what you just talked to it about. It can't do tool calling and computer use very well. It's just limited in these ways. You have to use it for a certain session. I don't know if we'll hit it this year, but certainly in 2027 you will hit a full human Turing test with speech. You'll be able to take a phone call from an AI system on your phone, and it'll be able to fool you guys. I'll be able to have a human call you and a robot call you, and I don't think you guys will be able to tell the difference. That's a 2027 event I feel pretty strong about.

好的。第三和第四是什么？

All right. What's the third and fourth?

过去 10 年，校园枪击案增加了 10 倍。到 2026 年，第一个能从 20 英尺外探测武器的全扫描系统将建成，并在 K-12 学校进行测试。我们 10 月开始建造第一个全尺寸系统，我想年底前能上线。不知道会不会在 K-12 学校。所以这个预测可能差一个季度。

Over the past 10 years, school shootings have increased by 10x. In 2026, the first full scanning system capable of detecting weapons from a 20-foot standoff will be built and beta tested in a K-12 school. We'll have our first full-scale system starting in October, and I think we'll bring it up before end of year. I don't know if it'll be at a K-12 school. So we might miss this one by a quarter.

那家公司有单独的 CEO 吗，还是你也是那家公司的 CEO？

Do you have a separate CEO running that one, or are you the CEO also of that company?

我有一位来自 JPL 和 NASA 的首席工程师，非常优秀，这基本上是一个纯工程项目。业务方面没什么可做的。我们有一些供应链的事情，但大部分就是：你能不能造出一个能探测武器的系统？这既是硬件问题，也是 AI 问题。这是一个大规模、深科技、深工程的问题。我的团队全是工程师，他们非常棒。我们其实做了一个很大的方向调整。本来现在应该已经上市了，但大约一年前我整个转向了技术系统。我们当时在建这个，我基本上找到了一种在硅和芯片上做所有事情的很便宜的方法，把价格降低了 90%，让它更容易扩展，效果更好，然后我们转向了。问题是，设计我们自己的芯片并流片需要大约一年时间。所以几个月前我们才拿到芯片，正在测试，效果很棒。现在我们需要生产更多，而再生产一批还需要六个月的周期。所以我们正在应对真正的硅片、非常难的芯片的长制造周期。我们最终会解决，但这不像你能在货架上买到的芯片。这些是定制设计的芯片，以前没人设计过。我们找了一家欧洲的特殊代工厂来制造，花了大约一年。

I have a chief engineer from JPL and NASA who's really good, and it's mostly a pure engineering project. There's really not much to do on the business side. We have some supply chain stuff and other things, but most of it is purely: can you build a system that can detect weapons? It's partly a hardware problem, partly an AI problem. It's roughly a large-scale, deep tech, deep engineering problem to solve. My whole team is all engineers; they're really good. We actually made a pretty big change of course. We would already be in market by now, but I pivoted the whole technology system about a year ago. We were building this, and I basically found a way to do everything very cheaply in silicon and chips, reduced the price by like 90%, made it much more scalable, made it work better, and we pivoted. The problem was that the fabrication times for designing our own chips and getting them out took about a year. So we just got those chips in a couple months ago, and we're testing them, and they're awesome. Now we need to make more, and there's another six-month lead time to make even more of them. So we're dealing with real silicon, long fabrication of very difficult chips, lead times now. We'll be out of this at some point, but it's not like chips you can go off and buy off a shelf. These are custom-designed chips that nobody's really ever designed before. We had a special fabricator in Europe that had to go make them, and it took about a year.

嘿，你现在事业上似乎火力全开。我其实想知道：你现在这种生活的代价是什么？因为你很乐观，看起来很兴奋，但代价是什么？

Hey, you are firing on all cylinders right now professionally, it seems. And I actually would like to know: what's the trade-off for the life that you're living right now? Because you're very optimistic, you seem excited, but what are all the trade-offs?

