驾驭 AI 浪潮：Spotify 的战略转型

也许一个有趣的起点就是最明显的地方。每个人都面临着这场巨大的技术变革。我的朋友拉维·古普塔称之为“AI 或死亡”的必然要求。也就是说，公司，即使是那些大型、性感、成熟的技术公司，也需要找到方法来拥抱和利用这项新技术，随着它的展开，否则就面临被淘汰。我很想听听你和 Spotify 是如何思考这个挑战的。我知道你们非常迅速地拥抱了它，而且你们很早就将机器学习和数据科学应用到了整个产品中。但这是一次巨大的转变，你和丹尼尔以及团队是应对这类转变最有思想的人之一，而且你们以前也做过。请详细地给我们讲讲你最初是如何感受到它的，你做了什么，在大公司里处理这种事情是什么感觉。

Maybe a fun place to begin is the obvious place. Everyone is facing this giant shift in technology. My friend Ravi Gupta calls this imperative AI or die. That companies, even the big sexy established technology companies, need to find ways to embrace and use this new technology as it unfolds or face elimination. I would love to hear how you and Spotify are thinking about this challenge. I know you've embraced it very quickly and you were very early to using machine learning and data and data science all over the product. But this is a big shift and you and Daniel and the team are some of the most thoughtful people about addressing shifts like this and you've done it before. Walk us through in some detail how you first felt it, what you did about it, what it's like to be at a big company and process something like this.

是的，这是个好问题。我认为这是正确的描述，至少从长远来看，我认为是“AI 或死亡”。就像之前是“智能手机或死亡”，再之前是“互联网或死亡”、“计算机或死亡”。这是那种转变之一，你知道，你无法选择是否采用它。它会发生在你身上。我会说，这是宏观趋势的典型体现。通常当这些宏观趋势来临时，我们内部有句话：你可以让宏观趋势吹在你脸上，但它不会改变方向。所以你需要重新定位自己，让风从背后吹来，这样你就能驾驭这股宏观趋势，或者，你知道，有些人称之为宏观浪潮，你在上面冲浪。所以我们经历过几次这样的浪潮。第一次真的是智能手机出现的时候。Spotify 在智能手机之前为互联网做好了充分准备，我们在桌面上有免费层级，我们在那里获取用户，他们创建播放列表，然后自我留存。然后移动收听（随时随地听）是 Spotify 的付费功能。当大多数用户是电脑、少数是智能手机时，这没问题。然后当智能手机起飞时，我们面临了生存危机，开始出现没有桌面的消费者。他们只有手机，所以他们没有免费体验，我们的整个模式就死了。所以那是我们必须重新定位整个商业模式的时刻之一，实际上，要弄清楚如何在移动端做免费层级，同时又不蚕食付费功能（即移动性），我们稍后可以谈谈我们是如何解决的。但那是其中一个例子，我认为这次是类似的。对我来说，最大的问题是：这需要改变商业模式，还是仅仅是一次产品变革？

Yeah, it's a great question. I think it is the right description, at least in the longer term, I think it is AI or die. Just like it was smartphone or die and before that internet or die, computer or die. This is one of those shifts that, you know, it's not your choice whether you adopt it or not. It's going to happen to you. It's the epitome of a macro wind, I would say. And usually when these macro winds come, we have a saying internally that, you know, you can have the macro wind blowing in your face and it's not going to change its direction. So you basically need to reposition yourself so you get the wind at your back and you can sort of surf this macro wind or, you know, some people call it macro wave that you surf. So we've been through a few of these. The first one was really the smartphone when that came along. Spotify was really well positioned for the internet before the smartphone where we had a free tier on desktop and that's where we sort of acquired users and they built a playlist and they retained themselves. And then mobility to listen on the go was a paid feature on Spotify. And that was fine when the majority was computers and the minority was smartphones. And then when smartphones took off, we faced an existential crisis where there started to be consumers who didn't have a desktop. They only had a phone, so they had no free experience and our entire model died. So that was one of those moments we had to reposition the entire business model actually and figure out how do we do a free tier on mobile that doesn't cannibalize the paid feature which was mobility and we can talk about how we figured that out later. But that was one of those examples and I think this is a similar one. The big question to me is does this require a business model change or is it just a product change?

我认为 AI 的另一个不同之处在于，它不会只触及一件事。它触及消费者产品，但也触及你作为一家公司的生产力和竞争力。所以有很多不同的角度可以开始，但正如你所说，我们在机器学习方面起步很早，我们的旅程是，你知道，我们看到用户来到 Spotify，然后他们开始创建播放列表，这让他们留存下来。但只有一定数量的人擅长创建播放列表，因为你必须把目录记在脑子里，包括新发行和旧目录。所以有些人自我留存得很好。然后我们试图通过让编辑为那些不擅长创建播放列表的人创建播放列表来扩展这种行为。我们看到人们利用社交来寻找灵感。最终机器学习开始出现，我们看到了为每个人构建一个音乐朋友的机会。所以那就是我们开始的地方。我们开始投资于此，并变得相当擅长。我认为有些人说 AI 只是机器学习，只是一个新词。这是一个有趣的问题，有什么区别？我认为人们过去所说的机器学习和我们现在所说的生成式 AI 之间的区别在于，统计机器学习有点像一种输出机制，我认为那个时代的典型代表是全屏的 TikTok 信息流。就像，你知道，用户界面会根据驱动它们的技术来塑造自己，以最大化指标，我认为那是最大化老式机器学习的统计探索-利用范式的用户界面。生成式 AI 带来的变化，我认为最大的转变是你可以接受自然语言输入。所以即使技术上它们都是机器学习，我认为生成式 AI 是一个新时代。最大的转变是它是双向的。如果你想想 Spotify，例如，作为消费者产品，它的外观几乎像老式的宽带。你知道，那种宽带，下行链路可能有 1 兆比特，但上行链路只有 150 千比特。所以下行带宽很大，但反馈不多。这就是大多数消费者服务的样子。你有，你知道，下行链路上的流媒体视频，每秒大量信息。但上行链路只有几次点击和滑动。这是非常非常窄的信号。这就是以前的机器学习时代所关注的。我认为在生成式 AI 时代，变化的是上行链路现在可以是英语。它可以几乎和下行链路一样丰富。我认为这要求我们所有消费者公司最终彻底重新思考产品。所以如果你按照我所说的进行推导，如果全屏 TikTok 信息流是机器学习范式的典型代表，即不对称的下行-上行链路范式，那么它同时成为生成式 AI 时代的典型代表的可能性有多大？我不这么认为。我认为消费者产品将发生根本性的变化。我无法准确预测如何变化。我认为在信息接收与信息给予方面，它们将变得更加对称。我认为如果你快进 5 到 10 年，几乎所有大型消费者产品都将在某种程度上成为对话，而不是你使用的这种服务。所以我们在产品方面的工作就是试图找出下一个范式是什么。我还不知道它到底是什么。我们正在试验，如果我确实知道，我现在可能不会告诉你。我会先保密一段时间。但这就是产品方面的情况。然后我们也可以谈谈生产力方面，在编码生产力方面有明显的收益，我们正在使用所有其他人都在使用的工具。

The other thing that is different, I think, about AI is that it's not going to touch one thing. It touches the consumer product, but it also touches your productivity and competitiveness as a company. So there are lots of different angles to start, but as you said, we were quite early with machine learning and the journey we had was, you know, we saw users coming on Spotify and then they started playlisting and that retained themselves. But it was only a certain amount of people who were good at playlisting because you have to know the catalog in your head, the new releases, the back catalog. So some people retained themselves really well. And then we tried to scale that behavior by having editors who created playlists for people who couldn't playlist that well. And we saw people using social to find inspiration. Eventually machine learning started happening and we saw this opportunity of sort of building a music friend for everyone. So that's where we started. We started investing in that and got quite good at that. I think some people say that AI is just machine learning, it's just a new word. And it's an interesting question, what is the difference? I think the difference between what people used to call machine learning and what we call generative AI is that the statistical machine learning was sort of an output mechanism and I think the epitome of that age is the full screen TikTok feed. It is like, you know, UIs shape themselves after technology that powers them to maximize metrics and I think that is the UI that maximizes the statistical explore-exploit paradigm of old school machine learning. What happens with generative AI, I think the big shift is that you can take natural language input. And so even if technically they're both machine learning, I think of generative AI as a new age. And the big shift is that it's two-way. If you think about Spotify, for example, as consumer product the way it looks, it's almost like an old school broadband. You know, the broadband where you had like maybe 1 megabit downlink but only like 150 kilobit uplink. So a lot of bandwidth down, but not a lot of feedback. This is what most consumer services look like. You have, you know, streaming video on the downlink, a lot of information per second. But the uplink is only like a few clicks and swipes. It's very very narrow signal. And this is what the previous machine learning age focused on. I think what changes in the age of generative AI is that the uplink can now be English language. It can be almost as rich as the downlink. And I think that requires all of us consumer companies to in the limit totally rethink the product. So if you just do the deduction of what I said, if the full screen TikTok feed is the epitome of the ML paradigm, the asymmetric downlink-uplink paradigm, what are the chances that that is also the epitome of this generative AI age? I don't think so. I think consumer products are going to change fundamentally. I can't predict exactly how. I think they're going to be much more symmetric in terms of information you receive versus information you give. And I think if you fast forward 5 to 10 years, almost all big consumer products are going to be a conversation to some extent rather than this service that you use. So really the job for us on the product side is to try to figure out what is the next paradigm. And I don't know exactly what it is yet. We're experimenting and if I did know, I probably wouldn't tell you right now. I'd sit on it for a bit. But this is where on the product side. And then we can talk a bit about the productivity side as well where there are the obvious gains in terms of coding productivity where we are using all the tools that everyone else is doing.

但作为大公司，与初创公司有一些不同，因为到目前为止，生成式 AI 在编码方面的影响最大的是编写全新代码，这在初创公司中占很大比例，而在大公司中只占很小一部分。大部分工作只是重构等等。我记得看到过一些统计，在大公司，你每天 8 小时中大约只有 1 小时在编码。所以不仅编码只占 1/8 的时间，而且在这 1/8 的时间里，全新代码的比例也很小。所以我实际上认为，在编码方面，最大的影响还没有到来。这涉及两个方面。这些模型正在变得足够大，能够理解像 Spotify 这样庞大而复杂的代码库。我们还没有达到让它们重构我们代码库的程度，它还没有那么深入的理解，但将来会的。那将是一个巨大的转变。另一个是自动同行评审，这即将实现。目前还不够好，不能完全信任。所以很多开发者坐着等待他们的代码被评审并返回。所以我认为我们正在看到这种增长。我认为未来几年我们会看到它大幅增长。但真正有趣的是开发者所做的其他 7 个小时，包括大量沟通、规划、与设计师合作、原型设计会议。我认为这些方面的影响实际上会和编码本身一样大，甚至更大。

But as a big company, there are a few differences from the startups because so far generative AI in coding has had the most impact when you write net new code, which is a lot of what you do as a startup and a tiny bit of what you do as a big company. Most of it is just refactoring, etc. And I think I saw some statistic that in a big company you basically code one out of every 8 hours in a day. So not only is coding only 1/8 of the time, of that 1/8 of the time net new code is very small. So I actually think the biggest impact is yet to come when it comes to coding. That's two things. These models are getting big enough to understand really large and complex code bases like Spotify's. And we're not quite there where these things can refactor our code base. It doesn't have quite a deep understanding, but it will. And that will be a big shift. The other is doing automatic peer review, which is just on the verge of working. It's not quite good enough that you can trust it. So a lot of developers sit and wait for their code to be in review and come back. So I think we're seeing that ramp. I think we're going to see it ramp a lot in the next few years. But then the really interesting side is these other 7 hours what a developer does, which is a lot of communication, planning, working with designers, prototyping meetings. Those things I think will actually have as big or even more impact than the coding itself.

我想从下行链路和上行链路这部分开始。关于消费者愿意在上行链路上投入多少精力，你学到了什么？看起来聊天界面，比如 GPT 这类，我们知道人们愿意进行大量的来回交流。因为那是原生界面，你去那里就期望写很多东西，从 Twitter 复制提示词等等。在像 Spotify 这样的应用里，人们有多大的意愿不偷懒，而是真正描述他们想要什么？关于人们懒惰的本性，与通过更丰富的上行链路获得他们想要的东西而愿意投入大量工作的意愿，你学到了什么？

I want to start with this downlink-uplink part. What have you learned about consumers' willingness to put a lot of effort into the uplink? It seems like the chat interfaces, the GPTs of the world, we know that people are willing to do a lot of back and forth. And because it's the native interface, you're going there expecting to write a lot of stuff, copy prompts from Twitter or whatever. In an app like Spotify, how willing are people to get not lazy and really descriptive about what they actually want? What have you learned about the nature of people's laziness versus willingness to put a lot of work in to get the thing that they want via that more rich uplink?

是的，这可能是生成式 AI 时代和双向上行链路范式对我们来说最令人兴奋的事情。以前，我们主要依赖你在创建播放列表时的一些显式输入。那是高价值信息。你坐在那里思考，这首歌和那首歌搭配得很好。所以如果你把它看作标注，即使你为自己创建播放列表，你也在某种程度上标注这些曲目之间的关系。你投入了大量精力。那过去是、现在也是我们在音乐推荐方面的巨大优势，尽管生成式推荐系统开始取代这些更传统的协同过滤系统。所以我们有了一些非常强的信号，你很少会投入大量时间制作一个描述你的播放列表数据集。但大多数时候我们只有跳过信号，而我们的挑战是手机在口袋里。所以即使我们有赞/踩，你也不会每次都拿出手机说我不喜欢这个因为那个，甚至做赞或踩都需要你拿出手机、解锁、打开 Spotify。你能从耳机里做的就是跳过。所以我们有跳过信号，但这是一个非常粗糙的信号。我们播放你的歌曲，你跳过了。那可能是因为你非常讨厌它，也可能是因为你喜欢它，但已经听了一百遍，你厌倦了。也可能你喜欢它，不厌倦，但你在健身房，所以爵士乐不合适。所有这些对我们来说都只是跳过。所以这是一个非常粗糙的信号。我们有很多这样的信号，但很粗糙。你永远不会通过它实现完美的个性化。现在，我们在生成式 AI 中发现，我们推出的首批服务之一，目前已在约 40 个国家上线，叫做 AI 播放列表。我们实际上使用了一个基于你的收听数据和世界知识等训练的 LLM。你可以直接用英语告诉我们你想要什么样的播放列表。以前，你可以把歌曲加入播放列表，也许你给它一个标题，我们可以猜测这可能是跑步播放列表。所以我们可以做一些事情。现在你可以说，我想要一个 EDM 跑步播放列表，我想要大鼓点，我想要 160 BPM。然后你会得到 LLM 的建议，然后你可以保留一些曲目，说这些不错，这些不好。现在优化它。我不喜欢这些艺术家，但我想要更多那样的。所以对我们来说，这是我们第一次获得用户心中真正想法的这种保真度。思考 Spotify 的一种方式是，我们总是试图在我们的服务器上重现你新皮层的一小部分。仅仅通过跳过的点击流很难做到。现在，当你告诉我们你心中所想，就更容易近似你这个人。所以这真的是我们第一次拥有这种信号。我喜欢这样思考：当我们做用户研究时，我们做两件事。我们做定量测试、A/B 测试，但在此之前，我们做定性测试。我们深入采访几个人以理解需求。然后我们构建产品，然后进行 A/B 测试看我们是否正确。生成式 AI 的承诺是，AI 实际上是对近 7 亿用户进行持续的深度定性用户研究。听起来很大，但如果你眯着眼看，它就是这样。

Yeah, so that's probably the most exciting thing for us of this generative AI age and the dual uplink paradigm. So previously we mostly relied on some explicit input when you playlist. That's like high-value information. You are sitting there thinking like this song goes really well with this song and that song. So if you think about it as labeling, even though you're playlisting for yourself, you're sort of labeling these tracks in relation to each other. And you're putting a lot of effort in it. And that was and is our big advantage in music recommendations, even though generative recommendation systems are starting to take over from these more old-school collaborative systems. So we had some really strong signal like that where you quite seldomly invested a lot of time in producing a data set that described you about playlisting. But most of the time we just had skips, and the challenge for us is the phone is in the pocket. So even if we had like a thumbs up/down, you're not going to take out the phone every time and say I didn't like this because of that, or even do a thumbs up/thumbs down requires you to take out your phone, unlock it, open Spotify. What you can do from your earphone is to skip. So we have the skip signal, but that is a very blunt signal. So we play your song and you skip it. That could be because you absolutely hated it. Could be because you love it, but it's a hundred times, you're tired of it. It could be that you love it, you're not tired of it, but you're at the gym. So jazz is not the right thing. All of those just look like a skip to us. So that's a very blunt signal. We have a lot of that signal, but it's blunt. And you will never get to perfect personalization through that. Now, what we find with generative AI, one of the first services that we've launched that is live now in like 40 countries is something called AI playlisting. We literally use an LLM that is trained on your listening data and world knowledge and so forth. And you can literally tell us in English what kind of playlist you want. Previously, you could playlist songs and maybe you put a title on it and we could guess like this is probably a running playlist. So we could do something. Now you can say I want a running playlist that is EDM. I want big drops. I want it to be 160 BPM. And then you get a suggestion from the LLM and then you can keep a few tracks and say these were good, these were not good. Now refine it. I don't like these artists, but I want more of that. So for us it's the first time that we get that kind of fidelity of what is actually in the user's mind. One way to think about Spotify is we always try to reproduce a small part of your neocortex on our servers. It was just very hard with a clickstream of skips. Now, when you tell us what is in your mind, it gets easier to approximate you as a person. So this is really the first time that we have that signal. Now, one way I like to think about it is when we do user research, we do two things. We do quantitative testing, A/B test, but before that, we do qualitative testing. We interview a few people deeply to understand the need. Then we build a product and then we A/B test to see if we're right. The promise of generative AI is AI is really a deep, ongoing qualitative user research with almost 700 million users all the time. And it sounds big, but if you squint at it, that's kind of what it is.

