代理经济支柱——对话Stripe数据与人工智能负责人Emily Glassberg Sands

大家好，欢迎收听《Litin Space》播客。我是Kernel Labs创始人Alessio，今天和我一起的是《Litin Space》的编辑Swiggs。

Hey, everyone. Welcome to the Litin Space podcast. This is Alessio, founder of Kernel Labs, I'm joined by Swiggs, editor of Litin Space.

你好，你好。我们再次回到演播室，今天请到了Stripe的Emily Sands。欢迎你。

Hello. Hello. We're back in the studio with Emily Sands from Stripe. Welcome.

谢谢。

Thank you.

那么Emily，你是Stripe的数据与AI负责人，这个头衔可不小。在实际工作中这具体意味着什么？

So, Emily, you're head of data and AI at Stripe. That's a that's a big title. What does that actually mean in practice?

Stripe正在为互联网构建金融基础设施。我们最初是支付基础设施，现在正帮助企业解决一系列问题——如何处理订阅等周期性支付？如何进行用量计费、收入确认、税务处理、资金流动、稳定币接收等等？我们处理的交易量约占全球GDP的1.3%。

So Stripe is building financial infrastructure for the Internet. We started out as payments infrastructure, and now we are helping businesses solve a whole range of problems. How do they accept recurring payments like subscriptions? How do they do usage billing, revenue recognition, tax, money movements, accept stablecoins, and more? And when you think about what we're looking at, we're looking at on the order of 1.3% of global GDP.

每年约有1.4万亿美元通过Stripe处理。这为我们提供了独特机会：既能利用这些数据理解经济动态和用户需求，又能将其反馈到产品中优化支付体验——比如减少欺诈、提高授权通过率、改善客户界面、优化结账流程等。我们的数据与AI团队核心任务就是帮助Stripe高效利用数据，这从基础层就开始着手。

About $1,400,000,000,000 a year is processed on Stripe. And so that obviously creates a very unique opportunity to use that data to understand both what's happening in the economy, what do our users need, but also feed it back into the product to power better payments experiences. So cut down on fraud, drive the right authorization, better customer facing experiences, optimizing the checkout suite, and more. So, anyway, our our data and AI org is just really focused on helping Stripe make effective use of our data. And that starts all the way at the foundation layers.

比如数据平台建设、数据工程实施、机器学习与AI基础设施搭建，直至最终的应用层实现。

Right? Like, what's the data platform? How do we do data engineering? What are our ML infrastructure, AI infrastructure? And then all the way up to the applied layer.

我们还有一个有趣的小团队。实际上规模很小，只有二十多人，但我们称之为实验项目组。它不局限于特定数据领域，前提是实验可以且应该发生在任何地方。但当前世界变化速度之快，常常会涌现出这些跨领域的机会，它们并非自然产生或易于在任何单一产品垂直领域立即抓住。

We also have a fun little group. It's actually quite small. It's just two dozen people, but we call it the experimental project team. And it's not data specific, but the premise is experimentation can and should and does happen everywhere. But there are often these sort of cross stripe opportunities that are being pulled out of us by our users just given the pace at which the world is changing that aren't natural or easy to jump on within any one product vertical today.

因此这些成员都是相当资深、经验丰富的工程师，他们能迅速抓住这些机会，从零到一实现落地。我们的商业代理工作就源于此，稍后可以讨论的代币计费系统也出自这个团队。这只是我们团队中一个有趣且被证明极具杠杆效应的侧面。

And so these are just quite senior, quite seasoned engineers who run at those opportunities and and zero to one them and get them the ground. So our agent to commerce work came out of that. Token billing, which we can talk about in a bit, also came out of that. And and that's just a fun sort of side angle within our group that's that's proven very high leverage.

是的。我很喜欢这个定位——Stripe的使命是为互联网构建金融基础设施，而你们的分支使命是为AI构建经济基础设施。这是个相当雄心勃勃的目标。你在Stripe已经四年了。

Yeah. I I like the framing that, you know, Stripe's mission is build financial infrastructure for the Internet, and your subset of that is build economic infrastructure for AI. And that's a pretty ambitious goal. You've been at Stripe four years.

AI是什么时候成为战略级议题的？因为你们显然早就在使用机器学习做欺诈检测之类的工作了。你们本身就是用户。

At what point did AI become a title level? Because, I mean, you were obviously using machine learning for like fraud detection and everything, like You were a customer. Yeah.

没错。我们基本上是在GPT-3.5问世时开始专门投资AI和LLM领域的。当时我们就意识到需要让每个人都能安全便捷地使用高质量的大语言模型，不仅是日常办公，更要能构建生产级应用。那大概是2023年初或2022年底，我们开始系统规划这件事。

Yeah. We started investing in AI or LLM specific experiences basically when GPT 3.5 hit the scene. We were like, okay, we need everybody to be able to have high quality, safe, easy access to LLMs, not just for their own, you know, day to day work usage, but actually to build, you know, production grade experiences. So that's sort of, what was that, early, like, January 2023 or late twenty twenty two. We we started reasoning about, okay.

这不仅是机器学习基础设施，更是AI基础设施；不仅是ML应用，更是AI应用。但直到过去一年半左右我们才真正明确：

Like, you know, it's not just ML infrastructure. It's also AI infrastructure. It's not just ML applications. It's also AI applications. But then it was really only in the last year and a half or so that we said, hey.

虽然我们早就有transformer等模型，但最近一年半才意识到需要打造自己的领域专用基础模型。实际上，我们可以从单任务ML模型转向更丰富、更密集的支付嵌入模型，从而支撑各种下游应用。这是个渐进过程，整个行业对ML和AI的界定仍有争议。但你说得对，Stripe使用ML已有十多年历史，早期就不只用于众所周知的Radar反欺诈系统。

I mean, we had, like, transformers or whatever before, but only in the last year and a half or so that we're like, hey, we actually need to have our own domain specific foundation model. And actually, we can move from these, you know, single task point solution ML models to, a much richer, denser payments embeddings that can then power the various downstream applications. So I think it was an evolution for us. And I think we could still debate what's ML and what's AI in the industry at large. But you're right that we're more than a decade into using ML at Stripe, you know, way back in the early days for not just radar, which I think people know about.

对吧？比如那些为我们的客户拦截欺诈的机器学习系统，同时还有用于我们内部运营的机器学习。其他支付服务提供商每天最多只能上线十几个用户，而我们每天要上线数千用户，你必须确保这些用户是可支持的、非欺诈的且有信用价值的，因为你正在处理他们的交易。仅这一点就需要机器学习，而且长期以来一直如此。

Right? Like, the the machine learning systems that block fraud for our customers, but also ML internally for our own operations. Like, other payment service provider is onboarding max a dozen users a day. We're onboarding thousands of users a day, and you have to make sure they are supportable and not fraudulent and are creditworthy because you are processing their transactions. And that alone requires machine learning and and long has.

是的。我想说，你知道，领域特定模型最初出现的方式很有趣，就像前基础模型时代那样，然后我们进入了这个基础模型时代。但在Stripe的规模下，我想你们还必须处理如此大量的推理请求，以至于可能又需要对这些模型进行领域专业化。

Yeah. I would say, you know, it's it's kind of interesting how the domain specific models first come out, like pre foundation models, and then we have this foundation model era. But then at the scale of Stripe, I imagine that you also have to just serve in so much volume of inference that then you might have to domain specialize them again.

这有点像屋顶上的铃铛，上面还有微调。我的意思是，我们每分钟要处理大约5万笔交易。所以

There's this kind of, like, bells on the roof. There's fine tunes on top. I mean, we are we see, like, 50,000 transactions a minute. And so

所有这些都要经过你的基础模型吗？

all that goes through through your foundation model?

是的。每一笔交易都要经过。例如，基础模型的一个功能就是检测信用卡测试攻击。你们知道什么是信用卡测试吗？

Yes. Every single transaction. So for example, the foundation model, one of the things that powers is detecting card testing attacks. Do guys know what card testing is?

知道。

Yeah.

好的。那么对于听众来说，信用卡测试者可能是在枚举卡片，也可能是随机猜测卡号。他们找到一张有效的卡后，有时会用于欺诈，但更多时候是将其出售。

Yeah. Okay. So for listeners, it could be like a card tester could be enumerating through cards or they could be random guessing cards. They find a card that works, and then sometimes they use it for fraud. More often, they sell it.

许多传统机器学习模型在检测信用卡测试方面表现不错，但测试者变得很狡猾。他们常用的手段是将测试交易隐藏在大型企业的海量交易中。想象一个大型电商公司，其交易量有多大。测试者可能会在其中混入100笔、200笔金额为几美分的测试交易。传统机器学习根本无法捕捉这种模式。

Lots of traditional machine learning models can do a pretty good job detecting card testing, but card testers have gotten clever. And one of the things they do is that they hide their card testing in the volumes of very large businesses. So if you think about a very large ecommerce company, you can think about how many transactions are there. A card tester might sprinkle, like, 100, two hundred, three, four, five cent transactions in testing. Traditional ML is, like, not gonna catch that.

而基础模型则不同。每笔交易都会生成密集嵌入向量。你能实时看到这些异常集群浮现，立即识别出是信用卡测试并予以拦截。整个过程在支付路径上完成，延迟不到100毫秒。

Then you have a foundation model. Each, you know, charge becomes this, like, dense embedding. You start to see these clusters sort of pop out, and you know in real time that they're card testing and and you can block them. So, yes, it is it is happening on the charge path in less than a 100 milliseconds of latency.

确实。基础模型是否允许在嵌入中使用更多数据？像数字模式、邮编与位置关系这些特征以前也能处理。现在有没有新增的数据维度？

Yeah. Have the foundation model enabled more data to be put in the embedding? I I think, like, you know, things like, you know, number patterns and, like, ZIP code versus, like location. I think those you could do before. What are there any new data points that you get?

我认为有两个重大突破：首先，构建小模型时通常只关注有明确标签的近期数据，可能还需要人工设计特征。但构建基础模型时，考虑到下游多种应用场景，我们会放入数百亿笔交易数据。

Yeah. So so I think there's I think there's two big things like one, you know, when you're when you're building a small model, you're usually like, oh, looking at the data that has reasonable labels. It's like recent history. You probably have some hand engineered features in many cases. If you're actually building an FM and you're imagining there's many downstream use cases, you're putting tens of billions of transactions in it.