大约 5 年前，我有了孩子和公司。我遇到一个问题，我把生活看成三个口袋。有工作，我非常在乎。家庭——我有三个孩子，现在都很小。然后还有“其他”那一桶：朋友来了、年度高尔夫之旅、单身派对、欧洲的婚礼等等。我觉得我需要做个决定：如果我想把任何一件事做好，我无法三者兼顾。我真正想做好的是家庭和事业。我想在这些方面做到 A+。所以我基本上停止了第三个桶。我不再参与那边任何事。所以，一个大学朋友，我的大一室友，来湾区待 10 天。他想见面，我很久没见他了。喝杯咖啡会很好。我就说：“哥们，我跟你说实话，我没时间，不能见你。”他说：“我可以配合你，我去找你。”我说：“我真的没时间。我离开这两个桶的每一分钟，都是离开家人或工作的每一分钟。”而且我能投入这两个桶的时间几乎有限。

About 5 years ago, I had kids and the companies. I had an issue where I think of my life as three pockets. I have work, which I care deeply about. My family—I have three kids, they're pretty young right now. And then I have the 'other stuff' bucket: friends in town, the annual golf trip, a bachelor party, a wedding in Europe, whatever it is. And I felt like I needed to make a decision: if I want to do any of these well, I can't do all three. What I wanted to do really well was family and business. I wanted to be A+ in those areas. So I basically stopped the third bucket. I don't do anything anymore over here. So, a friend from college, my freshman roommate, was in town for 10 days in the Bay Area. He wanted to meet up, hadn't seen him in a long time. It'd be great to get a coffee. I was just like, 'Man, I'm going to be real. I don't have any time. I can't meet you.' He said, 'I'll make myself available. Come to you.' I was like, 'I literally have no time. Every minute I'm away from one of these two is a minute I'm away from my family or work.' And there's almost a limited amount of time I can put in both those buckets.

我能问问你的工作流程吗？你开了个玩笑，说“我不使用 Slack”。如果你方便，现在能举起手机吗？你手机主屏幕上有什么？你的应用设置是什么？你有什么？

Can I ask you about your workflow? You made a joke: you're like, 'I don't use Slack.' If you're comfortable, could you just hold up your phone right now? What's on the home screen of your phone? What's your app setup? What do you got?

全是通知。

All notifications.

那你得打开它。

Well, you got to open it up.

哦，我的……

Oh, what's my...

所以，你有很多短信。

So, you have just tons of texts.

这是我的……那些我觉得都是 Slack 和短信。我是说，我用 Slack。只是白天处理不过来。我让 Hark 处理，然后 Hark 给我发短信说“嘿，这个重要”，我需要看链接。

This is like my... those are all I think Slacks and texts. I mean, I use Slack. I just can't get through it during the day. I have Hark going through it, and then they Hark text me, 'Hey, it's important,' and I need to look at it with a link.

那你的设置是什么样的？你日常怎么过？你用笔记本电脑吗，还是只用手机？

So, what's your setup like? What's your day-to-day? Do you use a laptop at all, or are you only on the phone?

我用笔记本电脑。是的，经常用。笔记本和手机。我会说我现在用 Hark 处理所有 AI 相关的事情，端到端。甚至跟踪工程项目的进度、招聘，所有一切。我跟踪它；它在我的邮件里，在我的 Slack 里。

I use a laptop. Yes, laptop a lot. Laptop and phone. I would say I use Hark now for all my AI stuff end to end. Even tracking stuff I'm doing on engineering projects, recruiting, all of it. I track it; it's in my email, it's in my Slack.

那你的待办事项清单呢？

What about your to-do list?

那都在 Hark 里。Hark 管理所有那些。

That's all in Hark. Hark manages all that.

那在 Hark 之前呢？

So, what about before Hark?

我的待办清单是在 Google 文档里做的。我有一个叫“重新规划”的文档，每周都会不断更新。我通常周日来更新一周的计划，在那里更新。

My to-do list was done in a Google doc. I had a doc called 'replanning,' and I would constantly keep updating it every week. I would come in on Sundays usually and update my plans for the week, and I updated there.

那健康呢？你为健康做了什么吗？

And what about health? Are you doing anything for health?

是的，我现在能接触到一些特殊的医生和资源，他们基本上让你做季度血液检查、全身扫描、心脏 CT 扫描等等。老实说，这非常不可思议。

Yeah, I've gotten access to some special doctors and things now where they basically send you through quarterly blood tests, whole body scans, CT scans of the heart, everything. And it's been honestly pretty unbelievable.

