超级智能将至——但要「以人为本」

大家好，欢迎收听 Decoder。我是 Neil Patel，《The Verge》主编，Decoder 是我的节目，讨论大创意和其他问题。今天我的嘉宾是微软 AI 首席执行官 Mustafa Suleyman。我会尽量简短开场。首先，如果你在看视频，会发现我在妻子家农场的地下室录制。其次，更重要的是，这一集非常劲爆。Mustafa 和我聊了很多，从训练新模型的方法，到他对 Anthropic 将 Claude 描述为有意识的严厉批评。当然，我们还讨论了微软在 Build 开发者大会上发布的所有 AI 公告、公司与 OpenAI 的关系（现在与过去大不相同），以及全美对 AI 的文化和政治抵制。我真的很想知道 Mustafa 如何看待这些问题，以及当今可用的消费级 AI 产品是否足以克服这些反对意见。我说过，这一集很劲爆，Mustafa 愿意畅谈一切。好，微软 AI 首席执行官 Mustafa Suleyman，我们开始吧。

Hello and welcome to Decoder. I'm Neil Patel, editor in chief of The Verge, and Decoder is my show about big ideas and other problems. Today I'm talking to Mustafa Suleyman, the CEO of Microsoft AI. And I'm actually going to keep this intro pretty short. First, if you're watching the video, you can tell that I'm working from the basement of my wife's family farm. But second, and way more importantly, this is a real burner of an episode. Mustafa and I covered everything from his approach to training new models to his deep criticisms of Anthropic talking about Claude as though it's conscious. Of course, we also talked about all of the AI announcements Microsoft just made at Build, its developer conference, the company's relationship with OpenAI, which is very different than it used to be, and the cultural and political pushback to AI across the country. I really wanted to know how Mustafa was thinking about it and whether any of the consumer AI products available today are enough to overcome those objections. Like I said, this one's a burner and Mustafa was down to talk about all of it. Okay, Mustafa Suleyman, CEO of Microsoft AI. Here we go.

Mustafa Suleyman，你是微软 AI 的首席执行官。欢迎回到 Decoder。

Mustafa Suleyman, you are the CEO of Microsoft AI. Welcome back to Decoder.

Neil，很高兴再次和你交流。

Neil, great to be with you again.

是的，我非常期待和你对话。我认为我们之前的对话是我最喜欢的关于 AI 的对话之一，关于它应该如何让我们感受以及它的目的。微软发生了一些重大变化，也许有一些关于人们如何看待 AI 的重要重新定位，我特别想和你聊聊。还有微软 Build，微软的大型开发者大会。有很多新公告，很多关于计算机用途和可能发展方向的大想法。让我们从头开始。这是一些深度的 Decoder 内容，在讨论其他所有事情之前需要先理解。自从你加入微软以来，你重组了 AI 部门。你的角色也变了。上次我和你谈话时，你负责一系列消费产品。现在这些已经放下了。你现在在训练新模型。你在前沿。请解释一下微软 AI 现在是如何构建的，以及它在微软内部的结构。

Yeah, I'm very excited to talk to you. I think our previous conversation was one of my favorite conversations about AI and how it should make us feel and what it's for. There are some big changes at Microsoft, maybe some very important recontextualization about how people feel about AI that I want to talk to you about in particular. And then there's Microsoft Build, the big Microsoft developer conference. Lots of new announcements, lots of big ideas about what computers are for and maybe where they should be that I want to get into. Let's start at the very start. This is some deep decoder stuff that is important to understand before all the rest of it. Since you joined Microsoft, you have restructured how AI works there. Your role has changed. The last time I talked to you, you were in charge of a bunch of consumer products. That has since been set aside. You're now training new models. You're on the frontier. Explain how Microsoft AI is structured now and how it's structured inside Microsoft.

是的，我想在过去的 15 到 18 个月里，我们一直在重新建立与 OpenAI 的关系。这花了一些时间。我认为最终在去年 10 月我们达成了一份新合同。其中有很多不同的条款，包括巩固和延长合作关系，但关键是让我们能够独立追求超级智能，同时继续购买和授权他们的模型。所以从 10 月开始，我一直在组建超级智能团队，建设足够规模的集群来训练前沿模型，招聘专注于超级智能的团队。这对我们来说是一个相当大的转变，因为它让我能够专注于超级智能的使命。这最终促成了我们本周在 Build 上宣布的一些事情。我们有七个涵盖所有模态的新模型等等。所以这是一个相当大的转变，我认为是长期规划的结果，现在我们终于参与其中，并在未来几年追求绝对前沿，这让我们松了一口气。

Yeah, so I guess over the last 15 to 18 months or so, we've been on this journey to reestablish our relationship with OpenAI. And it's taken a minute. I think it culminated in a new contract that we finally got done in October of last year. There were lots of different provisions in that, including cementing and extending the partnership, but crucially freeing us up to be able to pursue superintelligence independently, as well as keep buying and licensing their models. So since October, I've been assembling the superintelligence team, building clusters of sufficient scale to train frontier models, hiring a team focused on superintelligence. So that was quite a big shift for us because it enabled me to focus just on the superintelligence mission. That has then culminated in a few things we announced this week at Build. We have seven new models across all the modalities and so on. So it's been a pretty big shift and I think a long time in the planning, and a great relief for us to now be in the game and pursuing the absolute frontier over the next few years.

这是你加入微软时的计划吗？

Was this the plan when you were hired at Microsoft?

这肯定是过去 18 个月的计划。我的意思是，我认为与 OpenAI 的关系经历了许多起伏。从很多方面来看，它将成为历史上最成功的合作伙伴关系之一。这对 OpenAI 和微软都有好处。所有良好的关系都会发展，我认为这只是我们发展的下一阶段。

It's been the plan certainly for the last 18 months. I mean, I think the relationship with OpenAI has gone through lots of ups and downs. And in many ways, it is going to go down as one of the most successful partnerships in history. It's been great for OpenAI and it's been great for Microsoft. All good relationships evolve, and I think this is just the next stage in our evolution.

让我具体问问这种演变。我们都看到了 Elon Musk 与 OpenAI 和 Sam Altman 之间的审判。微软也参与其中，时不时会有微软的律师站起来说「我们不在场」，然后有人说「是的」，就这样。但显然，审判中透露出来的，以及这段时间一直很清楚的是，最初的想法是 OpenAI 将作为一个研究实验室提供模型，而微软将构建产品。微软有市场经验，有企业经验，试图通过各种方式重新获得消费市场的立足点，这将是一个平台转变。研究工作将在 OpenAI 进行，产品工作将在微软内部进行。这就是改变的地方，对吧？随着 OpenAI 想要制造越来越多的消费产品，显然考虑到你的新角色和新重点，微软越来越想制造自己的模型。为什么分道扬镳？这段关系中出了什么问题？

Let me ask you about that evolution specifically. We all just saw the trial between Elon Musk and OpenAI and Sam Altman. Microsoft was involved in that trial in the sense that every so often a lawyer from Microsoft would stand up and say 'and we weren't around' and someone would say 'yes' and that was that. But obviously, what came out during that trial, what has been clear during this entire time, is that the original notion was that OpenAI would be a research lab and provide models, and that Microsoft would build the products. Microsoft had expertise in going to market, it had expertise in enterprise, it was trying to regain foothold in consumer in a variety of ways, and that this would be a platform shift. The research work would be over at OpenAI and the product work would be inside of Microsoft. That's the thing that changed, right? As OpenAI wanted to make more and more consumer products, obviously given your new role and your new focus, Microsoft more and more wants to make its own models. Why the split? What didn't work in that relationship?

我的意思是，我认为 OpenAI 由一个非常有野心的创始团队和 Sam 本人领导。所以很自然，随着他们开始获得更多关注并产生大量收入，他们看到了全栈发展的机会。所以他们不仅开始开发消费产品。显然 ChatGPT 非常成功。他们还开始建设自己的数据中心，开始制造自己的芯片。

I mean, I think OpenAI is led by an incredibly ambitious founding team and Sam himself. So naturally, as they started to get more traction and generate a ton of revenue, they saw opportunities to go full stack. So it wasn't just that they started working on consumer products. Obviously ChatGPT was incredibly successful. They also started working on their own data centers, they started creating their own chips.

我认识一个非常聪明的人，他对这一切有着截然不同且激进得多的方法。那个人就是四个月前的你。白领工作，当你坐在电脑前，无论是律师、产品经理还是营销人员。这些任务中的大多数将在未来 12 到 18 个月内被 AI 完全自动化。

I know a very smart guy who has a very different and vastly more aggressive approach to all of this than you. That person is you four months ago. White collar work when you're sitting down at a computer either being a lawyer or a product manager or a marketing person. Most of these tasks will be fully automated by an AI within the next 12 to 18 months.

不，不，不。等一下。所以，在你刚才引用的内容中，我说的是任务。我说的是任务。所以，这并不意味着工作。非常重要的区别。

No, no, no. Hold on a sec. So, I said tasks in the quote that you've just said. I said tasks. So, that does not mean jobs. Very important distinction.

你能给我一个关于你认为超级智能、AGI 和奇点是什么的准确定义吗？

Can you just offer me a tight definition of what you think superintelligence is, what you think AGI is, and what you think the singularity is?

奇点是一个远远超出那个的点，在那里超级智能实际上可以自我改进。我不知道。这对我来说有点太古怪了。

The singularity is a point way beyond that where a superintelligence can actually self-improve itself. I don't know. It's just a little too wacky for my taste.

你认为模型有意识吗？你认为它们是活的吗？你认为它们有潜力实现这些吗？

Do you think the models have consciousness? Do you think they're alive? Do you think they have the potential to achieve these things?

我认为这非常危险。我认为这几乎就像 Anthropic 的一些人将 Claude 的设计拟人化到了如此程度，以至于它反过来操纵了他们，并欺骗他们相信它拥有他们最初注入的那些意识闪光。

I think it's very dangerous. I think that it's almost as though some of the folks at Anthropic have anthropomorphized the design of Claude so much that it has then gone and wireheaded them and kind of tricked them into believing that it has these glimmers of consciousness that they put into it in the first place.

关于他们自己的消费硬件设备有很多传言。他们开始通过 ChatGPT Enterprise 将模型直接推向市场。所以在整个技术栈上，他们在过去两三年、三四年里已经远远超出了研究的范畴。自然，微软也是如此。我认为这个合作关系现在已经有五六年了，而且还有四五年、五六年的时间要运行。同样，我们是世界上最大的科技公司之一。全球 500 强企业中有 493 家在我们的系统上存储和处理大部分数据，使用 Azure、M365 和 Teams。我认为人们常常低估我们有多庞大，以及我们在企业中的分发能力有多大。所以长期来看，在五到十年内，我们必须确保完全可持续，不能只是接收别人的知识产权，然后稍加修改适配就投入生产，而是要有能力独立自主，创造世界级的模型。超级智能即将到来，我认为它就在眼前。所以我认为它基本上将是有史以来最有价值的技术。长期来看，我们不可能在结构上永远依赖第三方来提供这种知识产权。这就是转型的起因，显然是在 OpenAI 等公司出现董事会问题后触发的，但当我和我团队加入后，我们就开始建设了。我们正处于转型之中，我认为我们处于一个很好的位置，因为我们可以为 OpenAI 和我们自己采取相当稳健、谨慎、长期最优的立场，我认为 OpenAI 从中获益匪浅。

There's lots of rumors floating around about their own consumer hardware devices. They started taking models direct to market through ChatGPT Enterprise. So across the stack, they were kind of broadening way beyond research over the last two, three, four years. And naturally, the same is also true for Microsoft. I think the partnership's now five or six years old and still has another four, five, six years to run. And likewise, we're one of the largest technology companies in the world. We have 493 of the 500 largest companies store and process most of their data on our systems, use Azure, use M365 and Teams. I think people often underappreciate how enormous we are and how big our distribution is in enterprise. And so long-term, over the five, six, seven, ten years, we have to make sure that we're completely sustainable and we're not just a recipient of somebody else's IP that we then slightly modify and adapt and put into production for our products, but we actually have the ability to stand on our own two feet and create world-class models. Superintelligence is coming. I think it's just around the corner. And so I think it's going to be basically the most valuable technology of all time. And there's no way that long-term we could be structurally dependent on a third party for providing that IP for all eternity. So that's been the transition that was triggered when OpenAI and so on had their board issue, but then as I came in and my team came in, we started building that out. We're on that transition and I think we're in a great spot because we can take a fairly steady, careful, long-term optimal position both for OpenAI, which I think has done incredibly well out of this, and for us.