有趣的是，这项新技术可以让你以多种不同的方式发展产品。我很想知道 Spotify 内部的机制，比如领导团队、产品团队，以及你们实际如何运行决定团队工作内容的过程。显然，你们有一个庞大的团队，但无论如何，你们的精力有限，能投入的能量单位有限。这项技术有巨大的应用空间，而你们拥有 7 亿用户的令人兴奋的优势。实际的会议流程、设置流程是怎样的？我问这个问题的原因是，许多公司都面临同样的挑战。令人兴奋，但也令人恐惧，他们需要在别人之前获得创新，以免被颠覆。那么，你们是如何得出可能尝试的想法的背景过程是怎样的？

It's interesting how many different ways you could take the product with this new technology. I would be really curious to know the apparatus inside of Spotify, like the leadership team, the product team, and literally how you run the process of deciding what to do with your team. You have a big team, obviously, but no matter what, you have limited effort, limited units of energy you can apply. There's this huge space of stuff you could do with this technology and the exciting advantage that you have of all these 700 million users. What is the literal meeting-by-meetings process, the setup process look like? And the reason I'm asking this question is so many companies face this same challenge. It's exciting, but also scary that they need to get the innovation before somebody else does and disrupts them. So what is the background process for how you arrive at the things you might try?

所以实际上有两件事。我们有一个非常结构化的过程，我想谈谈它是如何运作的，但我们也使用一些概念。多年来，我向公司引入了一些战略框架，其中一些我知道你很热衷，比如 Hamilton Helmer 的《七力》，以及 Tushar 写的捆绑框架。

So there are really two things. We have a very structured process that I want to talk through how it works, but there are also some concepts that we use. And over the years I've introduced some strategic frameworks to the company, some of which I know you're passionate about, like Seven Powers from Hamilton Helmer, and the bundling framework that Tushar wrote.

他在董事会里，对吧？

He's on the board, right?

他在董事会。而且是秘密武器。没错，非常好的秘密武器。另外我还发现 Felix Oberholzer 的《更好、更简单的战略》非常不错。

He's on the board. And secret secret weapon. Exactly. Very good secret weapon. And also I found Better, Simpler Strategy by Felix Oberholzer to be very good.

那本是讲什么的？我没听说过。

What's that one? I don't know that.

它讲的是“价值棒”的概念，其中包含支付意愿，这对我们非常重要。比如，我们衡量价值的方式就是支付意愿。但它还引入了出售意愿。想想你的员工，他们愿意以什么条件为你提供服务？每个人都专注于提高支付意愿，但你也可以提高，或者取决于你怎么看，降低出售意愿。事实证明，世界上最好的公司不一定是付钱最多的，而是那些使命最有意思、文化最好的公司。所以，因为我们是一项捆绑式服务，我们总是试图给用户越来越多的价值。你知道，我们加入大量音乐，那是价值。然后我们加入更多播客，那也是价值。现在我们加入书籍，那是更多价值。支付意愿和出售意愿这个框架对我们非常有用，它非常适合我们的业务。我们的工作就是让支付意愿与实际价格保持相当远的距离。这个差距就是你提供的消费者剩余。我们作为一项服务的目标就是确保 Spotify 始终是一笔超值的交易。你总会觉得支付意愿，也就是你感知到的实际价值，远超我们的定价。所以我们经常使用这个框架。所以，引入这些框架，不仅在业务部门，也在产品和技术部门，能让人们以更有结构的方式思考，让他们拥有共同的语言。我们可以谈论网络效应、摊销、品牌力等等。所以我花了很多时间让团队使用这些框架，这样我们就能有结构化的战略思维。

It's a concept of the value stick where you have willingness to pay, which is very important for us. Like, the way we measure value is the willingness to pay. But what it introduces is also the willingness to sell. If you think about your staff, like what is their willingness to sell their services to you? And everyone focuses on increasing the willingness to pay, but you can also sort of increase or, depending on how you think about it, decrease the willingness to sell. And it turns out the best companies in the world are not necessarily the ones that actually pay the most, it's the ones with the most interesting mission, the best culture, et cetera. So, because we're a bundled service where we try to just give users more and more value all the time. You know, we put in lots of music, that's value. Then we put in more podcasts, that's value. Now we put in books, it's more value. This framework of willingness to pay and willingness to sell is very useful for us. It just fits our business really well. And the job of us is to keep the willingness to pay quite far from the actual price. Like, that gap is how much consumer surplus you're giving. And our goal as a service is to make sure that this Spotify is just an amazing deal. You're always going to feel like the willingness to pay, the actual value you perceive, is way over the price that we have. So, we use that framework quite a lot. So, introducing these frameworks, not just in the business org, but also in the product and technology org, makes people think in more structured ways. It makes people have a vocabulary. We can talk about network effects, amortization, you know, brand power, all of these things. So, I spend a lot of time getting the teams to use these frameworks so that we have structured strategic thinking.

另外，我试图推动的一个论点就是，我非常喜欢苏格拉底式的辩论。和许多人一样，我惊叹于希腊人和罗马人仅凭讨论就取得了如此大的成就，即使他们没有科学，只有推理。强大的推理非常有用。所以我试图推动这种有点挑衅性的说法：说话很便宜，所以我们应该多说。这有点像对“快速行动，打破常规”的一种反制。

And the other thing I try to push as a thesis is that, you know, I'm a big fan of sort of Socratic debate. I'm amazed, like many other people, at how far the Greeks and the Romans came with just discussion, even though they didn't have science, just reasoning. Strong reasoning is very useful. So, I try to push this sort of provocative line of talk is cheap, so we should do a lot of it. Sort of as a counter to like moving fast and breaking things.

嗯。

Yeah.

就像说话这么便宜，所以我们实际上应该多做一些。如果，你知道，有时候它会带你到达，比如，希腊人提出了原子的概念，对吧？所以我们进行大量结构化的讨论和构思。我和我的领导团队这样做，而且经常是领导团队加一级，也就是副总裁层和总监以上。我有很多时间专门用来讨论概念。

It's like it's so cheap to talk, so we should actually do a bit more of it. If you, you know, sometimes it takes you to like, you know, it took the Greeks to the concept of the atom, you know? So, we do a lot of talking and ideation that is quite structured. And I do that with my leadership team and often the leadership team sort of plus one. Which means the VP layer and sort of the director plus. And I have a lot of time just for discussing concepts.

所以，回到我生活中的一位英雄，David Deutsch，他的书《无穷的开始》和《现实的构造》对我影响很大。他谈到一种叫做“好的解释”的东西。他对什么是好的解释有一个清单。它显然需要可证伪等等，但它还需要有延展性，需要能扩展。所以，一个还行的解释能解释这个现象，但它不能扩展到其他现象，对吧？不能上下扩展。真正好的解释能扩展，比如从解释地球如何运作扩展到太阳系、行星。但它也非常难以变动。我认为这一点经常被低估。如果你有一个解释，你可以用另一个解释同样的东西来替换它，比如，如果你用神来解释天气，你可以把这个神换成那个神，那可能就不是一个好的解释。它必须非常难以变动。如果你变动它，它就不再解释了。他说的最后一点是，解释不应该只是预测性的。那不是解释，那是模型。解释需要解释为什么。所以我试图推动我的团队，即使某件事在 A/B 测试中有效，我倾向于说，在你有一个好的理论解释它为什么有效之前，我不想发布它。因为如果你弄清楚了为什么，那就是模式识别和真正理解之间的区别。模式识别是有用的，那叫经验。你知道，我喜欢模式识别能力强的人，但如果他们能解释为什么有效，那就能扩展到整个组织。其他人可以利用这些知识。那要有价值得多。所以，这些就是我一直试图注入组织的一些概念。

And so, back to one of my heroes in life, David Deutsch, his book The Beginning of Infinity and The Fabric of Reality shaped me quite a lot. And he talks about something called good explanations. And he has a list of what a good explanation is. It obviously needs to be falsifiable and so forth, but it also needs to have reach. It needs to scale. So, an okay explanation explains this phenomena, but it doesn't scale to other phenomena, right? Doesn't scale up and down. Really good explanation scales, you know, from explaining how the Earth works to the solar system to the planets. But it's also very hard to vary. Which I think is often underestimated. If you have an explanation and you can switch it out for another explanation that explains the same thing, like, you know, if you explain the weather using gods, you can switch out this god for that god, it's probably not a good explanation. It needs to be very hard to vary. If you vary it, it doesn't explain it anymore. The last thing he says is that explanations should not just be predictive. That's not an explanation, that's a model. An explanation needs to explain why. So, I try to push my teams even if something works in an A/B test, I tend to say like I don't want to launch it until you have a good theory of why it works. Because if you figure out the why, it's the difference between pattern recognition and actually understanding something. Pattern recognition is useful, that's called seniority. You know, I love people with good pattern recognition, but if they can explain why it works, it scales to the entire org. Other people can use that knowledge. It's much, much more valuable. So, those are some of the concepts that I've tried to put into the org over time.

所以，我认为这很重要，因为它塑造了文化。然后我们有结构化的流程，我们每次执行 6 个月。我们有一个叫做“赌注”的流程，所有副总裁，大约 14 位。Spotify 的一个好处是他们规模很小，所有副总裁，整个公司都能装进一个房间。我们每周二开会 3 小时。整个公司完全同步，无论好坏，我们稍后可以讨论。但每 6 个月，这些副总裁会进行推介，真的像我们是风投、他们是初创公司一样推介。他们认为公司应该做的赌注以及为什么。这非常像初创公司的过程。你知道，你不能利用 Gustav、Alex 或 Daniel 可能喜欢你这个事实。就像，这是一次风投会议。你必须说服我们。所以他们推介。然后我和另一位联席总裁 Alex Nordstrom，我们根据这些推介决定一个排序，一个全局排序。这次我们有 44 个赌注。通常介于 30 到 50 之间。我们按从 1 到 44 排序。然后我们回到组织说：“现在，试着为这些配置资源。”他们从顶部开始，也许到 30 就说：“这是我们在接下来 6 个月能做的。”然后他们承诺这些事情。我们开始执行。这是自下而上创新和自上而下同步的良好结合，你不仅利用 Daniel，不仅利用我和 Alex，还利用所有副总裁和下面的层级来提出好主意，因为他们最接近用户。但然后有全局同步。我们对它们排序，确保它们符合单一战略，然后回到组织，他们承诺。而且我想你知道，如果你是自己说“我能做这个”的人，你会比老板说“你能做这个”时交付得更好，对吧？所以，这就是流程。但在此之前，我们有一个叫做原型阶段的阶段。所以，在之前的 6 个月里，我们用 Figma 和越来越多的生成式 AI 工具来原型设计 Spotify 在接下来 6 个月后应该是什么样子或可能是什么样子。

So, I think that's important because that shapes the culture. Then we have the structured process, which is we execute for 6 months at a time. We have something called a bets process where all the VPs, which is about 14. So, one of the benefits of Spotify is they're so small that all the VPs, the entire company can fit in one room. And we meet 3 hours every Tuesday. The entire company is completely synchronized, for good and bad, and we can talk about that later. But so, every 6 months, these VPs they pitch, literally pitch, as if we were a VC and they were a startup. The bets that they think the company should do and why. And it's very much like a startup process. You know, you don't get to use the fact that, you know, Gustav or Alex or Daniel may like you. Like, you know, this is like a VC meeting. You have to convince us. So, they pitch. Then me and the other co-president, Alex Nordstrom, we decide based on these pitches a stack rank, a global stack rank. This time we have 44 bets. Happens as usual between 30 and maybe 50. We stack rank them from 1 to 44. Then we go out to the org and say, "Now, try to resource this." And they start from the top and then maybe they get to 30 and say, "This is what we can do in the next 6 months." And then they commit to those things. And we start executing. And it's a good mix of sort of bottoms-up innovation where you leverage not just Daniel, not just me and Alex, but all the VPs and the layers below to come up with good ideas because they're closest to the user. But then there's global synchronization. We stack rank them, make sure that they fit a single strategy, then it's back to the org and they commit. And as I think you know, you're going to be much better at delivering something if you were the one who said I can do this than if your boss said you can do this, right? So, that's the process. But leading up to that, we have something called a prototyping phase. So, the previous 6 months, we prototype in the combination of Figma and increasingly gen AI tools what Spotify should look like after the next 6 months or could look like.

而且这个原型也有助于同步整个公司。我之前发现，当人们提交这些赌注时，每个人心里都有自己的理想功能。你开始构建，然后到后来你意识到你们其实并没有对齐。然后周期快结束时会有很多争吵，你知道，这个东西和那个东西不兼容，事情进展不顺利。我现在和 Alex Nordstrom 一起尝试做的是同步整个公司。Alex 和我没有各自的直接下属团队。我们作为一个单一团队每周二开三个小时的会。我们试图利用我们规模小这个优势，而不是把它当作相对于那些非常大的竞争对手的劣势。所以我们把所有东西都提前做成原型。所以，所有所谓的“争吵”都发生在你真正承诺做某事之前，而且你手里有东西可以拿着说，这就是如果我们成功的话 Spotify 会是什么样子。所以，这是文化输入和非常结构化的流程的结合，以确保它真正运作。

And this prototype also helps synchronize the entire company. What I found previously was that when people submitted these bets, everyone had in their mind what their great feature would be. You start building and then down the line you realize that you were not actually aligned. And then you get a lot of fighting towards the end of the cycle where, you know, this thing doesn't work with that thing and, you know, things don't work out so well. What I've tried to do now together with Alex Nordstrom, we synchronize the entire company. Alex and I don't have our direct reports team. We meet as a single team 3 hours every Tuesday. And we try to use the fact that we're small as an advantage instead of as a disadvantage versus our competitors who are, you know, very very large companies. And so we prototype everything up front. So, all the so-called, quote-unquote, fighting happens before you actually commit to doing something and you have something you can hold in your hand and say, this is what Spotify would look like if we pull this off. So, that's a combination of sort of cultural input and then a very structured process for actually making it work.

关于流程我有很多问题。第一个是周二那个三个小时的会议是怎么运作的？比如，那个会议的结构是什么？

I have so many questions about process. The first is how that 3-hour meeting on Tuesday works. Like, what is the structure of that meeting?

它叫 E-team，即执行团队。所以它非常专注于公司的执行。理念是，如果你有五个五天的工作周，平均而言，你被某件事卡住后，最多不超过两天半就能升级到我、Alex 和其他所有副总裁那里。所以，理念是你永远不应该被卡住超过最多两天半。因为我们是这样同步运作的，如果你被卡住，代价会非常高，因为其他人都在你的下游。所以，如果你像我们这样进行同步运作，升级流程非常重要，解决问题也非常重要。所以，那个会议很大一部分是，人们会说，你知道，我们这里偏离轨道了。我依赖于那边的某个人，但他没有做到他承诺的。而能让所有副总裁都在同一个房间里的好处是，你知道，我经历过很多会议，我相信你也参加过，人们会说，好吧，我们私下再谈。我稍后跟你聊。而我们说的是，你不允许说“私下”或“稍后”这些词，因为那个人就在房间里。所以，就像，你知道，我依赖于那边的 Anna 来做这件事，但 Anna 其实就在那里。然后 Anna 可以说，你知道，好吧，我不知道那件事，或者我会解决它。所以，就像实时解决。理论上很简单，但实践中非常强大。大多数公司不这么做。所以，这种不允许私下解决、不允许稍后解决、实时解决的理念，这就是为什么会议要三个小时。所以，这是那个会议的一个方面。

It's called the E-team, execution team. So, it's very focused on execution of the company. And the idea is that, you know, if you have five 5-day working weeks, there's never on average more than 2 and 1/2 days before you, if you're blocked on something, can escalate to me, Alex, and all the other VPs. So, the idea is you should never be blocked more than max 2 and 1/2 days. Because we run this synchronized ship, if you're blocked, it gets very expensive because everyone else is downstream of you. So, if you're running a synchronized operation the way we're doing, escalation processes are very important and resolution is very important. So, a big part of that meeting is people say, you know, we're off track here. I'm dependent on this man or woman over there who hasn't done what they said. And the beautiful thing about being able to have all the VPs in the same room is, you know, I've met so many meetings that I'm sure you've been in, people say like, okay, we'll take that offline. I'll talk to you later. And what we said is you're not allowed to say the word offline or later because that person is in the room. So, it's like, you know, I'm dependent on maybe Anna over there for this, but then Anna is actually there. And then Anna can say, you know, okay, I didn't know that or I'm going to solve that. So, it's like real-time resolution. Very simple in theory, but incredibly powerful in practice. Most companies don't do it. So, this notion of like, no taking it offline, taking it later, real-time resolution. That's why it's 3 hours. So, that's one thing of this meeting.

所以，这种不允许私下解决、不允许稍后解决、实时解决的理念，这就是为什么会议要三个小时。所以，这是那个会议的一个方面。

So, this notion of like, no taking it offline, taking it later, real-time resolution. That's why it's 3 hours. So, that's one thing of this meeting.