我们把支付的所有细节都输入模型，让基础模型自行判断哪些要素重要。更关键的是交易序列——可以把单笔支付看作单词，整个支付流就像语言数据。

You're putting, like, the entirety, like, every detail of the payment in it and and letting the FM reason about what are the components that matter versus not. So it is literally all the things. But I think what's even more interesting is, like, the last k. Like like, what matters is the sequence. What like, you can think of a payment sort of like a word, and so you can think of payments data kind of, like, language data.

重要的不是单个'单词'，而是其上下文关系。但支付数据的难点在于，你无法像听播客中Emily连续说20个词那样明确序列边界。关键序列可能是某个特定零售商从固定IP发起的交易，

And what matters isn't the word. Right? It is the word in relation to the words around it. But what's tricky about payments is you don't see, like, you know, Emily on a podcast saying 20 words and know those are the words. Like, the sequence that could matter could be, you know, this particular retailer charges from this IP.

也可能是周五晚上某张信用卡的所有消费。因此需要选择广泛的相关序列来捕捉最后k个关键要素。就像看电影时，需要关注哪些场景才能发现异常情况？

It could be anyone on a Friday night with this credit card. And so you kind of have to choose a broad swath of relevant sequences to capture kind of the last the last k that matters. So if you think about, like, like a movie, right, like, what what are the what are the scenes in a movie that you need to be watching to know if there's something anomalous happening?

是的。对于听众们，我想你们已经在多个场合讨论过这个问题。汽车测试的检测率从59%提升到了97%。是的，在更大规模上，这听起来非常有帮助。

Yeah. And for listeners, I think you've talked about this in in a number of places. The car testing detection numbers went from 59 to 97%. Yeah. On larger Which sounds pretty helpful.

没错，确实很有帮助。但另一个真正有帮助的是我们推出它的速度，对吧？有几家大型AI公司找到我们说，嘿，在卡片测试之后，他们觉得雷达在对抗欺诈性争议方面表现惊人。

Yeah. It was really helpful. But the other thing that was really helpful was like the speed at which we got it out, right? So like we had a couple of the large AI companies came to us and they said, Hey, This is, like, after card testing, they're like, hey. Radar is amazing for fighting fraudulent disputes.

这正是它训练的目标，对吧？那些导致欺诈性争议的交易。但我们还有大量可疑交易虽然没引发争议，我们仍希望标记它们——因为它们本质上是机器人流量。

And that's what it's that's what it's trained on. Right? Transactions that result in fraudulent disputes. But we have all these sus, like suspicious transactions that don't result in fraudulent disputes, but we still want them flagged. We want them flagged because even though they don't result in a fraudulent dispute, even though we get paid for them, like, they're bots.

这不是优质流量。它们会搞乱我们的数据。出于各种原因...我们可以聊聊AI公司面临的部分欺诈问题。所以希望你们能提供包含所有SaaS交易的管道，即便它们能产生收入，因为我们很可能要封禁这些。

It's not good traffic. Like, they're messing up our numbers. All sorts of different reasons. We can talk about some of the fraud that AI companies are facing. And so we want you to send us a pipeline with all the SaaS transactions even if they're gonna be revenue generative because we're probably gonna wanna block them.

整个过程只用了几天。我们建立了FM嵌入聚类和良好的文本对齐，可以开始标注了——这些聚类看起来可疑是因为它们在枚举登录流程的某些环节，那些则是因为其他原因显得可疑。

And it was, like, literally days. We're like, okay. Like, FM embeddings, clusters, you know, good textual alignments. You can start to label them, and you're like, this is the clusters, you know, that look sketchy because they're enumerating some component of the login flow. These are the these are the ones that look sketchy because they're enumerating some components of the login flow.

这些由于XYZ原因显得可疑。然后AI公司就能明确说：这批要封禁，这批不封。这样你们就能更快识别不仅是新欺诈手段，还包括全新类型的可疑交易。

These are the ones that look sketchy for x y z other reasons. And then the AI companies could literally say, okay, this batch, we wanna block. This batch, we don't. So it just allows you to move faster on identifying not just new fraud vectors, but like whole new types of suspicious transactions.

AI如何改变了欺诈规模？以前我运营软件网站时也遇到同样问题，有人买20美元的软件后发起退单。但现在有人能用信用卡注册OpenAI API并消费1万到1.5万美元——现在的欺诈形态是怎样的？

How has the scale changed with AI? So before, you know, I used to run some software website and we would have the same issue. People buy the software and then get us charged back, but it's like $20. Like today, you could like, you know, use the credit card and sign up for the OpenAI API and spend $10,000 $15,000 Like, what's the shape of the fraud So

友好欺诈并非盗用信用卡信息，而是诸如拒付滥用、免费试用滥用、退款滥用等行为。虽然使用的是我的账户凭证，但实际上并未为企业创造增值收益。这种现象已存在一段时间。事实上，若询问企业高管，约47%的支付业务负责人会表示他们面临的最大欺诈挑战就是友好欺诈。

friendly fraud is like not stolen card credentials, but something like nonpayment abuse, free trial abuse, refund abuse. So it's me. They're my credentials, but I'm not actually creating accretive revenue for the business. This has happened for a while. And actually, if you if you ask business leaders, like, think something like 47% payment leaders, like 47% of them will say that their biggest fraud challenge is friendly fraud.

我认为这对SaaS行业影响较小，原因有二：其一，你能窃取什么？总不会是计算机推理能力吧；其二更重要的是，即便你盗用某些服务，像Salesforce这类企业提供该服务的边际成本近乎为零，因此不会彻底破坏你的单位经济效益。

I would say this was, like, just much less of an issue for SaaS for two reasons. One, like, what were you stealing? You weren't stealing computer inference or whatever. And two, more importantly, even if you were stealing some service, like, the marginal cost of providing that good or service for Salesforce or whomever was like near zero. And so it didn't totally crush your unit economics.

如今我们身处GPU昂贵、推理成本高昂的时代，免费试用滥用、退款滥用或普遍拒付行为——当你累积这些费用却拒不支付时，对AI企业而言简直是生存威胁。前几天我与一位小型AI创始人交谈，我们正在构建一套专门针对此类欺诈的雷达扩展工具，所有人都告诉我这是个严重问题。

Now we're in a world where GPUs are expensive, inference costs are high, and free trial abuse or refund abuse or general nonpayment abuse. Right? You you rack up these charges and you never pay is, like, existentially threatening for AI businesses. I was talking to a a small AI founder the other day because we're we're building sort of a suite of radar extensions that are explicitly targeted at this type of fraud. And everyone tells me it's a huge issue.

因此每当我与企业交流时，都会深入询问他们具体面临什么问题。第一位告诉我这不是问题的先生让我很惊讶——他说自己完全关闭了免费试用，并在用户证明支付能力前严格限制信用额度。

And so with every company I talk to, try digging in on, for you, what exactly is the issue? And there's the first guy who told me it's not an issue. And I was like, oh, fascinating. What are you doing? He's like, well, I completely shut down free trials and I dramatically throttle credits until you've proven ability to pay.

我当即指出：你声称欺诈不是问题，但这实际上是在扼杀自身收入。后来我们合作帮助他们恢复了免费试用，目前正在推进中。值得注意的是，由于边际成本问题，这确实是AI企业面临的困境，但不仅限于此。

And I was like, well, you don't think fraud's an issue, but it's like, totally, like, you're choking your own revenue. Right? So anyway, we we worked with them. We got free trials back on, and and that's that's in flight. What's interesting, this is definitely a problem for AI companies because of the marginal cost, but it's not only a problem for AI companies.

想想广告主的情况。假设你是社交媒体平台，广告主入驻后——

If you think about, like, advertisers. Right? If you're like you're like a social media platform. Right? Advertisers come in.

你允许他们开始投放广告，采用后付费模式。当累积消费后拒不支付时，真正的损失不在于窃取计算资源（本案中），而在于占用了本可由付费企业使用的广告位。

You let them start advertising. They do post hoc billing. Right? So you rack up some spend and then you pay. And if you don't pay, that actually is expensive, not because you've stolen compute in this case, but because you've taken ad slots from businesses who would have paid.

有趣的是，我不知道该不该提这事，但我想可以说。好吧，我朋友前几天办了张Robinhood信用卡。嗯哼，对吧？

Fascinatingly, I don't know if I should mention I think I can say it. Okay. So my friend got a Robinhood credit card the other day. Mhmm. Yeah?

实际上，在办理Robinhood信用卡时，他被推销说还能获得试用卡。我当时就想，哦，快详细说说。什么是试用卡？试用卡基本上就是印有你名字、有效期24小时就会作废的卡片，这样你就能注册免费试用而无需被扣费。对于善意消费者来说，这听起来没问题。

And literally, as part of getting the Robinhood credit card, he was marketed that he could also get free trial cards. And I was like, oh, tell me more. What are free trial cards? Free trial cards are basically, like, cards with your name on them that are good for twenty four hours and then expire so that you can sign up for free trials without ever having to get charged. So in the hands of a well meaning consumer, that sounds fine.

但到了骗子手里，这对AI经济极具破坏性。我们刚宣布了免费试用服务，能在源头拦截大部分试用滥用行为。我们正在研发针对退款滥用的类似方案。退款滥用最严重的场景之一——你提到大规模业务时说过——有些AI公司提供企业级套餐，月费600美元或1000美元甚至1万美元，对应数十万信用点或他们提供的任何单位。

In the hands of a fraudster, that's, like, extremely disruptive to the AI economy. We just announced our free trial offering. We can catch the majority of free trial abuse at the source. We're working on the analog for refund abuse. One of the places where refund abuse is really painful, you mentioned this in the context of large volumes, like some of these AI companies will have, like, enterprise grade plans where it's like $600 or a thousand dollars or $10,000 a month for hundreds of thousands of credits in whatever units they're providing.