有什么让人难以置信的？

What was unbelievable about it?

你拿回来的数据量以及它的全面性。比如，你可以花大约 100 美元做一次心脏 CT 扫描，我觉得基本上可以预防心脏病发作。你可以做全身核磁共振，实现早期癌症检测。大量的血液检查能发现一些异常，你可以去修复它们，改善健康。所以大概有十几种这样的检查。

The amount of data you get back and the thoroughness of it all. For instance, you can get a CT scan of your heart for about 100 bucks, and I think you can basically prevent heart attacks. You can get a full-body MRI and have early cancer detection. A lot of blood work can find anomalies that you can go fix for better health. So there are maybe a dozen of those.

是的。但解决所有这些问题的办法，可能都是你不愿意做的事情。比如你可能愿意吃天然食品，但像是起床、散步、锻炼，这些都不在你的关注范围内。

Yeah. But the solution to all those things are probably things you're unwilling to do. It's like you're probably willing to eat whole foods, but it's like get up, go for walks, exercise, and that was outside of your buckets of focus.

是的。不幸的是，我一直没有足够的时间来充分锻炼。但是，你知道，吃对东西。我现在吃得挺好的。

Yeah. Unfortunately, I haven't been able to have enough time to exercise enough. But, you know, eat right. I eat pretty well now.

嗯。

Yeah.

我是说，听着，总得有所取舍。我不能整天坐在这儿吃东西，然后去工作。我喜欢工作。我想去把这些企业做成功。

I mean, listen, something's got to give. I can't sit here all day eating and then go work. I love it. I want to go crush these businesses.

你在我们的准备文档里写道：“我在前三个创业公司上全力以赴，几乎每年都跌到谷底。”你能描述一下你说的“谷底”是什么意思，以及你应对谷底的方法是什么？当你陷入低谷时，你内心会和自己说什么，或者你有什么创业策略？

When you wrote in our prep doc, you said, "I went all in on my first three startups and I pretty much hit rock bottom every year." Can you describe what you mean by rock bottom and what is your method of dealing with it? What's the conversation you have with yourself or the entrepreneurial strategy when you hit those lows?

是的，我基本上在将近 15 年的时间里总是资金紧张。在 Vettery，我们早期经历了几次转型。最终在 2015 年筹集了一笔 50 万美元的可转换票据。那时，我想我借了 5 万或 10 万美元的贷款。我没有给自己发工资。我当时在纽约市，穷得叮当响，账户都是负的。我们筹集了可转换票据，情况看起来并不好。大约 6 个月后，我们在 Veter 推出了市场平台，然后它就彻底起飞了。一年后，我们以 1.1 亿美元的价格卖掉了公司。我觉得从 2012 年到 2017 年那段时期，我基本上负债累累，事情不顺利，非常艰难。

Yeah, I basically almost for 15 years was always running out of money. At Vettery, we had a couple of pivots early on. We ended up raising a $500,000 convertible note in 2015. At that point, I think I took out a $50,000 or $100,000 loan. I was not paying myself a salary. I was in New York City. I was so broke, I was in the negative. We raised a convertible note. It did not look great. And I think it was about 6 months later we launched the marketplace at Veter, and it just completely took off. Then a year later we sold for $110 million. And I think that period from 2012 to 2017 was just like I had basically debt. Things weren't working and it was hard.

那你内心的独白是什么？你对自己说什么？

And what's the inner monologue? What do you tell yourself?

内心的独白就是：这太糟糕了，超级痛苦。在那种时候，你只能一天一天地过，熬过今天。当事情变得那么糟糕时，你得列一个待办清单，然后一步步完成。唯一的出路就是穿过它。所以你需要列一个清单，然后一天一天地过。你不能一周一周地看。你看周五，你得熬到第二天。硬撑过去。我当时在训练超级马拉松，而我讨厌长距离跑步。我读了一个故事，讲一个人，他帮了我。他说：“你只需要选一个目标，不管它是一百英尺还是半英里远，即使你还有 49 英里要跑。选一个半英里外的目标，告诉自己，到了那里再考虑放弃。”然后你到了那里，你会想：“好吧，也许我还能再撑一点。”然后你再选一个只有 200 码远的目标，你会想：“好吧，到了那里我再考虑放弃。”我觉得这完全就是我的想法。然后我卖掉了 Veter，接着做 Archer。我当时想：“天哪，我刚赚了 1.1 亿美元，我们让所有冒险者翻了 12 倍。”然后我想：“我们要融资，会没事的。”然后所有人都说：“你在干什么？”