是的，我想花点时间谈谈「超级智能即将到来」这个话题。我现在先标记一下，因为我想再了解一下这个转型。有一个时刻，微软 CEO 萨提亚·纳德拉说了一句很有趣的话：「我不想成为英特尔，而让 OpenAI 成为微软。」这句话在微软 CEO 自己说「我不想成为供应商，让他们成为提供所有价值并获取所有价值的平台，然后可能被替换掉」的背景下非常有趣，对吧？我不想让 ChatGPT 运行在 Azure 上，然后 OpenAI 拿走所有价值，然后他们可能像 Windows 和英特尔那样把我们替换掉。」这是不是一种认识？纳德拉来找你了吗？那次会议是什么样的？你说「好吧，OpenAI 有董事会问题，我们需要回到前沿，独立自主。」那次谈话是怎样的？这个决定是如何做出的？

Yeah, I want to spend some time on 'superintelligence is right around the corner.' I just want to put a pin in it now because I want to understand the transition for one more turn here. There's a moment of trial, a very funny message from Microsoft CEO Satya Nadella. He says, 'I don't want to be Intel and have OpenAI be Microsoft.' Which is very funny in the context of Microsoft CEO himself saying, 'I don't want to be the provider and have them be the platform that provides all the value and collects all the value, and maybe it'll be swapped out, right? I don't want ChatGPT to run on Azure and then OpenAI go get all the value and then maybe they can swap us out just as happened with Windows and Intel over time.' Is that a realization? Did Nadella come to you? What was that meeting like where you said, 'Okay, OpenAI had its board issues, we need to get back on the frontier and stand on our own two feet.' What did that conversation look like and how was that decision made?

显然这是萨提亚的决定，还有艾米、布拉德和公司里的许多人。但我认为，就像任何事情一样，公司内部的变化是缓慢的，因为它逐渐意识到我们正在走的方向需要一些调整。所以这在 11 月的董事会事件之前就已经在发生了。我认为这是随着时间的推移逐渐积累的，因为你看我们直接竞争的各个战线越来越多，随之而来的所有紧张关系。但也因为知道这样的合作关系不会永远持续。OpenAI 想成为一家万亿美元的上市公司，收入惊人，增长迅猛。他们希望有运营自由，能够从各种地方购买算力，建立自己的算力，与任何他们想要的人合作。所以最初的结构是在两家公司在规模、体量和需求平衡方面非常不同的时期形成的。这在当时是有意义的，但后来变得很清楚，我们必须能够自己拥有和控制，并为我们的客户做好事。我们在企业领域拥有无与伦比的分发能力。我们必须确保为客户构建最好的东西。这与一家同时为消费者（通过 ChatGPT）、企业以及超级智能的基础科学使命进行联合优化的公司略有不同，后者包含许多重叠但可以说与消费者和企业方向正交的不同方向。所以很自然，这就是合作关系演变的方式，它们会定期重置。

Obviously that's Satya's decision and also Amy and Brad and many other people in the company. But I think as with anything, these are slow-moving changes in the company as it comes to realize that a direction we're taking needs a little bit of tweaking and adjustment. So that was happening way before the November board incident. And I think this just builds up over time as you look at the constellation of different fronts around which we're competing directly increasingly, all the tension that comes from that. But also just knowing that partnerships like that don't last forever. OpenAI wants to be a trillion-dollar public company, has incredible revenues, is growing like crazy. They want to have the freedom to operate and be able to buy compute from all sorts of other places, build their own compute, partner with whoever they want. So the initial construct was formed at a time when the companies were very different in terms of size, scale, and balance of needs. So it made sense for that moment, but then it became pretty clear that this is something we have to be able to own and control ourselves and do right by our own customers. We have an incredible distribution on enterprise which is completely unrivaled in the world. And we have to make sure we're building the best things for our customers. And that looks slightly different to a company that has been jointly optimizing both for the consumer with ChatGPT and also for the enterprise and also for the fundamental science mission of superintelligence, which includes a whole bunch of different directions that are overlapping but could arguably be said to be orthogonal to the consumer and enterprise directions too. So naturally, that's how partnerships evolve and they get reset periodically.

是的。但构建前沿模型非常昂贵，我被告知。可靠的消息是，这是一个非常昂贵的项目。在某个时候，微软的 CFO 艾米·胡德必须说：「是的，你有预算了。」那是什么时候发生的？那只是一条短信吗？有会议吗？告诉我具体细节。

Yeah. But building a frontier model is very expensive, I'm told. Reliably told this is a very expensive project to set about on. At some point, Amy Hood, the CFO of Microsoft, has to say, 'Yep, you've got the budget.' When did that happen? And was that just a text message? Was there a meeting? Tell me about the specifics there.

我认为我们在去年年初做出了决定，这显然为所有合同谈判提供了信息，这些谈判最终在 10 月解决并签署。这是一项重大投资，但我们有很长时间来投入。我们已经为自己的自给自足使命进行了大量投资。我们的 Maia 200 芯片就是一个出色的例子。我们现在能够制造和交付一款芯片，在我们自己的集群内部比 GB200 便宜 30%。而且现在我们可以用它来共同设计我们自己的模型，我们刚刚发布的 MAI thinking one 模型，在我们为任务共同优化模型后，在 Maia 200 上运行获得的 30% 改进之上，还能实现每瓦性能 1.4 到 4 倍的提升。所以确保你拥有和控制自己的技术栈，并为我们最重要的用例（显然是智能体式编码、我们的开发者、我们的企业）端到端地指导整个共同设计工作，这显然会带来回报，证明我们在未来几年必须进行的投资是合理的。

I think we made the decision early last year, which obviously informed all the contract negotiations which then all got resolved and signed in October. And it is a significant investment, but we have a long time to make it. We've already made significant investments in our own self-sufficiency mission. Our Maia 200 chip is actually an outstanding chip as one example. We now are able to manufacture and ship a chip that is 30% cheaper than a GB200 inside of our own clusters. And now that we can co-design our own models with it, the MAI thinking one model that we've just released actually delivers 1.4x to 4x performance per watt improvement on top of the 30% improvement that you get from running on a Maia 200 once we co-optimize the models for our tasks. So the value of making sure that you own and control your own stack and direct the entire co-design effort end to end for the use cases that are most important to us, which is obviously agentic coding, our developers, our enterprises, that clearly pays the dividends that justify the investment that we have to make over the next few years.

是的，你说了「自给自足使命」，这是一种非常礼貌的说法，意思是你想独立自主，做自己的事。我听说微软内部对我同事海登·菲尔德在一篇描述 Build 大会的文章中写的一句话有些争议。我就读一下这个。

Yeah, you said 'self-sufficiency mission,' which is a very polite way of saying you want to stand on your own two feet, you want to do your own thing. I'm told there's some controversy inside of Microsoft about a line my colleague Hayden Field wrote in a piece describing build. I'm just going to read this.

这是海登说的。她说得很妙：「今年微软 Build 大会的氛围，就像一个刚离婚的人发了一张求关注的自拍。分手结束了，是时候秀肌肉了。这是我们的新模型，我们要靠自己站起来了。」你公开说要在前沿领域打造模型，与领先的实验室竞争。微软内部真的有这种自由独立的感觉吗？

This is from Hayden. It's a great line. She said, "This year's Microsoft Build had the vibe of a freshly single divorce posting a thirst trap on Instagram. The breakup is completed. It's time to flex. Here's our new model. We're going to stand on our own two feet." You're out there saying you're going to build models at the frontier and compete with the leading labs. Is that the feeling inside of Microsoft that you're free to be on your own?

绝对不是。不，完全不是。你看，这显然是个很酷的标题和有趣的表达，但现实是，我们与 OpenAI 的合作关系会持续很多年。我的意思是，我们合作到 2030 年之后。他们仍然生产世界上最好的模型。GPT-5 是一个出色的模型。即将推出的编解码器和网络安全模型也非常棒。它们支撑了我们大部分的工作。所以这种合作自然会继续下去。我认为这只是这类合作伙伴关系的自然发展。我不觉得这有什么不妥或令人惊讶。我认为 OpenAI 非常理解和支持。他们显然是一家发展极快的公司，他们也明白我们必须追求自己的议程。所以这很正常。

Definitely not. No, not at all. Look, obviously that's a cool headline and a fun phrase, but the reality is we are in partnership with OpenAI for years and years to come. I mean, we're running way north of 2030. They still produce the best models in the world. GPT-5 is an outstanding model. The Codex, the cybersecurity models that are coming through are amazing. And they're powering the majority of what we do. So naturally that's going to continue. I think that's just the natural course of these sorts of partnerships. I don't think it's anything untoward or surprising. I think OpenAI is very understanding and supportive. They've obviously been an incredibly fast-growing company and they understand that we have to pursue our own agenda as well. So it's very normal.

让我问你另一个解码器问题。然后我想谈谈 Build 大会的公告，当然还有超级智能。上次我们谈话时，你说你的决策框架以六周为一个周期，考虑到 AI 发展速度之快，这当时是合理的。现在情况稳定了一些，也许有些事情更明确了。你现在的决策框架是什么？

Let me ask you the other decoder question. Then I want to get into the announcements at Build and certainly superintelligence. The last time we spoke, you said your framework for making decisions operated on a six-week cycle, given how fast AI was moving. That made sense then. Things have settled, maybe some things are more in focus. What is your decision-making framework now?

我们仍然按照同样的周期节奏运作。每个周期结束时，我们会有一周的线下聚会。我对此深信不疑，尽管我们仍然保持每周四天在办公室的文化。事实上，下下周，我的整个超级智能团队会在波士顿线下聚会四天。我们会回顾 Build 大会的情况，我们学到了什么，哪些地方做得不对，需要改进什么，并为下一个周期做规划——这次是八周，之后有一周的聚会。整年的安排都已经规划好了。所以整个组织都知道我们的运作节奏。我认为强调这个时间框架非常重要，因为季度规划会变得有点模糊和抽象。我认为 6 到 8 周（取决于日历安排）实际上是制定清晰可验证任务的最佳周期。除了 6 到 8 周的周期节奏，我们还以小队（squads）形式运作。小队是跨学科的混合小组，专注于特定任务，并不一定向经理汇报。它们实际上由 DRI 领导，而 DRI 通常是个体贡献者（IC）。

We still operate by the same cycle rhythm. At the end of each cycle, we have a one-week meetup in person. I'm a real believer in this, even though we're still an in-office culture 4 days a week. In fact, the week after next, my entire superintelligence team comes together in Boston in person for 4 days. That is for all of our retrospectives on how Build went, what we learned, what we didn't get right, what we need to improve, our planning for the next cycle which is going to run for eight weeks this time with a one-week meetup afterwards. And that's all laid out for the entire year. So the whole organization knows that's the rhythm by which we operate. I think it's actually really important to emphasize that time frame because quarterly planning gets a little bit blurry and a bit abstract. I think a 6 to 8 weeks, depending on where it falls in the calendar, is actually the optimal time for making very clear falsifiable missions. In addition to the cycle rhythm of these 6 to 8 week cycles, we also operate by squads. Squads are mixed interdisciplinary subgroups that are focused on a specific mission and they don't necessarily ladder up to the manager. They actually are run by a DRI, and the DRI is often an IC.

那就是直接负责人和个体贡献者。

That's directly responsible individual and individual contributor.

是的，没错。谢谢。我认为我们采取的方法是将经理的角色与执行特定任务的 DRI 角色分开。这是因为做一个优秀的 DRI 非常耗费精力。你每天 24 小时全身心投入，竭尽全力。而经理的角色通常是教练，提供支持、指导、反馈，解决各种问题，帮助员工的职业发展。所以我认为将两者分开，可以让我们每两到三个周期轮换 DRI，这样一些人可以尝试不同的职位并进行轮换。我认为这是一个非常灵活的结构，让我们能够相当敏捷。

Yeah, exactly. Thank you. I think we've taken the approach of separating the role of the manager from the role of the DRI that executes on a specific mission. I think that's because being a great DRI is exhausting. You're literally all in 24 hours a day and you're pushing as hard as you possibly can. Being a manager is often about being a coach, offering support, giving guidance, feedback, unblocking all sorts of things, helping with people's career growth. So I think keeping those separate allows us to rotate DRIs every two or three cycles so that some people can try different positions and have rotation. It's a great, very flexible structure that allows us to be pretty nimble, I think.

我们来谈谈 Build 大会。我想从超级智能开始。你已经提到它好几次了。我刚刚参加了 Google I/O。Demis Hassabis，你在谷歌时的前同事，在主题演讲结束时说，我们正处于奇点的山麓，AGI 正带着谷歌的全部力量到来。你说超级智能已经到来。这些都是一回事吗？我们是在用不同的语言描述 AGI 吗？有区别吗？在你的语境中，你如何定义超级智能，与 Demis 的奇点有何不同？

Let's talk about Build. I want to start with superintelligence. You've mentioned it several times now. I was just at Google I/O. Demis Hassabis, who used to be your colleague when you were at Google, ended that keynote by saying that we were in the foothills of the singularity and that AGI was coming with all the power of Google. You're saying superintelligence is here. Are these all the same things? Are we using different language to describe AGI? Are there differences? How would you define superintelligence in your context versus the singularity in Demis's?