那个会议里我们还有另一个原则，除了不允许私下解决之外，就是你实际上不能带你的直接下属来，无论好坏。理念是，如果你带很多直接下属来，会发生两件事。一是副总裁就不会被迫去深入了解细节。所以，我实际上是试图强迫副总裁们自己解决问题，因为我想让他们深入细节。所以，你不允许带任何人来解释你的事情。你必须足够了解它，能自己解释清楚。另一个好处是，随着时间的推移，这个群体会变得非常紧密，因为你不会一直换人。所以，你们会建立更强的默契。人们可以坦诚相待。没有人害怕。这是一个非常强大且高效的团队。所以，这是我们做的很多工作。另一部分是战略和展望未来。比如说，你知道，我们一直在做的一些事情现在已经公开了。我们想引入音乐视频。然后团队去研究，说，这在许可、产品方面需要什么？对公司、对损益表有什么成本影响？他们来向 E-team 展示，说，这就是我们想做的。这是我们认为需要的时间。所以，这既是保持引擎运转、永不停歇，也是为未来做规划。但我们并不真正在那个会议上做详细规划。那太大了，无法做详细规划。那发生在焦点小组里，由专家组成的小组进行。然后他们来向那个团队展示。并不是只有我们公司这么做。我知道 Airbnb 也做类似的事情。我和 Brian Chesky 谈了很多。我认为 Netflix 可能也曾有过类似的做法，但他们现在有点分成内容和业务和产品了。我认为重要的是，这既涉及业务也涉及产品。我们在那里大量讨论产品。所以，Spotify 的业务人员，他们对 AI、对什么是 monorepo 了解很多。他们参与技术讨论，但另一方面，我所有的产品人员、人事和工程师，他们完全了解损益表是什么样的。他们知道我们的目标。他们知道，你知道，他们知道毛利率是什么，运营什么，他们什么都知道。所以，这相当独特，这给了他们一种 CEO 的视角，我认为这种视角在许多公司消失了，因为我们给人们贴上角色标签，比如，你是产品人员，所以你不应该理解财务。那不是真的。如果你是 CEO，你必须理解所有这些东西。

Another principle we have in that meeting, except nothing goes offline, is you actually can't bring your direct reports for good or bad. The idea is that if you bring in a lot of direct reports, two things are going to happen. One is the VP is not going to get forced to know the details as much. So, I'm trying to literally force the VPs to solve it themselves because I want them to be in the details. So, you're not allowed to bring anyone else in to explain your thing. You have to be on top of it enough to explain it to yourself. The other benefit of that is over time this group gets very tight because you don't switch people out all the time. So, you build stronger rapport. People can be honest. No one is afraid. It's a very strong and high-functioning team. So, that's a lot of what we do. The other part is strategy and looking forward. So, let's say that, you know, something we've been working on for some time now that's public. We wanted to introduce music videos. And the team goes off and says, what does that take in terms of licensing, product? What is the cost implications for the company, for the P&L? They come and present to the E-team like, this is what we want to do. This is how long we think it should take. So, it's a combination of keeping the engine running and never stopping and also planning for the future. But we don't really plan in that. It's too big to have detailed planning. That happens in focus rooms with smaller groups with experts. And then they come and present to that team. It's not like we're the only company that does this. I know Airbnb does something similar. I spoke a lot to Brian Chesky about it. I think Netflix may have had something similar at a time, but they're now sort of divided into content and business and product. What I think is important about this is it's both the business and the product side. And we talk a lot of product there. So, the business people in Spotify, they know an awful lot about AI, about what a mono repo is. They're there for the participation on technology, but on the flip side all my product and people and engineers, they know exactly what the P&L looks like. They know our goals. They know, you know, they know what gross margin is, what operating they know everything. So, that's quite unique and that gives them sort of a CEO perspective that I think disappears in many companies because we put on these roles of like, you're a product person, so you're not supposed to understand finance. That's not true. If you're the CEO, you have to understand all of it.

如果听众对这个赌注看板流程感到好奇，你可以提交项目，这似乎是一种非常优雅的资本分配方式。关于这个流程的利弊，你会给他们什么建议？而且我知道你已经做了很长时间，它如何随着时间演变，以反映对什么使其成功或失败的经验教训？

If people are listening and are curious about this bets board process where you can submit projects and it seems like a really elegant way to allocate capital. What advice would you give them about the pros and cons of this process? And I know you've been doing it a long time, how it's changed over time to reflect the learnings of what makes it work or fail.

所以，这个概念本身其实非常简单。它来自看板。它有点来自开发者社区。而它实际上最终来自汽车制造业。就像，它其实就是堆栈排序的概念，理论上很容易，实践中很难。很少有人能说这个实际上比那个更重要。他们只会说，这些都很重要，两者都是。当你追问他们时，他们会说，不，它们同等重要。但那样就没有排序了。所以，真正的秘诀是进行堆栈排序，说，你有两个宝贝，但如果你必须杀掉其中一个，按相反顺序你先杀哪个？所以，很容易，但要让整个公司都同意这一点就很难了。但一旦你有了它，它就会消除，它给组织带来很多清晰度。因为当你说这三件事同等重要时，会发生什么？但它们并不是真的同等重要，从来都不是。

So, the concept itself is actually really straightforward. It comes from the Kanban board. It kind of comes from the developer community. And it actually, which actually comes from car manufacturing eventually. It's just like, it's really the concept of stack ranking which is very easy in theory and very hard in practice. Very few people manage to say this is actually more important than that. They're just saying like, these things are very important, both of them. And when you press them, they say like, no, they're equally important. But then they're not ranked. So, the real secret is to stack rank and say like, you have your two darlings, but if you have to kill one of them, which do you kill first in reverse order? So, very easy, but hard to do across the entire company to agree on that. But once you have it, it removes, it gives so much clarity to the org. Because what happens when you say like, these three things are equally important, but they're not really, they never are.

你必须做出选择。你把决策往下推，那么对这个事负责的副总裁就会开始和另一个事的副总裁打架。如果你作为领导不给出明确性，你的组织就会陷入内斗。而且人们都很友善，他们会以为是自己不喜欢对方。所以，只要进行排序并在全公司保持完全透明，就意味着如果我来找你说，你知道，我需要你做这个，你说，好，但我在做那个，我们看看看板说，哦对，我们应该做这个。这很简单，但非常有效。

You're going to have to choose. You just push the decision down the org. And this VP who's on the hook for that thing is going to start fighting this VP who's on the hook for the other thing. And if you as a leader don't bring clarity, you're going to set your org up for fighting. And people are very nice. They're going to think that they don't like each other. So, just the stack ranking and being completely transparent across the entire company means that if I come to you and I say, you know, I need you to do this, and you say, yeah, but I'm doing this, we look at the board and say, oh right, we should do this. It's very simple, but very effective.

当你这么做的时候，你会遇到一堆问题，理论上的问题，比如，一旦你完成了这个 BET 看板，你是全球调配资源吗？你会不会逐个开发者说，我们尽量推进？如果每个人都可以被抢走，那规划过程就是地狱。我的副总裁们对自己将拥有多少资源毫无预估。你基本上剥夺了所有副总裁的权力。这在某种意义上很有效，因为你做到了完美的全球资源调配，但规划过程极其低效。

And when you do that, there are a bunch of things you run into, theoretical questions of, okay, once you have this BETs board done, do you resource it globally? Do you go through every developer and say, let's just try to get as far as we can? That planning process is hell if everyone is up for grabs. None of my VPs have any estimate of what resources they will have. You basically disempower your entire VPs. And it's effective in a sense because you do perfect globally perfect resourcing, but it's incredibly inefficient to do the planning.

那么问题就是，你怎么把它分成块？然后我们有这样的结构：我们有一个平台组织，负责云端 GCP、开发者工具、安全等等。然后我们有一个体验组织，负责整个消费产品，涵盖移动端、车载、桌面等。然后有一个个性化组织，因为这对我们非常重要，负责所有 AI 和推荐，并在书籍、音乐、播客、视频等之间做平衡。然后我们有三个业务垂直领域：音乐、播客和书籍。所以，他们有自己的资源。

So, then the question is, how do you divide it into blocks? And then we have the structure we have: we have a platform organization that works with GCP in the cloud and the developer tools and security and all of that. Then we have an experience organization responsible for the entire consumer product across mobile, car, desktop, etc. Then a personalization organization, because that's so important to us, that does all the AI and recommendations and balances between books, music, podcast, video, etc. And then we have three business verticals: music, podcast, and books. So, they have their own resources.

我们做的是，先要求他们在不互相抢资源的前提下，用现有资源尽可能推进。然后我们尽量推进，因为你需要给他们可预测性，以便他们能规划自己的工作。然后在这个过程结束时，你可能会在全球范围内调动一些人，以确保不会出现某个重要项目缺两个人的情况。那对公司不是最优的。所以，你可能会调动一些人，但大体上我们尽量让人们保留资源。

And what we do is we start by asking them to resource as far as they can with the resources they have without stealing from each other. And then we get as far as we can because you need to give them predictability for them to be able to plan their own work. And then at the end of that process, you may move some people around globally to make sure that you don't have something really important with two people missing. That's not optimal for the company. So, you may move some people around, but largely we try to let people keep the resourcing.

所以，你会遇到很多这样的问题。但我要说这个模型最大的风险，如果你完美同步，听起来不错。这个模型的缺点是规划成本非常高。所以你必须非常擅长规划。我们不得不构建自己的工具。我们尝试了一些外部规划工具，但不够好。如果规划不起作用，相对于执行，开销会增长得非常快。我们执行 6 个月，以免开销变得太大。但我们不能拖到一年，那样你就无法反应。一个季度太短，规划开销相对于执行太大。所以，规划是你必须真正擅长的事情。我不会说我们非常擅长，但我们一直在进步。这是我最关心的事情，确保规划规模合理。

So, lots of those problems that you run into. But I would say the biggest risk with this model, it sounds nice if you're perfectly synchronized. The drawback of that model is that the planning is very expensive. So, you have to be really good at planning. And we've had to build our own tooling. We tried some external tooling for planning. That wasn't good enough. And if the planning doesn't work, the overhead just grows very quickly versus execution. And we execute for 6 months in order for the overhead to not get too big. But we can't go to a year. Then you can't react. A quarter is too short. It's too much planning overhead versus execution. So, the planning is the thing you have to get really good at. And I'm not going to say we're really good. But we're getting better all the time. It's the thing that I care the most about, making sure that the planning is reasonably big.

如果你能为我们做到这一点，那就至关重要，因为 Spotify 的整个产品策略是我们拥有庞大的分发渠道，单个应用接近 7 亿月活跃用户。我们的整个策略基本上是我们多年前在它流行之前就决定的，但当时你看到中国应用开始构建超级应用。而西方世界是每个用例一个应用。我们有点采用了中国超级应用的理念，并说最难的事情将是获得安装量。你可以看到 App Store 的平均安装数下降到平均不到一个。所以，分发成为最重要的事情。

If you can do it for us, it's critical because the whole of Spotify's product strategy is that we have large distribution closing in on 700 million MAUs for a single application. And our entire strategy is basically we decided this many years ago before it was popular, but you saw the Chinese apps starting to build super apps. Whereas the Western world built one app per use case. We kind of adopted the Chinese super app idea and said the hardest thing is going to be to get installs. You could see the average number of installs from the App Store dropping like below one on average. So, distribution became the most important thing.

然后我们选择，当我们做播客，后来做书籍和视频时，把它构建在同一个应用中，因为这样我们可以利用自己的分发。但这有缺点。你必须有一个组织，因为那样一切都相互依赖。你要向 App Store 发布一个应用，每个人都是利益相关者。所以，你不能分而治之。你不能说，好吧，图书团队，你们可以先行，或者音乐团队，你们做这个。不，每个人都必须等待每个人。所以，由于我们的消费策略，公司需要同步。因为需要同步，我们需要一个非常强大的规划流程。所以，这有点是我们消费策略的结果。

And then we chose, when we did podcast and later books and videos, to build it in the same application because then we can leverage our own distribution. But that has drawbacks. You have to have an organization because then everything is dependent on each other. You're going to ship one app to the App Store and everyone is a stakeholder. So, you cannot divide and conquer. You cannot say, well, the book team, you can run ahead, or the music team, you do this. No, everyone has to wait for everyone. So, because of our consumer strategy, the company needed to be synchronized. And because it needed to be synchronized, we needed a really strong planning process. So, it's kind of an outcome of our consumer strategy.

我要说的是，这不是唯一正确的方案。这对我们来说是合适的。我们擅长做全局变更，比如改变整个用户界面，因为我们同步。但在快速尝试新事物方面，我们可能比其他公司慢得多。因为它需要经过大量规划等等。我认为你无法在规划上做到完美。你最好的期望是，在重要的事情上做得相当好，在不太重要的事情上做得不那么好。

And what I would say is it's not the right one. It's the right one for us. We're good at doing global changes, like changing the entire UI, because we're synchronized. But we're probably much slower than other companies at quickly trying something. Because it needs to go through a lot of planning and so. I think you can't win in planning. The best you can hope for is to be quite good at the important things and not so good at the less important things.

我要回到你刚才说的一个非常有趣的观点，关于采用一些最前沿的工具。让我们以 Cursor 为例，现在大家都熟悉这家公司，估值 100 亿美元。似乎每个软件工程师都在用 Cursor 来提升自己。但你的表述很酷，是的，当然，但那主要是新代码。那只是他们时间的八分之一的一小部分。所以，在饼图里，那是很小的一块，被 Cursor 在大公司里解决了。你能描述一下你认为这会如何发展吗？因为感觉公开市场，尤其是，非常好奇。我想私募市场也是。非常好奇 AI 公司、产品和工具将如何解决这个更大的饼图部分，听起来还没有被直接触及。

I'm going to come back to something very interesting you said around the adoption of some of the tooling that's at the most cutting edge. So, let's take Cursor as an example of a company that now everyone's familiar with, $10 billion valuation. Seems like every software engineer is using Cursor to make themselves better. But the way you framed it was so cool that yeah, sure, but that's new code primarily. That's a fraction of 1/8 of their time. And so, there's all in the pie chart, there's a very small sliver that's being addressed by Cursor at big companies. Can you describe how you think this will play out because it feels like the public markets, especially, are very curious. I guess private markets, too. Very curious about how AI companies and products and tools will address this, you know, this bigger much bigger part of the pie that sounds like really hasn't been hit too directly yet.

有几件有趣的事情，我认为不太明显。一是过去每个开发者都开始使用 Cursor。但现在，我开始看到更多非开发者使用 Cursor。部分原因是行业开始就一个名为 MCP（模型上下文协议）的协议达成共识。这意味着如果你把你的内部服务包装成 MCP，你就可以用英语与你的基础设施对话。所以，如果你现在是开发者，你可以坐在 Cursor 里说，你知道，我要……或者如果你是设计师，例如，或产品人员，假设你想在 Spotify 中原型一个功能。

There are a couple of things that are interesting that I don't think are super obvious. One is used to be that every developer started using Cursor. But now, I'm starting to see a lot more non-developers using Cursor. And that's partially because the industry is starting to agree on this protocol called MCP, model context protocol. Which means that if you take your internal services and you wrap them in an MCP, you can speak English to your infrastructure. So, if you're a developer now, you can sit in Cursor and say, you know, I'm going to... Or if you're a designer, for example, or a product person, let's say you want to prototype a feature in Spotify.

一个工作流是，你拿现有的 Spotify，双击截图，上传到 Cursor 里说：“用 HTML 把这个做成可点击的。”然后如果你的服务都封装在 MCP 里，你理论上可以说：“现在，把这个接上，比如我的喜欢音乐流。”就算你不是开发者，你也能做原型，因为基础设施现在通过 MCP 用英语封装了。我觉得这很重要。所以，我认为你会看到比开发者多得多的人用 Cursor。我开始看到，我有一个产品经理，她在瑞典，她是新西兰人，不会说瑞典语。她用 Cursor 报税，成功把瑞典税务局封装进了 MCP。她不是开发者，对吧？所以，我认为它会扩展到开发者之外。

One workflow is you take the existing Spotify, you double click and screenshot it, you upload that into Cursor and say, "Wire this up clickable in HTML." And then if your services are wrapped in MCP, you could theoretically say, "Now, wire this up to, you know, our my liked music feed or something." And you can prototype even though you're not a developer, because the infrastructure is wrapped in English language now through MCP. I think that's an important thing. And so, I think you're going to see many more people using Cursor than just developers. I'm starting to see I had one of my PMs who is in Sweden. She's from New Zealand. She doesn't speak Swedish. She did her taxes in Cursor. Managed to wrap the Swedish tax authority in an MCP. Not a developer, right? So, I think it's going to grow outside of developers.