The inner monologue is like this really sucks. Super painful. At that point you just got to go day by day. You just got to make it through the day. When things get really bad like that, you got to build a punch list and you just got to get through it. The only way out is through. So you need to build a punch list and you need to get to day by day. You can't go week to week. You look at Friday, you got to get to the next day. Pile through it. I was training for this ultramarathon and I hate really long distance running. I read this story about this guy who kind of helped me. He said, "All you got to do is pick something, it doesn't matter if it's 100 feet or half a mile in the distance, even though you have 49 miles left to go in the race. Just pick something half a mile away and tell yourself once you get there, then you'll consider quitting." Then you get there and you're like, "Okay, maybe I have a little bit more," and you pick another thing just 200 yards away. You're like, "Okay, I'll consider quitting when I get to that." I was like, "Great." I think it's exactly how I thought about it. So then I sold Veter, and then I was doing Archer. I was like, "Oh man, I just made $110 million. We 12xed all the adventure guys." Then I was like, "We're going to raise money. It'll be fine." And everybody's like, "What are you doing?"

为 Archer？

For Archer?

是的。所有人都说：“你在干什么？我们不会投这个的。你在说什么？”

Yeah. Everybody's like, "What are you doing? We're not going to fund this. What are you talking about?"

你怎么对抗那种内心的独白，所有人都说你愚蠢、错了、这很傻？让你干脆去做软件。

How do you fight that inner monologue where everyone says you're stupid and wrong and this is silly? Just go do software.

这来自于信念。我知道我是对的，因为我做了功课。我理解它。我在一线。

It's coming from a place of conviction. I know I'm right because I've done the work. I understand it. I'm on the floor.

是的。但胜算仍然对你不利，对吧？

Yeah. But the odds are still against you, right?

但这就是游戏。当你玩这个游戏时，你签了约，周围 95% 的人都会失败。我记得在 Vettery，我们从纽约大学孵化器开始。我当时非常兴奋。我们被录取了一个学期，那里大约有 50 家公司。我们在苏豪区起步，很棒，我们玩得很开心。如果你回头看，我想五年后，我和另外一个人是仅有的两个赚到超过零美元的人。还有一个团队，48 家公司归零了。我当时就想：“天哪。”如果你在这个游戏里待得够久，每个人都会死，到处都是这样。20 年来一直如此。我一直在观察，你看到 TechCrunch 和 X 上人们融资的消息，但随着时间的推移，这些都消退了。这真的很残酷。所以我不得不买房子，把所有的钱都投进 Archer，然后我还有股票锁定期。所以即使我来到 Figure 时，股票正在解锁，我不得不用 Archer 的股票来资助 Figure，因为我没有其他现金。股票在下跌，就像一把下落的刀。那时，它从大约 10 美元跌到 2 美元，后来涨了很多。但我不得不拿房子做二次抵押贷款，才能资助 Figure。

But that's the game. When you play this game, you sign up and 95% of everybody around you will fail. I remember at Vettery, we started at the NYU incubator. I was so excited. We got in for one of the semesters, and there were about 50 companies there. We started in Soho. It was great. We had a great time. If you look back, I think five years later, me and one other guy are the only two people that made greater than zero dollars. One other team, 48 companies went to zero. I was just like, "Holy cow." If you're around this game for long enough, everybody dies, and that's everywhere. It's been like that for 20 years now. I've been watching around you, you see all the TechCrunch stuff and things on X about people raising money, and over time that all just fades away. It's really brutal. So I had to buy a house and put all the rest of the money into Archer, and then I had a stock lockup. So even while I was coming over to Figure, the stock was unlocking, I was funding Figure with stock from Archer because I had no other cash. The stock was coming down, literally like a falling knife. At that point, it went from about $10 to $2, and since then it's gone up a lot. But I had to take a second mortgage out of my house to even fund Figure.