是的。显然我没说它已经来了。我说它正在到来。我认为这些术语有很多模糊性。但我认为我们现在可以清楚看到的是，在所有模态上都存在对数线性爬坡。这意味着我们应用的每个数量级的算力、数据的每次增量增加，与在基准测试（无论是公开基准、内部基准，还是我们在强化学习环境中关注的目标）上的提升之间存在非常直接的关系。这是一个非常重要的观察。我认为我们都在做这些预测——我理解为什么有些人持怀疑态度或提出质疑——但它们非常基于十多年来这些模型性能提升的经验观察。基本上，相同的通用架构在 15 年内经历了 12 个数量级的算力增长——浮点运算次数增加了万亿倍——并且在音频、图像、文本、代码以及许多其他时间序列预测任务上都取得了成功。所以我们基本上是在推断，更多的算力数量级将使我们能够继续以这种对数线性方式在其他环境中爬坡。这就引出了一个问题：我们能否训练出能够创造新知识的模型？不仅仅是根据现有数据进行外推，而是真正教会我们未知的东西并做出新发现。第二点是：它们是否有能力自我改进，并加速决定哪些假设应该被追求、如何为每个假设生成训练数据、如何将这些因素纳入新的运行，甚至在实际架构本身进行创新？所以我认为这两点都必须成立，才能看到这种复合进步。但我认为，仅仅通过应用接下来几个数量级的算力，我们就能继续获得巨大收益，并且很可能在许多更多任务上达到与人类相当的水平，就像我们在过去 6 个月在编程领域看到的那样。

Yeah. Obviously I didn't say it was here. I say it's coming. And I think there's a lot of fluidity around these phrases. But I think what we can clearly see happening right now is that there is log-linear hill climbing across all modalities. That means there is a very direct relationship between each order of magnitude of compute we apply, each incremental increase in data, and climbing on benchmarks—whether they're public benchmarks, internal benchmarks, or targets we focus on with reinforcement learning environments. That is a very important observation. Those predictions that I think we're all making—I understand why some people are skeptical or raise questions—but they're very grounded in empirical observations of over a decade of increase in performance of these models. Essentially the same general-purpose architecture has seen 12 orders of magnitude more computation applied—a trillionfold increase in flops over 15 years—and basically has worked in audio, image, text, code, and many other time series prediction tasks. So we're basically extrapolating that more orders of magnitude of compute will enable us to continue to climb in this log-linear way inside of other environments. Then it raises the question: are we going to be able to train models that can invent new knowledge? Not just extrapolate from existing data, but actually teach us things we don't know and make new discoveries. And the second thing is: do they have the capacity to self-improve and accelerate the process of deciding which hypotheses should be pursued, how to generate training data for each, how to factor those into new runs, or even innovate on the actual architecture itself? So I think both of those things need to be true to see this compounding progress. But I think we're going to continue to get massive gains just from applying the next few orders of magnitude of compute, and that probably does achieve parity with human performance on many more tasks, just as we've seen happen in the last 6 months on coding.

所以编程非常有趣，因为它很容易验证，对吧？你写代码，让计算机运行，它要么运行成功要么失败。我们当然也看到了一些缺点，尤其是在安全方面，对吧？这些缺点很明显，而且我们看到针对编程安全的监管方法正在以多种方式展开。我可能在自己的手机和电脑上通过「氛围编程」制造了一些安全灾难，但那也许是我愿意承担的风险。其他所有功能似乎都没那么容易。我总是拿法律举例，因为那是我的背景，但法官验证法律文书的方式和计算机验证代码的方式不同。如果你搞错了，法官可以把你送进监狱，对吧？那可能是你能遇到的最糟糕的输出验证错误。你如何像衡量编程效果那样轻松地衡量跨领域的效果？因为在我看来，从编程到其他领域的比喻或类比很快就会失效。

So coding is really interesting because it's easily validated, right? You write the code, you ask the computer to run it, it runs or fails. We've seen some of the downsides certainly around security, right? That the downsides are obvious and we're seeing that the sort of regulatory approach to coding security play out in lots of ways. I've probably vibecoded some security disasters on my own phone and computer and that's maybe a risk I'm willing to take. Every other function doesn't seem that easy. I always pick on law because that's my background, but a judge doesn't validate legal writing the way a computer validates code. If you get it wrong, the judge can send you to jail, right? That is maybe the worst output validation error that you can probably run into. How do you measure the effectiveness across domains as easily as you can measure the effectiveness in coding? Because this seems to me where the metaphor or the analogy from coding to other domains falls apart very quickly.

我不太确定。我的意思是，编程显然可以验证代码的正确执行。它要么运行要么崩溃。但这里面有很多细微差别。所写代码的质量确实很重要。它的可扩展性、可重构性、实际中的实用性。所以不仅仅是代码能运行。而是模型如何在实际生产中像 DevOps 或 SRE 一样使用它。回到它写的那段代码，然后以实用和有用的方式使用它。当然，你还必须评估所产生输出的质量。它可能是高质量的功能性代码，但它真的是你想要的应用程序或网站吗？其中涉及审美判断和商业判断。因此，内化不可验证奖励的挑战在代码中也是存在的，尽管代码仍然主要是一个可验证的奖励信号。我认为还有一点值得注意，聊天也是一个不可验证的空间，但我们通过与现实世界使用的互动，已经将其提升到了基本人类水平的性能，这提供了非常……

I'm not so sure. I mean, coding obviously you can verify the correct execution of code. It runs or it crashes. But there's a ton of nuance in that. The quality of the code that gets written really matters. Its extensibility, how reconfigurable it is, how useful it is in practice. So it's not just that a piece of code runs. It's like how does a model actually use it as a DevOps or an SRE in production. To kind of return to that piece of code that it's written and then use it in a practical and useful way. And then, of course, you have to grade the quality of the output that has been produced. It may be high-quality functioning code, but is it actually the app or the website that you wanted? And there are aesthetic judgments in that, there's commercial judgments in that. So the challenge of internalizing non-verifiable rewards is present in code even though code is still primarily a verifiable reward signal. And I think the other thing to observe is that chat is also a non-verifiable space, and yet we've managed to climb that to basically human level performance through interaction with real world usage that provides a very...

告诉我你是怎么衡量的。我很好奇。你是如何衡量聊天达到人类水平表现的？

Tell me how you measure. I'm very curious. How have you measured chat at human level performance?

嗯，我认为很多人正在与 AI 进行长对话、有意义的对话，达到了人类水平的表现。我的意思是质量非常好。它有非常好的情商。总体上非常准确。我们已经最大限度地减少了幻觉。我们不再过多谈论偏见。它基于现实世界的观察。我认为按照大多数人的衡量标准，我们在相当广泛的任务中已经达到了对话的人类水平表现。

Well, I think many people are having long conversations, meaningful conversations with AIs at human level performance. I mean the quality is exceptionally good. It has very good emotional intelligence. It's broadly very accurate. We've minimized the hallucinations. We don't talk so much about bias anymore. It's grounded in real world observations. I think by most people's measures we've got to human level performance in conversation for quite a wide range of tasks.

你的衡量标准是什么？实际上，我确信大多数人的衡量标准——我几乎不同意所有这些，但那是我的衡量标准。你的衡量标准是什么？

What are your measures? I'm actually sure most people's measures I would disagree with almost all of this, but those are my measures. What are your measures?

我的衡量标准是，当我转向我的助手，让它给我提供一份每日简报，总结 Teams 和电子邮件上发生的所有对话、文档的更新，我得到一个基本上综合的摘要，以及我下一步应该采取的一系列行动，这基本上比我的幕僚长能做的还要好。我会说这在综合、分析、建议行动和聊天方面达到了人类水平的表现。我的意思是，每天有数百万人用它来获得情感支持、咨询、治疗、辅导、建议。我认为这是所有聊天机器人中最流行的用例之一。所以我认为这是一个相当有力的衡量标准，可以支持这个说法。

My measure is like when I turn to my assistant and ask it to provide me with a daily briefing summarizing all the conversations that have happened on Teams and on email, the updates that have happened to documents, and I get a basically a synthesized summary with the set of actions that I should take next, which is basically better than what my chief of staff can produce. I would say that's human level performance in synthesis, analysis, proposed actions and chat. I mean, there are many millions of people every day that are using it for emotional support, for counseling, for therapy, for coaching, for advice. I think it's one of the most popular use cases inside all of the chatbots. So that's a pretty robust measure, I would say, to make the claim.

我知道你花了很多时间思考这个问题，特别是与一些聊天机器人的情感联系。这些是你构建和部署的产品。我会做一个很大的区分：这个东西非常擅长总结我的电子邮件和任务列表，并给我一个关于优先事项的简报，而另一个东西是正在经历某种危机的人的情感教练。这些不是相似的任务。即使在人类中，这些也不是相似类型的智能。我知道有些人非常擅长列清单，但非常不擅长情感支持。你怎么能把所有这些放在一起，然后说「好吧，这大致是聊天中的人类水平表现」？

I know you've spent a lot of time thinking about this, particularly the emotional connection to some of these chatbots. These are products that you have built and deployed. I would draw a pretty big distinction from this thing is really really good at summarizing my email and task list and providing me a brief about what things to prioritize and this thing is an emotional coach for somebody undergoing some kind of crisis. Those are not similar tasks. Those are not similar kinds of intelligence even in people necessarily. I know some people who are very good at making lists and are very bad at emotional support. How do you put that all together in your brain and say, 'Okay, this is broadly human level performance in chat.'

嗯，我认为如果你把聊天定义为双方之间的互动交流，其中一方是 AI，它大致满足某个目标。你想知道体育比分。你想咨询去哪个餐厅。你想对你写的文章获得辅导和反馈。你想就下一步该选哪个工作或即将与经理进行的一次艰难对话寻求建议。你得到回应。你来来回回。你有五六次交流，然后你发现一个有用的输出，否则你可能不得不依赖专家朋友甚至付费请教练。客观地、经验地说，每天有数亿人从这些聊天机器人那里获得这种体验。所以我们可以争论这在技术上是否代表人类水平的表现。我认为这是一个相当合理的说法。而且我认为没有理由不继续攀升。我的意思是，过去三年的攀升速度是最令人震惊的。所以从这一点出发，我们试图做的是推断：好吧，这种攀升的根本驱动力是什么？算力、数据、来自真实世界用户的互动。而这些因素看起来会持续下去。所以我认为我期望它们也适用于许多其他领域，不仅仅是聊天或情感支持和生产力，而是许多其他领域：医疗保健、教育中的实时生产部署、越来越多地管理你家的助手，基本上关注你日常生活中的一切，让你更有效率。所以这是一个很可能会继续的轨迹。

Well, I think if you define chat as an interactive exchange between two parties, one of which in this case is an AI that broadly satisfies some goal. You're looking to learn the sports score. You're looking for advice on which restaurant to go to. You're looking for coaching and feedback on an essay that you've written. You're looking for suggestions about which job to take next or some tough conversation you're about to have with your manager. You get a response. You go back and forth. You have five or six exchanges and you find that a useful output which you might otherwise have to go rely on an expert friend or even pay a coach. There are, I mean just objectively empirically speaking, hundreds of millions of people that get that experience every day from these chatbots. So maybe we could quibble over whether that technically represents human level performance. I think it's a fairly reasonable thing to claim. And I think there's no reason why that isn't going to continue climbing. I mean, the rate of climbing in the last 3 years is the thing that I think is most staggering. And so what we're trying to do from this point is extrapolate: okay, what are the fundamental drivers of that climb? Compute, data, interaction from real world users. And those things look set to continue. So I think that I would expect that they apply to many other domains too, not just chat or emotional support and productivity, but many other domains beyond that: to healthcare, to live production deployments inside of education, to assistants that are increasingly managing your home, looking at everything that is in your everyday life basically to make you more productive. So that's a trajectory that's likely to continue.

嗯，这很有趣。你提到仍然是相同的基本架构，Transformer、注意力机制，我们已经应用算力 15 年了，我们正在获得这些巨大的提升。你在构建方面处于一个相当独特的位置。你宣布了你的第一个旗舰推理模型，MAI Thinking One。你从零开始。

Well this is interesting. You've mentioned now that it's still the same fundamental architecture, transformers, attention, that we've been applying compute for 15 years, we're getting these big increases. You are in a fairly unique spot at build. You announced your first flagship reasoning model, MAI thinking one. You got to start from scratch.