这些正是遭遇退款滥用的重灾区。虽然免费试用滥用多来自消费者的小额交易（但积少成多），而退款滥用往往涉及超大额订阅——用户全额使用后再取消。我们能监测到这些使用行为。

And those are the ones that are getting hit with refund abuse. So a lot of the free trial abuse is like the consumers, the little dollars, but it adds up. But a lot of the refund abuse is like very, very large subscriptions. You use it in full, and then you go and cancel. And we see the usage happening.

而且我们能验证操作者身份。所以我们有很多应对手段。我认为我们能解决这个问题，但目前确实给整个生态造成巨大痛苦。那位创始人说他已解决欺诈问题的片段恰恰说明——这种痛苦太深刻了。

So, like and we can verify that it's the person. So there's there's a lot we can do here. And I think I think we can burn it down, but it's, like, clearly creating a lot of pain for the ecosystem today. And just that vignette of that founder who told me he'd solved fraud. Just you know, I think that just speaks to, like, it's so painful.

他们宁愿放弃收入也不愿处理这些问题。

They will literally give up revenue to to not have to deal with it.

你刚才提到要将Radar扩展到支持新型AI商业模式。我觉得Stripe整体上对赋能这些AI商业模型的支付很感兴趣。本质上，人们到底需要什么？

I think you teased a little bit about how you're extending radar to serve new AI business models. And I think Stripe in general, I think is interested in like enabling payments for these AI business models. But basically like what do people want?

是的。

Yes.

那么现实与不现实的界限在哪里？我不确定这是否是个合适的说法。

And maybe what's realistic versus what is not realistic? I don't know if that's a term.

确实如此。我把Stripe看作AI公司的骨骼系统。看看《福布斯》AI 50强榜单，所有在线盈利的企业都通过Stripe变现。他们用我们做什么？大多数公司，比如Cursor和Lovable这些案例你都见过。

For sure. So, you know, I think of Stripe as like the skeletal system for AI companies. So if you look at the Forbes AI fifty, all of the Forbes AI fifty who monetize online, monetize through Stripe. And what do they use us for? You know, most of these companies, I mean, you've seen it with like the the cursor and lovable examples.

对吧？他们用极精简的团队建立规模化业务。所以希望一次性通过Stripe获得多层次的经济基础设施和金融技术栈，无需额外雇佣人力。这些AI公司从成立第一天就全球化运营。

Right? They build these very scaled businesses with very lean teams. And so they want to go, like, all in kind of on Stripe to get many layers of the sort of economic infrastructure, financial infrastructure stack in one go without needing to hire humans to do it. So they use us for payments. These AI companies are are going global from day one.

我们统计过Stripe上收入最高的100家AI公司，第一年结束时中位数已覆盖55个国家，第二年就超过100国——这比三年前的SaaS浪潮全球化速度快一倍。几乎所有公司都采用我们的优化结账系统（内置100种支付方式），使用雷达反欺诈系统。目前市场仍在摸索供需平衡点，因此变现模式迭代非常频繁。

Like, we were looking at the top 100 grossing AI companies on Stripe, and, like, the median was in 55 countries at the end of their first year and over a 100 countries at the end of their second year, which is, like, twice as global as the SaaS wave from three years before. So they they almost all adopt our optimized checkout suite, which comes with a 100 payment methods out of the box, very global reach. They almost all adopt radar and our fraud suite. And then one of the things that's been really interesting is, you know, I think the the market's still trying to figure out what's the intersection between supply and demand. And so there's a lot of iteration across monetization models.

比如固定订阅费？按量付费？还是信用额度消耗模式？每种选择都涉及收入结构和反欺诈策略。

Like, is it a fixed fee subscription? Is it pay as you go usage? Is it this credit burn down model? And there are revenue implications. There are also fraud implications.

但同样重要的是单位经济效益。我们最近有个发现：随着大语言模型进步，越来越多AI公司成为'包装商'（rapper，注意是w-r拼写）。这绝非贬义说法。

But equally important, there's, like, unit economic implications. And I think one of our recent ahas was, you know, as the LLMs have gotten better, more and more AI companies are rappers. And I don't I don't say that in a sorry. Rappers are w r, not r. And I don't say it in derogatory way.

我用Aravind Srinivas曾经对我说过的方式说，就像，我很自豪自己是从说唱歌手起步的

I say it in the same way Aravind Srinivas once said to me, like, I am proud to have started as a rapper

因为复杂性

because Complexity

是个说唱歌手。找到产品市场契合点，打造出色产品，快速行动，不被研究和他人可能提供的底层模型拖慢脚步。如果愿意，我以后可以再做那些事。但很多AI企业都是包装层，对吧？他们的服务都有内在的LLM成本。我们知道LLM模型供应商起起落落。

is a rapper. Find product market fit and build an amazing product and move really quickly and not get slowed down in the research and other people could provide the underlying models. I could do that later if I wanted or not. But because a lot of these AI businesses are wrappers, right, their services have an inherent LLM cost underlying them. We know that LLM model providers are ebbing and flowing.

底层模型时好时坏。这些模型的价格也随时间波动。这导致你最终服务的定价变得非常复杂，特别是当这项服务如此依赖上游LLM时。两周前我们推出了一个叫'令牌计费'的功能，本质上是个API，让你能实时追踪和定价推理成本。这有什么用呢？

The underlying models are getting better or worse. The price of those models are getting better or worse over time. And so that actually leads to a lot of complexity in how you price your final service when that final service is, like, so dependent on an upstream LLM. So one of the things we launched two weeks ago now, we call it token billing, but it's basically you it's an API that lets you track and price to inference costs in real time. And what does that do?

这么说吧，如果你的服务建立在底层LLM上，而模型成本突然下降80%——我们都见过这种情况——你肯定不想维持原价，因为竞争对手会趁机杀入。但反过来更危险的是，如果底层LLM成本突然变成三倍（这种情况也出人意料地发生过），不及时调价就会导致单位经济直接崩盘。所以令牌计费就是个解决方案。

It's like, well, if if your service is built on underlying LLM and the model cost drops 80%, which we've all seen it happen. Right? You don't wanna keep your price where it is because the competition's gonna swoop in. But then conversely, and I think more threatening, if the if the cost of the underlying LLM three x's, which it also sometimes does, surprisingly, you could have unit economics that are literally underwater if you don't if you don't adjust your price. So, you know, token billing is an example.

但我们看到这些AI公司在尝试基于用量的计费方式，基于结果的计费也很有意思。

But but we're seeing these AI companies, you know, iterate across usage based billing, outcome based billing is kind of an interesting one.

Stripe会介入这方面吗？因为这并不在你们常规业务范围内。

Does does Stripe get involved there? Because that's not really within your normal wheelhouse.

我们确实有。所以我们有支付系统，还有一个计费套件。几乎所有AI公司都在使用我们的计费套件。计费包括固定费用订阅，但也支持基于使用量的计费，比如计量计费。是的。

We do. So our we have so we have payments, and then we have a billing suite. And almost all the AI companies use our billing suite. Billing includes things like fixed fee subscriptions, but it also supports usage based billing, so, like, metered billing. Yeah.

你可以用各种不同的单位来定义使用量。我们有许多客户，包括Intercom，他们实际上是根据结果来定义的。在这个案例中，就像支持案例的解决数量。完全正确。

And you can define usage in all sorts of different units. And we have a number of customers, including Intercom, who actually define it in terms of the outcome. So in this case, it's like support cases. Resolutions. Exactly.

我觉得这真的很有趣。我接受的是经济学训练。之前在电梯里我跟你说过，我认为物理学家是最好的科学家或机器学习工程师，但我当时并不知道，所以我成了经济学家。我经常思考，是什么让市场高效或低效？在AI领域，我担心的一点是，当人们在看到价值之前就要为产品付费时，把产品推向市场非常困难。

And, you know, I I think it's really interesting. I'm an economist by training. I told you earlier in the elevator that I think physicists make the best scientists or MLEs, but I didn't know that at the time, so I'm an economist. And I think a lot about, okay, like, what makes the market efficient or inefficient? And one of the things that I worry about in AI is it's incredibly hard to take a product to market when someone has to pay for it before they see the value.

这在AI领域尤其如此，因为很多买家，尤其是企业买家，不知道如何评估底层技术。所以如果你能通过不仅仅说'你只需为实际使用付费'来打开局面——我的意思是，按使用付费确实有帮助，因为他们不需要预先承诺大额合同，但他们可能会担心：如果员工大量使用却对业务没有实际帮助怎么办？但如果你能提供一个明确对他们有利的实际成本定价方案，就更容易打开市场。

And that's especially true with AI because a lot of the buyers, especially enterprise buyers, don't understand how to evaluate the underlying technology. And so if you can get your foot in the door by saying not just, oh, you'll only pay for what you use. I mean, pay for what you use is kinda helpful because they're not committing upfront to some huge contract, but they can come in with a fear like, well, what if my employees use it a lot and it's not actually helping the business? Mhmm. And if you can come in with, like, an actual cost sort of pricing function that is clearly profit positive for them, it's a lot easier to get your foot in the door.

对吧？你知道人工解决一个支持工单的成本是x。我承诺向你收取的费用会低于x。从质量角度看，尝试我的服务显然对你更有利。总之，我们看到很多基于结果的计费方式。

Right? So you know that a human resolving a support ticket is x. I promise that I will charge you, you know, less than x. Like, it seems conditional on quality strictly better for you to try out my service than that. So, anyway, we see we see a lot of outcome based billing.

我们在AI公司看到的另一个有趣现象是稳定币的使用。这个趋势出现得更早，但他们的使用案例很有意思。比如你现在去vZero注册账户，实际上可以用稳定币支付。我们看到很多希望实现全球覆盖的AI公司采用这种方式，也包括那些定价很高的AI公司。比如YC孵化的Shadeform就是个很好的例子。

The other thing we see an interesting amount of from AI companies is Stablecoins. And this one's this one's earlier, but their use cases are interesting. So, like, if you if you go to vZero now and sign up for an account, you can actually pay in stablecoins. We're seeing this a lot for AI companies that want to have very global reach, but also for AI companies that have very high price points. So like Shadeform, the YC startup is a great example.