我们问过你的一个哲学，你说：“我相信做难事在很多方面比做容易的事更容易。”你能解释一下吗？

We asked you one of your philosophies and you said, "I believe that doing hard things is easier in many ways than doing easier things." Could you explain?

每个人都在尝试做容易的事。当你做更难的事情时，通常竞争会更少。可能难事意味着如果成功，潜在的市场规模会非常大，退出回报也会非常大。你有这种风险回报权衡。可能想从事难事的人，大概是世界上最好的超额完成者，他们想去那里工作。一般来说，难事对投资有这种二元回报。

Everybody's trying to do easy things. When you work on harder things, you have generally less competition. Probably a hard thing means it could be a potential really big TAM, really big exit if it works. You have this risk-reward trade. You have folks that probably want to work on hard things, probably the best overachievers in the world want to work there. Generally, hard things have this binary payoff for investments.

他们真的很想资助那些东西，因为我们可以为投资组合带来 100 倍的回报。而且我认为 Scaling 有一个非线性曲线。难点在于：我觉得很多困难的事情并不是难 10 倍或 100 倍。有时候只是难两倍、三倍、四倍，也许五倍，但绝不是 100 倍。所以你可能获得 100 倍的回报，但难度可能只增加三四倍。比如在机器人领域，建造四足机器人（像四腿机器狗）与人形机器人相比，人形机器人可能只难三倍，也许四倍，仅此而已。但我觉得人形机器人并没有像那些机器狗那样的真正市场，对吧？我认为那只是一个小众领域，我不觉得它是一门真正的生意。而且目前我认识的人里，没有谁真的想在那上面花大量时间。所以你做仿人机器人——好吧，难三倍，但回报可能是百万倍。对投资者、对想在那里工作并获得股票和分享上涨收益的人来说，回报可能是百万倍甚至十亿倍。你为什么要去做四腿机器狗呢？机器狗能带来什么规模化的经济价值？如果你真正理解的话，我觉得每个人都在试图做容易的工作，结果反而变得非常困难。看看今天市面上那些 OpenClaw 之类的 AI 垃圾吧，全是垃圾，都不好。它们都会消失的——我认为它们没有一个能长期存活。可能会有一些整合，比如人才收购之类的，但最终都会完蛋。

They really want to fund those things because we could have a 100x return for the portfolio. And I think there's a nonlinear curve to scaling. Here's the difficulty: I think a lot of the hard things are not 10 or 100 times harder. Sometimes they're two, three, four, maybe five times harder, but not 100 times harder. So you might have a 100 times better payoff, but it might be only three or four times harder. For example, in robotics, building quadruped robots like four-legged dog robots versus humanoids. Humanoids are probably three times harder than that, maybe four. That's it. But there's really no real market for humanoids like there is for those dogs, right? I think it's just a niche thing. I don't think there's a real business for it. And I don't know anybody at this point who really wants to spend a lot of time on that. So you do humanoids—it's okay, three times harder, but it's probably a million times higher payoff. Probably a million or a billion times higher ROI for investors, for humans who want to work there, get stock, and participate in the upside, and for everything else. Why would you ever want to work on four-legged dogs? What economic value can a robot dog bring at scale? If you really understand it, I think everybody's trying to do the easy work, and it just becomes really difficult. Look at all the AI slop like OpenClaw harnesses out there today. It's all crap. It's not good. They're all going to go—I don't think any of them will make it long term. You might have some consolidation here and there for acqui-hires and stuff, but that's going to go all the way.

老兄，你说话总是这么绝对。这难道没给你惹过麻烦吗？

Dude, you talk in so many absolutes. Has that not gotten you in trouble ever?

我不知道。我是说，记住我的话。你们后来用过 OpenClaw 吗？

I don't know. I mean, mark my words. Like, have you guys even used OpenClaw since then?

没有，我不知道怎么用。

No, I don't know how to.

你用 OpenClaw 吗？

Do you use OpenClaw?

我从来都不信任 OpenClaw，没去设置过。我当时……

I never trusted OpenClaw to set it up. I was...