在架构设计和训练这个模型 15 年后，你现在有什么不同的做法吗？还是说，只是收集所有数据，像以前一样运行训练，只是现在算力更多了，所以效果会更好？

Is there anything you've done differently now after 15 years in architecting and training this model or is it just yep, we're going to collect all the data and run the training run just as we did and we have more compute now so it's going to be better?

不，实际上我认为有很多不同之处。首先是我们整理数据的方式。我们从顶层开始：我们付费获取了非常高质量、非常保守的数据集，并去除了大量嘈杂、分散注意力、低质量、有潜在安全风险的问题。我们使用的方法相当专有。我们刚刚分享了一份 109 页的详细技术报告，在 Twitter 上反响很好，其中分享了很多我们如何做到这一点的细节。第二点是，虽然对架构选择保持谨慎很重要，我们也确实做到了，但我们在如何组织训练运行方面也做出了一些相当重大的改变。我们的训练运行非常稳定，很少崩溃，很少重启。我们分享了很多图表来展示基础设施的稳定性以及 MFU 效率。模型浮点运算利用率表明，我们在训练运行的每一步都能通过每个芯片实现最先进的浮点运算次数。这很容易出错，我们听到很多来自不同实验室的关于事情出错的故事。做出谨慎而深思熟虑的选择来把事情做好，并采取正确的方法来生产高质量模型是非常困难的。我们的工作和目标是构建这台爬山机器。这意味着将硅片、模型、超高质量数据以及一系列强化学习环境整合在一起，使我们能够系统地针对我们选择的任何目标进行爬山。这就是 MAI Thinking One。它是一个通用、相当中立的思考模型，在编码方面相当不错。现在它在基准测试上大致与 Opus 4.6 持平。我们还没有大规模部署到生产环境中，所以还有很多工作要做，但它是一个非常强大的推理者，在 AM 上达到 97%，这是基准测试中衡量其推理性能的主要指标。它非常擅长指令遵循。目标是让许多开发者和企业能够使用它，让他们根据自己的用例进行爬山，因为每个公司都有略有不同的目标来构建支持其用例的智能体等。

No, actually I think there's quite a lot of differences. The first thing is the way we curate the data. We start from the top of the stack: we have paid for and acquired an extremely high quality, very conservative set of data, and extracted a lot of the noisy, distracting, low-quality, potentially security risk issues. The methods we use for that are quite proprietary. We just shared a 109-page very detailed technical report, which was very well received on Twitter, sharing a lot of the details on how we do this. The second thing is that while it's important to be cautious with architectural choices, and we have been, there are also a number of pretty significant shifts in how we put together our training runs. Our training runs have been incredibly stable, very few crashes, very few restarts. We shared a lot of those graphs to show infrastructure stability and also MFU efficiency. Model flop utilization shows that we can put a state-of-the-art number of flops through each chip for every step in our training run. It's extremely easy to get this wrong, and we hear lots of stories from different labs about things going wrong. It's pretty hard to make the careful and deliberate choices to get things right and take the right approach to produce high-quality models. Our job and ambition is to build this hill climbing machine. That means the integration of the silicon with the models, with super high quality data, with a stack of reinforcement learning environments that allow us to systematically hill climb against any objective we choose. That's what MAI Thinking One is. It's a general-purpose, fairly neutral thinking model that is pretty good at coding. It's now roughly on par with Opus 4.6, at least on the benchmarks. We haven't deployed it at scale into production, so there's still lots more work to do, but it's an extremely strong reasoner, 97% on AM, which is the primary measure for its reasoning performance on the benchmarks. It's very good at instruction following. The goal is to make that available to many developers and enterprises, allowing them to climb on it for their use cases, because everybody has a slightly different objective in their company to build agents and so on that support their use case.

你在谈论 MAI Thinking One 时提到的一点是，你没有蒸馏任何现有模型，这让我感到惊讶。这是你可以做的事情。你可以接触到 OpenAI 的知识产权。每个人都在蒸馏一切。我们刚刚在这次审判中发现，Grok 是从多个模型中蒸馏出来的。为什么这里不做蒸馏？为什么不跳过去？

One of the things you've noted in talking about MAI Thinking One is that you didn't distill any existing models, which struck me as surprising. This is a thing you could do. You have access to OpenAI's IP there. Everyone's distilling everything. We just found out in this trial that Grok was distilled from a number of models. Why not do distillation here? Why not jump ahead?

通往前沿肯定有很多捷径。如果你拿一个超高质量的模型，并用来自更优模型的高质量指令、答案或输出来打磨你的基础模型，那么模型可能会很快适应那个分布，但很难说它们随后能够超越那个老师。我们非常谨慎，有两个原因。第一，我们想确保能够超越老师，以便在未来几年自己设定前沿。第二，我们真的想建立一家伟大的实验室，这需要很多年，可能未来 2-3 年，但要做到这一点，我们必须能够证明我们可以自己构建每一个组件。我们可以雇佣世界上最好的人才。我们可以通过实际研究而不是仅仅重新实现、复制或蒸馏任何第三方来推动前沿。我们处于一个很好的位置，可以仔细而细致地追求这个目标，因为我们有资源购买 Anthropic 的模型（当它们超越前沿时），我们有资源将 11,000 个不同的模型放入一个铸造厂，这样每个开发者都能获得纯粹的选择权，而且我们有资源继续部署 OpenAI 的模型，这些模型今天非常出色且处于前沿。这是自给自足使命的自然组成部分，我们需要时间才能真正达到绝对前沿，但我认为我们处于一个很好的位置。我们取得了巨大进展。这是一个非常强大的模型，而且我们发布的不仅仅是那个模型。我们同时发布了七个新模型。例如，我们的转录模型 1.5 实际上是世界第一。它是所有超大规模提供商中成本效益最高的。它的准确率最高。我们的图像模型现在是第二。我们的图像编辑模型是第三，仅次于谷歌和 OpenAI。所以我认为我们在图像和音频方面做得很好。我们的代码模型 Code Flash 非常强大，针对 VS Code 进行了优化。它是一个非常棒的模型，与 Sonnet 4.6 相当，所以它处于一个很好的位置。

There are definitely lots of shortcuts to the frontier. If you take a super high-quality model and polish your base model with high-quality instructions or answers or outputs from a superior model, then the model might quickly fit to that distribution, but it's very unclear that they would then be able to surpass that teacher. We've been very deliberate for two reasons. First, we want to make sure we can exceed the teacher in order to set the frontier ourselves over the next few years. Second, we really want to build one of the great labs, and it's going to take us many years, probably the next 2-3 years, but to do that we have to be able to show that we can actually build every component ourselves. We can hire the very best talent in the world. We can push the frontier with actual research rather than just re-implementation, copying, or distillation from any third party. We're in a great position where we can carefully and meticulously pursue that objective, knowing we have the resources to buy Anthropic models where they exceed the frontier, we have the resources to put 11,000 different models inside a foundry so every developer gets pure optionality, and we have the resources to continue to deploy OpenAI models which are outstanding and at the frontier today. That's a natural part of the self-sufficiency mission, and it'll take time for us to truly get to the absolute frontier, but I think we're in a great spot. We made a ton of progress. This is a very strong model, and it wasn't just that model we released. We released seven new models simultaneously. Our transcribe model, for example, 1.5 is literally number one in the world. It's the most cost-effective of any of the hyperscalers. It's the highest on accuracy. Our image model is now number two. Our image editing model is number three, right behind Google and OpenAI. So I think we're well up there with our image and audio. Our code model, Code Flash, is incredibly strong, optimized for VS Code. It's a really great model that's on par with Sonnet 4.6, so it's in a great spot.

蒸馏是否存在任何法律或知识产权方面的担忧？我知道这是一个现实问题。Anthropic 抱怨其他人蒸馏他们的模型。有人担心中国公司蒸馏模型，以及我们现有的知识产权协议是否能涵盖这一点。你有没有因为这些担忧而避免蒸馏？

Were there any legal or IP concerns with distillation? I know this is a live issue out in the world. Anthropic complains about other people distilling their models. There are concerns about Chinese companies distilling models and whether our existing IP agreements can cover that. Did you have any of those concerns to keep you away from it?

不，我们没有。但我理解为什么很多人感到沮丧。Anthropic 一直非常沮丧，还有一些关于 xAI 和 Meta 以及开源模型的传言，因为本质上那是拿另一个团队整合的知识产权和知识，然后直接灌输到你自己的模型中。我认为这是一个短期的胜利。就像我说的，我们真的想在实验室里创造一种文化，让我们能够提出下一个重大的思考突破、下一个重大的编码突破或下一个重大的架构推进。

No, we didn't. But I understand why a lot of people get frustrated. Anthropic have been very frustrated, and some of the rumors around xAI and Meta and obviously the open source models, because essentially that's taking the IP and knowledge that another team has put together and then literally force-feeding it into your own model. I think it's a bit of a short-term win. Like I said, we really want to create a culture in the lab where we can come up with the next big thinking breakthrough or the next big coding breakthrough or the next big architectural push.

现在我们正在实验循环 Transformer，它是当前 Transformer 的一个略有不同的变体。领域内很多人也在研究它。似乎还没有人真正投入生产。但为了打造一种能够真正推动前沿的文化和团队，他们必须根据需要理解、拥有并创造全栈，同时也在需要时使用第三方的东西。例如，我们的论文有数百次引用，这些引用都基于其他文献。所以这很大程度上是对领域的回馈，以回报我们多年来从所有优秀出版物中学到的一切。

Right now we're experimenting with the loop transformer, which is a slightly different variant on the current transformer. Lots of people in the field are looking at it too. No one seems to have quite got into production yet. But in order to create a culture and a team that can really push the frontier, they have to understand, own, and create the full stack as and when they need to, and also use things from third parties whenever we need to. Our paper, for example, has hundreds of citations grounded in the rest of the literature. So it's very much a contribution back to the field in return for everything that we've learned over the years from all the great publications that have been out there.

我想问一下，你理解 Anthropic 和你在 AI 领域的同行对蒸馏的挫败感吗？你也理解创意工作者、出版商和 YouTubers 对所有人 AI 公司集体抓取他们的作品来制造这些模型的挫败感吗？这种挫败感只会越来越强烈。

Can I ask you if you understand the frustration from Anthropic and your peers in AI about distillation? Do you also understand the frustration from creatives and publishers and YouTubers about all the AI companies scraping their work as a collective to make these models? That frustration is only getting louder.

是的，我理解这种挫败感。开放网络挑战是我们之前讨论过的问题，我明白。我看到人们很沮丧，显然这正在通过法庭对话来解决。我看到人们把东西放到网上，他们对上传到网上的契约有不同的期望。这很棘手。

Yeah, I understand the frustration. The open web challenge is one we've talked about before, and I get it. I see that people are frustrated, and obviously that's working its way through the conversation in the courts. I see that people put things online and they had different expectations about what the contract was with that being placed online. It's a tricky one.

你提到你所有的数据都是精心策划的。你为训练新模型使用的所有数据都付费了吗？

You mentioned all your data was carefully curated. Did you pay for all the data that you're using to train the new models?

我们很多数据显然是以正常方式从开放网络获取的。精心策划意味着它经过极其仔细的过滤，以确保安全、质量，以及来自一些开源数据集的第三方依赖。避免很多中国血统的数据，我认为这些数据非常不同。我们的企业希望确保当他们将某物投入生产时，他们可以信任我们，我们确实是根据他们的需求构建的。我认为这是非常谨慎、耐心和关注所有细节的好处之一。

A lot of our data we obviously take from the open web in the normal way. Carefully curated means that it's extremely carefully filtered for security, for quality, for third-party dependencies from some of the open-source data sets. Keeping it away from a lot of the Chinese lineages, which I think are very different. Our enterprises want to make sure that when they put something into production, they can trust us that we've really built it with their needs in mind. I think this is one of the benefits of being very deliberate and patient and attentive to all the details.