但我认为这正说明了许多大公司里实际发生的情况，这也是为什么初创公司能跑得更快。所以，如果你想想 Spotify 这样的公司，它有大量的基础设施。你知道，有 15 年播放历史的数据库。你知道，家庭计划里有谁。那是一个服务器。这个数据集在某个地方。你的品味图谱是一个数据集，等等。现在，大型 AI 公司来了，给了你这个推理引擎。你知道，其中一些是开源的。所以，基本上免费，你就得到了接近 AGI 的东西。现在，你有了这个你以为会极其昂贵的东西，却几乎免费得到了。这是一份礼物。你开始用它。你遇到的第一个问题是什么？你说，比如，“我的音乐收听在过去一年里怎么变化的？”这没有作为 API 暴露出来，因为在以前的机器学习世界里，那些数据，15 年前的收听数据，在某个冷存储里。工程师得做一个 SQL 任务，可能要花一周才能拉出来。然后你会训练一个模型，然后你会把它放回冷存储。如果现在你想实时推理这些数据，你需要把所有数据实时暴露为 API。实际上，我最大的工作不是 AI 工程，而是老式工程，把我们拥有的所有这些数据暴露出来，这样你就能让一个推理引擎作为产品人员为你推理，或者实际上作为消费者，可能实时推理我自己的数据。

But I think this points to what is actually happening in many of these big companies, which is why the startups can move faster. So, if you think of a company like Spotify, it has tons of infrastructure. You know, you have the database with play history going 15 years back. You have, you know, who is in the family plan. That's one server. This one's data set somewhere. Your taste graph is a data set and so forth. Now, here comes the big AI companies and they give you this reasoning engine. You know, some of them are open source. So, basically for free, you get what is getting close to AGI. So, now you have this thing that you thought would be incredibly expensive and you get it almost for free. It's a gift. You start using it. What is the first problem you run into? You say like, you know, "How has my music listening changed over the last year?" That's not exposed as an API, because in the previous machine learning world, that data, the listening data 15 years back, it's on cold storage somewhere. And an engineer would have had to do like an SQL job that may have taken a week to pull it up. Then you would have trained a model, then you would have put it back in cold storage. If now you want to be able to reason over that in real time, you need to expose all your data as APIs in real time. And actually, my biggest job to enable AI is not AI engineering. It's old school engineering, exposing all this data that we have so that you can have a reasoning engine reason for you as a product person, or actually for me as a consumer potentially over my own data in real time.

所以，我认为这就是正在发生的事。所以，现在有了标准的 MCP，你可以把 API 封装进去，而且许多公司，至少我们，在努力暴露所有这些数据，这意味着业务人员、律师、产品人员、设计师将能够使用 Cursor 而无需编码。他们实际上至少可以原型设计或与真实服务对话。所以，我认为这就是我们正在走的旅程。它从开发者开始。但我认为，当你暴露基础设施并封装成 API 时，它会扩展到外部。它必须扩展到外部。

So, I think that's what's happening. So, the combination of now there's a standard MCPs that you can wrap these APIs in, and many of these companies, at least us, trying to expose all of this data, means that a business person, a lawyer, a product person, a designer will be able to use Cursor without having to code. And they can actually at least prototype or talk to real services. So, that I think is the journey that we're on. It started with developers. But I think as you expose the infrastructure and wrap it in APIs, I think it's going to go outside. It's going to have to go outside.

有没有一种说法，虽然会丢失很多信息，但可以总结一下，我们在开发者身上看到的情况将会发生，甚至可能，我很惊讶他们用的是 Cursor。这很有趣。但用其他类似工具。是的。而且我们更多的工作将会是，感觉就像我们在和一个团队合作，和团队说话，用自然语言做原型、尝试东西。这会慢慢扩散到整个，不仅是软件开发者的工时，还有其他每个职能领域的工时。

Is there a way to say that and lose a lot, but summarize it, that what we've seen happen with developers is going to happen maybe even it's I'm surprised that it's Cursor that they're using. That's quite interesting. But with other similar tools. Yeah. And that just more of our work is going to be, it's going to feel like we're working with a team, speaking to a team, using natural language to prototype things, to try things. And that will diffuse slowly through the entire not only the hours of the software developer, but the hours of each of the other functional areas.

是的，我想是这样。很难看到它会落在哪里，因为你现在处于某个位置，但我们相当确定那个位置在这条曲线上。所以，你可以相当确定你现在看到的工作流不会保持不变。这实际上是问题之一。比如，我们要为现在看到的东西构建多少，而你知道模型会更有能力，很快会有不同的工具。所以，你不想过度拟合当下。

Yeah, I think so. It's hard to see where it's going to land, because you're somewhere right now, but we're pretty certain that that somewhere is on this curve. So, you can be pretty certain that the workflows you see right now are not going to be the same. And that's actually one of the problems. Like, how much are we going to build for what we see right now, when you know the models are going to be more capable, they're going to be different tooling very soon. So, you don't want to overfit too much to the moment.

一个现代公司的合理视图是，它的所有数据都实时暴露，你上面有像 Cursor 或其他工具。也许不同技能用不同工具。也许一个工具，Spotify 的授权团队可能有一个不同的工具来推理所有合同，并快速说：“我们认为我们能在那个市场做这个吗？我们需要授权什么才能做？”但产品团队也可以问那个授权引擎，“我们有大约 15 年的合同，包括当前的和以前的。”所以，这个 AI 对音乐授权样子的洞察比 Spotify 里任何一个人都多，如果你那样训练它的话。所以，可能会有针对不同技能的略微定制界面。我不确定哪个会胜出，但我认为它会看起来像那样。

A reasonable view of a modern company is that all of its data is exposed in real time, and you have some tool on top like Cursor or something else. Maybe different tools for different skills. Maybe a tool more the licensing team at Spotify may have a different tool to reason over all the contracts and quickly say like, "Do we think we can do this in that market? And what do we need to license to do this?" But also the product team could ask that licensing engine, "We have like 15 years of contracts, both current and previous." So, this AI has a lot of insight into what music licensing looks like more than any single person in Spotify if you train it that way. So, there will probably be slightly custom interfaces for different skills. I'm not sure which is going to win out, but I think it's going to look something like that.

现在，我们看到人们在做什么，他们在分享他们用于工作流的提示词示例。然后是他们用过的原型。这感觉非常像某个时间点。有点粗糙，你知道，各种东西。如果你要给外面的世界校准一下，很少有人有你这样的内部视角，你对这项技术感到兴奋，你试图拥抱它，你只能以我们描述的方式这么快地拥抱它。比如在 1 到 10 分的尺度上，你会给目前为止它对你的影响打多少分？以及它可能会变得多疯狂。比如，人们非常兴奋，认为这会彻底改变一切。还有一些人实际上担心它可能有多强大。从实际现实世界的角度来看，你能给我们校准一下吗？作为少数几个既对它兴奋又每天面对现实的人之一。

Right now, what we see people doing is they're sharing examples of prompts they used for the workflows. And then prototypes that they've used. And that feels like very much a point in time. It's kind of hacky and, you know, different things. If you were to calibrate the world out there, so few people have the inside view that you do, where you're excited by this technology, you're trying to embrace it, you're only able to embrace it so fast in the ways that we've described. Like on a 1 to 10 point scale or something like this, what score would you give how much this is impacting you so far? And like how crazy this might get. Like, people are very excited that this is going to literally change everything. And there's some people that are actually worried about how powerful it might be. From a practical real world standpoint, could you calibrate us a little bit as someone one of the few people that like actually is both excited about it and also faces reality on a daily basis?

如果你想尽可能现实一点，你拿开发者用例来说。我见过其他大公司的研究，如果你实际衡量开发者的时间，加速大约是 7% 或什么的，这听起来非常令人失望，因为所有这些事情。比如，净新增编码只占一小部分。净新增只占一小部分，等等。所以，我认为现在，就实际影响而言，它有点被过度炒作，至少对这些大公司来说。但我认为它会变成相反的情况。我认为现在人们相对于实际影响过于兴奋。但我认为相反的情况会发生。我认为长期来看它会产生巨大影响。我现在看到人们在做什么，这取决于情况。我的意思是，我个人经常使用它。

If you want to be as realistic as possible about it, you take the developer use case. I've seen studies from other big companies that if you actually measure out of a developer's time, the speed up is like 7% or something, which sounds very disappointing because of all these things. Like, the net new coding is a small part. Net new is a small part of that and so forth. So, I think right now, it's a bit overhyped in terms of actual impact, at least for these big companies. But I think it's going to turn into the opposite. And I think right now, people are overexcited versus the actual impact. But I think the opposite is going to happen. I think it's going to have tremendous impact over the longer term. What I see people doing right now, it depends. I mean, I use it a lot personally.

我看到很多开发者、产品人员和设计师一直在用 AI 来提高生产力。事实上，把东西放进引擎里，让它总结，诸如此类，这些事一直在发生。我很难估计这已经让他们的效率提升了多少，但确实有提升。不过我认为真正大的影响来自于你用这项技术重塑这些公司。现在，我们只是把它附加在现有系统之上。但正如我所说，你必须重塑并重建公司，以适应这种推理引擎能实时推理整个公司数据的工作方式。但这实际上需要大量的重新调整。这就是为什么初创公司领先。他们不需要重建，他们没有 15 年的数据。所以，他们可能在未来，这就是为什么他们觉得‘不，不，Gustav 错了，影响已经很大了。’我认为对初创公司来说确实如此。我认为对大公司来说需要更长时间。像我们这样的大公司，我们必须振作起来，加速前进，以免落后。

I see a lot of my developers and product people and designers use it all the time for productivity purposes. In fact, you know, putting things into an engine, asking it for a summary, and so forth. Those things happen all the time. It's hard for me to estimate how much that speeds them up already. It certainly does. But I think the really big impact comes as you reshape these companies from this technology. Right now, we're just tacking it on top. But as I said, you have to reshape it and rebuild it for this work where a reasoning engine can reason in real time over the entire company's data. But that requires actually a lot of retooling. That's why startups are ahead. They don't have to rebuild. They don't have 15 years of data. So, they're probably in the future, which is why they feel like, 'No, no, Gustav is wrong. The impact is really big already.' And I think it is for a startup. I think it takes a bit longer for big companies. And big companies like us, we have to shape up and accelerate in order to not be behind.

不过，我相信他们都希望能有 7 亿月活跃用户来做实验。

I'm sure they would all like to have 700 million monthly active users to experiment with, though.

是的，这是好处之一。

Yeah, that's one of the benefits.

关于这个话题，你提到了移动时代的经历，不仅一切都因移动而改变，而且商业模式也需要改变。到目前为止我们主要谈了产品，关于产品还有更多要问的。但谈谈商业模式吧。比如，在什么样的世界里，由于这项技术，Spotify 的整个商业模式需要改变？你又是如何评估这类事情的？

On that topic, you mentioned going through mobile and the experience of not only was everything changing as a result of mobile, but actually the business model also needed to change. We've really talked about product so far, and there's more to ask about product. But talk about business model. Like, what would be the world in which as a result of this technology, Spotify's whole business model needs to change? And how do you go about evaluating something like that?

这是个很好的问题。我们见过一些商业模式变革的例子。我倾向于告诉我的产品团队，每个人都说世界被技术颠覆和改变。我认为这在某种意义上是正确的，因为根本驱动力是技术本身。技术是不断给予的礼物。它给了你计算机、互联网、智能手机、机器学习、AI、量子计算。这些礼物几乎按计划不断到来，而且间隔越来越短。以前，科技公司不叫科技公司。顺便说一句，它们被称为汽车公司。但它们是科技公司，或者你知道，制药公司是当时最先进的技术。但因为这些浪潮相隔太远，它们称自己为汽车公司。它们从未成为无处不在的科技公司。它们有点过度适应了那个时代。我认为在 90 年代某个时候，大约在 Google、Amazon 等公司出现时，这些浪潮开始来得如此之快，以至于人们试图把它们固定下来，比如 Amazon 是一家图书公司。它们说，‘不，不是真的。我们在卖书，但这里还有其他东西。’然后他们说，‘好吧，你是一家什么都卖的公司。’它说，‘不，不是真的。现在我们在这里卖 Amazon Web Services。’所以，我认为这些公司是第一波将技术作为战略的公司。之前的公司只把一波浪潮作为战略。然后，你知道，IBM 出现了，把计算机作为战略，或者首先是内存等等。我认为我们正在看到第一波通用科技公司。有趣的是，这可能意味着它们可能——我的意思是，公司几乎总是过一段时间就会消亡。这些可能是第一批永不消亡的公司，因为它们是无处不在的科技公司。无论技术礼物是什么，只要努力让公司快速适应它，弄清楚产品和商业模式。所以，我认为这很有趣。这也是我对 Spotify 的看法。是的，我们是一家音乐公司，然后是播客公司，然后是图书公司，然后是视频公司。但真正重要的是试图预测技术，弄清楚它能做什么，然后调整产品，通常还有商业模式。

It's a great question. And we've seen a few of those examples of business models. And I tend to tell my product teams that everyone says that the world is disrupted and changed by technology. And I think that's true in the sense that the underlying force is technology itself. And it's this gift that keeps on giving. It gives you computers, internet, smartphones, ML, AI, quantum computing. And these gifts keep coming almost on a schedule, and they actually come closer and closer. Previously, technology companies were not called technology companies. As a side note, they were called car companies. But they were technology companies, or you know, pharmaceutical that was the state of the art technology right then. But because these microwaves came so far apart, they called themselves a car company. They never became ubiquitous technology companies. They kind of overfitted to that. I think somewhere in the '90s, around Google, Amazon, etc., these microwaves started coming so fast that people tried to pin them down as, you know, Amazon is a books company. And they were like, 'No, not really. We're doing books, but here's other stuff we're selling.' And then they're like, 'Okay, you're the everything store company.' It's like, 'No, not really. Now we're selling Amazon Web Services over here.' So, I think these companies are the first set of companies to have technology as the strategy. The previous ones took one wave as the strategy. And then, you know, IBM comes along and does computers as a strategy or first memory and so forth. I think we're seeing the first wave of general technology companies. Which, interestingly, might mean that they could be—I mean, companies almost always die after a while. These could be the first companies that never die because they're ubiquitous technology companies. Whatever the technology gift is, just try to have a company that can quickly wrap around it, figure out the product and business model. So, I think that's interesting. And that's how I think about Spotify. Yes, we're a music company, and then a podcast company, and then a book company, and then a video company. But it's really about trying to anticipate technology, figure out what it can do, and then adapt the product and often the business model.

所以，我说移动是我们需要改变商业模式的情况之一。我认为当这些技术礼物之一出现时，技术本身会带来巨大变化。比如，你知道，盗版，大混乱。但真正的变化发生在有人也弄清楚了商业模式时。所以，我告诉我的产品团队，技术可以做好事。技术和新的商业模式才能真正改变世界。但没有商业模式，很少有大规模变革。你可以摧毁很多东西，但你从未真正创造价值。所以，移动是我们第一次需要在移动端推出免费层而不蚕食付费层的情况。我们当时做的是查看数据，发现 50% 的高级用户在用随机播放模式。所以，我们说，‘如果我们把随机播放作为一个功能，免费提供呢？这应该是高级消费的 50%，非常有价值，但不会是任何人高级消费的 100%，所以不会蚕食。’我们设法创建了一个层级，你可以把你最喜欢的歌曲放在播放列表里，按播放，把手机放进口袋，然后免费在后台无限听。所以，那是商业模式创新加上技术创新。

So, I said that mobile was one of these things where we needed to change the business model. And I think what happens when one of these technology gifts comes along is there is a big change when the technology happens. Like, you know, piracy. Big havoc. But the real change happens when someone also figures out the business model. So, I tell my product teams like technology can do good things. Technology and a new business model can really change the world. But without a business model, there's seldom large-scale change. You can destroy a lot of things, but you never really create value. So, mobile was the first where we needed to figure out the free tier on mobile without cannibalizing our paid tier. And what we did there was we looked at our data and saw that 50% of premium users were listening in shuffle mode. So, we said, 'What if we take shuffle as a feature, give that away for free. It should be 50% of premium consumption is very valuable, but it's not going to be 100% of anyone's premium consumption, so no cannibalization.' And we managed to create a tier where you could playlist all your favorite songs in a playlist, press play, put the phone in your pocket, and listen forever for free in the background. So, that was a business model innovation along with technology.

最近的一个例子是有声书，你知道，在美国有声书是按本购买的。当然，我们在流式播放那本书方面做了一些不错的创新。但真正重要的不是流式播放一本书。你很久以前就能流式播放音频了。真正的创新是商业模式，能够将有声书捆绑到 Spotify Premium 中，并把它从一种——几乎就像音乐。音乐也是按首购买的，而且相当小众。一旦我们把它变成访问模式，你知道，没有边际成本，它就变得大得多。我们也是这么看待有声书的。所以，我们见过几个这样的例子，并设法采用了它们。

The most previous one was audiobooks where, you know, there were audiobooks in the US a la carte. And sure, we did some nice innovation around being able to stream that book. But the real thing is not stream a book. You've been able to stream audio for a long time. The real innovation was the business model, to be able to bundle audiobooks into Spotify Premium and take it from a sort of—It's almost like music. Music was also a la carte and quite niche. And once we made it an access model, you know, with no marginal cost, it got way larger. And that's how we think about audiobooks as well. So, we've seen a few of those and managed to adopt them.

回到你的问题，AI 会做到这一点吗？我们需要改变商业模式吗？

To your question, is AI going to do that? Do we need to change the business model?

我不确定。我认为有一个明显的不同之处，那就是之前的风险投资模式，从芯片和硅片一路走来，是你进行大额前期投资，然后摊销，最终边际成本几乎为零。软件就是这样运作的。但 AI 不是这样运作的。边际成本很高，你需要覆盖它。所以，你可以说这应该改变每个人的商业模式。你需要以某种方式要么通过广告非常有效地变现，要么向用户收费。你看到 OpenAI 是订阅产品。我认为你会看到更多这样的产品来覆盖成本。所以，边际成本是一个全新的东西。对 Spotify 来说，这很有趣，因为我们就像一家总是有边际成本的科技公司。

I'm not sure. I think there's one glaring thing that is different, which is the previous sort of VC model coming all the way back from chips and silicon was you make a big upfront investment and then you amortize and you get to almost zero marginal cost. That's how software worked. It's not how AI works. The marginal cost is high, and you need to cover it. So, you could say that that should change everyone's business model. You're going to need to somehow either monetize very effectively through ads or charge users. And you see OpenAI being a subscription product. And I think you're going to see more of those to cover the cost. So, the marginal cost is a net new thing. For Spotify, it's interesting because we're like the one technology company that always had a marginal cost.