是的，我以前用过，现在不用了。它不太好。这股浪潮已经过去了。我只是想说，作为创始人，你能做的最重要的事情就是想清楚你真正要做什么，因为你将在接下来的 10 到 15 年里做这件事。它会决定整个轨迹，概率轨迹——就像是对潜在结果做概率加权决策。

Yeah, I used to use it. I don't use it anymore. It's not very good. The wave is over. I'm just trying to say that the most important thing you can do as a founder is to think through what you're actually going to go do, because you're going to spend the next 10 or 15 years doing it. It'll map the whole course, the probability course—it's like a probability-weighted decision of potential outcomes.

嗯，但你说的是一个非常特殊的游戏。比如，你说我们要成为万亿美元公司，否则就破产。这是二元的。大多数生意不是二元的。你玩的是二元结果的游戏，你喜欢这样，但对很多人来说不是这样。对很多人来说，如果他们能建立一个每年 1000 万美元的酷生意，那就是一个巨大的成功。

Well, but you're talking about a very particular game. For example, you said we're going to be a trillion-dollar company or we're going to go bankrupt. It's binary. Most businesses are not binary. You're playing the game where binary is the outcome, and that's what you like, but it's not like that for a lot of people. For a lot of people, if they can build a really cool $10 million a year business, that's a massive home run.

是吗？如果他们能做得那么好，当你 70 或 80 岁时回头看，你会问同一个人：‘嘿，你建立了一个很酷的 500 万或 1000 万美元的生意，你做了 30 年，期间没有做任何其他事情，没有尝试其他东西。你就只做那个生意。’你会回到 30 年前尝试更大的一搏吗？你会比 Vettery 做出不同的选择吗？Vettery 就是那样的。

Is it? If they can do that well, and you look back when you're 70 or 80, would you have asked the same person, 'Hey, you built a really cool $5 or $10 million business, you did it for 30 years, you didn't do anything else, you didn't try anything else while you were doing it. You just worked on that business.' Would you have gone back 30 years ago and tried to take a bigger swing? Would you have taken a different swing than Vettery? Vettery was like that.

Vettery 是我的桥梁。我在 Vettery 内部待了大约七年。我们实际上在内部为自己构建了一个营销自动化工具。一年后我意识到，‘哦，天哪，看看这个。这是 Outreach.io，一家十亿美元的公司。我们在一两年前就在内部构建了它。’然后看着各种不同的事情发生，我心想，‘天哪，我们实际上在内部做过一些这样的工作。它的价值不如外面的一些团队。’对于初创公司来说，决定把时间花在什么上至关重要，假设有一个……而且我确实认为初创公司是二元的。即使是那些做到 1000 万美元的人，可能还有另外 90% 的人在尝试时没有成功。所以我认为这很难。而且，老兄，向那些做到 500 万或 1000 万美元生意的人致敬。那很难，尤其是在资本很少或没有资本的情况下做到。

Vettery is my bridge. I sat inside of Vettery for like seven years. We literally built a marketing automation tool for ourselves internally. Then a year later I was like, 'Oh man, look at this. It's Outreach.io and it was a billion-dollar company. We built that internally a year or two prior.' Then watching all this different stuff happen, I was like, 'Man, we actually did some of this work internally. It's value is less than some other groups out there.' This whole decision of what you spend time on is super critical for startups, assuming there's a... and I do think startups are kind of binary. Even guys that get to $10 million, there's probably another 90% of those folks that just didn't make it when they're out there trying. So I think it's just hard. And dude, kudos to guys getting to five or ten million in business. That's hard, especially doing that with maybe a little bit of capital or no capital coming in.

嘿，让我快速问你一下你见过的其他东西。我相信因为你正在做非常有趣的工作，你会遇到其他在无关领域做有趣事情的创始人。所以，非人形机器人，但同样酷，对未来的有趣一瞥。我想你可能比我们看到了更多未来，肯定比大多数听众多。你能告诉我们你见过的、听过的或读到的某个创始人做的事情，让你觉得‘哦，是的，你们意识到未来实际上会是这样，只是对我们其他人来说还没有均匀分布’吗？

Hey, let me ask you real quick about your other stuff you've seen. I'm sure because you're doing really interesting work, you meet other founders that are doing interesting things in unrelated spaces. So, non-humanoid robots, but equally cool, interesting peek at the future. I think you've probably seen more of the future than us, and definitely more than most of the listeners. Can you give us anything that you've seen or heard or read about a founder you've met that's doing something that's like, 'Oh yeah, you guys realize the future is actually going to look like this, and it's just not evenly distributed for the rest of us yet.'