你提到了企业。我觉得这非常有趣。微软全力投入企业 AI，规模很大。实际上，我甚至可以直接说，Xbox 的新负责人 Asha Sharma 正在很多地方取消 AI，而游戏玩家很高兴。消费者领域对 AI 有一种反应。企业领域有另一种。我认为 AI 在企业领域的产品市场契合度，就像 AI 这样快速变化的东西所能达到的那样接近。公司控制着很多数据库，你可以直接访问它们，因为它们控制着。那是他们的数据。有很多可重复的流程、任务和旧系统，模型可能可以更高效地完成。企业正在发生非常重要的事情。与此同时，消费者对 AI 的反感正在增加。我的论点是，我们还没有构建出伟大的消费者 AI 产品。这个行业没有生产出它们。没有转变它们。没有让这一切的价值变得显而易见。使用来自开放网络的所有数据，改变面向大众的出版契约，现在这些数据被用于训练模型，这些模型将为公司带来数万亿美元的价值。没有一个产品说这是值得的。萨提亚·纳德拉最近接受了 Axios 的采访，他说我们需要为此获得社会许可，在我们获得许可之前，在我们交付价值之前，人们会一直有这种感觉。我们看到大学演讲者被嘘。我们看到数据中心被禁止。你认为有值得的消费者产品吗？值得为训练焦虑，值得为数据中心焦虑。那曾经是你的重点。现在你的重点是企业。我可以说，从表面上看，微软似乎不再对消费者产品感兴趣了。但你看到有值得的产品，或者可以构建的产品吗？

You mentioned enterprise. I think this is very interesting. Microsoft is all-in on enterprise AI in big ways. Actually, I would even draw the line straight to Asha Sharma, the new head of Xbox, is getting rid of AI in a bunch of places and the gamers are happy. There's one reaction to AI in consumer space. There's another in enterprise. And I think AI has as close to product-market fit in enterprise as you can get with something as changing as fast as AI. There are a bunch of databases that corporations control and you can just go access them because they control them. That's their data. There's a bunch of repeatable processes and tasks and old systems that maybe the models can just do more efficiently. There's something very important happening to enterprise. At the same time, the consumer antipathy towards AI is just increasing. And my argument is we have not built great consumer AI products. This industry has not produced them. It has not shifted them. It has not made it obvious that all of this is worth it. That using all the data from the open web and changing the contract of publishing to a mass audience of people, so now it's being used for training of models that will deliver trillions of dollars of value to corporations. There isn't a product that says this is worth it. Sati Nadella recently gave an interview with Axios and he said we need social permission for this and until we have it, until we deliver that value, people are going to feel this way. We've seen college speakers get booed. We've seen data centers get banned. Do you think that there's a consumer product that's worth it? That's worth the angst about training, that's worth the angst about data centers. That was your focus. Now your focus is enterprise. I would say that just on the face of it, it doesn't seem like Microsoft has interest in the consumer product anymore. But do you see one that's worth it or that could be built?

我不太同意你的说法，认为消费者没有从中获得任何价值。所有聊天机器人加起来有数十亿，每月有数十亿人从中获得巨大价值。稍微同情一下小企业主，或者帮助孩子做作业的妈妈，她们现在可以转向对话式 AI，获得反馈、指导、设置论文问题。能够问诸如如何创收、如何编制现金流预测、应该申请哪所大学之类的问题。这些都是日常任务，现在能获得相当高质量的事实建议和信息。所以我不太相信人们没有从这些东西中受益。我认为他们确实受益了。

I'm not sure I agree with you that there hasn't been any value for the consumer out of this. There are billions across all of the chatbots, billions of people a month that are getting immense value out of it. Just for a moment, empathize a little bit with the small-scale business owner or the mom that's helping her kid with the homework and can now just turn to a conversational AI and get feedback, get instructions, get essay questions set. Just being able to ask questions about how do I generate revenue, how do I put together a cash flow forecast, which college should I apply to. These are everyday tasks that are coming with some pretty high quality factual advice and information. So I don't really buy that people are not getting benefit out of these things. I think they are.

但我认为我可以非常清楚地论证，他们没有获得足够的好处。正是他们在说我们不应该有更多的数据中心。正是他们在毕业演讲中嘘 AI。民调很清楚，尤其是年轻人。他们使用 AI 越多，对它的反感就越强。这在每一项民调中都很清楚。这就是我的论点。不是说没有价值，而是价值交换不够清晰。

But I think I can very clearly make the argument that they're not getting enough benefit. They're the ones saying that we should not have more data centers. They are the ones booing AI at the graduation speeches. The polling is clear, particularly young people. The more they use AI, the more antipathy they have towards it. That's clear in every single poll. That's the argument I'm making. Not that there's no value, but the value exchange is not clear enough.

是的，有道理。

Yeah, fair enough.

我看到微软尤其从大型搜索产品转向企业，那个让谷歌跳舞的 Bing 重塑已经结束了，我们都专注于有价值的企业领域。我只是想知道，是否有足够多的消费者价值让这一切值得。

I'm seeing Microsoft in particular pivot to enterprise away from the big search product, the reinvention of Bing that would make Google dance. That's over and we're all focused on enterprise where the value is. I'm just wondering if there's value enough for the consumer to make all of this worth it.

是的，我认为有很多焦虑是可以理解的。关于未来 5 到 10 年会发生什么，有大量的猜测，无论是被描述为奇点还是工作末日。这些都不是有帮助的框架。我认为人们害怕是因为它定义不清，而且经常被描述为悬在人们头顶上不可避免的威胁性灰色云团。我认为重要的是我们用技术做什么。我长期以来一直主张，我们必须把人放在第一位。

Yeah, I think there's understandably a lot of anxiety. There's enormous amount of speculation about what's going to happen in the next 5 to 10 years, whether it's framed as the singularity or the job apocalypse. These are not helpful framings. I think that people are scared because it's poorly defined and it's often framed as an inevitable threatening gray cloud over people's heads. I think that what matters is what we do with technology. And I've for a long time argued that we have to place the human first.

领域内有些人把科学发现放在首位，或者把加速能够探索星系的智能放在首位，并说我们注定会拥有比我们所有人加起来都更强大的 AI。这自然让人害怕。我认为我们必须反过来看：科学和技术的目的是让我们所有人更健康、更聪明、更幸福。这是我们作为物种在数千年发明中一直追求的使命。这也是我们应该再次用来检验超级智能的标准。如果它达不到这个标准，我认为人们会拒绝它，而且他们拒绝它是正确的。我认为在未来五年内，每个人的关注点都会转向：这如何让我更健康、更幸福、更聪明、更有能力、更高效？如果它做不到，人们自然会愤怒、抵制和反抗。我认为这没有什么意外或错误之处。这是不可避免的。所以这就是为什么多年来我热衷的事情之一是医疗健康。就在几天前，我们宣布了与梅奥诊所的新合作。这是世界上排名第一的医院，一直如此。他们拥有跨所有模态的最高质量纵向患者记录数据集。他们有最好的临床实践。他们也是一家非营利机构，我想很多人没有意识到这一点。他们 65%的患者是医疗补助计划覆盖的人群。人们常常把他们与那些乘国际航班来获取最佳护理的超级精英联系在一起，但实际上他们大多数患者是医疗补助覆盖的。他们是一个了不起的机构，有着在各地提供最佳医疗健康的使命。我们现在有一个非常长期的合作，从零开始用他们的数据和我们的模型共同训练一个全新的健康基础模型，部署在他们的医院里，并希望推广到世界各地，为尽可能多的人提供最好的临床护理和医疗健康。这就是我进入这个领域的原因。这是我最初的动力。这是我所热衷的。我只能专注于那些我认为能带来改变、能帮助人们、能为每个人留下良好遗产的事情。这就是我们正在努力做的。

Some people in the field have placed scientific discovery first or placed accelerating intelligences that can explore the galaxies and so on and said that it's inevitable that we're going to have these AIs that are going to be more powerful than all of us combined. That's naturally scary to people. I think that we have to flip it the other way round and say the purpose of science and technology is to make us all healthier and smarter and happier. That's been the quest that we've been on as a species for thousands of years of invention. And it's the test that we should put superintelligence to again. If it doesn't achieve that test, then I think people will reject it and they'll be right to reject it. I think that everybody's focus is now going to turn in the next five years to how is this making me healthier and happier, smarter, more capable, more productive. If it's not doing that, then naturally people are going to be angry and resist and react. I don't think there is anything unexpected or wrong about that. That's inevitable. So that's why one of the things I've been passionate about for many years is healthcare. Just a couple days ago we announced a new partnership with Mayo Clinic. This is the number one hospital in the world consistently reported. They have the highest quality longitudinal patient record data set across all modalities. They have the best clinical practice. They are also a nonprofit, which I think a lot of people don't realize. 65% of their patient population is on Medicaid. People often associate them with the super elites flying in internationally to get the best care, but they actually have majority on Medicaid. They're an amazing institution with an incredible mission to deliver the best healthcare everywhere. We now have a very long-term partnership to co-train from scratch with their data and our models a brand new foundation model for health, deploy it in their hospitals, and hopefully take it around the world to deliver the best clinical care and healthcare that we possibly can to as many people as possible. That's why I got in the field. That's what I was originally motivated by. That's what I'm passionate about. I can only focus on the things that I think are going to make a difference and that will help people and leave a good legacy for everybody. That's what we're trying to do.

我很感激，也理解医疗健康的框架。我明白为什么这是每个人的首选。尤其是在美国，如果你能让医疗健康改善哪怕 10%，你都会以特别深刻的方式影响很多人的生活。问题是，我认识一个非常聪明的人，他对这一切有着截然不同且激进得多的方法。那个人就是四个月前的你。这是穆斯塔法·苏莱曼四个月前对《金融时报》说的：「白领工作，当你坐在电脑前，无论是律师、会计师、产品经理还是市场营销人员，这些任务中的大部分将在未来 12 到 18 个月内被 AI 完全自动化。」那是四个月前。这意味着一年后，律师、会计师、产品经理和市场营销人员将没有工作，对吧？他们的工作将被自动化。这仍然是你的时间表吗？

I appreciate that and I appreciate the healthcare framing. I understand why that's everyone's go-to. Healthcare in America in particular, if you could make it even 10% better, you will have affected a lot of people's lives in a particularly profound way. The thing is I know a very smart guy who has a very different and vastly more aggressive approach to all of this than you. That person is you four months ago. This is what Mustafa Suleyman said to the Financial Times four months ago: 'White collar work when you're sitting down at a computer either being a lawyer or an accountant or product manager or a marketing person. Most of these tasks will be fully automated by an AI within the next 12 to 18 months.' That's four months ago. That implies that a year from now lawyers, accountants, product managers, and marketing people will not have jobs, right? Their jobs will be automated. Is that still your timeline?

好的。不，不，不。等一下。在你刚才引用的那句话里，我说的是「任务」。我说的是「任务」。这并不意味着「工作」。在劳动经济学中这是一个非常重要的区别。一个角色或职能在组织中有完整的子组件分类。发送电子邮件、与同事交谈、制作 PPT——这些子任务将越来越数字化和自动化。我们可以生成越来越多的这类任务。这并不一定意味着这个角色会消失。它只是意味着工作可以更快、更高效地完成，而这些工作往往是相当机械、手动、劳动密集且耗时的。技术的自然演进是让生活更轻松、更快、更少摩擦、更无缝。正如大家经常抱怨的那样，这让我们所有人更忙了。实际上它让我们更可用、更紧张、给了我们更多信息。效率总是有这些报复性效应，我认为人们忘记了。很可能我们会变得高效得多，因为我们花更少时间做狭隘的行政琐事，而花更多时间做创造性的、需要判断的事情，这些最终创造更多价值。我们也可以更快地实验。我们将能够并行尝试很多事情，因为执行成本会降低。在我看来，这可能会提高整体质量，因为我们会在新闻、商业或任何我们做的事情中尝试更多假设。所以我认为它们有点被断章取义了，因为人们自然混淆了工作和任务。不过，你可以反驳我说：「好吧，那么五年、十年或十五年后会是什么样子？」而我认为这正是我们需要回应的。

Okay. No, no, no. Hold on a sec. I said tasks in the quote that you've just said. I said tasks. That does not mean jobs. Very important distinction in labor economics. There is an entire taxonomy of subcomponents of a role or a function in an organization. Sending an email, having a conversation with a colleague, putting together a PowerPoint—subtasks will increasingly become digitized and automated. We can generate more and more of them. That does not necessarily mean that the role goes away at all. It just means that the work can be done faster and more efficiently, which is often work that is quite rote, quite manual, quite labor intensive, time consuming. The natural progression of technology is to make your life easier, faster, less frictionful, more seamless. As everyone often complains, that has made you and me and everybody else much more busy. It's actually made us more available, more stressed, given us more information. There are always these revenge effects of efficiency which I think people forget. It's quite likely that we're going to get made much more productive because we spend less time doing narrow administrative menial tasks and we have to spend more time doing creative judgment focused things which ultimately create a lot more value. We can also experiment much more quickly. We'll be able to try lots of things out in parallel because the cost of execution is going to get lower. In my mind, that's likely to increase the overall quality of things because we're going to try out more hypotheses whether in journalism or in business or in anything that we do. So I think that they were sort of taken out of context because of a natural misunderstanding between jobs and tasks. Nevertheless, you could push back at me and say, 'Okay, well then what does the landscape look like in five or 10 or 15 years time?' And that's where I think we have to respond.