他们接受稳定币支付。稳定币现在实际上占了他们交易量的20%。他们使用稳定币的案例基本上就是面向全球市场和高成本业务。全球化意味着ACH（自动清算所）支付不是可选方案。对吧？

They accept stablecoins. Stablecoins are now actually, like, 20% of their volume. And their use case for stablecoins is basically, like, very global and very high cost. And so the global means ACH isn't an option. Right?

ACH通常是美国用户用于低成本支付的选择。如果使用国际信用卡这类方式，面对大额交易时，你实际上要支付高达4.5个百分点的国际信用卡费用。这部分直接侵蚀了你不愿放弃的利润空间。现在20%的交易量通过稳定币完成——我们实际做了实验，其中一半是完全增量业务，也就是说如果不开放稳定币支付，他们原本只能获得90%的收入。

ACH is usually what folks in The US would go to for low cost. If you're gonna use something like, you know, international cards, though, on, like, a very large basket cost, you're talking about paying four and a half percentage points, literally, just to international card costs. And so that's just taken a bunch of your margin that you don't that you don't wanna give away. So now 20% of the volume comes through stablecoins. We actually an experiment with them, and half of that is fully incremental, which is to say they would only have 90% of revenue they had had they not opened up Stablecoins.

另一半交易量是从其他支付方式转向稳定币的。在成本方面，稳定币支付的成本是1.5个百分点，相比信用卡的4.5个百分点，这让他们多赚了几个百分点的利润，自然乐见其成。稳定币正是我们看到AI公司快速采用的另一项技术。我常说极客就爱买极客的账。

The other half is a shift from other payment methods to stables. And then on the cost side of the house, you know, the cost of stables for them is, like, 1.5 percentage points versus 4.5 percentage points. So that's a couple extra percentage points in their pocket, which they don't mind either. So stablecoins is is another thing that we're seeing AI companies adopt pretty quickly. I like to say that nerds buy from nerds.

好。

K.

这里存在一种良性的网络效应。懂吗？当AI用户都拥有稳定币钱包时，他们转到其他AI服务商时依然能使用稳定币支付。这就形成了自我强化的循环。

And so there's a nice sort of network effect there. Right? Like, if people who use AI have Stablecoin wallets, when they go to the next AI provider, they have a Stablecoin wallet. Right? It's sort of self reinforcing.

我们的消费级产品LINK也能看到这种现象。LINK用户已突破2亿，这个网络规模不小。但更有趣的是在AI领域，这个网络非常非常密集。比如Lovable...对，接受...

We also see this with LINK, which is our consumer product. Like, LINK just passed 200,000,000 consumers, so it's not a small network. But what I think is more interesting is in the case of AI, it's a very, very dense network. So Lovable Yeah. Accepts.

LINK支付占Lovable平台交易量的58%。每三个在Lovable购物的用户中，就有两人会使用LINK一键支付——因为他们早已拥有LINK账户。这个数据既体现了LINK网络的密集度，也反映了AI网络的密集程度。

LINK. 58% of Lovable's volume flows through LINK. So for every three people who are buying on Lovable, two of them are buying with one click LINK checkout because they already have a LINK account. And I think that just, like, gives you a flavor of of the the density of the link network, but also the density of the AI network.

确实。你们是否会关注网络密度的传统衡量指标？作为经济学家，这确实是我最先想到的问题。

Yeah. Is there a are there classical measurements of network density that you keep an eye on? I mean, because obviously, as an economist, like, that's the first thing I go to.

是的。我是说，赫芬达尔指数与其说是特别关注网络密度，不如说是当我们审视所有通过Stripe的交易流时，它们的集中程度如何。你可以从多个维度来看待这个问题——比如商户集中度（交易是集中在某些商户还是分散在多个商户之间），或者行业集中度（是集中在某些行业还是跨行业分布）。

Yeah. I mean, the Herffindale index is less about network density in particular and more about as we look at all of the transactions that are flowing through Stripe, how concentrated are they on I mean, you can look at it along a lot of dimensions. You can merchants, how concentrated are they on certain merchants versus spread across merchants? Yeah. How concentrated are they on certain industries versus spread across industries?

还有地域集中度（是集中在某些地区还是广泛覆盖多个地区）。我们确实在追踪这种集中度。对我们而言，其中部分指标实际上与网络密度相反——我们追求的是多样化。我们希望涉足不同行业、不同市场，实现全球覆盖。虽然当前浪潮是AI（我对AI极其乐观），但我们真正想要的是广泛推动互联网经济总量的增长，而不仅限于AI领域。

How concentrated are they on certain geos versus a broad range of geos? And, yeah, we definitely track that concentration. Now for us, some of that is actually the inverse of network density, which is we want diversification. Like, you know, and we wanna be exposed to many different industries and many different markets and have global reach because the current wave is AI, and I am incredibly bullish on AI, but we really wanna be growing the GDP of the Internet broadly. And that's not constrained to only the AI domain.

是的，非常棒。我们是否该转入ACP的话题了？我不知道如何更自然地过渡，但感觉智能体最终确实会相互进行商业活动。我想这就是衔接点——正是金融基础设施与AI的完美交汇处。

Yeah. Excellent. Should we move into the ACP? I don't know how to transition it better than feel like agents do want to eventually do commerce between themselves. I guess that's the transition and that that that, is the perfect intersection of financial infrastructure and AI.

那么，能否请你讲讲ACP的来龙去脉？我认为这可能是下半年最重要的发布之一，也是OpenAI与Stripe之间极具战略意义的合作。

So maybe, could you tell us the story of ACP, right? Like, I I think this is one of the biggest launches of, I guess, like in the second half of the year. And like, I guess a really important strategic move between OpenAI and Stripe.

没错。我们讨论了很多关于AI公司的话题，其中一个重要分支就是AI商业——智能体商业。退一步看，我们花在ChatGPT这类大众消费工具和Replit、Vercel等AI开发工具上的时间越来越多，自然希望这些智能体工具能越来越多地代我们执行操作。

Yeah. So, you know, we talked a bunch about AI companies in general. One important slice of AI companies is AI commerce, agentic commerce. And, you know, I think just zooming back, like, we're all spending more and more time in some combination of broad consumer based tools like ChatGPT and AI dev tools like Replit or Vercel or whatever. And we want those agents, those tools to increasingly take action on our behalf.

早期我们在ChatGPT的Operator功能中见过雏形。但关键领域是让它们代我们购物——有时是推荐产品，更多时候是直接完成全流程交易。几周前我们与OpenAI联合发布了智能体商业协议（ACP），本质上是企业如何与智能体对话的通用标准。

And I think, you know, we saw an early version of this in ChatGPT with operator. But an important area we want them to take action is buying on our behalf. You know, sometimes it's recommending products, but often it's like literally getting it all the way over the wall. So a couple weeks ago, we announced our agentic commerce protocol, which is joint with OpenAI. And it's basically just a shared standard for how businesses can talk to agents.

想象一下：过去是人类向企业购买，现在中间多了个智能体。这从根本上改变了金融基础设施的运作方式——支付流程需要革新，反欺诈检查也需要重新设计。

So if you think about it like it used to be that a human was buying from a business. Now there's an agent that's sitting in the middle. And that fundamentally needs to change how the financial infrastructure works. Like, checkout needs to look different. Fraud checks need to look different.

支付流程需要有所不同。但同时，商家也在努力探索如何通过各类代理高效展示他们的产品目录、库存、品牌和定价，以获取新的需求流。这确实是一个全新的领域。因此，代理商务协议的核心就是为代理提供一套共享语言，让他们能从商家那里获取产品库存、价格信息以及品牌呈现方式。我们还构建了共享支付凭证，让代理可以代表买家传递必要支付信息，因为代理不愿承担风险。

Payment flows need to look different. But, also, merchants are trying to figure out how they can efficiently expose their product catalog, their inventory, their brand, their pricing through a range of agents to have access to that new stream of demand. And it's kind of a brave new world. So the agented commerce protocol is really about that shared language for agents to get from merchants what products they have available at what prices, how they want the brand to appear. And then we also built a shared payment token, which basically allows the agent to pass over the required payment credentials on behalf of the buyer because the agent doesn't wanna bear the risk.

代理实际上并不想介入交易过程。而商家则希望最终能直接向消费者收款并建立直接关系，以便处理退货等事宜。因此共享支付凭证也是重要组成部分。此外，欺诈防范是另一个关键要素，对吧？

The agent doesn't actually wanna be in the middle of the transaction. And the merchant wants to, you know, undertake the charge and actually have the direct relationship with the consumer at the end of the day for returns and more. And so the shared payment token was also an important component. And then fraud was another important component. Right?

这是好机器人还是坏机器人？就在不久前，企业的最佳做法还是屏蔽所有机器人。但现在许多机器人是有益的。你肯定不想切断这部分需求。因此我们在共享支付凭证中传递的信息包括交易可信度评分，让商家能做出正确决策。

Is this a good bot or a bad bot? There was a day not very long ago when the optimal thing to do as a business was to block all bots. Now many bots are good bots. You do not want to cut off that demand. And so what we pass over as part of the shared payment token includes scores on the goodness of of the transaction so that the merchant can can make the right decision.

这种模式的体现之一就是ChatGPT的即时结账功能，你们有人用过这个买东西吗？

One of the ways this manifested was in instant checkout in ChatGPT, which have you guys bought anything from this?

还没有。说实话我试过。它只是推荐商品，但我总想自己完成最后一步——亲自结算。你懂吧？

Not yet. Honestly, I I've I've tried. Like, it just recommends things, but I always want to take over the last mile of Yeah. Checking it out myself. You know?

对对对，我很难完全交出控制权。

Yeah. Yeah. Yeah. It's hard for me to, like, hand over control.

好的。我觉得他们其实也还在不断优化推荐算法。昨晚我女儿上课，我就带儿子出去吃饭。他说学校要演话剧，需要穿西班牙店主服装。我试着搜索'儿童西班牙店主戏服'，结果系统给我推荐了Etsy上1300美元的波列罗外套。

Okay. Yeah. And and and it I think they're also still, like, iterating on their recommendations to some extent. I last night, my daughter was at a class, and so I took my son out to dinner, and he told me he has a school play, and he is supposed to dress as a Spanish shopkeeper. And so I tried to search for, like, kids' Spanish shopkeeper outfit, and they recommended me a $1,300, like, $1,300 bolero off Etsy.