我喜欢思考这个问题：30 年后世界会是什么样子，一切将走向何方。我认为我们有一个能源问题——不是能源消耗，而是能源生产。也许两者都有，但最终我们作为一个物种如何产生更多的能源？我认为这里有一个长期趋势，你应该去顺应并真正帮助推动。有很多工作将这与人类的生活水平联系起来。所以，关于下一代能源是什么有很多工作：是太阳能、风能、核能？然后在这些里面还有聚变、裂变等不同的分支。我认为这是一个非常令人兴奋的领域。这需要很长时间，但你需要真正伟大的企业家来解决这些问题。我认为 AI 将在未来 10 或 20 年主导很多事物，对我们所有人都是如此。我认为它将比互联网大 100 倍。我们都经历过互联网。我认为它会非常非常大。AI 将吞噬整个互联网，把它全部吃掉。

I like looking at and trying to think through this problem of what the world is going to look like in 30 years, where everything is headed. I think we have an energy problem—not energy consumption, but generation. Maybe both, but ultimately how do we generate more energy as a species? I think there's a secular trend here that you want to go ride and really help. And there's a lot of work done correlating this to standards of living for humans. So, there's a lot of work on what the next generation is: is it solar, wind, nuclear? And then there are a bunch of different traits inside of that for fusion and fission and the rest. I think it's a really exciting area. It would take a long time, but you need really great entrepreneurs solving that stuff. I think AI is just going to dominate a lot of stuff in the next 10 or 20 years for all of us here. I think it's going to be 100 times bigger than the internet. We all lived through the internet. I think it's going to be so, so big. AI is going to eat the whole internet. It's going to eat it all up.

而且我认为这无论在物理层面还是数字层面都会是一个巨大的趋势。你现在有没有在关注一些还不是主流的产品或公司，你觉得它们是你所说的这种趋势的好例子？

And I think it's going to be an extremely large trend both physically and digitally. What are there any products that you're looking at or companies that you're looking at now that are not already the mainstream that you think are good examples of what you're talking about?

我是说，我们在 Harig 就在做这些。现在还不明朗。我们仍然处于一个很难说谁会做好的阶段。很多这类东西都处在一个模糊地带，还没有真正的突破。有过一些早期的胜利和突破，但接下来还有一段路要走，我们现在正处在这个阶段。我觉得未来一两年我们会更清楚它到底是什么样子。但我用过市面上所有的 AI 设备，没有特别兴奋的。我不知道你们在市场上有没有看到这类东西，但我没有那种“哇，这真是个超棒的产品”的感觉。

I mean, we're working on this stuff at Harig. It's unclear. We're still in this spot where it's really not clear who's going to do well here. We're in this foggy area for a lot of this stuff, there's been no breakout here. There's been early wins and early breakouts, but there's a next leg here that we're going to go through, and we're in it now. I think we'll know more in the next year or two what that really looks like. But I've used every AI device out there. Haven't been super thrilled. I don't know if you guys are seeing stuff in the market for these types of things, but I haven't been like, man, this is a crazy great product.

我喜欢那些小东西。我喜欢 Whisper。Whisper Flow 在很大程度上改变了我的沟通方式。

I like the small stuff. I like Whisper. Whisper Flow has pretty meaningfully changed how I communicate.

嗯。

Yeah.

那真的很酷。我觉得那是我过去六个月里最大的亮点。

That's been pretty cool. I think that's been my big standout the last six months.

最后一个问题。有没有什么让你敬佩的人？

What about last question? What about people who inspire you?