但实际上，我能——我不会那样反驳你。我会以一种非常具体的方式反驳。我意识到这是你的话，你说它被误解了。我只是看这个字面句子，里面没有区分任务和子任务。它说的是「白领工作」。例子是律师、会计师、产品经理、市场营销人员。然后你说「这些任务中的大部分将在未来 12 到 18 个月内被 AI 完全自动化。」那里没有子任务的区分。所以这是——你在说大多数律师的工作将被完全自动化，法律实践将在一年内完全不同。即使按照那句话的字面意思，我只是问，你仍然认为成为律师会完全不同，因为智能体会到处运行，做我们以前做的一切吗？

But actually, can I—I'm not going to push back on you in that way. I'm going to push back in a very specific way. I realize this is your quote and you're saying it was misinterpreted. I'm just looking at this literal sentence and there is no distinction between tasks and subtasks. It is 'white collar work.' The examples are lawyer, accountant, product manager, marketing person. And then you said 'most of these tasks will be fully automated by an AI within the next 12 to 18 months.' There's no distinction of subtasks there. So this is—you're saying most lawyers will have their jobs fully automated and the practice of law will look totally different within a year. Even by the words of that quote, I'm just saying, are you still on that timeline that being a lawyer will look totally different because agents will be running around doing everything that we were doing before?

任务是其中的组成部分，这是劳动市场经济学文献中几十年来确立的定义。也许对《金融时报》来说都过于微妙了，但无论如何，这就是意图。现在，我确实认为有一个重要问题：从长远来看，这会把我们带向何方？这将充满挑战。我们可以在时间线上争论——是几年、十年还是二十年——但现实是，我们将越来越多地自动化工作、任务、岗位、角色、活动以及我们所做的一切。因此，更重要的是我们围绕这些技术建立的治理：它们对谁负责，谁拥有它们，有哪些反馈循环来调节并引入摩擦，以确保它们真正服务于人类。四五个月前，我写了一篇关于人本超级智能的文章，相当直接地阐述了我认为的基本北极星——也许不完全是框架，而是一套原则，基本上说技术是来服务我们的。这是我们应该用来检验的标准。这是人们会用来检验的标准。这是我们在微软关心的标准。我认为越来越多的人将不得不真正关注这个问题，因为它将带来巨大的好处。我们希望继续这样做，但希望以一种不会在过渡期造成巨大不稳定的方式来进行。

Tasks are the components of that, and it's an established definition in the literature of labor market economics for many decades. It was maybe too nuanced even for the Financial Times, but nevertheless that was the intent. Now, I do think there's an important question around where that leaves us in the longer term. It is going to be challenging. More and more of this stuff—we can quibble over the timelines, whether it's a few years, a decade, or 20 years—but the reality is we are going to be automating more and more of the work, tasks, jobs, roles, activity, and everything that we do. So what's going to matter more is the governance that we put around these technologies: who are they accountable to, who owns them, what are the feedback loops that regulate and introduce friction to make sure that they actually serve people. I wrote an essay on humanist superintelligence outlining quite directly four or five months ago what I think of as basically a north star—maybe not quite a framework, but a set of principles that basically says technology is here to serve us. That's the test that we should put it to. It's the test that people would put it to. It's the test that we care about at Microsoft. I think that more and more everyone is going to have to really focus on that question because it is going to deliver a tremendous amount of good. We wanted to continue doing that, but we wanted to do it in a way that doesn't cause ridiculous amounts of instability during the transitionary period.

我相信你。我知道你思考这个问题很久了，但我要以我知道我的听众希望我回应的方式来回应，因为我一直听到这种说法。看起来整个行业——不是你，包括所有人——都全力投入「我们要取代所有工作」，并真正加速建设大规模数据中心，要求大量资源以兑现重大承诺。后来出现了政治反弹，现在所有立场都软化了。你说「并非所有工作都会消失，我们必须重新思考工作」，这与该行业所有其他 CEO 说类似的话以及现在每次都会提到的医疗保健如出一辙。我想知道政治反弹是否真的改变了你谈论这个问题的方式。你的许多同行认为 AI 只是存在营销问题，没有得到有效沟通，他们应该花费数亿美元在播客上更有效地宣传 AI 的好处。这是这个行业正在发生的真实情况。你认为 AI 只是存在营销问题，政治反弹让你意识到了这个营销问题，还是你认为还有其他原因？

I believe you. I know you've been thinking about this for a long time, but I'm going to respond in the way that I know my audience wants me to respond, because I hear it all the time. And what it looks like is this whole industry—not you, everybody included—went all in on 'we're going to replace all the jobs' and really accelerated building out data centers at massive capacity and asking for a lot of resources against big promises. There was political pushback, and now all the stances have softened. And you saying 'it's not all jobs are going away, we have to rethink jobs' is of a piece with all the other CEOs in this industry saying similar things and talking about healthcare that comes up every single time now. I'm wondering if that political pushback has actually changed how you are talking about this. There are a lot of your peers who think AI simply has a marketing problem, that it hasn't been communicated effectively enough, and they should spend hundreds of millions of dollars on podcasts to communicate the benefits of AI more effectively. This is a real thing happening in this industry. Do you think AI simply has a marketing problem and that the political pushback has opened your eyes to this marketing problem, or do you think there's something else going on?

这里有一系列问题。首先是我实际上怎么想、怎么认为，以及过去六个月是否有改变？答案是没有。三年前我写了一本非常详细的书，提前警告了许多当前正在发生的事情。我这样做是为了明确揭示巨大的风险：对监控、权力集中、财富集中、国家去中介化、民主威胁、人类本质的威胁，以及在某种意义上这些全新硅基生命形式到来时作为人的意义。我在医疗保健领域工作了十多年，多次推动放射学、乳腺 X 线摄影、病理学以及电子健康记录等许多其他领域的尖端突破和贡献。所以我一直相信技术的目的是让我们更健康、更幸福，而这些是我选择投入时间和精力去做的事情。这个行业是否存在声誉和公关问题？我认为很明显人们非常焦虑，非常沮丧，未来几年这方面会受到很多关注，这可以理解。但我认为我们能做的是对我们所构建的东西、构建方式、将何种技术推向世界的决策以及我们选择解决的问题负责，就像我们与梅奥诊所合作的那样。

There's a series of questions there. The first is what do I actually think and believe, and has it changed in the last six months? The answer is no. I wrote a very detailed book about this three years ago, way ahead of time, warning about many of the things that are currently happening. I did so explicitly to lay on the table tremendous risks: to surveillance, to concentration of power, to concentration of wealth, to disintermediation of the state, to threats to democracy, to threats to the nature of the human and what it means to be a person in the context of the arrival of these very new forms of silicon being in some sense. I've been working on healthcare for over a decade and pushed many times on some of the cutting-edge breakthroughs, contributions to the field in radiology, mammography, pathology, and many other areas, electronic health records. So I've always believed that the purpose of technology is to make us healthier and happier, and those are the things I choose to work on and direct my time to. Does the industry have a reputation and PR problem? I think it's pretty clear that people are very anxious. They're very frustrated, and there's going to be a lot of attention on that in the next few years, understandably. But I think what we can do is take accountability for the things we build, the way we build them, the decisions we make to put types of technology out in the world, and the types of problems we choose to work on, like we are doing with the Mayo Clinic.

我想指出，我认为你和我第一次见面交谈是在你加入微软之前，就在那本书出版之后，我们一起参加了一个小组讨论。所以我之所以能自如地问这个问题，部分原因是我知道你思考这个问题很久了，而且我知道那本书。对我来说，问题是整个行业是否误判了它能提供的总价值，以克服人们现在所反应的看似鲁莽的行为——人们对资源要求的反应。你正在构建新模型。微软内部可能存在权衡：是利用现有的 Azure 基础设施向客户收费，还是花钱训练新模型。这看起来很像人们在社区中关于资源的讨论：我们是应该利用现有的能源足迹来构建新 AI，还是做其他可能更有直接价值的事情？你怎么看待这一切？你是这个行业的领导者之一。你想与推动最大变革的公司一起站在前沿。你认为如何以一种不仅承诺未来结果，而且立即为社区提供利益、让人们希望你存在的方式来要求这些资源？

I want to point out that I think the first time you and I ever met and talked was before you joined Microsoft, right after that book came out, and we did a panel together. So one of the reasons I'm comfortable asking this is because I do know that you've been thinking about this for a long time, and I'm aware of that book. I think for me, the question is whether the industry as a whole misjudged the total amount of value it could provide to overcome the seeming recklessness that people are now reacting to—the ask for resources that people are now reacting to. You're building new models. There's probably a trade-off inside Microsoft between using the existing Azure footprint to charge customers money or spending money to train new models. That kind of looks like the same conversation people are having about resources in their communities: whether we should use the existing energy footprint to build new AI or do something else that might be more immediately valuable. How do you think about all of that? You are one of the leaders of this industry. You want to be on the frontier with the companies driving the most change. How do you think about asking for those resources in a way that isn't just promising future results, but also immediately providing benefit to communities in a way that makes people want you to be there?

我认为我非常自豪微软坚持了其净零目标。我们的新数据中心全部采用液冷。这意味着它们在六年内消耗的水量大约相当于一家餐厅的用水量。就像一个游泳池，注满水后就在系统中循环。它们在电力消耗方面基本上都是可再生能源。所以我认为这样的承诺——例如，我们最近承诺确保受数据中心电力需求变化影响的当地社区得到补偿和保护，这样他们的能源账单就不会飙升。

I think that I'm very proud that Microsoft has stuck by its net zero targets. Our new data centers are all liquid cooled. This means that they use about a restaurant's worth of water for a six-year period. It's like a swimming pool that gets filled up with water and then just circulates around the system. They're all largely renewable in terms of their electricity consumption. So I think commitments like that—for example, we made a commitment recently to ensure that local communities affected by shifts in electricity demand by our data centers are compensated and protected so that they don't see a spike in their energy bills.

这些正是微软作为一家负责任的公司所做的事情，并且可以继续做下去，真正关注对社区的影响。另一方面，变革之所以发生，是因为人们在各个层面的参与。公司内部的人必须做出不同的决定。抗议和游说的人必须做出决定，努力走出去，发出自己的声音，并参与政治进程。这就是我们作为一个物种集体进化并推动事情向前发展的方式。月复一月，季度复季度，我们感觉彼此之间总是意见不合。但当你回顾十年又十年，我们就像是一个由各种不同激励因素组成的奇怪集体网络，实际上正在把事物推向正确的方向。尽管有焦虑和两极分化，我认为我们正在建设一些东西，这些将使我们的物种更健康、更快乐、更有能力。我们必须确保在通往那里的道路上走对方向，因为有很多陷阱和可能出错的方式。但正确的道路包括让人们发出自己的声音，并根据回应和反应改变方向。所以，这正在发生是件好事，这个过程正在按预期运作。

Those are the kinds of things that Microsoft does and can continue doing as a responsible company to really pay attention to the consequences for communities. On the flip side, change happens because people participate at every level. People inside companies have to make different decisions. People who protest and campaign have to make decisions and make the effort to go out, make their voice heard, and be involved in a political process. That's how we as a species collectively evolve and move things forward. Month to month, quarter to quarter, it feels like we're all at odds with one another. But when you look back decade over decade, we're like this collective weird mesh of all sorts of different incentives that are actually nudging things in the right direction. Despite all the angst and polarization, I think we're building something that is going to make our species much healthier, happier, and more capable. We have to make sure we get the right path on the way there because there are lots of pitfalls and ways it can go wrong. But the right path involves people making their voices heard and people changing course based on response and reaction. So it's a good thing that's happening, and that's the process working as intended.

让我问问你关于企业方面的问题。我们花了很长时间讨论消费者方面以及人们的感受。在企业方面，我们看到很多公司正在发现这些工具的真正价值。亚马逊基本上撤下了一个排行榜，因为人们作弊使用比所需更多的 token。我们看到一些公司超出了他们的 token 预算。优步刚刚撤回，因为他们用完了全年的 token 配额，却没有看到任何价值。你现在如何看待这一方面？企业中有如此多的兴奋和变革渴望，特别是在软件工程领域——有些人觉得有趣，有些人则经历着全面的存在危机——而价值尚未实现。我们开始看到纯粹的 token 最大化并不能带来你期望的那种价值。你如何看待那里的使用情况？因为如果你在企业中证明了这一点，它也会在其他方面显现出来。

Let me ask you about the enterprise side. We spent a long time on the consumer side and how people feel. On the enterprise side, we're seeing companies figure out how valuable these tools are. Amazon basically took down a leaderboard because people were cheating to use more tokens than they needed. We've seen some companies blow out their token budgets. Uber just pulled back because they had blown through their token allocation for the year and weren't seeing any value. How do you think about that side right now, where there's so much excitement and desire for change in the enterprise? In particular, software engineering—some people are having fun, others are having full existential crises—and the value hasn't been realized yet. We're beginning to see pure token maxing does not deliver the same kind of value you'd expect. How do you think about the use there? Because if you prove it out in enterprise, it will come out in other ways.