这部戏没那么重要。

The play is not that important.

要知道，重要的不是我喜爱这个孩子，而是这部戏本身并不那么重要。但有很多很棒的商品可以购买。在最初与ChatGPT推出的即时结账功能中，你可以从美国Etsy卖家那里购物。很快将有超过100万家Shopify商家加入，包括一些大品牌如Glossier和Viore。本周Salesforce也宣布参与其中。

Know play is not I love the child, but the play is not that important. But there's a but there is a lot of great stuff you can buy. And so in the in the initial instant checkout launch with ChatGPT, you could buy from US based Etsy sellers. There's over 1,000,000 Shopify merchants coming soon, including some really big ones like Glossier and Viore. This week, Salesforce announced that they're also in.

然后，我最喜欢的是，就在一两周前，你可能会问：大型零售商会加入吗？最近几天，沃尔玛和山姆会员店已签约，同意通过ChatGPT和Agent电商协议销售他们的库存。这无疑是大型零售商参与的最大信号。

And then, you know, this this my favorite is that, like, a week ago, two weeks ago, you could have asked, hey. Are the largest retailers gonna get on board with this or not? In the last couple of days, Walmart and Sam's Club have just signed up to also make their inventory purchasable through ChatGPT and and the Agent Ecommerce protocol, which, like, I don't think that there is a bigger signal on a big retailer being up for it

比钱包还大。确实。

than wallet. Biggest one. Yeah.

是的，这相当令人兴奋。对我们和整个生态系统来说，Agent电商协议的重要之处在于它不局限于Stripe。我之前提到的共享支付令牌或Agent电商协议，无论你的支付提供商是谁都能使用。

Yeah. Yeah. So that's that's pretty exciting. And then, you know, one of the things that is important to us and to the broader ecosystem about the agented commerce protocol is it's not about Stripe. So that shared payment token I talked about or the agentic commerce protocol, like, that works no matter who your payments provider is.

我们可以将共享支付令牌传递给任何其他支付服务提供商。你不必通过Stripe处理。这也不仅限于OpenAI。就像你我看到新模型不断上线一样，我们希望能在不同模型间灵活切换。

We can pass the shared payments token over to any other PSP. You don't have to process on Stripe. It's also not just about OpenAI. And so in the same way that, like, you and I are seeing, you know, new new models come online all the time. We wanna be able to move across models flexibly.

未来会有各种新型Agent购物体验不断涌现，我们希望商家能一次性轻松集成所有上线的Agent。这正是ACP协议提供的价值——它是一个标准协议，无需为每个Agent定制集成。你们看到Karpathy的推文了吗？他基本上重现了...

There's gonna be sort of new agentic buying experiences coming online all the time, and we wanna make it easy for merchants in kind of, like, one shot to integrate with all of the agents as they come online. And that's what the ACP really provides because it's a standard protocol versus needing to do custom integrations per agent. Should you guys see Karpathy's tweet about the he he basically, like, recreated

NemoChat？嗯。好的。

NemoChat? Mhmm. Okay.

是啊。你看，如果他能用不到100美元和8000行代码做到这一点，你可能会觉得很多公司很快都会推出自己的版本。

Yeah. So, like, if he can do that, 8,000 lines of code, less than a $100, like, you might think that a lot of companies are gonna roll their own really soon.

用Nanochat？那会很有趣。我我我还没想到这个关联。我们

Using Nanochat? That would be interesting. I I I haven't made that connection yet. Let's

我们看看。

Let's see.

我们看看。

Let's see.

但我认为这个基本前提是，虽然可能是赢家通吃，但现在还不清楚谁是赢家。顺便说一句，为了商业效率，我希望不是赢家通吃。因此，许多商家实际上需要通过多个代理销售产品，这某种程度上是ACP的前提。看到早期进展我们非常高兴，老实说，已经有很多商家和AI平台想要加入，简直应接不暇。所以我觉得我们走对了路。

But I think this basic premise that, like, it may be a winner take all, but it's not yet clear who the winner is. And by the way, like I hope for the efficiency of commerce that it isn't a winner take all. Therefore, many merchants need to actually be having their products sold through many agents is kind of the the premise of ACP. And we're just delighted to see the early traction and, you know, have just been flooded, honestly, with both merchants and AI platforms wanting to wanting to join. So I think I think we're on the right path.

是的。这就是你们的协议类型。

Yeah. And this is the your protocols kind of.

是的。我们和Crunch AI做过一期节目，他们为代理商提供网页重写服务，HIMSS是他们的客户开发对象。我觉得对品牌来说这很简单，就像'我不在乎来源，只要你买我的产品，我们就是朋友'。

Yeah. We we did an episode with Crunch AI, which does web rewriting for agents, and they have HIMSS as a customer dev skims. And I I think every brand is like I mean, for brands, it's very easy. It's like, hey, I don't really care where it comes from. If you buy my thing, we're friends.

我认为你消除了最令人头疼的部分——欺诈问题。我有个关于稀缺资源（比如活动门票）的好机器人和坏机器人的问题。我觉得这会...

I think you take away the part which is the most annoying to think about, which is the fraud. I do have a question on the good bots, bad bots when it comes to scarce releases, like tickets for events and things like that. Think that's Oh, going to so

掌握得真好。天啊。

good to master. Oh my god.

没错。但现在如果任何人都能用代理程序直接上网购买，那排队机制怎么办？因为所有人都会瞬间到达。对吧？所以...

Exactly. It's like, well, but now if anybody can just have an agent that just goes on the website to buy them, now it's like, how do you do the queue? Because everybody gets there instantly. Right? So like

这家公司...前几天晚上我和他们的CEO共进晚餐，你可以把它想象成某国的StubHub（具体国家不便透露）。他在销售大约3000张Bad Bunny演唱会门票，结果有40万...

the company, but I had dinner the other night with the guy who's the CEO of you can basically think of it as, like, StubHub for Country X. I won't say what Country X is. And he was selling, I think it was, like, 3,000 Bad Bunny tickets, and he had 400,000

天啊。

Oh my god.

人来抢购Bad Bunny门票，但几乎全是机器人。这完美印证了我们之前讨论的——问题不仅在于欺诈争议。那位CEO告诉我，他们有黄牛用机器人抢票，然后高价转卖，但这并不会导致退单。

People come to buy the Bad Bunny tickets, except almost all of those people were actually bots. And this is this is a great example of what we were talking about earlier around, like, it's not just about the the fraudulent dispute. So the conversation I was having, he was like, you know, we have scalpers scalpers who have bots. They they end up scalping the tickets They come and they buy. It doesn't result in a chargeback.

比如，他们确实付款了，但这些人并不是我们想要的支付者。回到我们之前关于可疑交易的讨论，欺诈行为其实有很多种类型。如果把欺诈仅仅定义为导致争议的交易，这种看法就过于狭隘了。他想要拦截所有黄牛党，那些在注册时枚举邮箱地址的人等等。他还跟我说了件有趣的事——由于系统架构的特殊性（虽然部分原因具有普适性），他们需要多维度考量。

Like, they pay, but they're not the people that we want paying. And so to our conversation earlier on suspicious transactions, like, are lots of different types of fraud. And thinking of fraud as just things that result in a fraudulent dispute is actually overly narrow. And he wants to block, you know, all the scalpers, everyone who's, like, you know, enumerating through email addresses in their sign up and and and. The other thing he said to me, which was interesting, is he for various reasons and some of this is actually like the the the nuances of how his system is built, but I think some of it generalizes.

他希望在用户到达结算页面前就能捕捉到欺诈信号。那么问题来了：如何在用户未输入支付凭证时就能识别客户？我们有很多方法可以优化这方面。他有个具体原因听起来很平常：一旦票券进入购物车——

He wants to have those fraud signals before they even get to the checkout page. And so how can we understand the customer independent of them entering their payments credentials? And there are a bunch of ways we can and we can get better there. His particular reason why is he's got this is a little mundane, but once the ticket goes into the cart Oh,

是

it's

系统会有10到15分钟的锁定状态。虽然最终在扣款环节成功拦截了欺诈，但这也意味着那些优质客户在此期间被误伤了。总之，我认为现在有很多令人兴奋的工作可以做，这不仅得益于Stripe网络的规模，更因为AI技术能帮我们识别更广泛的欺诈模式。传统确定性系统只能设置'拦截/放行'这种二元规则，

can't be touched for, like, ten or fifteen minutes. And so even though, you know, it's successfully blocked at the actual charge time, it's, like, been held from from those those other good customers. So, anyway, I I think there's this there's a lot of exciting work to be done that's actually, like, increasingly possible, not just because of the scale of Stripe network, but also because of AI around understanding an expansive set of fraud vectors. And, you know, if if you think about traditional deterministic systems, you know, you'd write rules like block, don't block. Right.

但现在我们可以基于大语言模型，用人类可读的方式描述'为什么这笔交易值得警惕'。目前由人工处理，未来则会是智能代理在模型输出基础上进行决策推理。我觉得未来3到6个月就会实现这种工作模式。

Now you can think about, okay, actually, foundation model, text alignment, like, human readable description of, like, why we're worried about this charge. And then today, a human tomorrow, an agent sitting on top of that and decisioning, like, reasoning over the model outputs. I think that's the world we'll be in in the next three to six months.

确实。我们在推出这类功能时必须谨慎，因为用户被AI拒绝却得不到解释时，他们有理由感到愤怒——尤其是当他们本应获得某项权益时。必须建立申诉流程或设置二级人工审核机制，

Yeah. I think we need to we have to be careful about rolling those kinds of things out because people get very upset and justifiably so, and they are denied something that they they they should have by a bot that won't explain itself. Yes. Right? There needs to be like an appeals process or like some like tier two human, like

当然。

For sure.

请问我能和经理谈谈吗？当然可以。

Can I speak to the manager, please? For sure.