我真的很敬佩那些全身心投入自己手艺的人。你看迈克尔·乔丹的纪录片，他就是那种“我就是要成为世界上最好的”的人。我觉得创业也是一样。首先，你知道，我从未见过史蒂夫·乔布斯，但天哪，我听过的那些故事和其他一切，他就是一个令人难以置信的运营者和产品型创始人。我也和杰夫·贝索斯很熟。他投资了 Figure，来过这里很多次。我觉得杰夫在很多事情上都是一个很好的共鸣板。上周 Jensen 又来了，我们关系相当近，我觉得 Jensen 也是一个令人难以置信的运营者。他非常亲力亲为，过去 30 年管理 Nvidia 和他的组织的方式非常独特，我觉得他做了很多非常好的事情。

I think I really admire the folks that are fully dedicated to their craft. You watch the Michael Jordan documentary. He's just like, I just want to be the best in the world at this. I think for startups, it's the same thing. And first and foremost, you know, I never met Steve Jobs, but my lord, the stories I've heard and everything else, the guy was just an unbelievable operator and product-led founder. I've also gotten to know Jeff Bezos pretty well. He invested in Figure and he's been here a lot of times. I think Jeff has been a really good soundboard for a lot of things we've gone through. I had Jensen here last week again. We are fairly close, and I think Jensen's just an unbelievable operator as well. He's very hands-on, has a very unique way of managing Nvidia and his organization for the last 30 years, and I think he's done a lot of really good things.

杰夫给了你什么有意义的建议？

What advice did Jeff give you that was meaningful?

杰夫上次来的时候说：“听着，你处在一个非常有趣的时期，因为你不知怎么搞的做到了这一步。未来一两年，你要么找到突破的方法，真正把它做大，要么就不行。现在对你来说就是关键时刻。你得全身心投入，想办法突围，让这个东西运转起来并规模化。你处在一个非常有趣的节点。我不知道你是怎么走到这一步的，也不知道你为什么走到这一步，但你已经在这里了，你需要想办法——你现在在大舞台上了，你接下来的大动作将决定成败。”我觉得他说得基本没错。我们现在有机器人在用 AI 油门自主做这些事情，这太疯狂了。我觉得四年前你会说，不可能，绝对不可能。

Jeff said when he was last here, he's like, "Listen, you're at a really interesting period because you figured out how to do this somehow. In the next year or two, you're either going to figure out how to break through and really get this working in a bigger way or you won't. This is game time for you now. You got to just get wired in and figure out how to break out and make this thing work and scale it. You're at a really interesting point. I don't know how you got here and I don't know why you got here, but you're here and you need to figure out how to, you're on the big field now, and your next big push is going to make or break it." Which I think he's largely right. I think we've got robots now doing this stuff autonomously with AI throttles, which is crazy. I think four years ago you'd have been like, no way. No way you could.

老兄，四年前我去你办公室，你只有一个膝盖在动，我当时想，哦，那是个膝盖，挺酷的。那只是个膝盖。你只有大概五个工程师。你说，这家伙刚造完特斯拉 X 或者 Cybertruck 什么的。这家伙做了件了不起的事。这家伙治好了癌症。看看膝盖怎么动，脚踝还有背屈。我们就坐在那儿看着那个膝盖。那已经是最酷的事了。

Dude, four years ago I came to your office and you just had a knee working, and I was like, oh, that's a knee, that's cool. All it was was a knee. You had like five engineers. You're like, this guy just got done building the Tesla X or Cybertruck or something. This guy did this amazing thing. This guy cured cancer. Look how the knee moves and the ankle has dorsal flexion. And we were just sitting around looking at this knee. And that was like the coolest thing.

我知道，老兄。现在我们有 AI 在做人形机器人。我们接入摄像头，它在机载做推理，输出所有关节的动作。这太不可思议了。而且疯狂的是，它真的能工作。下一步就是让它在更大规模上运转。所以我不觉得那已经很好了。我觉得这些都是值得敬佩的很好的人，他们深深热爱自己的手艺，并且非常在乎。

I know, man. And now we have AI that's working on a humanoid robot. We're taking in cameras. It's doing inference on board. It's outputting all the joints. It's unbelievable. And it's crazy. It works. And the next leg up is just making that work at higher scale. So I don't think it's been great. I think those are some really good folks to look up to that really love their craft deeply and really care.

好了，Brett，我觉得你该回去工作了，朋友。

Well, Brett, I think it's time for you to get back to work, my friend.

太好了。谢谢大家。很高兴再次见到你们。

Great. Thanks, guys. It was good to see you again.

非常感谢，老兄。好了，就这样。收工。

Thank you so much, dude. All right, that's it. That's a pop.