多一次播放对唱片公司来说就是边际成本。所以，我们成长在一个如果免费层做得太成功，就可能一夜破产的世界。这对 Twitter 或 Facebook 来说从来不是问题，所以风投会说：“尽管疯吧，变现的事以后再说。”Spotify 永远不能那样做，因为我们可能一夜破产。所以，我们总是得操心变现，操心免费层和付费层转化之间的平衡，以及免费层的变现。所以，对我们来说，好处是我们相当习惯商业模式中的边际成本。

One more stream was a marginal cost to labels. So, we grew up in a world where if we were too successful on the free tier, we could go bankrupt overnight. Which was never true for Twitter or Facebook, which is why VCs said like, "Just go crazy. Worry about monetization later." Spotify could never do that because we could go bankrupt overnight. So, we always had to worry about monetization and the balance between free tier and paid tier conversion and free tier monetization. So, the good thing for us is we're fairly used to marginal cost in our business model.

所以，我不认为你会看到那些情况。很可能有些消费者会想要海量的推理。因为那是边际成本，作为消费者，你可能得为此付费。所以，我认为你会看到消费产品根据你想要的推理量进行更多分层。但对我们来说，这并不新鲜。我们已经有好几个层级了。

So, I don't think you're going to see those things. It's very likely that some consumers are going to want tons and tons of inference. And because that's a marginal cost, you're probably going to have to pay somehow for that as a consumer. So, I think you're going to see more tiering of consumer products based on how much inference you want. But for us, that's not that new. We've had several tiers already.

我很好奇，因为我投资了一家叫 Etch 的公司，它将成为压低推理成本的公司之一。就像计算的历史一样，你会看到随着推理成本单位成本越来越低，带来巨大的消费者剩余和消费者利益。但反作用力是我们会使用更多，你知道，更多的推理 token，更多别的。所以，这让我想知道，你能想象更好的模型在多大程度上对你有用？比如，如果我们今天就冻结推理和模型能力，我们可能还有十年以上的消化期，来思考如何用这些模型做出更好的产品、更好的功能等等。你能想象一个好 10 倍、再好几个数量级的模型会开启许多你目前无法实现的功能吗？这是可能的吗？还是你认为我们基本上已经拥有了所需的一切？因此，推理可以预期会非常便宜。

I'm curious because I'm an investor in a company called Etch that's going to be one of these companies that pushes down that inference cost. And like the history of compute, you're going to see this incredible consumer surplus and consumer benefit that comes from cheaper and cheaper unit by unit inference cost. But the countervailing force is that we would just use more of it, you know, more reasoning tokens, more whatever. So, it makes me wonder how much more you can imagine better models being useful to you. Like, it seems like if we just froze reasoning and model capabilities today, we probably still have decade plus of digestion to do of how we could use these models to make better products, better features, whatever. Can you imagine like a 10 times better, another couple orders of magnitude better models opening up lots of features that you can't currently do? Like, is that a thing? Or do you think we kind of have what we need? And therefore, inference we could expect to be really cheap.

所以，我既认同产品过剩的观点，即如果我们冻结，会有巨大的产品过剩。我认为我们会看到产品推出，看起来很棒，持续好几年，然后才会耗尽我们现有的东西。所以，我认同这一点。但我也认同算力没有极限。你最终会达到计算物质（computronium）。但如果你看看计算物质的物理学……

So, I both subscribe to the product overhang idea that there's a huge product overhang if we froze. I think we would see product shipped that look amazing for several years before we exhausted what we have. So, I subscribe to that. But I also subscribe to that there is no limit for compute. You eventually you get to computronium. But if you look at this the physics of computronium

什么是计算物质？

What's computronium?

它是宇宙理论上能进行的最小的、最大的计算量。

It's the smallest, it's the most computation a universe could do, you know, theoretically.

嗯。

Yeah.

我们离那个极限还很远。所以，我认为我们会一路走到那里才会停下。而且我认为我们会非常有创造力。有一个很好的类比，我不知道是谁提出的，但我想 Ben Evans 经常提到。你知道，当电子表格出现时，想法是一样的，现在所有会计师都要失业了。实际发生的是，我们无法想象如果计算成本降到零会发生什么，你可以想象做那件事的价值会归零，因为世界上有那么多会计师。所以，实际发生的是我们开始做多得多的会计工作。当电子表格没有成本时，你会开始做模型来预测这个资产或商品的未来，永远预测下去。我们想出了更多可以做的电子表格工作。而且它比以往任何时候都大。我认为我们会看到完全一样的情况。我认为从财务角度看，当某样东西的成本下降时，需求通常会增加超过下降幅度。我认为这必然发生在智能上。它就像终极之物。说“不，我有足够的智能”并不有趣。我认为我们会为把推理花在多么平凡的事情上而感到羞愧。就像，你知道，“我的咖啡明天能不能热 1 度？”如果提问真的没有成本，我想人们会问的。

We're very far from that limit. So, I think we're going to go all the way there before we stop. And I think we're going to be very inventive. There is a nice analogy that I think, I don't know who came up with it, but I think Ben Evans talks about it quite often. You know, when the spreadsheet came along, the idea was the same, you know, now all the accountants are going to go out of business. What happened was we could just not imagine if calculation cost went to zero, what's going to happen is you could imagine that the value of doing that is going to go to zero because there were so many accountants in the world. So, what happened was we just started doing massively more accounting. When there's no cost to spreadsheeting, you're going to start do models to predict the futures of this asset or good or something, you know, into the future forever. We just came up with so much more spreadsheeting that you could do. And it's bigger than ever. And I think we're going to see exactly that. I think from a financial point of view, when the cost of something drops, the demand usually increases more than the drop. And I think that's bound to happen with intelligence. It is like the ultimate thing. And to say like, "No, I have enough intelligence." It's not interesting. I think we're going to be ashamed of how mundane things we spend inference on. It's like, you know, "Could my coffee be like 1° warmer tomorrow?" If it's truly no cost asking the questions, I think people will.

也许现在是时候问问你，和后花园里的 David Deutsch 坐在一起，和他谈论《无穷的开始》这个概念。计算物质让我想到你对这个话题的兴趣，以及你那里的回答，不，不，这里没有终点。就像，我们会继续前进。我们会继续学习，继续部署我们的新技术。你能谈谈他、那本书、它为什么影响了你，以及你和他的对话吗？

Maybe now's the time to ask you about sitting in the back garden with David Deutsch and talking to him about this concept of The Beginning of Infinity. Computronium made me think of your interest in this topic that like and your answer there that no, no, there's no end point here. Like, we're going to just keep going. We're going to keep learning, keep deploying our new technology. Can you talk about him, that book, why it influenced you, your conversation with him?

是的，David Deutsch 自从我读了《无穷的开始》以来就是我的英雄。然后他写了另一本书叫《真实世界的脉络》。他被认为是量子计算之父。显然，量子计算是技术将给我们的礼物之一，而且我认为它很快就会变得非常真实。所以，我一直很感兴趣，因为量子计算或量子力学是这个星球上最疯狂的东西。你知道，我们生活在我们认为是现实的这个现实中，但如果你去到底层，这不是现实。它只是我们生活在其中的某种三维投影。所以，《真实世界的脉络》对我影响很大。而且他是 Everett 诠释的信奉者，他相信多重世界情景。所以，是的，那本书让我大开眼界。然后，《无穷的开始》可能是他最著名的书，而《真实世界的脉络》实际上是关于量子计算以及量子计算机如何工作的。《无穷的开始》非常哲学化，他在那里有很多大想法。他是一个非常积极的人。而我现在，在他 70 多岁的时候，终于有机会在牛津的花园里采访了他。他健康状况不太好，所以必须在户外，保持距离。每个人对未来都非常悲观，你知道，有太多可能出错的问题，所有这些都可能出错，气候。他实际上对未来非常乐观。他清楚存在风险。但是，你知道，他预见到我们走向星辰。我问他，你认为一百万年后的我们在哪里？他说，嗯，也许我们离太阳系这么远，但还没完全到那里。他非常确定我们会到达那里。他是一个非常积极的人，当我问起他的生活时，他对自己的生活非常满足。他非常非常快乐。所以，即使在这个年纪，他仍然是一个鼓舞人心的人，你知道，我希望我在那个年纪能像他一样。但是，这本书有几个概念我试图在 Spotify 应用，其中之一是他谈到的解释的力量。

Yeah, so David Deutsch has been a hero of mine since I read The Beginning of Infinity. And then he wrote another book called The Fabric of Reality. He's considered the father of quantum computing. And obviously, quantum computing is one of these gifts that technology is going to give us, and it's about to get very real, I think, very soon. So, I've always been interested because quantum computing or quantum mechanics is the most insane thing on this planet. You know, we live in what we consider this reality, but if you go to the bottom layer, this is not reality. It's just some sort of three-dimensional projection that we live in. So, that The Fabric of Reality had a big impact on me. And he's a believer in, he's an Everettian. He believes in multiple worlds scenario. So, yeah, that book blew my mind. Then, Beginning of Infinity is maybe his most famous book, whereas Fabric of Reality is really about quantum computing and how a quantum computer works. Beginning of Infinity is very philosophical and he has a bunch of big ideas there. He's a very positive person. And I've now, at 70-plus, I finally got to interview him in his garden in Oxford. He's not of great health, so had to be outdoors, you know, distanced. And everyone is very negative on the future, you know, there's so many problems that could go wrong and all these could go wrong, climate. He's actually very positive about the future. He's clear that there are risks. But, you know, he sees us going out there into the stars. And I asked him like, where do you think we are in a million years? And he's like, well, maybe we're this far outside of the solar system, but not quite there. He's very certain we're going to get there, so. He's a very positive person and when I asked him about his life, he's very content with his life. He's very very happy. So, he's just an inspiring person still at this age, you know, I wish I will be like him at that age. But, this book has a few concepts that I've tried to apply at Spotify and one of them is he talks about the power of explanations.

他认为人类心智是无限可扩展的。他不认为我们的理解存在极限，因为有了解释。我觉得这一点很多人——我同意这一点，但很多人不同意。当然，有些事情我们永远无法理解。他的观点是：不，我们的理解没有极限。我们是唯一打破这个障碍的物种，因为我们有解释。其他物种有模式识别。它们能做事，能学会“这样有效”的模式。也许有一些文化传递，比如观察别人做那个模式。鸟能看到另一只鸟。有些物种能教它们的孩子，但它们从不产生解释。而且他对“好的解释”有一个定义。他深受哲学家卡尔·波普尔的影响。波普尔是他的“家神”。所以，他从波普尔那里取了一点，也从科学那里取了一点。所以，他说显然一个好的解释必须是可证伪的。但他说了另外几件事，我觉得事后看来很明显，但事前并不明显。他说一个好的解释必须能扩展，必须有影响力。他是什么意思？他说有些解释解释得很局部。比如，你可以有一个关于太阳绕地球转的解释，它解释了一堆东西，但它不能扩展到其他行星，对吧？更好的解释是地球绕太阳转。它只是能更好地扩展到不同尺度。所以，一个好的解释必须能上下扩展。一个好的解释必须与所有先前的解释兼容。但最有趣的是，他说一个好的解释必须难以变动。我觉得这非常明显，但对人们来说也非常不明显。那么，他说“好的解释必须难以变动”是什么意思？他的意思是，比如，如果你对地球天气的解释是“现在托尔生气了，所以有雷声”，那是一个解释，但它太容易变动了。你可以说，好吧，现在是别人生气了。他们也有锤子。太容易变动，而且得到同样的结果。一个好的解释，如果你移动其中一个参数，整个东西就不再具有预测性了。那么你可能接近真相了。我觉得这非常有趣，因为人们喜欢的多数阴谋论的问题在于它们太容易变动了。你只要把那个角色换成另一个疯子做了疯狂的事，仍然会产生同样的结果。所以，如果在一个阴谋论中更换人物太容易，那它可能不是真的。所以，我认为这是非常有力的。好的解释需要非常难以变动。

And he thinks the human mind is infinitely scalable. He does not think there's a limit to what we can understand because of explanations. And I think this is something that a lot of people... I agree with that, but a lot of people disagree. Certainly, there are things we could never understand. His view is no, there is no limit to what we can understand. We are the only species who broke that barrier because we have explanations. Other species have pattern recognition. They can do things and learn the pattern that this works. There's some cultural transfer maybe of looking at someone else doing that pattern. A bird can see another bird. Some species can teach their kids, but they never produce explanations. And he has a definition of a good explanation. He's very inspired by Karl Popper as a philosopher. It's his house god. So, he takes a bit from Popper and he takes a bit from science. So, he says obviously that a good explanation has to be falsifiable. But he says a few other things that I think are obvious in retrospect, but not before. He says that a good explanation has to scale, has to have reach. What does he mean with that? He says that some explanations explain something quite locally. You can have an explanation about, for example, the sun revolving around the earth, which explains a bunch of stuff, but it doesn't scale, right? To other planets. A better explanation is to have the earth revolving around the sun. It just scales better to different scales. So, a good explanation has to scale up and down. A good explanation has to be compatible with all the previous explanations. But most interestingly, he says that a good explanation has to be hard to vary. This I find very obvious, but also very non-obvious to people. So, what does he mean with a good explanation has to be hard to vary? He means that for example, if your explanation for the weather on the planet is that now Thor is angry, so there's thunder there. It's an explanation, but it's too easy to vary. You can say like, well, now someone else is angry. They also had a hammer. It's too easy to vary and get the same result. A good explanation, if you move one of the parameters, the entire thing is not predictive anymore. Then you're probably close to the truth. And I think this is so interesting because the problem with most conspiracy theories that people love is they're so easy to vary. You can just exchange that character for another crazy person did something crazy and still going to produce the same thing. So, if it's too easy to change people in a conspiracy theory, it's probably not true. So, I think that's something very powerful. Like good explanations need to be very hard to vary.

所以，这是我试图在我的组织里灌输的东西。而且我认为这里有一个有趣的元观点，那就是人们作为产品人问我，你知道，产品开发中有多少是魔法，多少是科学。我试图挑衅地说，我认为恰好是 100% 的科学和 0% 的魔法。人们被激怒了，因为这有点暗示没有技巧。所以，我这么说就是为了挑衅。我的意思是，当然，人们在这个神经网络中会有模式识别。他们见过很多例子。这就是我们所说的高级资历。人们见过很多东西。他们会本能地比别人更快得出正确的结论。所以，那是有价值的。我想要很多高级资历。所以，我不放弃高级资历。它带来很多价值。你可以节省很多时间和很多错误。但你说它是魔法的原因是因为那个人无法解释它。它实际上不是魔法。它只是科学。只是你不够聪明，无法解释自己。如果你能想得更远，解释它，并为你所看到的提出一个解释，就像大卫·多伊奇那样，那对公司来说价值大得多。如果你有一个理论，而不是说“不，修补我的直觉是这样的，你不够聪明理解不了，所以我不告诉你。照我说的做。也许我对，也许我错，但这对你没有多大帮助，你知道，当我离开公司时，你会接手，你会想，我不知道他们为什么那样做。你必须发展自己的直觉和自己的模式识别。但如果我能提出一个解释，比如，我认为人们的心理行为，就像卡尼曼的损失厌恶或前景理论。我认为人们失去某物的价值是得到它的 1.5 倍，所以，我们不应该只是推出功能并测试它，因为移除它要难 1.5 倍，成本更高。然后你有一个理论，它可以在大约一周内传遍公司，现在每个人都知道了。所以，我真的很想强迫我公司里的人尝试，即使我们看到某件事在 A/B 测试中有效，我试着告诉他们，我不想推出它，直到你至少有一个理论解释它为什么有效。即使它非常明显，也有很大的压力要推出它，因为有参与度和变现价值。但如果你，我希望你至少有一个理论。因为随着时间的推移，公司会建立一个理论，一个消费者理论。如果你有一个强大的消费者理论，那么你就能预测那些非常不可能的事情。大卫·多伊奇还说的是，模式识别会迭代地让你更多地在同一条路上，但它永远不会从地心说一路跳到日心说模型。只有解释能带你去量子物理。完全反直觉。没有模式识别能让你想到，也许它同时是波和粒子。

So, this is something I've tried to instill in my org. And I think there's an interesting meta point here, which is people ask me as a product person, you know, how much of product development is magic and how much is science. And I try to be provocative in saying I think it's exactly 100% science and 0% magic. And people get provoked because it kind of implies that there's no skill. So, what I mean with that, I say it to provoke. What I mean is that certainly, people are going to have pattern recognition in this neural network. They've seen a lot of examples. That's what we call seniority. And people have seen a lot of things. They're going to get instinctively to the right conclusion faster than others. So, that is valuable. And I want lots of seniority. So, I don't discard seniority. It brings a lot of value. You can save a lot of time and a lot of mistakes. But the reason you call it magic is because that person can't explain it. It isn't actually magic. It's just science. It's just you are not smart enough to explain yourself. If you could think even further and explain it and come up with an explanation for what you see, the way David Deutsch does, it's so much more valuable for the company. If you have a theory, instead of saying like, no, patching my intuition is this, you're not smart enough to understand it, so I'm not going to tell you. Just do what I say. Maybe I'm right. Maybe I'm wrong, but it's not very helpful for you, you know, when I leave the company, you're going to take over, you're like, I have no idea why they did that. You have to develop your own intuition and your own pattern recognition. But if I can come up with an explanation, which is, you know, I think the psychological behavior of people, you know, it's like Kahneman's loss aversion or prospect theory. I think people value losing something one and a half times the value of getting it, so therefore, we should not just launch feature and test it because it's 1.5x harder and more expensive to remove it. Then you have a theory and it can spread across the company in like a week and now everyone has that. So, I really want to force people in my company to try to even if we see something working in an AB test, I try to tell them I don't want to launch it until you at least have a theory of why it works. Even if it's super clear, there's a lot of pressure to launch it cuz there's like engagement value and monetization. But if you, I want you to at least have a theory. Cuz then over time, the company builds up a theory, a consumer theory. And if you have a strong consumer theory, then you can predict things that were very unlikely. What David Deutsch also says is that pattern recognition will iteratively get you more on the same path, but it's never going to jump all the way from the geocentric to sort of the heliocentric model. Only an explanation can take you to quantum physics. Entirely unintuitive. No pattern recognition gets you to like maybe it's a wave and a particle at the same time.