是的，确实如此。而且，人类也会做出错误判断。有时他们犯错的比例甚至比语言模型更高，因为他们无法处理那么多信息。不过我同意。

And yes. For sure. And well, humans make bad calls too. Sometimes they make bad calls at higher rates than LMs because they can't reason over as much information. But I agree.

绝对需要申诉流程。但当我们真正审视不良行为者时，会发现极少数的不良行为者却造成了大量问题，对吧。

Definitely appeals process. But then also, like, when we actually look at bad actors, it's like a tiny, tiny share of bad actors accounts for a huge volume Right.

你只是在清除那些问题。

You're just rooting those out.

的不良后果。是的。所以最终结果是，守规矩的用户反而体验更差——可能无法享受免费试用、使用额度受限、需要预付款，或是被迫支付更高价格来弥补坏账成本。我认为这里存在区分良莠的机会，我...

Of bad outcomes. Yeah. And so what ends up happening is actually good actors have worse experiences, which could mean they don't have access to free trials or they're gated and how many credits they can use prepayments, or they're just charged a higher price because they're covering for the cost of the bad guys. And I and I actually think there's an opportunity to, like, to the sheep from the goats. And I'm like, I

不知道。

don't know.

但它们都很好，都很可爱。

But they're both good. They're both cute.

我觉得绵羊本该很可爱，但我一直印象深刻。你们去过黄石公园吗？对，从猛犸入口进入黄石国家公园时，会看到一面巨大的岩壁。那些山羊就正好在那上面。

I think sheep sheep are supposed to be cute, but I've always been impressed. Have you guys been to Yellowstone? Yeah. You go into Yellowstone National Park at, like, the Mammoth Entrance, and there's this, like, sheer giant, you know, wall of rock. And there's these mountain goats like exactly.

它们几乎是以90度角在攀爬。

They're, like, scaling it at Yeah. 90 degrees.

是啊。像爬山我能理解，但水坝...我不懂它们为什么那么喜欢水坝。看起来太危险了，还有小羊羔在上面走。

Yeah. Like mountains I get, but, like, dams, I I don't know why they love dams so much. It's like it looks so dangerous and, like, there's babies just walking on there.

太危险了。简直像AI生成的画面，但这是真实的。

So dangerous. Yeah. It almost looks like AI generated the image, but it's real.

是真的。千真万确。好吧，我要开启经济学家模式了，毕竟你显然还自认为是个经济学家。

It's real. It's real. And okay. I'm gonna do like economist corner just because like, you know, you you clearly still identify as an economist.

我确实自认为是经济学家。这挺奇怪的。

I do identify as an economist. It's very strange.

对对对。那当你遇到类似'X的StubHub'这种情况时，难道不会想推荐拍卖机制吗？明明有那么多拍卖机制可以出清市场。

Yeah. Yeah. Yeah. Well, so like when you encounter those like the StubHub for x, don't you feel the temptation to recommend an auction? Like and there's so many auction mechanisms that can clear the market.

这显然是个市场清算问题。这里的市场解决方案是什么？

This is clearly a market, sort of clearing problem. What's the market solution here?

我不确定我通常会倾向于推荐拍卖。我更倾向于提出许多深入的问题，探究他们为何如此设计系统，然后尝试集思广益，看看是否存在更高效的路径。是的，我对大多数价格匹配发现建议都是这种感觉。大多数市场其实效率低下，因此我认为存在许多改进空间。

I don't know that I am usually tempted to recommend an auction. I am usually tempted to ask a lot of very probing questions about why they've designed the system the way they've designed it and then try to brainstorm, yeah, whether there's a more efficient path. But yeah, I feel that way about most pricing matching discovery recommendations. Like, most markets are just inefficient. And so I think there's a lot of opportunity to make it better.

我加入Stripe的原因之一，也是过去四年在这里最珍视的一点是：当我们看到提升市场效率的机会时，我们可以直接投入实践，而不必先考虑如何变现。你可以理解为双方利益高度一致——任何帮助Stripe上企业成长的举措都会促进我们自身发展。

One of the reasons I joined Stripe and one of the things I have loved about being at Stripe for the last four years is when we see those opportunities to make the market more efficient, we can actually invest in doing them without optimizing them, you know, without monetizing them directly. So you can think of it like incentives are very aligned. Anything we do to help the businesses on Stripe grow helps us grow because, you know, they Yeah.

你们的优势。通过我们来实现。是的。

You advantage. Through us. Yeah.

所以我们经常这样做。比如指导商家优化结账流程，我们甚至会直接为他们更新结账系统；或是优化支付受理、自动化重试机制。总之，在一家无需为助商工具操心市场推广的公司很舒心——只需专注于帮助平台商家做得更好，这本身就是良性循环。

And so we do this all the time. Like, here's how to, you know, improve your checkout, and we just, like, update the checkout for them or, like, optimizing their payments acceptance or automating their retries. So, anyway, I I think it's just it's it's very nice to be at a company where you don't have to worry about the go to market for something that helps the businesses that run on you. You just have to help the businesses that run on you do better, and that in and of itself is is a good outcome

因为他们非常

because they're very

有利于公司发展，并证明了投资的合理性。

for your company and justifies the investment.

是的，他们积极性很高。有时候就像营收和利润的关系。

Yeah. They're very incentivized. It's like top line and and bottom line sometimes.

没错。顺便说一句，这并不全是因果关系。但去年使用Stripe的公司增长速度是标普500指数的七倍。而且你知道，再次强调，并非全是因果关系，还存在选择效应，但确实——

Yeah. And by the way, like, this isn't all causal. But last year, the companies on Stripe grew seven times faster than the S and P 500. And, you know, again, like not all causal and there's selection effects, but definitely

某种程度上这只是规模效应。

some It's just of for sizing effects.

有些不错的顺风因素。确实。

Some good some good tailwinds. Yeah.

对。我是说，确实。而且我很喜欢经济学家用的那个术语——'减重'。基本上就是消除摩擦，意味着同时增加生产者和消费者的剩余价值，Stripe也因此受益。这感觉真的很棒。

Yeah. So I mean yeah. And I think the economist term I really like is their weight lost. And like basically, eliminating friction, it means improving surplus for both producer and consumer and Stripe benefits as a result. It's just like, it's nice.

多方共赢。这种通过协议实现的效果让人感觉很温暖。我认为选择以协议形式发布是个有趣的决定。就像你说的，这将形成多对多关系，有时候Stripe甚至不会参与其中。

Everyone wins. It's a it's a good good fuzzy feeling. Coming out to the protocol. I think it's an interesting decision to actually release it as a protocol. Like you said, it's gonna be many to many, and, like, sometimes Stripe is not involved.

你还提到了Stripe Link和Stripe Checkout，这些都是Stripe的产品对吧？它们不是协议。所以我觉得选择以协议形式发布而非产品，是个非常关键的决定。我很好奇内部是否有过争论，或者关于这个协议化决策背后有什么内部故事。

You also mentioned, like, Stripe Link and Stripe Checkout, and those are Stripe products. Right? Those are those are not protocols. So I think, like, it's a very interesting and pivotal decision to choose to release it as a protocol as opposed to not. I was wondering if there's like any internal debate or is there any internal color about like the decision behind choosing a protocol.

知道，

Know,

Stripe在我任职的四年里一直发展迅速。我认为这种速度还在加快。加速的原因是客户、用户需求的变化，以及市场在AI领域对企业和买家需求的变化都在加速。因此像ACP、代币支付这样的新产品和解决方案正被市场迫切需求。我们发现市场上存在一个空白。

Stripe has moved fast for the entire four years that I have been there. I think it is accelerating. And I think it is accelerating because customers, users, changes in what the market needs, both what businesses need and what buyers need in the world of AI is accelerating. And so new products, new solutions like ACP, like token billing are literally being pulled out of us. And what we saw was a hole in the market.

消费者和我们讨论过的开发者都希望通过代理进行购买。代理已准备好为他们购物，商家也准备好让代理代表消费者购买，但市场还没找到实现方式。这不仅仅是Stripe的问题，而是关乎互联网GDP的增长。

Like consumers and we talked about developers too, but, like, consumers wanna be buying through agents. Agents are ready to buy for them. Merchants are ready to let agents buy on behalf of consumers, and yet the market can't figure out how to make it work. And that's not about Stripe. That's about growing the GDP of the internet.

这是确保商业流通的问题。所以当时没有任何争议，整个生态系统都需要这个。我们会与OpenAI合作突破障碍。顺便说一句，正如我所说，我们还处于早期阶段。

That's about making sure commerce can flow. And, you know, so there was no debate. It's like the ecosystem needs this. Like, we will pair with OpenAI to get it over the wall. And and by the way, like, as I said, we're early days.

ACP目前获得了巨大关注。但即使它没有，或者六个月后出现其他方案，我们真正需要的是一个共同标准。我们不需要把自己的名字刻在这个标准上。我认为这绝对是正确的方向。有人问我，代理式商务是否会直接取代现有的互联网商务？

You know, ACP is seeing a ton of traction. But if it wasn't or if something different comes out six months from now, like, all we want is a shared standard. We don't need to have, like, our name on the shared standard. I think it's absolutely absolutely the right thing the right thing to be doing. You know, I have had folks ask me, like, is agentic commerce just a straight substitute for the commerce that's already happening on the Internet?

是的，就像高级API一样。

Yeah. It's just like fancy APIs.

没错。我认为这不是简单替代，而是实际上扩大了商业活动的范围。第一个让我产生这个想法的是Dwarkash。他提出时我确实需要反应一下。

Exactly. And I think the answer is it's not a straight substitute. I think it is actually, like, expanding the aperture of what commerce will get done. And the first person to put this bug in my head was Dwarkash. And when he said it, it actually like took me a second.

比如，我原本不相信这个观点。但我觉得它是正确的——如果你观察人们将收入用于消费的比例与其收入水平的关系，会发现高收入人群的消费占比要低得多，而低收入人群的消费占比则高得多。这背后有很多原因，但一个重要原因是：对超高收入人群而言，消费的最大成本不是金钱支出，而是消费行为所耗费的时间成本。因此我对代理商务如何为高收入人群拓宽消费渠道非常感兴趣，因为它消除了最昂贵或最具约束力的限制因素——时间成本。

Like, I didn't believe it. But I think it's right, which is if you look at the share of income that is spent on consumption as a function of how much income you have, you see that high income people spend a much lower share of their income, low income people spend a much higher share of their income. There are many reasons for that. But one reason is the biggest cost to very high income people consuming is not the dollar cost, it's the time cost of consumption. And so I'm very interested in how agentic commerce can open the aperture for spending by high income people because it's removing the most costly or binding constraint, which was their time.