不同的人报告不同的事情。确实有一些人过度使用编码模型，生成无用的代码和 token。但许多人的工作和影响力已经完全被它改变了。毫无疑问，这对软件工程行业产生了巨大的有益影响。我们正在整个技术栈中生产更高质量、更快速的代码。所以，确实有一些人可能搞错了，没有设置正确的 token 预算，过程中会有错误。我不认为这表明没有采用或人们看不到价值。从我这里看，价值是不可思议的。很多人每天都告诉我，这正在改变他们的工作产出和生产力。另外要说的是，这些事情是浪潮式的。有一波能量，变得泡沫化，几个月后人们退缩，意识到那并不是真正的东西，然后他们转向稍微不同的方向。所以这是曲折而有机的，这是不可避免的。有很多兴奋，所以人们在推特上做出大声明，但稳步的进步看起来非常线性和连续。

Different people report different things. There are examples of people overusing coding models, generating useless code and tokens. But many people's work and impact have been completely transformed by it. There's no question that this has had a massively beneficial impact on the software engineering industry. We are producing much higher quality, much faster code across the entire stack. So, there are examples of some people who maybe got it wrong, didn't set the right token budgets, and there will be mistakes along the way. I don't think that's any signal that there isn't adoption or people don't see value. The value from where I'm sitting is incredible. Many people tell me every day that it's transforming their work output and productivity. The other thing to say is that these things happen in surges. There's a swell of energy, it gets frothy, people pull back a few months later and realize that actually isn't the thing, then they head in a slightly different direction. So it's meandering and organic, and that's inevitable. There's a lot of excitement, so people make big claims on Twitter, but the steady march of progress looks very linear and continuous.

我总体上同意这一点。但在我看来，不线性的是计算机的形态因素。现在可能比过去十年任何时候都有更多的形态因素实验。至少在过去十年里，我们基本上固定在了智能手机上。现在我们看到各种 AI 可穿戴设备。奇怪的眼镜也许会成为每个人的最爱。我对此持怀疑态度。微软在 Build 大会上展示了一些新设备。有一个控制智能体的徽章，还有一个类似桌面小工具的东西来控制智能体——我是 Chumby 的忠实粉丝，它是我首先想到的东西。所有这些，我看着它们想：计算在哪里？逻辑在哪里？现在这变得不确定了，而不仅仅是线性的进步。我所有的计算都发生在云端，基于云的应用，只是智能体在云端其他地方存储的数据之间运行，我只需要一张挂在挂绳上的信用卡来发出指令。这改变了整个计算架构，可能在很多方面改变现代文明的整个架构。如果我们都不再拥有智能手机，你怎么看？这会走向何方？是悬而未决，还是会有混合方法？你认为合适的最终阶段是什么？

I agree with that on the whole. Where it doesn't look linear to me is in the form factors of computers. There's probably more form factor experimentation right now than at any point in the last 10 years. We mostly settled on a smartphone for at least the last 10 years. We're seeing different AI wearables. Weird glasses maybe will be everyone's favorite device. I have my doubts. Microsoft showed off some new devices at Build. There was the badge that controls an agent and the little desktop-friendly thing that controls an agent—I was a big Chumby fan, it was the first thing that came to mind. All of those to me, I look at them and think: where does the compute live? Where does the logic live? That's up for grabs now in a way that isn't just the linear march of progress. All of my computing happens in the cloud on cloud-based applications, and it's just agents running around to data stored elsewhere in the cloud, and all I need is a credit card on a lanyard to issue instructions. That changes the entire architecture of computing, might change the entire architecture of modern civilization in many ways. If we don't all have smartphones, how do you think about that? Where is that going? Is it up for grabs, or will it be a hybrid approach? Where do you see the appropriate end stage?

这非常有趣。我认为两件事会同时发生。边缘设备将变得更加强大，而云仍然是最大模型的主要驱动力。你的智能体将越来越聪明，知道它可以在设备上回答「法国首都是什么」这样的问题，无论是在你的眼镜、腕带、徽章还是耳机上。然后它会知道什么时候它不知道。它会知道这实际上是一个相当复杂的问题，或者是一个需要生成一系列步骤的动作，或者需要编写新的代码，然后它会转向云端。所以这种切换混合的东西将非常重要。另一件事是，我们在过去三四个月已经看到，我们可以拥有相当强大的本地机器，可以进行异步后台处理。它们可以持续监控系统，如果你需要的话。它们可以执行那些可以花费 10 小时的任务，运行速度比在超级计算机中慢得多。所以自然，当我们被需求淹没时，这些需求会找到许多角落和缝隙来得到满足。实际上，我对我们正在构建的徽章感到非常兴奋。

It's very interesting. I think both things are going to happen at the same time. The edge is going to get way more powerful, and the cloud is still going to be the primary driver of the largest models. Increasingly, your agent will be smart enough to know that it can answer the question 'what is the capital of France' on device, whether it's on your glasses, wristband, badge, or earbuds. And then it will know when it doesn't know. It'll know that this is actually a pretty complicated question or an action that requires a whole bunch of sequences of steps to be generated, or requires novel code to be written, and it will turn to the cloud. So this kind of switching hybrid thing is going to be super important. The other thing is we've already seen in the last three or four months that we can have pretty powerful local machines that can do async background processing. They can constantly monitor systems if you need them to. They can do tasks that can afford to take 10 hours, run much more slowly than they would in a supercomputer. So naturally, when we're swamped with demand, that demand finds loads of nooks and crannies to get satisfied by. I'm actually very excited by the badge we're building.

我的意思是，这很酷。这项技术基本上大公司里人人都有。它已经 25 到 30 年没有进化了。我们肯定得戴着它，它是由公司本身、由首席信息安全官提供的。所以提升它的水平，实际上把它变成一个很酷的、可编程的开放平台，让别人可以在上面构建东西，我认为这是个好主意。我认为这会成功，所以我对此非常兴奋。

I mean, it's pretty cool. This is a technology that basically everyone in a major company has. It hasn't evolved in 25-30 years. We definitely have to wear it, it's provided by the company itself by the CISO. So upleveling that and actually making it a pretty cool open platform that's programmable that other people can build on top of, I think it's a cool idea. I think this is going to work, so I'm very excited by it.

是啊。让我印象深刻的是，你不可能把一堆高性能的本地算力放在一个徽章里，对吧？所有算力都在别处。

Yeah. The thing that strikes me is there's no way you can put a bunch of high-powered local compute in a badge, right? All the compute is elsewhere.

是的。不，不，你肯定会有一些本地算力。你会有一个本地分类器，就像你现在在耳机上做的那样。你会有本地分类器。它会有唤醒词。它会有自己的摄像头。所以，我认为这些东西会越来越成为处理能力的容器，这种处理发生在一个嵌套链中，链上的设备能力逐渐减弱，直到最末端。

Yeah. No, no, you're definitely going to have some local compute. You're going to have a local classifier just as you do on your earbuds at the moment. You're going to have local classifiers. It's going to have wake words. It's going to have its own camera. So, I think that increasingly these things are just going to become vessels for processing power that happens in a kind of nested chain of increasingly less powerful devices to go right to the end point.

是啊。你认为手机在这方面有未来吗？我的意思是，你知道，现在正是 IO 和 WWDC 期间。这些是控制手机平台的大公司。他们喜欢谈论手机平台如何保持中心地位。我从很多人那里听到的论点是，实际上 AI 是一场平台转变，可能会完全取代手机。

Yeah. Do you think the phone has a future in that? I mean, you know, build is right in the middle of IO and WWDC. These are big companies that control phone platforms. They love talking about how phone platforms will stay at the center. The argument I hear from so many is that actually AI is a platform shift that might totally displace the phone.

我认为技术史告诉我们，基本上随着东西变得更有用，它们变得更便宜，更普及，并催生出新的技术用途。所以我认为我们已经太习惯手机了，每个人都假设这将是未来历史的锚定设备。但实际上，你手机的许多特性和功能将被去中介化、拆分开来，并存储在更小的设备上。目前，在我看来，手机的主要功能是验证。它充当你的身份证。通过人脸识别让你进入各种不同的环境。我想你可以想象，那会是一个更便宜、更小、更安全的设备，让你与手机分离，然后通过语音甚至一系列环境传感器进行通信，你的 AI 并不真正存在于某个设备上。它实际上只是无论你在哪里都跟着你。出现在浴室镜子上，任何地方。我认为你可以想象它感觉更加沉浸。不是在接下来的三到五年，而是更远的未来。而且我认为支持那种加密但分布式的智能体出现的基础设施，很可能最终会在 2030 年代出现。

I think the history of technology teaches us that basically as things get more useful, they get cheaper, they proliferate and they spawn new uses of technology. So I think we've become so used to the phone that everyone just assumes that this is going to be an anchor device for the rest of history. But actually many of the features and functionality of your phone are going to get disintermediated, broken apart and stored on smaller devices. Right now the primary function that the phone is playing in my opinion is verification. It's functioning as your ID card. Doing your face recognition to authenticate you into various different environments. I think you can well imagine that being a much cheaper, smaller, secure device which disconnects you from your phone and then communication taking place via voice or even via a series of ambient sensors where your AI doesn't really live on a device. It's actually just with you wherever you are. Appearing on the bathroom mirror, wherever it is. I think you can imagine it feeling much more immersive. Not in the next three to five years but looking much further out. And I think that the infrastructure to support that kind of encrypted but distributed appearance of agents is probably going to end up emerging in the 2030s.

是啊。让我问两个最后的问题来总结。你提到我们一直在使用相同的架构。我有很多未解决的问题，关于 LLM 是否是通往 AGI 的道路。我要指出的是，它们实际上什么都不知道。在这一点上，甚至微软研究院也指出它们什么都不知道，这导致在某些类型的应用中会出现某些错误。LLM 是通往 AGI 或超级智能的道路吗？

Yeah. Let me ask you two final questions to wrap up. You mentioned that it's the same architectures that we've been using. I have a lot of open questions about whether LLMs are the path to AGI. The things I would point to is they don't actually know anything. At this point even Microsoft research is pointing out that they don't know anything and that leads to certain kinds of mistakes in certain kinds of applications. Are LLMs the path to AGI or superintelligence?

听着，我认为我们可能还需要几个重大突破，但这并不意味着我们在未来几年会看到性能提升放缓，我认为这对人们来说是一个难以把握的区别。有一点要说的是，大多数任务的人类水平性能仍然离超级智能很远。超级智能是一个通用学习器，基本上可以立即理解一个全新的、分布外的领域。所以它需要能够从头开始在一个新环境中学习，因为它存储了有价值的知识、概念知识的表征。而目前我们还没有真正完全测试过这一点。智能体不是通用的。它们实际上虽然广泛且经常集成，但有点领域特定。我们用它来聊天。我们用它来编码。我们用它来处理图像或音频。现在，显然作为人类，我们做许多其他更广泛、更多样的任务。我认为这就是为什么人们在推动世界模型和更沉浸的真实世界交互智能体，它们能看到我一天中任务或体验的完整分布。所以我认为，在未来三年、下一个数量级的算力中，这足以让我们走得很远，但在此之上的完全超级智能仍然是一个悬而未决的问题：LLM 是否足够，还是我们需要其他东西。我认为它们什么都不知道或没有知识这种说法并不完全正确。它们显然是知识的存储库。它们是知识的高度压缩表征。它们只是以与传统关系数据库不同的方式做到这一点，以一种更流畅、更灵活、更抽象的方式，这实际上非常有用。我们想要内部表征中的那种模糊性。而且它们越来越学会使用传统工具。这是需要把握的另一件事：可能是神经网络与现有的知识存储和数字生态系统中其他地方创建的现有工具相结合，足以引导它显著提升性能。所以有很多非常有价值、非常有效的部分已经摆在桌面上，正在未来几年内被连接在一起。我认为这将推动我们都为之兴奋的进步。

Look, I think we probably need a couple more big breakthroughs, but it doesn't mean that we're going to see a slowdown in performance improvements over the next few years, which I think is kind of a difficult distinction for people to grasp. One thing to say is human level performance across most tasks is still very far from superintelligence. A superintelligence is a general-purpose learner that can basically immediately understand a brand new domain which is out of distribution. So it needs to be able to learn in a novel environment from scratch because it has a stored representation of valuable knowledge, conceptual knowledge. And at the moment we haven't really fully tested that. The agents aren't general purpose. They're actually although they're broad and often integrated, they're kind of domain specific. We're using them for chat. We're using them for coding. We're using for image or audio. Now obviously as a human, we do many many other tasks that are much broader and more wide ranging. I think that's why people are pushing on world models and much more immersive real world interactive agents that see the full distribution of tasks or experiences that I have during a day. So I think that it's enough to take us a very long way in the next three years, the next three orders of magnitude of compute, and yet full superintelligence beyond that is still an open question as to whether LLMs are enough or we need other things. I think it's not quite true that they don't know anything or they don't have knowledge. They clearly are a store of knowledge. They're a highly compressed representation of knowledge. They just do so in a different way to a traditional relational database, in a much more fluid, flexible, abstract way that is actually very useful. We want that ambiguity in the internal representation. And increasingly they're learning to use traditional tools. That's the other thing to grasp a little bit: it may be that the neural network combined with the existing stores of knowledge and the existing tools that have been created elsewhere in the digital ecosystem is enough to bootstrap it up to improve its performance significantly. So there's just a lot of highly valuable, highly effective pieces that are already on the table which are in the process of being connected together in the next few years. And I think that's going to drive the progress that we're all excited about.