内部有没有一个很好的解释的例子，它导致了某种从地心说到日心说式的跳跃？比如，实际上是怎么发生的？有什么例子？

What's an example internally of a great explanation that then led to some that the geo to heliocentric type of jump? Like how did how what's an example of how that actually played out?

我认为一个很好的例子，某种程度上是公开的，就是我告诉过你的免费层级。当我们面临我们只有付费移动层级的前景时。你实际上要付费才能在 Spotify 上获得移动性。现在，智能手机在扩展。用户没有电脑。我们需要一个免费层级。那里的竞争对手是 YouTube。他们是前台、按需、带视频。所以，模式识别，显而易见的事情会是说，让我们也那样做。它被证明了。但我们实际做的，特别归功于一个叫查理·赫尔曼的人，是从第一性原理出发推理，说，好吧，让我们看看我们对 Spotify 的使用情况。如果我们把我们的许可限制在同样的事情上，有多少？它只在前台工作。一旦你锁屏，音乐就停止了。

I think a good example that is sort of public is the free tier that I told you about. When we faced the prospect of we only had a paid mobile tier. You actually paid to get mobility on Spotify. Now, smartphones are scaling. Users don't have a computer. We need a free tier. The competition there was YouTube. They were foreground, on demand with video. And so, the pattern recognition, the obvious thing would have been to say like, let's do that. It's proven. But what we did instead and specifically attributed to a person named Charlie Hellman, was to reason around it from sort of first principles and say, okay, let's look at our usage of Spotify. How much of it if we limited our license to the same thing? It only works in the foreground. As soon as you lock the screen, the music stops.

有多少收听是在前台进行的？结果发现，当时大概只有 9% 左右。所以，91% 的使用场景是在后台。我们可能得做点别的东西。用户的需求可能就是后台收听。然后你再看 App Store，有没有办法在 App Store 里免费后台听音乐？最接近的是 Pandora，但那是电台，你没法听你最喜欢的歌。所以，我们就说，我们想要一个消费级产品，让你可以把手机放口袋里，永远免费听你最喜欢的歌。问题在于，这几乎是一个高级用例。如果我们直接推出这个，它会蚕食我们的高级订阅。那我们该怎么办？然后我们看了高级用户的使用情况，发现高级用户大约 50% 的时间在随机播放他们的播放列表。他们使用点播功能，比如搜索、点击、播放特定歌曲，占 50% 的时间，但另外 50% 的时间他们在随机播放播放列表。所以，我们想，如果我们把这个拿出来呢？这看起来即使你有按需点播，你也会自愿随机播放，这是一个很大的用例。我们把它免费提供。那应该意味着没有高级用户会转回免费，因为他们仍然想要那 50% 的点播，但你免费提供了很多价值。所以，我们试着建模一个消费者需求，围绕它推理，想出了这个随机后台层级。这非常非常非常反直觉。连公司内部的人都说，这是个糟糕的主意。但我们有点相信数据，我自己甚至也怀疑过。我说，你看，看看这个点播。我们是不是该试试时间限制之类的？很多人就想让我们试试长期免费试用。但免费试用的问题是，即使你知道，诺基亚，我想诺基亚有音乐，他们试过一年免费试用。但即便如此，用户知道，如果我现在开始投入建立播放列表，一年后我的播放列表投资就会消失。所以他们从不开始投入。所以，我们采用了这个随机层级，这让增长爆发了。直到今天，这是我们与其他服务的差异化。这是唯一一种能让你把手机放口袋里永远免费听音乐的方式。

How much of the listening is in the foreground? Turns out back then, it was like 9% or something. So, you have 91% of the use case being in the background. We probably want to get something else. The user need there is probably background listening. And then you look at the App Store. Is there a way to listen to music for free in the background in the App Store? The closest thing was Pandora. But that was radio. You could not listen to your favorite songs. So, then we said we would like a consumer product where you can listen to your favorite songs with your phone in your pocket forever for free. So, the problem with that is that's almost a premium use case. If we just launch that, it's going to cannibalize our premium tier. So, what do we do? Then we looked at the premium usage and we saw that premium users about 50% of the time, they were shuffling their playlist. They were using on-demand features, you know, searching and clicking and playing specific songs 50% of the time, but they were shuffling playlist 50%. So, then we thought, what if we take this that seems to be something that even when you have on demand, you voluntarily shuffle. It's a big use case. We give that away for free. That should mean that none of the premium users convert back to free cuz they still want their 50% on demand, but you're giving a lot of value away for free. So, we tried to model a consumer need, reason around it, came up with this shuffle background tier. It was very, very, very unintuitive. Even the people inside the company said like, that's a terrible idea. It's a terrible idea. But, we kind of trusted the data and I was even skeptical of it myself. I was like, look, look at this on demand. Shouldn't we try like time caps or a lot of people just want us to try long free trials. But, the problem with the free trial is even if you know, Nokia, I think, Nokia comes with music, they tried a year-long free trial. But even then, the user knew that if I start investing in playlist now, a year from now, my playlist investment is going to disappear. So, they never started investing. So, we went with this shuffle tier and this is what made growth explode. And to this day, that's our differentiation against the other services. It's like it's the only way to listen to music for free forever with your phone in your pocket.

太有意思了。所以，这是一个理论化和解释，而不是模式识别的例子。我很想聊聊与音乐行业关系的演变。这家公司毫无疑问彻底改变了音乐，这非常有趣，也非常酷。你知道，回想早期，你在这里待了很长时间。回想早期，它产生的影响令人惊叹。从投资者的角度来看，许多人一直关注的一件事就是业务的毛利率。就像转移定价问题，无论你做多大，你总是会有这个问题，你知道，拥有知识产权的音乐行业总是会拿走同样的一块肉。谈谈你如何看待这种变化随时间推移。看起来这对他们来说是一段良好的关系，但对 Spotify 来说也是一条非常耐心的道路。也许给我们一些见解，看看这是如何运作的，以及你如何看待它。

Fascinating. So, that's an example of theorizing and explaining rather than pattern recognition. I'd love to talk about the evolution of the relationship with the music industry. It's a company that unquestionably has wholesale changed music, which is so interesting and so cool. You know, thinking back to the early days, you've been here a long time. Thinking back to the early days, it's amazing the impact that it's had. And from an investor's perspective, one of the things that many were always keyed in on is just the gross margin of the business. Just like how much transfer pricing problem are you always going to have that no matter how big you get, you know, the music industry that owns the IP is just going to always take their same cut of the meat. Talk about how you've thought about that change over time. It seems like it's been both a good relationship for them, but also a very patient path for Spotify. Maybe just give us like the insight into how it's worked and how you think about it.

是的，当然。我的意思是，我在 Spotify 成长的过程中，正值瑞典盗版猖獗的时代，那是最糟糕的市场。有一个著名的引述，来自英国唱片公司高管对瑞典唱片公司高管说的，大概在 2000 年代初，瑞典唱片公司高管展示了其中一家瑞典公司的损益表，然后说：“这不是一门生意，这是一种爱好。”当时就是这么糟糕。而这实际上就是 Spotify 能出现的原因，因为音乐行业愿意在瑞典冒险。我想把很多功劳归于音乐行业。他们对 Spotify 冒了很多风险。Spotify 冒了巨大的风险，巨大的资本风险。我们做了很多保底，我们承担了很多风险，但当然他们也冒了很多风险。所以，你知道，我认为音乐行业当然值得成功。Spotify 也是。自从我加入，大概在 2012 年左右，我开始说，我的团队，研发团队和整个 Spotify，我们是音乐行业的研发部门。起初人们说：“你什么意思？”我说：“你看，这是一个整个行业都没有研发部门的行业。”就像，你知道，手机有研发部门，叫苹果或谷歌。其他行业都有很多研发。但音乐行业没有研发投入。我认为事实证明确实如此。如果你看发展轨迹，今年是 Spotify 自成立以来首次盈利。人们说，你知道，有很多讨论说 Spotify 是否分享了足够的收入。我们分享大约 70%。但事实是，另外 30% 我们并没有保留。我们把所有这些都投资回了音乐行业，甚至更多。所以我们亏损了 15 年。我们只是投资、投资、再投资。所以非常有耐心。而与此同时，实际上音乐行业一直在盈利。Spotify 一直在亏损。所以，我认为可以公平地说，我们确实是音乐行业的研发部门。我们投资了，我们亏损了 15 年，而音乐行业一直在获利。现在，这不可能永远持续。我们需要盈利。我们必须盈利。否则我们无法成为音乐行业的研发部门，除非我们能拥有最好的机器学习工程师、最好的产品人员、开发者等等。要做到这一点，你需要盈利。事实证明这些人很贵，因为他们很抢手。所以我们只是，你知道，我们是一家非常非常有耐心的长期公司，我们投资了很长时间。但只是时间问题。大约两年前，我们决定现在是时候盈利了，有点像是掌握自己的命运。在能够投资自己方面。所以，是的，我们盈利了，但我们实际上几乎把所有这些都投资回更多的人、更多的产品、更多的人工智能。所以我们仍然，现在我们有了自己的投资工具，而不是必须最初向私人投资者或市场要更多钱。所以这就是我的看法。真的就是音乐行业的研发部门。我认为我们做得很好。今年我们支付了超过 100 亿。这比大约 10 年前的 10 亿有所上升。就像稳步增长。音乐行业比以往任何时候都大。

Yeah, for sure. I mean, I grew up as Spotify grew up in the era of piracy in Sweden, which was the worst market. And there's this famous quote from a UK label exec to a Swedish label exec, you know, around early 2000, where the Swedish label exec showed a P&L of one of these Swedish companies and said, "That's not a business. That's a hobby." That's how broken it was. And that's actually why Spotify could happen because the music industry was prepared to take risk in Sweden. And I want to give a lot of credit to the music industry. They took a lot of risk with Spotify. Spotify took an enormous amount of risk, enormous amount of capital risk. We MG'd a lot. We ate a lot of the risk, but certainly they took a lot of risk. So, you know, I think the music industry certainly deserves the success. As does Spotify. I've called my team since, you know, I joined in 2000 somewhere around 2012 or something, I started saying that, you know, my team, the R&D team and all of Spotify, we are the R&D department of the music industry. And first people were like, "What do you mean?" And I'm like, "Well, look at it. It's an entire industry that doesn't have an R&D department." Like, you know, mobile phones has an R&D department. It's called, you know, Apple or Google. Everyone else has a lot of R&D. But there's no R&D spend in the music industry. And I think that's turned out to be true. And if you look at the trajectory, this year is sort of the first year of profitability for Spotify since its founding. People say that, you know, there's a lot of talk about is Spotify sharing enough of the revenue. You know, we share about 70%. But the truth is the other 30% we haven't kept. We've invested all of that in the music industry and then more. So we were unprofitable for 15 years. We just invested, invested, invested. So a ton of patience. And at the same time actually the music industry has been profitable. Spotify has been unprofitable. So we've taken, I think it's fair to say we are literally the R&D department of the music industry. We invested and we're, you know, had losses for 15 years and the music industry has been gaining profit. Now, that is not sustainable forever. We need to get profitable. We needed to get profitable. So we can't be the R&D department of the music industry unless we can, you know, have the best machine learning engineers, the best product people, developers, etc. For that, you need to be profitable. It turns out these people are expensive because they're sought after. So we just, you know, we are very, very patient and long-term company and we invested for a long time. But it was just time. About 2 years ago we decided now it's time for us to become profitable to sort of take control of our own fate. In terms of being able to invest in ourselves. So yes, we're profitable, but we're actually investing almost all of that back into more people, more product, more AI. So we're still, now we just have our own investment vehicle instead of having to ask private investors initially or the street for more money. So that's how I think about it. Really as the R&D department of the music industry. And I think we've done a good job. This year we paid out over 10 billion. And that's, you know, up from 1 billion I think in, you know, almost 10 years ago. It's like just steadily increased. The music industry is bigger than it ever was.

人们还在谈论 CD 时代的鼎盛时期。事实是，音乐产业比那时更大了。所以这是有史以来最好的时期。比以往任何时候都好，钱也比以往任何时候都多。蛋糕既更大又更高，但也在被切分。但那是因为更多人尝试了。如果我们说“不，2020 年之前的创作者是好的，但 2020 年之后就不该有人尝试了”，那感觉非常不对。新创作者应该能尝试做音乐。这就是现状。

People still talk about the heyday of the CD era. The truth is the music business is bigger than it was back then. So this is the best it's ever been. It is better than ever. More money than ever. The pie is both bigger and higher, but it's also getting sliced up. But that's because more people take a shot. And it feels very wrong for us to say like, "No, the creators up until 2020, they were good, but no one should be able to try after 2020." New creators should be able to try to do music. So that's the dynamic.

我认为思考这个问题的一个方式是，人们经常谈论每次播放的支付金额等等。Spotify 应该每次播放分享更多。当其他公司说他们每次播放分享更多时，有两件事在发生。行业不是按播放付费的，而是按订阅者付费。但我们的用户参与度是竞争对手的两倍多。所以如果你拿同样的 10 美元，你在 Spotify 上听的时间是两倍，那么每次播放的金额就是一半。所以那些每次播放金额更高的公司，是因为他们的产品更差。我们从唱片公司那里了解到，我们的参与度是竞争对手的两倍，流失率只有一半。所以这有点像一种诅咒，在每次播放模式下，我们的产品越好，每次播放的金额看起来就越低。但我们看的是总金额，我们在那方面领先所有人。我们占据了这些支付的绝大部分。所以我认为如果从整体来看，这个模式是有效的。我们进行了大量投资，现在行业获得了巨大回报。Spotify 现在也盈利了。

And I think a way to think about this is people talk about the per-stream payouts and so forth a lot. And Spotify should share more per stream. There are two things that are happening when other companies say that they share more per stream. The industry doesn't pay per stream. They pay per subscriber. But we have more than twice the engagement of our competitive services. So if you take the same $10 and you listen twice as much as Spotify, the per stream is half. So these are the companies that have higher per stream because they have a worse product. We've learned from the labels that we have twice the engagement and half the churn of competing services. So that's sort of a curse where the per-stream model just the better we are as a product, the lower the per stream is going to look. But we're looking at the aggregate number and we're leading everyone else there. We're the vast majority of these payouts. So I think if you look overall, the model is working. We took a lot of investments and now the industry is getting a huge return. And Spotify also is profitable now.

而做大这个蛋糕的方法是，你知道，现在我们接近 3 亿付费订阅者，接近 7 亿月活跃用户。我认为世界上大约有 5 亿付费订阅者。所以我们占了其中近 3 亿。但相对于世界人口，那只是 5 亿。如果你看像瑞典这样的市场，平均而言，你可以从公开数据看到转化率大约 40%。但如果你看成熟市场，我不会给你确切数字，但要高得多。而新兴市场则较低。所以平均是 40%。但那不是全球平均值，而是低转化和高转化市场的混合。到目前为止，在我们的历史上，随着时间的推移，一切都开始变得越来越像瑞典。所以解决方案就是更快地扩展。更好的免费层级，让更多人加入并转化为高级用户。我们认为应该有数十亿人为音乐付费。那才是真正做大蛋糕的方法。

And the way to grow this pie is to, you know, now we are closing in on 300 million paid subscribers, closing in on 700 million MAUs. There's about 500 million paid subscribers I think in the world. And so we're almost 300 of those. But that's like half a billion out of the world's population. If you look at markets like Sweden, you know, on average you can just look at the public numbers convert about 40%. But if you look at the mature markets, I won't give you the exact number, but it is much higher. And if you look at the emerging, it's lower. So the average is 40. But that's not the average across the world. That's a blend of low and high converting. And so far throughout our history, everything starts to look more and more like Sweden the more time passes. So the solution to this is just to scale it faster. Better free tier that gets more people on that converts to premium. We think there should be billions of people paying for music. And that's how you make the pie truly bigger.

收入分成实际上是一个转移注意力的东西。所以假设我们今天分享大约 70%，或者为了简单起见说 2/3。即使我们是慈善机构，支付 100%，那也只是你今天得到的 1.5 倍。所以如果你认为每次播放 X 美分太少，即使我们是慈善机构，也只是 1.5 倍。解决方案不是收入分成。我们放弃了绝大部分。解决方案是快速扩大付费听音乐的人数。如果你只看数字，你只需要继续前进，它会达到数十亿付费用户，然后是几十亿。那样音乐产业就会非常庞大。我只是觉得音乐产业被低估了。最终它会比看起来大得多。

The rev share is actually a red herring. So let's say that we share, you know, 70% today, ish. Or let's say 2/3 to make it easier. Even if we were a charity and we paid out 100%, that would only be, you know, 1.5x what you get today. So if you think like, you know, X pennies per stream is too little, even if we were a charity, it would be 1.5. The solution is not the rev share. We're giving away the vast majority. The solution is to quickly scale the amount of people paying for music. And if you just look at the numbers, you just have to keep going and it's going to get to billions of users paying. And then several billions. Then the music industry is absolutely massive. I just think the music industry is like undervalued. Terminally it's going to be much bigger than it looks.