这样我们实际上是在向经济中注入真正增量的、非替代性的、而是额外增加的货币。显然，这会带来第一、第二、第三阶的连锁效应。

And we're actually then truly pumping incremental, not substituted, but additive dollars into the economy. And obviously, like, you know, first, second, third order effects of that.

是啊。有时候这会导致人们购买1300美元的戏服这类消费

Yeah. Well, it results in sometimes buying $1,300 costumes and

我可没买1500美元的戏服。相信我，我宁愿花一小时时间也不愿买1300美元的戏服。

I didn't buy a $1,500 costume. Trust me. I would rather spend an hour than buy a $1,300 costume.

嗯，你知道的，只要在Stripe多工作几年，你就能理解了。我觉得有个很有意思的点——作为开发者工具领域的人，我热爱协议设计，曾参与设计并辩论过多个协议。有哪些关键分歧点呢？比如有人强烈主张某个方案但最终被否决，或者至今仍悬而未决的问题。我先说一个例子，然后你可以补充：我提到过Solana和Circle都赞助过我的会议，他们也在尝试构建代理协议。

Well, you know, just work a few more years to Stripe, you'll get it. So I think one like, there's some, I think, interesting well, I love protocols, you know, as a developer tools person, I've been involved in designing a few of them, debating a few of them. What are the forks in the road that, you know, like someone else was like discussing something really strongly and we decided against it, or maybe it's still an open question. I'll give you one and then maybe you can volunteer another. So the I mentioned that both Solana and Circle have sponsored my conferences before, they're also trying to be build a protocol for agents.

实际上他们都为代理配备了钱包功能，而非支付代币。我认为让代理拥有独立银行账户是个有趣的选择。你们没有采用这种方案。这是个决策考量点吗？还是有其他更重要的因素？

And both of them actually give agents a wallet, right, as opposed to a payment token. And I think having an agent with their own bank account effectively is an interesting choice. You didn't go for that. So is that a decision factor or there a different one that you want to focus on?

让我分两点说明。关于商务协议，核心在于企业如何标准化地展示产品、库存、价格和品牌，供代理呈现给消费者或代消费者购买。这本质上涉及产品目录的表述方式和定价机制。当前版本只是最基础框架，未来会持续演进。

Let me parse two things. When I think of the commerce protocol, that's primarily around what's the standardized way that businesses expose their products and their inventory and their prices and their brands and make those available to agents to expose to the consumers and or to buy on behalf of the consumers. That really comes down to, like, how should a product catalog be expressed? How should prices be expressed? I think the current version is like the bare bones version, and it will continue to evolve.

比如说，你可以想象，从市场清算的角度来看，商家也应该作为其中的一部分，明确说明他们愿意给代理的提成比例。对吧？就像，嗯，一点点

For example, like, you could imagine, like, a market clearing perspective, that the merchant should also be, as part of that, articulating the cut that they're willing to give to the agent. Right? Like like a Yeah. Little bit

在经典案例中，比如对于人工代理，我会给他们一个预算，说明他们的谈判目标和最高支出限额。

In a in a classic, like, if the human agent, I give them a budget, like what their negotiation target is and what their max spend could be.

对，对，完全正确。所以在这种情况下，我会觉得，好吧，随便吧。我没买的那件波莱罗简直是糟糕透顶的推荐。

Yeah. Yeah. Exactly. And so in this case, I'd I'd be like, okay, well, whatever. The the bolero that I didn't buy is terrible recommendation.

但他们有很多好推荐，那个实在太差了。你知道，这东西在Etsy上卖1300美元，但我愿意给促成交易的代理x金额的佣金，可以在1300美元之上或之下。总之，我认为各种参数会不断演变。但基本要素是，你必须能确定性地向代理展示哪些信息，让他们了解产品库存、品牌呈现，有时还包括尺码数量等，这些都需要提供给代理或最初发起的人类。共享支付令牌则略有不同，它更像是，好吧，实际怎么完成交易呢？

But they have many good recommendations, but that was a terrible one. You know, this costs $1,300 on Etsy, but, you know, I'm willing to give x to any agent who facilitates the transaction, either on top of or underneath the the 1,300. So anyway, I think there's gonna be an evolution of, like, the various parameters that should be included. But, like, the the basic set was what do you have to be able to deterministically expose to an agent so that they understand what's available and what representation of the product and brand and, you know, sometimes its size and number and whatever, like, has to be made available to the agent and or the human that's initiating in the first place. The shared payment token is a little bit different, which is like, okay, How do you actually get the transaction done?

比如资金如何流动？即使在Stripe，这一直在演变。共享支付令牌是我们构建并推出的技术，已应用于ChatGPT即时结账的后台实现。但一年前Perplexity推出了旅游搜索预订代理，你们看到了吗？

Like, how does the money flow? And even at Stripe, like, that has been evolving. So shared payment token is is what we built and launched and have in the background of the instant checkout implementation with, ChatGPT. But a year ago, Perplexity launched a travel search and booking agent. Did you guys see this?

没错，那也是由Stripe支持的。那里的支付流程稍有不同。我们有个发卡产品可以生成虚拟卡，在那类流程中，代理会获得一张一次性虚拟卡代表你消费。

Yep. That is also powered by Stripe. And there, the payment flows are a little bit different. So we have an issuing product which allows you to issue virtual cards. And what happens, in that sort of flow is the agent gets issued a onetime use virtual card to spend on your behalf.

你知道，人们对此很紧张，但我想提醒他们：当我用DoorDash点Phil's咖啡时，DoorDash就会给司机发一张6美元的一次性虚拟卡代我支付——信不信由你。现在只是把人类代理换成了AI代理。就像我的DoorDash司机从未见过我的信用卡信息，消费不能超过6美元且必须在我家附近限定时间内使用一样，Perplexity旅游搜索预订代理中的AI代理也是如此。目前仍有代理通过Stripe的虚拟卡实现进行交易，各有优缺点。

And, you know, people get very get very jumpy about that, but I like to remind them that, like, when I order from DoorDash my Phil's coffee, right, DoorDash is issuing a one time use virtual card to the driver for $6, believe it or not, to spend on my behalf there. And so now you're just inserting, you know, AI agent instead of human agent. And in the same way that my DoorDash driver never saw my card credentials and couldn't spend, you know, more than $6 and had to spend it in a constrained time window near my house. Same thing for the AI agent in the Perplexity travel search and booking agent. And we still have agents doing commerce through Stripe using the virtual card implementation and their pros and cons.

所以我不认为所有东西都会变成虚拟卡。我认为都会变成共享支付代币，都会变成代理钱包。稳定币会是个有趣的方向，钱包整体来说也会是个有趣的方向。

So I don't think that, like, it's gonna be all virtual cards. It's gonna be all shared payment tokens. It's gonna be all agent wallets. I think stablecoins will be an interesting direction. I think wallets in general will be an interesting direction.

我认为存储余额会很有意思，特别是当我们讨论微交易时。对吧？我们讨论了很多关于购买商品的事情。商品价格通常足够高，值得用卡类交易方式。但如果你要购买AI服务、购买推理结果或内容，你需要能进行5.1美元、25.5美分这样的交易，而这些在信用卡世界里效率极低。

I think stored stored balances will be interesting, especially as we're talking about kind of micro transactions. Right? So we're talking a lot about buying, like, goods. Buying goods are usually priced high enough that it's worth sort of like a card transaction type approach. But if you're talking about buying AI or buying some inference or buying content, you wanna be able to make $5.10, $25.50 cent transactions, and those are hyper inefficient in the card world.

因此我认为代理间支付将推动我们走向下一个前沿。但ACP与共享支付代币的运作方式或资金流动是不同的。我认为一方面它会持续进化，因为市场需求和技术可能性都在发展；另一方面，它也不需要以同样的方式标准化。

So I think agent to agent payments are gonna push us a little bit to the next frontier here. But ACP, again, is like distinct from how the shared payment token works or how the money flows. And I think that will, a, continue to evolve just because the market needs are evolving and the technology possibilities are evolving. But b, will also like, also doesn't need to be standardized in the same way.

接收端呢？我想知道，我的代理能为我赚钱吗？

What about the receive side? I guess, can my agent make money for me?

你的代理能为你赚钱？哦，我在想相反的情况——我很乐意给代理3美元让它去为我做深度研究。所以我们正在想办法实现...

Can your agent make money for you? Oh, I was thinking the opposite, which is like, I'd be happy to give an agent, you know, dollars 3 to go out and do deep research for me. And so we're trying to figure out how to enable

哦，那是研究。不，我说的是支出方面。我认为在支出方面我们有个很好的心智模型。特别是因为我们也有人类代理，知道如何支出。赚钱显然是任何创始人的原始动力。

Oh, that's research. No, was like, know, like spending, I think like there, we have a good mental model of how to spend Yeah. Especially because we have human agents as well, how to spend, making money is is is obviously the the original draw of of of, strive for any any founder. Yeah. Yeah.

我只是有点好奇你在这方面的想法。

I'm just kinda curious what's what's your thinking there.

让您的MCP服务器轻松实现盈利。这是其中一方面。我们还看到越来越多新企业开始使用AI开发工具，如Replit和Vercel。因此我们希望在这些工具和工作流中也能轻松实现盈利功能，无需跳出当前流程去创建Stripe账户、进行认证等操作。几周前，我们发布了可认领沙盒环境。

Make it easy to monetize your MCP server. So that's, like, one thing. We also are seeing an increasing number of new businesses get going in AI dev tools, like Replit and Vercel. And so we wanna make it really easy to spin up monetization also within those tools and within that flow, not be taken out of flow and go and create a Stripe account, authenticate, and whatever. A couple weeks ago, we released claimable sandboxes.