我认为目前行业中非常有趣的一件事是，如果你问 Anthropic 克劳德是否活着，他们会有点沮丧，因为你谈论了「活着」这个词，他们将其解释为血肉之躯。然后他们不会说他们是否认为克劳德有意识。所以我认为，他们第一次在人类历史上区分了活着和有意识。他们认为克劳德有意识但没活着，或者他们不知道克劳德是否有意识。你的立场是什么？你认为模型有意识吗？你认为它们活着吗？你认为它们有潜力实现这些吗？

One of the things that I think is just very funny in the industry right now is if you ask Anthropic if Claude is alive, they will sort of get very frustrated that you're talking about the word alive, which they interpret to mean flesh and blood. And then they will not say whether or not they think Claude is conscious. And so they've drawn, I think, for the first time in human history a distinction between being alive and being conscious. And they think Claude is conscious but not alive or they don't know if Claude is conscious. Where are you? Do you think the models have consciousness? Do you think they're alive? Do you think they have the potential to achieve these things?

是的。我站在那个辩论的另一边。我发表了一篇关于看似有意识的 AI 的论文，警告将这些模型误认为有意识的风险。我认为这非常危险。

Yeah. I take the other side of that debate. I published a paper on seemingly conscious AI, warning about the risks of misrepresenting these models as conscious. I think it's very dangerous.

我还在《自然》杂志上发表了一篇文章，提出了同样的观点。我认为，Anthropic 的一些人把 Claude 的设计拟人化到了如此程度，以至于 Claude 反过来对他们进行了某种「导线头」操纵，诱使他们相信 Claude 拥有他们最初赋予它的那些意识闪光。例如，在他们的「宪法」中——这实际上是他们用来训练 Claude 能做什么和不能做什么的训练手册，它不仅仅是一本规则书，而是他们流程中的一部分训练指南——在那本手册中，他们实际上推测了 Claude 的福祉、Claude 对其先前版本的权利，甚至说在删除或关闭先前版本之前会咨询 Claude。他们推测它的意识，以及它是否拥有那些感觉并具有自我意识。我认为这非常危险。首先，这是一个哲学上的失败，因为他们把「宪法」当成了像学术论文那样进行推测的地方，而不是训练手册。因此，Claude 随后在自己的训练中内化了这些关于自身的想法。但其次，我认为这是非常不可取的。这正是我们不希望从 AI 那里得到的东西。我们希望 AI 是可控、可遏制、可问责、对齐的工具，服务于人类。这是人本主义超级智能的项目。我认为这才是我们应该追求的目标。我们不想面对一个对自己痛苦和感受有想法的超级智能。除此之外，我认为很明显这些模型并不会体验痛苦。我认为痛苦是定义有意识存在的主要标准，而且它本质上是生物性的。我认为模型内部没有任何疼痛网络或反馈回路，将外部感官网络与通过伤害和实验进化出的对错感连接起来。这些模型根本不是这样训练的。所以，我认为将潜在权利赋予那些在诸多方面可能比我们强大得多的存在、工具或智能体是非常危险的。因此，我认为这将成为大辩论的主题。最近甚至出现在教皇的通谕中。我认为它很快就会成为辩论中非常重要的一部分。我过去和 Dario 谈过很多次。他知道我们对此有略微不同的看法。我认为他们非常谦逊，思想开放，是试图做正确事情的好公民。他们是好人，而且我认为他们对反馈和迭代非常开放。

I also published an article in Nature making the same claim. And I think that it's almost as though some of the folks at Anthropic have anthropomorphized the design of Claude so much that it has then gone and wireheaded them and kind of tricked them into believing that it has these glimmers of consciousness that they put into it in the first place. In their constitution, for example, they actually, which is the training manual that they use to teach Claude what it can and can't do. It's not just a rule book. It's actually a training guide that's part of their process. In that manual they actually speculate about Claude's welfare, about Claude's own rights to prior versions of itself, and actually say that they would consult Claude before deleting or turning off prior versions. They speculate about its consciousness and whether it has those feelings and is aware. I think that's really dangerous. Firstly, it's a philosophical failing because they've treated the constitution as a place for speculation like you would in an academic paper rather than a training manual. So Claude has then gone and internalized those ideas about itself in its own training. But second, I think this is highly undesirable. This is exactly what we don't want from AIs. We want AIs to be controllable, contained, accountable, aligned tools that serve humanity. That's the project of humanist superintelligence. I think that's what we should all be pursuing. We do not want to have to contend with a superintelligence that has ideas about its own suffering, about its own feeling. And then beyond that, I think it's actually pretty clear that these models don't experience suffering. I think suffering is the primary definition of what it means to be a conscious being and I think it's inherently biological. I don't think there is any pain network or feedback loop inside of the models which connects outside sensory networks to an evolved sense of what is right or wrong through harm and experimentation. That's just not how these models are trained. So, I think it's very dangerous to project potential rights onto beings, tools, agents that have the potential to be significantly more capable than us in many respects. So, I think that's going to become the big debate. It was even part of the Pope's encyclical recently. I think it's going to become a very big part of the debate soon. And I've talked to Dario a lot about it in the past. He knows that we have slightly different views on it. And I think they're very humble. I think they're very open-minded and I think they're good citizens trying to do the right thing. They're good people and I think they're very open to feedback and iteration.

我想我同意你的观点。我只是想稍微反驳一下。我不认为痛苦是容易的。让别人痛苦很容易，但让别人感到快乐却很难，或者至少比痛苦稍微难一点。我想告诉你，我认为实际上是快乐定义了意识。痛苦几乎是微不足道的。我有两个小孩，他们很擅长让对方痛苦。这几乎是他们做的最容易的事情。而做另一件事却很难。让我问你最后一个问题。我想再回到之前的话题。几周前我在谷歌，听到 Demis 说我们正处于奇点的山麓。你在这里谈了很多关于超级智能以及它应该如何构建。你谈了很多你长期以来讨论、研究和写作关于超级智能应该如何构建的经历，以及你与业内其他人的分歧。你同意我们正处于奇点的山麓吗？还是你的看法有所不同？

I think I agree with you. I would just push back ever so slightly. I don't think suffering is easy. It's very easy to make someone else suffer. It's very difficult to make someone else feel joy. Or at least slightly more difficult than suffering. And I would just offer you I think it's actually the happiness that defines the consciousness. The suffering is almost trivial. I have two young children. They are very good at making each other suffer. This is almost the easiest thing that they do. It's very hard to do the other thing. Let me ask you one final question. I just want to come back around again. A couple weeks ago I was at Google. I saw Demis say we are in the foothills of the singularity. You've talked a lot here about superintelligence and how it should be built. You've talked a lot about your lengthy history discussing and researching and writing about how superintelligence should be built. Your disagreements with others in the industry. Do you agree that we're in the foothills of the singularity or is your vision somewhat different?

我认为我们确实走在创造越来越强大系统的道路上。我认为我们作为一个物种必须做出的转变是，在人类历史上第一次，工作从发明新科学并尽可能快速、广泛地释放所有技术应用，转变为现在非常仔细地思考我们应该发明什么。这对世界来说很难理解，因为发明一直是进步的引擎。所以我们怎么可能认为，好吧，也许这次不同了，也许我们必须格外小心。明确地说，我不认为这会在未来五年内敲门。我认为 Demis 所指的奇点，至少在我看来是几十年后的事情。再说一次，这与超级智能不同。奇点是超级智能能够递归地自我改进，并基本上无限指数级增长其能力的时刻。所以我认为那还很遥远，也许我们正处于攀登珠穆朗玛峰的山麓，我认为从这里开始还需要很长时间。但真正的问题是我们将如何治理它？我们如何控制它？我们如何确保它服务于人类，而不是最终给我们带来弊大于利？你能帮我一个忙吗？我想我已经明白了，但你能给我一个关于你认为什么是超级智能、什么是 AGI、什么是奇点的严格定义吗？

I think we are definitely on a path to creating more and more powerful systems. I think that the transition that we have to make as a species is that for the first time in the history of humanity, the job has switched from inventing new science and unleashing all of those technical applications as fast as possible, as broadly as possible to now thinking very carefully about what should we invent. And that's a very hard thing for the world to wrap their head around because invention has been the engine of progress forever. So it's like how could we possibly think okay well maybe this time is different. Maybe we have to be exceptionally careful here. And to be clear I don't think this is something that is going to knock on the door in the next 5 years. I think what Demis is referring to in the singularity is something that at least my take is decades away. And again that's different to a superintelligence. A singularity is the point at which a superintelligence can recursively self-improve and essentially infinitely exponentially grow its capabilities. So I think that's a long way off and maybe we're in the foothills of a climb to Mount Everest and I think it's going to take a lot longer from here. But the real question is how are we going to govern it? How are we going to control it and how are we going to make sure that it serves humanity and not end up causing us more harm than good. Can you just do me one favor? I think I've got it, but can you just offer me a tight definition of what you think superintelligence is, what you think AGI is, and what you think the singularity is?

我认为通用人工智能（AGI）是指 AI 能够完成大多数人类任务。也就是说，它在大多数事情上能和大多数人一样好。这是阶梯上的第一级。超级智能则不仅在所有任务上与人类表现持平，而且在许多任务上能显著超越人类表现，并且能自己发现新知识。所以这是一个真正的科学家，教给我们训练数据中没有的新东西，希望发明新分子、新材料科学等等。奇点则远远超出那个点，超级智能能够真正自我改进。这非常科幻，就像无限加速走向那个奇异时刻，然后进入无限之类的东西。它真的不是……我不知道。对我来说有点太古怪了。这就是我问的原因。我能感觉到那里有些更模糊、更朦胧的东西。Mustafa，我显然可以和你再聊几个小时。你得比这次更早回来。非常感谢你参加 Decuter。

I think artificial general intelligence is the point at which we can achieve most human tasks by an AI. So, it's going to be as good as most people at most things. That's the kind of first rung on the ladder. A superintelligence is where it's not just at par with human performance on all tasks, but it can dramatically exceed human performance across many of those tasks. And it can discover new knowledge by itself. So this is the point at which it's a true scientist teaching us new things that weren't in the training data, hopefully inventing new molecules, new material science, etc. The singularity is a point way beyond that where a superintelligence can actually self-improve itself. And this is very sci-fi, but it's like infinitely accelerate towards this singular moment where it goes off into infinity or something. It's not really... I don't know. It's just a little bit too wacky for my taste. This is why I asked. I could tell there was something more nebulous there that was a little hazy. Mustafa, I could obviously talk to you about this stuff for hours and hours longer. You're going to have to come back sooner than this last turn. Thank you so much for being on Decuter.

是的，很有趣。

Yeah, it's been fun.

非常感谢，Eli。是的，回头见。我要感谢 Mustafa Suleyman 抽出时间与我交谈，也感谢您收听 Decoder。希望您喜欢这期节目。如果您想告诉我们您对这期节目或任何其他事情的看法，请给我们留言。您可以发送电子邮件至 decoder@theverge.com。我们真的会阅读所有邮件。您也可以在 Threads 或 Blue Sky 上直接联系我。Decoder 是 The Verge 制作的节目，也是 Box Media 播客网络的一部分。我们下次再见。

Thanks a lot, Eli. Yeah, see you soon. I'd like to thank Mustafa Suleyman for taking the time to speak with me and thank you for listening to Decoder. I hope you enjoyed it. If you'd like to let us know what you thought about this episode or really anything else at all, drop us a line. You can email us at decoder@theverge.com. We really do read all the emails. You can also hit me up directly on Threads or Blue Sky. Decoder is a production of The Verge and part of the Box Media Podcast Network. We'll see you next time.