我很好奇你怎么看待播客世界。这是我们此刻正在做的事情。我已经做了很长时间。现在似乎我们进入了一个有趣的新时代，当我刚开始做的时候，我记得它相当，我会称之为低地位。比如，2016 年我告诉人们这件事时，他们要么不知道那是什么，要么觉得有点傻。而现在，尤其是在美国，随着选举期间发生的事情以及播客在选举中的重要性，似乎它已经达到了一个临界点，基本上任何可能有理由开播客的人现在都有了或正在推出一个。企业营销策略是开播客，沟通策略是上播客。所以它的重要性和可见性真的爆炸了。Spotify 在其中扮演了什么角色，将来会扮演什么角色？还有，你怎么看待播客及其重要性？

I'm curious how you're thinking about the podcasting world. This is something that we're sitting here doing right now. I've been doing for a long time. Now it seems we've entered this interesting new era where when I started doing this, I remember it was quite I would call it like low status. Like, when I told people about it in 2016, they either didn't know what it was or thought it was kind of silly. And now, especially in the US with what happened around the election and the importance of podcasts in the election, it seems as though it has hit some tipping point where basically anybody that might make sense to have a podcast now has one or is launching one. And it's the corporate marketing strategy is to get a podcast. And it's the communication strategy is to go on them. So it's really exploded in importance and visibility. What role has and will Spotify play in all this? And just like what do you think about podcasting and its importance?

所以我们进入播客的原因，关于 Spotify 和 Daniel 以及我的最亲密伙伴 Alex，我非常珍视的一点是，作为一家公司，我们有很多可以走的路。我认为你的商业模式在某种程度上引导你。如果你主要是广告商业模式，你会被引导向任何额外的参与度。幸运的是，我们主要是基于订阅的商业模式。所以我们更关注留存。如果你想继续为我们付费，你每个月都会用钱包投票。所以我们不会不惜一切代价追求参与度。

So the reason we went into podcasting, one thing that I'm very precious about when it comes to Spotify and Daniel and the other co-person, Alex, who's my closest partner, is that there are many ways we could go as a company. And I think your business model to some extent steers you. If you're an advertising business model mostly, you're going to be steered towards any additional engagement. Fortunately for us, we're mostly a subscription-based business model. So we focus more on retention. And you're going to vote with your wallet every month if you want to keep paying for us. So we're not as steered towards engagement at any cost.

所以在 Spotify 待了很长时间，当这件事发生时，让我对 Spotify 感觉非常好的一点是，当人们使用它，在 Spotify 上失去一个小时时，他们感觉非常好。如果你在音乐上失去一个小时，你出来时会觉得那是一个好小时。我强烈推动公司做播客的原因之一是，我自己在使用它，我们很多开发人员也在使用。我看到它每年在黑客周被黑客进产品里。就像人们希望它在那里。我们就说，我们的开发人员就像是世界的一个小样本。如果他们是一个好的样本呢？所以那是一个推动。我们看到人们在内部使用它并破解它。但让我们决定做它的原因是这种形式。世界上的一切都变得越来越短。人们都是碎片化的。注意力持续时间在下降。而有一种相反的力量，那就是长篇讨论，深入的。人们用完整的句子说话，你知道，关于量子物理或其他什么。那感觉对世界非常重要和有益。

So having been at Spotify, you know, for a long time when this happened, one of the things that made me feel very good about Spotify was that when people used it when they lost an hour on Spotify, they felt very good about it. If you lost an hour on music, you come out feeling that was a good hour. One of the reasons I really pushed I pushed quite hard for podcast in the company was that I was using it myself and a lot of our developers were using it. And I saw it being hacked into the product at hack week every year. It's like people wanted them there. And we just said like our developers is like a small sample of the world. What if they're a good sample of the world? So that was one push. We saw people using it internally and hacking it. But what made us decide on it was that it was this format. Everything in the world was getting more and more short form. People were bite-sized. And attention spans were going down. And there was this counterforce, which was long-form discussions, deep. People spoke in full sentences, you know, about quantum physics or whatever. And that just felt like something very important and good for the world.

所以我们研究了它。我们看到它似乎从一个小基数开始增长。我们看到最大的竞争对手有点在方向盘上睡着了。我们基本上采用了彼得·蒂尔的想法，你知道，早期进入小市场并依靠有机增长，比试图在成熟市场占据一小部分份额更好。在成熟市场获得 1% 看起来风险更小，但人们忽略的是，成熟市场的获取成本是巨大的。对吧？而新市场的获取成本通常很小。

And so we looked at it. We saw that it seemed to be growing from a small base. We saw the biggest competitors sort of being asleep at the wheel. And we did basically the Peter Thiel idea of, you know, it's better to go after small markets early and better on organic growth than to try to take a small share of a mature market. It looks more like less risk in the mature market to get like 1% but the thing people miss is the cost of acquisition in a mature market is just massive. Right? Where as the cost of acquisition in a new market is usually small.

所以我们决定去做，因为我们觉得这与音乐是一致的。比如，如果你在一个深度播客上花了一小时，你出来时会觉得自己学到了东西。这也是我们进入有声书领域的原因，因为它与之一致。我们想成为这种有营养的服务。对此有两个试金石。一个是，如果你在 Spotify 上花了一小时，你出来时的感觉，与你在浴室里刷手机浪费一小时的感觉相比如何？在一种情况下，你觉得自己吃了很多糖果，虽然有很多能量，但那是坏卡路里。而在 Spotify 的情况下，你觉得自己学到了东西。我们有的另一个试金石是，看到很多父母限制孩子的屏幕时间，说去用 Spotify 吧。这表达了他们对它的感受，也表达了我们的感受。所以这是我们进入播客的原因之一，部分是出于哲学考虑。但我们也看到了一个即将增长的小市场的市场机会。我们看到了早期采用者的需求，你知道，他们试图挤进来，然后我们押注于利用我们自己的分发渠道，将其与音乐结合，说市场现在就这么大，但如果我们能把播客暴露给听音乐的人呢？我们能扩大市场吗？这就是我们下的赌注，事实上有声书也类似。在美国，有声书是一种非常小众的行为。我不知道，大概有 1000 万到 1100 万人按单付费购买，而问题在于商业模式限制了它。当你按单付费时，你不会去探索新书，你知道，每本书 15 美元的成本。就像在音乐中，当你为每首歌付 0.99 美元时，你不会去为你的睡眠配乐。每 3 分钟 0.99 美元太贵了。但如果我们有一种无边际成本的模式，让你可以随意探索呢？有声书是否比看起来大得多？是不是商业模式错了？所以再次，我们看到一个看起来相当小的市场，但你在北欧可以看到，那里有访问模式，它正变得非常主流。所以我押注于市场，但这是同样的哲学讨论。比如，这是好卡路里吗？有营养吗？这符合 Spotify 的使命吗？即成为你去那里让自己感觉良好而不是感觉糟糕的地方。

So we decided to go for it because we thought it was something that was in line with music. Like if you lose an hour in a deep podcast, you come out feeling like you learned something. And this is the reason we also went into books. Because it's in line with that. We want to be this nutritious service. There are two litmus tests for this. One is if you lose an hour on Spotify, how do you come out feeling versus if you lose an hour doom scrolling in the bathroom, how do you feel about that? In one case you feel like you ate a lot of candy. Like you had a lot of energy in you but it's bad calories. In the case of Spotify, you feel like you learned something. The other litmus test that we have is seeing a lot of parents restricting screen time for their kids and saying go to Spotify instead. Which means that expresses how they feel about it and how we feel about it. So that was one of the reasons to go into podcast. It was partially philosophical. But we also saw the market opportunity of a small market that was poised to grow. And we saw need in early adopters, you know, trying to hack it in and then we did this bet on leveraging our own distribution combining it with music saying that the market is this big right now but what if we could expose podcasts to people who listen to music? Could we grow the market? So that's the bet we did and the truth is audiobook is something similar. In the US audiobooks was a very niche behavior. I don't know, 10 11 million or something people who paid for them a la carte and the idea was that's a limitation because of the business model. When you pay a la carte you're not going to explore new books at, you know, $15 per book cost. Just as in music, when you pay $0.99 per song, you're not going to soundtrack your sleep. It's too expensive at $0.99 per 3 minutes. But what if we had like a no marginal cost model where you can just explore. Is audiobook much bigger than it looks? Is the business model that is wrong? So again, it was seeing a market that looked pretty small but you can see in the Nordics where you have the access model that it's getting very mainstream. So I bet on the market but it was the same philosophical discussion. Like is this are these good calories? Is this nutritious? Is this in line with Spotify's mission of being like the place where you go when you want to feel good about yourself instead of when you want to feel bad about yourself.

我记得很多年前我第一次和丹尼尔谈话时，沿着西区高速公路散步，他谈到了 Spotify 需要比免费更好的概念。这真是个酷的想法。如果你想想播客，它与音乐非常不同。当有人在 Spotify 上听这个节目时，你不欠我什么。你怎么看待播客，然后显然书籍可能更像音乐，我想听听你的看法。你怎么看待人们听或看 Spotify 上的内容供应，以及它如何影响你的商业模式和捆绑包？

I remember when the very first time I ever talked to Daniel walking along the Westside Highway here many years ago, he talked about this notion of Spotify needing to be better than free. And it was kind of a cool idea. If you think about podcasting, it's very different than music. That you don't when someone listens to this show on Spotify, you don't owe me anything. How do you think about the way that podcasting and then obviously books is a little bit maybe more like music and I'd like to hear how you think about it. How do you think about if that's the supply of the stuff that people are listening to on Spotify or watching on Spotify, the ways in which that affects your business model and the bundle?

我们开始之前不知道播客和后来的有声书是否会对其他媒体类型产生蚕食效应。但事实证明不会。所以最简单的思考方式是，Spotify 现在是一个捆绑包。你支付一定价格，或者你在这里基于广告，然后你获得一堆价值。我们的工作是努力增加你获得的价值，让你更重视它。然后随着时间的推移，也许我们可以通过提价来获取部分价值。我们提价了几次，这也是我们现在盈利的部分原因，但那是因为我们有如此多的用户价值盈余。那是因为我们不断堆叠价值。价值有两件事。价值是像个性化这样的功能，你知道，只是一个非常好的产品。但另一个价值是不同类型的媒体。所以我们看到的是，一个使用音乐的用户，有一定的消费量，当你添加播客时，它只是更多。它不是固定的，看起来像是一场无限游戏。至少目前是这样。我们还没有用完背景时间。然后当你添加有声书，它只是更多的留存、更多的时间和更多的支付意愿。所以这就是我们如何将其视为商业模式。然后在后端，它们有非常不同的商业模式。我认为我们可能是世界上后端最复杂的公司之一，因为你知道，音乐是一种基于池的版税模式。播客，如你所知，主要是基于广告。但现在我们也有这个 Spotify 合作伙伴计划，如果你付费，在高级层中没有 Spotify 广告，所以你得到更多不中断的内容。所以这是另一种商业模式，是高级捆绑包的一部分，然后你有有声书，你知道，出版业以第三种非常不同的方式运作。我们在高级层中也包含一定的时间，如果你超过它，就需要充值。关于 Spotify 的复杂之处，我认为没有被充分认识到的是，在前端它是一个应用，一个消费者。你只是在它们之间切换。但你在用户界面中点击的位置会触发不同的商业模式等等，含义非常不同。所以要对一家公司进行财务建模实际上相当困难。我们必须预测你的用户行为。你点击的位置很重要，你知道，我们有个性化，它在成本等方面有不同的影响。所以我们不得不建立一个系统。我们称之为 Spotify 机器，这就是为什么我说我有一个体验组织，这个体验组织的任务是确保所有这些复杂性、所有这些理论上可能被设置为相互竞争以修复其损益的团队，永远不会影响到用户。有一个人是消费者体验的负责人。那个人的工作是确保当你在移动端、桌面端、汽车和扬声器之间切换时，一切都有意义。它就像一道闸门，你知道，防止组织阻碍用户，在背后保护用户。但在个性化方面也是如此。我有一个个性化组织，因为你有同样的激励，你知道，编排音乐与播客与书籍。你知道，每个人都想夺取市场份额等等。所以这是同样的问题。我们必须为用户优化，并某种程度上保护用户免受团队和商业模式的内部激励的影响。所以这使得 Spotify 成为一家相当独特的公司。我们前端是一个东西，后端是许多不同的东西，有不同的产品。

We didn't know before we started if podcasting and later audiobooks would be cannibalistic to the other media types or not. But it turns out it's not. So the easiest model to think about it is Spotify is a bundle now. You pay some price or you have the advertising based here and you get a bunch of value. And our job is to try to increase the value you get so you value it more. And then over time maybe we can capture some of that value by price raising. We price raised a few times which is part of why we're profitable now but that's because we had such user surplus in value. That's because we kept just stacking value. And value are two things. Value are features like personalization and you know, just a really good product. But the other value is different types of media. So what we see is that have a user that uses music that has a certain amount of consumption when you add podcast, it's just more. It's not a fixed it's an infinite game looks like. At least for now. We haven't run out of time in the background yet. Then when you add audiobooks, it's just more retention more time spent and more willingness to pay. So that's how we think about it as a business model. Then on the back end, they have very different business model. I think we may be one of the most complex companies in the world on the back end because we're you know, music is a sort of a pool based royalty model. Podcast as you know is advertising based largely. But now we also have this Spotify partner program where you don't have Spotify ads in the premium tier if you're paying so you get more uninterrupted. So that's another business model which is part of the premium bundle and then you have audiobooks which you know, the publishing industry works in a third way very different. Where we also have certain amount of time included in the premium tier and then a top up if you run over that. One of the really complicated things about Spotify I don't think is appreciated is on the front end it's one app, one consumer. You just go between them. But there are very different implications of where you click in that UI in terms of triggering different business models and so forth. So to model a company financially is actually quite hard. We have to predict your user behavior. Where you click matters and you know we have the personalization that has different impacts in terms of cost and so forth. So we've had to build a system. We call it the Spotify machine and that's why I said I have one experience organization and the job of this experience organization is to make sure that all of this complexity, all of these teams who theoretically could be set up to compete with each other to fix their P&L that never ships to the user. There's one person who is the responsible person for the consumer experience. And that person's job is to make sure that as you go between mobile and desktop and car and speakers, the thing makes sense. It's like the gatekeeper against, you know, the org holding them back from the user behind them protecting the user. But it's also the same in personalization. I have a personalization organization because you have the same incentives of, you know, programming music versus podcast versus books. You know, everyone wants to take market share and so forth. So it's the same problem. We have to optimize for the user and sort of protect the user from the internal incentives of teams and business models. So that's that makes Spotify a pretty unique company. We're like one thing on the front end and we're many different things on the back end with different products.

如果设想一下，比如说五年后，你尽可能大胆地梦想 Spotify 会从今天的位置走向何方，请为我们描绘那幅图景。

If you think about, let's say, 5 years from now and you dream as big as you can possibly dream for where Spotify might go from where it is today to where it will be in 5 years. Paint us that picture.

当然，我希望我们已经突破了十亿用户的大关，但作为订阅服务，我希望我们正成为全球最大的媒体订阅之一，并不断为其增加更多价值。所以希望音乐比以往任何时候都更繁荣。我希望有声书能像在斯堪的纳维亚那样成为主流现象，在那里，听有声书的人数几乎与听音乐的人数相当。我认为这对世界来说是一件大好事。但我也希望我们增加了更多这样的垂直领域。我不能透露具体是什么，但你知道，就是订阅模式、捆绑模式，这些我们之前没怎么谈过的。

Certainly I hope we've cracked, you know, the billion user line but you know, as a subscription, I hope we're becoming one of the biggest media subscriptions in the world and we add more and more value to that. So hopefully music is bigger than it ever was. I'm hoping that audiobooks is a mainstream phenomenon as it is in Scandinavia where, you know, it's almost as many people that listen to music listen to audiobooks. I think that would be a net good for the world. But I also hope we've added a few more of these verticals. I can't say what they are but they're, you know, the subscription model, the bundling model that we didn't talk so much about.

一个好的捆绑产品的关键是什么？而且我也很好奇，你说你尝试过只在平台上提供的独家内容，现在少了。是什么驱动了这样的决定？对于那些可能想在别处创建捆绑产品的人，你怎么看？

What's the key to a good bundle? And I'm also curious, you know, you said you experimented with exclusive content that was only available on platform and less of that now. You know, what drives a decision like that and how do you think about other people that might want to create a bundle somewhere else?

当我们审视播客时，你看看像 Netflix 这样的公司，它有漂亮的商业模式，而且执行也极其出色。在我们看来，那可能很有趣。你知道吗？我认为当你是一家处理大众化内容的产品公司时，你总是会羡慕：如果我们能通过内容实现差异化呢？那样生活就会变得超级简单。你总是觉得别人在做的事情很容易，而自己的事情很难，但实际上做别人的事情通常也非常难。所以我们尝试在播客中采用独家模式来差异化服务，但我认为这最终是一个糟糕的赌注，因为播客的宏观趋势是制作成本非常低。乔·罗根最初是在他的拖车里录制的。制作成本很低。在此基础上搞独家，某种程度上是适得其反。整个要点更像 YouTube，内容非常便宜，所以你可以获得大量内容。你不必非得选对。一旦你进入独家游戏，你就必须选对。你必须成为内容挑选者。这是一项非常难的技能，Netflix 做得非常好。对吧？但我们有这个机遇。我们不必挑选内容。我们只需获取所有内容，用机器学习为你提供你想要的，为我提供我想要的，而且没有制作古装剧那样的资本密集型需求。这是我们做出的一个糟糕的战略决策。我们还大量押注名人。他们是名人，但并不总是优秀的播客主持人。而那些真正优秀的播客主持人，是通过这种有机体系成长起来的。所以，你知道，有两种方法可以永远正确。一种是永远猜对。另一种是犯错时改变主意。所以我们决定改变主意，说这看起来像是联合供稿的时代。创作者实际上希望无处不在。他们创作视频或音乐，希望到处都有。好吧，让我们拥抱这一点。我们实际上是一个平台。在音乐方面，我们一直是一个平台。我们从未玩过独家。我们说我们要最大的目录。书籍我们也在做最大的目录。让我们在播客中也拥抱这一点。于是我们调整了战略。这为我们节省了大量成本。这也是我们做得好的部分原因，同时也大大改善了我们的目录，现在我们的播客观看量处于非常好的轨道上。所以，这是一个糟糕战略的例子，我认为重要的是承认错误并改变主意。真正的代价是你试图为自己过去的决定辩护的时候。

When we looked at podcast, you know, you look at something like Netflix and it's this beautiful business model and insanely good execution as well on top of that. And it looked to us like that could be interesting. You know what? I think when you're a product company that works with commodity content, you always had this envy of like what if we could differentiate through content? Then life is going to be super easy. You always think the other thing that someone else is doing is easy and your thing is hard and it's usually like very hard to do the other thing. So we tried exclusivity in podcast as a way to differentiate the service but I think it was ultimately a bad bet because the macro trend for the whole thing with podcast was that the production cost was so low. Joe Rogan was initially sitting in his trailer. Like the production cost was low. And then go in and do exclusivities on top of that is kind of counter purpose in a way. The whole point is more like YouTube in that this is very cheap content so you can get a lot of it. You don't have to be right. As soon as you go into exclusivity game, you have to be right. You got to be a content picker. And that's a very hard skill that Netflix does extremely well. Right? But we had this opportunity. We didn't have to pick content. We just get all of it and use machine learning to serve you what you wanted and me what I wanted and there wasn't this capital intensive need there that there is in producing like costume dramas. This is a bad strategic decision that we did. We also betted a lot on celebrities. And they are celebrities but they're not always good podcast hosts. And the podcast hosts that were really good, they grew up through this organic system. So, you know, there are two ways to always be right. One is to always guess right. The other is to just change your mind whenever you're wrong. So, we decided to change our mind and say this looks like the age of syndication. Creators actually want to be everywhere. They create a video or they create music. They want to be everywhere. Okay, let's embrace that. We're actually a platform. In music we were always a platform. We never played with exclusivity. We said we want the maximum catalog. Books we're doing maximum catalog. Let's just embrace that in podcast as well. So, we pivoted strategy. And that saved us a lot of cost. Which is part of what we're doing well and it's also improved our catalog greatly and now we're on a really good trajectory with our podcast viewing. So, it was an example of a bad strategy and I think the important thing is to admit it and change your mind. The real cost is when you try to defend your past decisions.

在你的生活中，除了 Spotify 之外，你做什么最能让你做好准备或让你有能力在 Spotify 做到最好？

What things do you do outside of Spotify in your life that most prepare you or make you capable to do the best job that you can in Spotify?

我认为世界变化非常快，所以我很多时间都在努力跟上正在发生的事情。所以，我花了很多时间——最近我和家人在里斯本度假，我花了很多时间陪他们游览里斯本，那是一座美丽的城市。然后我请求他们给我放一天假，不工作也不陪家人，就为了放纵自己。这次是，你知道，回去尝试写点代码，用所有这些新工具，保持对最新动态的了解。有时是阅读，比如物理或数学之类的东西。这是跟上时代和激发自己思维的结合，跟上时代很难，因为它变化太快，但也要在精神上刺激自己。我在任何时候都必须有让我兴奋的事情。可以是新事物，比如 AI 及其意义。但也可以是我不知道的古老事物，比如学习更多物理或数学知识。所以，有一段时间，是的，我读了很多哲学，因为它就是一个有趣的领域。你知道，你会思考智能、意识等所有大问题，你可能花十年时间阅读所有这些。现在我觉得有点精疲力竭了。我认为我喜欢当你开始阅读时，你会想，‘是的，我要解决这个问题。’然后结果我没解决。人们一直在试图解决意识问题，但它如此深奥有趣。它让我在很长一段时间里对生活保持兴奋。我在学校从来不是数学很好的人。我还可以，但不出色，但我发现自己随着年龄增长对数学越来越兴奋。

I think the world is moving very fast so a lot of my time is just trying to keep up with what is happening. So, I spent a good deal of time—I was on a vacation in Lisbon with my family recently and I spent a lot of time with them seeing Lisbon which is a beautiful city. Then I asked them for like one day off from work and off from the family to just indulge myself. This time it was, you know, going back to trying to code a bit, use all these new tools, stay on top of what's happening. Sometimes it's reading, you know, like physics or math or something. It's a combination of keeping up with what is happening which is hard because it moves so fast but also stimulate myself mentally. I have to have something that I'm excited about at any point in time. And it can be new things like AI and what it would mean. But it can be age-old things that I just didn't know like learning more about physics or math or something. So, for a while, yeah, I've read a lot of philosophy for a while because it's just an interesting area. You know, you think through all the big questions of intelligence and consciousness and all of those things and you can spend like 10 years there just reading all of that. And now I feel like tapped out a little bit. I think I like when you start reading you're like, 'Yeah, I'm going to crack this.' And then turns out I didn't crack it. People have been trying to crack, you know, consciousness for a while but it's so deeply interesting. It kept me like excited about life for a very long time. I was never like a big math person in school. I was okay but not great but I found myself getting very excited about math the older I got.

所以，当你开始读一点哲学时，你会接触到像哥德尔不完备定理和构造性数学这类东西，它们与工作关系不大，但能让你保持活力。而且我实际上和我的产品人员和工程师聊过，结果发现他们中很多人都对这些东西深感兴趣。所以，我身边有非常有趣的话题可以和大家聊。这就是我保持活力的方式。然后我还会做运动。我和我的孩子们一起练巴西柔术，这非常有收获。

So, as you start reading a bit of philosophy, you get into things like Gödel's incompleteness theorem and constructive mathematics, which are loosely related to work but keep you energized. And I actually talked to my product people and engineers, and it turns out that a lot of them are deeply interested in these things. So, I have something very interesting to talk to people around me about. That's how I keep energized. And then I do sports. I do Brazilian Jiu-Jitsu with my kids, which is very rewarding.

这告诉了你什么？

What does that tell you?

谦逊。你进来时以为自己能做点什么，结果被一个只有你一半体型的人彻底碾压，而对方连汗都不出。然后你会想：“好吧，技术很重要。”是技术，是杠杆。

Humbleness. You come in and you think you can do something, and you get absolutely smashed by someone half your size, and they're not even sweating. And you're like, 'Okay, technique matters.' It's technique. It's leverage.

是啊，巴西柔术的美妙之处在于，腰带体系是真实存在的。

Yeah, the beautiful thing about Brazilian Jiu-Jitsu is that the belt thing is real.

这背后有一个很长的故事，但核心是，一个日本人把柔术带到了一个巴西家庭，那里有一群经常打架的兄弟。其中一个兄弟相对于其他人来说发育不良，他不够强壮，所以打不过他的兄弟们。于是，他开始学习日本柔术，琢磨如何利用物理原理，也就是杠杆。慢慢地，他开始打败他所有的兄弟，这就演变成了巴西柔术。所以，这就像他必须解决一个问题，他不能使用力量。然后，这个家族举办了所有这些比赛来真正检验它。从进化产品的角度来看，他们说：“好，任何人都可以来这里。空手道、踢拳，尽管来试。开放。”他们在地下室里打斗，就这样不断进化这项运动，证明它是真实的。我们很多武术都像魔法和秘密，从不检验技能。所以，巴西柔术在实践中非常有效。

There's a long story behind it, but the net is that a Japanese person brought Jiu-Jitsu to a Brazilian family, and there were a bunch of brothers who fought a lot. There was one brother who was just underdeveloped versus the others. He was not very strong, so he could not beat his brothers. So, he started taking Japanese Jiu-Jitsu and figuring out how he could use physics, just leverage. And slowly, he started beating all his brothers, and that became Brazilian Jiu-Jitsu. So, it was literally like he had to solve the problem. He could not use power. And then, this family put up all these competitions to really test it. From an evolutionary product point of view, they said, 'Okay, anyone come here. Karate, kickboxing, just try it. Open.' They fought in these basements, just evolving the sport, proving that it was real. A lot of our martial arts are like magic and secret, and they never test their skills. So, it works very well in practice.

我喜欢的另一点是，我练过很多其他武术，比如拳击和泰拳。那些作为锻炼很好，但用于自我保护就不太好。你不能一拳打在别人脸上，你会被起诉的。这种被称为“温柔的运动”的武术的美妙之处在于，你控制别人，约束他们，并且可以调整暴力程度。这就是为什么警察使用柔术而不是泰拳，因为你可以调节对他人的暴力，并在不伤害他们的情况下控制他们。所以，这就是为什么我认为每个人都应该尝试并练习它。它有助于自律，因为你会变得谦逊。它实际上也很有用，而且你可以用它而不伤害他人。

The other thing I like about it is that I've done a lot of other martial arts, like boxing and Thai boxing. Those things are great as exercise, but for self-protection, they're not very good. You cannot punch someone in the face; you're going to get sued. The beautiful thing about this martial art, which is called the gentle sport, is that you control people. You constrain them, and you can adapt the level of violence. This is why police use Jiu-Jitsu and not Thai boxing, because you can regulate the violence to the other person and control them without hurting them. So, that's why I think everyone should try it and practice it. It's good for self-discipline because you get humble. It's also actually useful, and you can use it without harming other people.

人们可能不知道这一点，因为他们怎么会知道呢？但 Spotify，尤其是你和 Daniel，可能对我以及我多年来如何思考建立我们的业务产生了最大的影响，当然作为公司也是如此。其中很多都归结于那些你看不到的东西。我们今天聊了很多，我特意聊了很多。比如赌注板，隐藏在美丽消费体验背后的复杂性。多年前你是第一个向我描述赌注板概念的人，我们非常有效地运用了它。还有从 Daniel 那里学到的关于如何思考对用户重要的东西的许多教训。我认为 Spotify 不仅是一个令人难以置信的产品，而且也是一个产品反映背后公司的例子。我认为通过听这样的对话来研究它是值得的，因为它为我提高了雄心的标准和卓越的标准，即如何构建一个公司来反映其独特的需求。还有运营者的品格和纪律。所以，和你做这件事非常有趣，非常感谢你多年来的所有教训。

People probably don't know this, because how would they? But Spotify, you and Daniel especially, have been probably the most influential people and certainly the company on me and how I've thought about building our businesses over time. And a lot of that comes back to the stuff that you don't see. We've talked, I've purposely talked about a bunch of it today with you. The bets board, the complexity that's hidden behind a beautiful consumer experience. You were the first person years ago to describe the bets board concept to me, and we've used that very effectively. And so many lessons from Daniel on how to think about what matters to users. And I think Spotify is not only an incredible product, but it's also one where the product is a reflection of the company behind it. And I think it's one worth studying by listening to conversations like this one, because it has raised, for me, the bar of ambition and the standard for excellence of how a company should be constructed to mirror the needs that it has, its unique needs. But also just the character and the discipline of the people running it. So, it's been so fun to do this with you, and thank you so much for all the lessons over the years.

嗯，你知道我最后要问每个人的问题。别人为你做过的最善意的事是什么？

Well, you know the closing question that I have for everyone. What is the kindest thing that anyone's ever done for you?

让我真正在岗位上表现出色的是 Daniel 允许我承担很多风险。实际上，我在 Spotify 搞砸了很多不成功的事情。我从未觉得自己会因此被解雇。他实际上鼓励了这一点，而且我得到了第二次机会。正是这让我有了更高的抱负，而不是因为害怕失败而退缩。所以，我认为是一系列这样的事情，比如被允许搞砸事情，可能对我的职业生涯产生了最大的影响。

The thing that made me really excel in my role was being allowed to take a lot of risk by Daniel. So, I've actually screwed up a bunch of things in Spotify that didn't work. And I never felt that I was going to get fired for it. And he actually encouraged that, and I got like a second chance. And that's what made me have the higher ambition instead of holding back for risk of failure. So, I think it's a series of those things, like being allowed to mess up things, that has probably had the biggest impact on my professional career.

你能举一个你犯过的严重错误的例子吗？以及在这个过程中，他和组织是如何让你感觉到的，从而使你能够重新鼓起勇气再次承担更多风险？

Can you give an example of a bad mistake that you made, and how he and the org made you feel through that process so that you could be re-emboldened to take more risk again?

我对新的用户界面很感兴趣，多年前我带领公司非常艰难地进行了一个当时非常前卫的界面项目。这个想法是 Spotify 直接开始播放内容。你向上滑动进入下一个流派，向左或向右滑动在同一流派内获取其他内容。现在你会说这听起来几乎像 TikTok。这发生在 Musically 之前。但发生了两件事。它非常前卫。它在你没有要求的情况下开始播放内容，所以人们感到不满。但我非常坚持，因为我确信即时性，而且这个想法是你可以通过非常低摩擦的界面找到你想听的内容。这可能是一个不错的主意，但那时机器学习还没成熟。它根本行不通。你就是无法达到目标。我们构建了这个东西，叫做 Moments。这个用户界面。我们在后端使用了编辑，这完全行不通。所以，这个想法远远领先于当时的技术。而且它花了很多钱。我们实际上发布了它。有一段我们展示这个用户界面的视频等等。人们幸运地忘记了它。但它就是不行。我们做了 A/B 测试，看起来还可以，于是我们就发布了。然后我们发现 A/B 测试在线上时有个 bug，实际表现比我们预期的差得多。所以，我们不得不回滚，而我让整个组织进行了这次冒险，在一个竞争非常激烈的行业里让我们损失了大约一年的时间。那是一个被解雇的好机会。但我没有被解雇。Daniel 说：“我理解这些解释。不，我理解。我同意这些想法和观点。错误是什么？”错误在于机器学习还不存在。我们不够好，无法让你通过足够的滑动到达那里。他更像杰夫·贝索斯，你知道，他衡量投入，而不是产出。

I was interested in new user interfaces, and many years ago I took the company very hard on a journey for an interface that at the time was very provocative. The idea was that Spotify just starts playing things. You swipe up to get to the next genre, and you swipe left or right to get other things within the same genre. Now you would say that sounds almost like TikTok. This was before Musically. But two things happened. It was very provocative. It started playing things without you asking, so people were upset. But I pushed pretty hard because I was convinced that immediacy and the idea was that you just sound your way to what you want to hear in a very low friction interface. And it was maybe a decent idea, but it was before machine learning. It just did not work at all. You could just not get there. And we built this. It was called Moments. The UI. We used editors on the back end, which did not work at all. So, the idea was far, far ahead of where the technology was. And it cost a lot of money. We actually announced it. There is a video of us presenting this user interface and so forth. People luckily forgot it. But it just didn't work. We had A/B tested it, and it looked okay, which is what we launched. Then we discovered there was a bug in the A/B test when it was live, and it actually underperformed drastically what we had. So, we had to roll it back, and I'd taken the entire organization on this excursion that lost us like a year or something in a very competitive business. That was a good opportunity to get fired. And I didn't. Daniel was like, 'I understand the explanations. No, I understand. I agreed with the thoughts and the ideas. What was the mistake?' And the mistake was that the machine learning was not there. We were not good enough to get you there in enough swipes. And he was more like Jeff Bezos, you know, he measures the inputs, not the outputs.

比如，如果输入很糟糕，想法模糊又愚蠢，那确实是个问题。但即使有好想法，你也不可能永远正确。所以我听他说过，杰夫·贝索斯有句话，大意是：我评判你，是看你拥有的输入，而不是输出，因为评判输出的问题在于，你可能只是运气好，即使你并不优秀，也能靠运气得到晋升。而如果你看输入，你知道，只要多给机会，如果想法和执行是有结构的，最终你会做对的。所以，只是多给机会而已。这实际上让我承担了更多风险，而不是降低风险。但我有很长一段时间感到非常非常非常受伤。公司内部有些关于某些时刻的笑话。

Like if the inputs are bad, if the ideas are fuzzy and stupid, that's a problem. But you're not going to be always right even with good ideas. So, and I heard him say this, Jeff Bezos quote of like I focus on judging you by the inputs you had, not the outputs, because the problem with judging the outputs is you could just get lucky and get promoted even though you're not very good just by luck. Whereas if you look at the inputs, you know, if you just give more chances, if they're structured ideas ideation and execution, eventually you're going to get right. So, just got more chances. And that made me actually take more risk instead of scaling down on the risk. But I felt very very very burned for a long time. There are jokes internally about moments, you know.

多么有力量的故事和心态，值得我们所有人学习。真是绝妙的收尾故事，Gustav。非常感谢你抽出时间。

What a powerful story and mindset for us all to adopt. Such a great closing story, Gustav. Thanks so much for your time.

谢谢邀请。荣幸之至。

Thanks for having me. Pleasure.