你们有没有注意到我们已经在Vercel上线了这个功能？

Have you guys seen we've been in Vercel and seen it or

不，我知道你们有沙盒产品，但是

no. I I know you have a sandbox product, but

我不

I don't

了解。我们有一个沙盒产品，您可能通过Stripe的开发者体验接触过。现在这个沙盒产品可以被调用、使用和配置。您甚至可以在拥有Stripe账户前，就在沙盒中创建产品、设置价格、进行测试扣款和生成客户数据。或者您可能已有之前业务的Stripe账户，但尚未将其与当前业务关联。

know about So we have a sandbox product, which you've seen from, like, being within the developer experience on Stripe. Now that sandbox product can be, like, invoked, used, manipulated. You can create your products and your pricing and run test charges and generate customers and in that sandbox before you even have a Stripe account. Or maybe you have a Stripe account from your last business, but before you've linked it to this business. Right?

我们称之为可认领沙盒。实际上Vercel和Replit——记得五月份在Stripe Sessions活动上我和Replit讨论过我们的沙盒产品。他们向我解释人们如何尝试端到端构建业务，而支付集成环节是流程中最棘手的部分之一。于是我们开始探讨能否将沙盒环境直接集成到他们的平台。大约四个月后，他们就实现了这个功能并正式上线。

And and we call them claimable sandboxes, and Vercel and Replit were actually remember talking to Replit at Stripe sessions in May about our sandbox product. And they were explaining to me how, you know, people are trying to build businesses end to end, and, like, one of the wonky parts of the flow is setting up the payments integration and get going. And so we started talking about, like, okay, could you actually have, like, the sandbox environment there? And it was, whatever, four months later, and they had it. They launched it.

我们与他们以及Vercel共同推出了这个功能。实际上Guillermo前几天发了一篇很酷的帖子，提到我们的Stripe集成目前已成为他们vZero平台上使用量第三高的集成——而发布才两周时间。但本质上，这些用户都不是为了好玩才创建vZero的，他们不是为了建个网站玩玩，而是要创建真正的业务。

We launched it with them and also with Vercel. Actually, Guillermo had a cool post a couple days ago that our Stripe integration is now one of their top I think it's like the third highest integration that they're seeing in vZero, and we're only two weeks into the launch. But it's basically, like, all of these people aren't going to create vZero just for fun. They're to create something just a website or whatever just for fun. They're going to create a business.

因此我们让他们能够轻松完成商业支付。是的，这是后端部分。实际运作方式就是，你直接在vZero平台上，开设你的植物商店，创建产品，设定价格，进行测试收费，最后点击按钮。当你对所见满意时，可以随时返回迭代。点击按钮后，你的标签页会在Stripe打开，你可以登录现有账户或创建新账户，认领那个沙盒环境。

And so making it really easy for them to do the business payments Yeah. Back end part of that. And how it actually works is literally, like, you're right there in vZero, and you, you know, you have your plant shop, and you create your products, and you set your prices, and you run your test charges, and you click a button at the end. Like, when you like what you see, you can go back and iterate later. But you click a button, and your tab opens in Stripe, and you can either sign into the account you have or create a new account, and you claim that sandbox.

这个沙盒环境可以直接上线变成你的业务。每天都有许多人将其投入实际运营，我们见证了许多前所未有的新业务诞生。最有趣的是，当我浏览企业名单时——虽然不能现场点名——但其中相当比例是AI公司。如今创建的初创企业中有很大一部分都是AI公司。

And that sandbox, you can take it live and it becomes your business. And lots of people are taking it live every single day, and we're seeing new businesses get created that were never before. And one of the things that's really fun to see, was I going through the list of businesses. I probably can't name them live, but, like, a chunk of them are AI companies. Like, a chunk of the startups being created today are AI companies.

但也有相当比例是非技术背景的创始人，如果没有vZero平台的支持，他们可能连Stripe的基础操作都难以开展。

But a chunk of them are, like, nontechnical founders who may have actually struggled to, like, get going on Stripe had they not had it kind of within that v zero.

是的，低代码技术是强大的赋能工具。

Yeah. Low code is a huge enabler.

低代码确实是强大的赋能工具。我们在自身入驻流程中做了优化使其支持低代码，比如低代码订阅等功能。内部我们称新入驻系统为'未来入驻体验'，它会引导你建立目标商业模式并自动配置沙盒环境。但更酷的是，如果你在vZero平台完成部分操作后转入Stripe，未来入驻系统已经通过Vercel了解了你的意图和偏好。

Low code is a huge enabler. And, you know, we've we've done some good work in our own onboarding experience to make it low code, and we have, you know, low code subscription and whatever. We have our our new onboarding experience internally. We call it future onboarding experience, and it kind of walks you through what's the business model you're trying to create and then sort of stands up the sandboxes for you. But what's cool is now you do a bunch of that in, say, v zero if you're in v zero, and then you come over, And and your future onboarding experience has already learned, like, your intent and your preferences and all of that from Vercel.

这样你进入系统时，沙盒环境已经完成了x%的初始化工作。

And so you're dropped in x percent of the way through with the sandbox already spun up.

那么外界对Stripe运用AI的认知是这样的。那Insight方面呢？你提到3.5版本是你们开始重视的转折点，最初内部有哪些应用场景？现在Stripe又是如何运用AI的？

So this is like from the outside how people perceive AI as Stripe. What about Insight? So you mentioned 3.5 was kind of like the moment you took it seriously. What were the first internal use cases? And then how do you use AI as Stripe today?

是的。最初的内部用例就是自下而上的实验性探索，对吧？所以我们创建了GoLLM，它本质上就是个类似ChatGPT的界面，可以对接多种模型。最早的版本其实算不上是能构建生产级系统的LLM代理，纯粹就是个ChatGPT类工具。

Yeah. First internal use cases were, you know, bottoms up experimentation, right? So we created, we call it GoLLM, but it's like just a chat GPT like interface where you can engage with a bunch of different models. It was the very, very first version actually wasn't like an LM proxy where you could build production grade systems. It was literally just like ChatGPT like stuff.

后来我们开发了预设功能，主要是用于提示词保存和共享。这样用户就能分享模板——比如'这是我筛选目标客户并生成触达话术的方法'，或者'用Stripe风格重写营销内容'之类的。一夜之间就涌现出数百个预设模板，因为大家都玩疯了。随后我们推出了LLM代理，为工程师提供生产级LLM访问权限，早期很多应用场景都围绕商家分析展开。

And then we had this, like, preset feature, which was, like, prompt saving and sharing. And so you could share your temp oh, you know, this is how I figure out what customers to reach out to and generate reach outs or, you know, rewrite my marketing content in Stripe Tone or whatever. And you had sort of hundreds of presets that came on, like, overnight because everyone was into it. And then we generated so then LLM Proxy was like, okay. Now production grade access for engineers to these LLMs, and a lot of the early use cases there were actually around merchant understanding.

刚才提到过，每天有数千家商户入驻Stripe，我们需要了解他们是谁、卖什么、是否符合卡组织规范、信用如何、是否存在欺诈等等。LLM在这些环节大有可为，这些都是我们早期的应用场景。快进到现在——我今早看数据面板时正在规划2026年的LLM成本——每天有8,500个Stripe用户在使用我们的LLM工具。

So I mentioned a little bit ago, but we have, like, thousands of merchants that come on to Stripe every day, and we have to understand who are they, what are they selling, is it supportable through the card networks, like, they credit worthy, are they fraudulent, and there's a lot that LLMs can do there. So those are some of our earlier earlier use cases. Fast forward to today, I mean, you know, I I actually I was looking at the dashboard earlier because I was planning for for 2026 and some of our LLM costs. 8,500 stripes a day use our LLM based tools. Okay.

要知道Stripe总共才约1万用户，这基本上等于全员使用了。而且大家的创新应用越来越丰富——上周我刚和本地支付方式（LPM）团队交流过。

There's, like, only 10,000 stripes. Like, not everyone is in every day. Like, it's basically it's basically everyone. And, you know, I think people are getting, like, pretty creative in in the applications. I was talking last week to the LPM team, so local payment methods.

虽然你我可能更关注信用卡支付，但本地支付方式至关重要，因为Stripe商户基本都在做跨境销售。在海外市场时，像日本的JiroPay这类...

You know, you and I think a lot about cards or whatever, but local payment methods matter because the businesses on Stripe are almost always selling internationally. And when you're in other countries, having local payment methods, you know, JiroPay if you're in

对，我来自东南亚。

Yeah. I'm from Southeast Asia.

确实各地都有。等等，你最喜欢哪种？

Yeah. It's all over. Wait. What's your favorite?

嗯，我知道。我是说，大概有GrabPay之类的吧。我也不确定。

Well, I know. I mean, there's like GrabPay, I guess. I don't know.

没错。确实如此。比如，如果你看不到本地支付方式，只看到信用卡支付，可能你根本没有信用卡。或者如果只看到信用卡支付，你会觉得这个服务不是为你本地化设计的。而当你看到符合当地市场的支付方式时，会感觉更有归属感。

Yeah. Exactly. And like, if you don't see a local payment method, like, if you only see cards, you might not have cards. Or if you only see cards you might not feel like it's, like, localized to you or meant for you. Whereas if you see, like, in market regional payment methods, you feel much more connected.

从统计数据来看，转化率会高得多。而且很多时候，手续费对商家也更合理。所以我们投入了大量资源整合本地支付方式。大多数Stripe商户都使用我们的优化结账套件，该套件开箱即用支持100多种支付方式。但我们最常收到的需求就是希望增加x支付、y支付、z支付，因为我还想拓展到h k国家市场。

You're much more likely to convert statistically, and oftentimes, the fees also, make more sense, for for the merchant. So we've invested a bunch in integrating local payment methods. Most businesses on Stripe use our Optimize Checkout Suite. Our Optimize Checkout Suite comes with over a 100 payment methods out of the box. But one of the most requested features we get is payment method x, payment method y, payment method z, because I also wanna be in country h k.

那么问题来了——真正的挑战是什么？挑战在于每接入一个新支付方式，都需要两三个月工程师工时。虽然不算世界末日，但Stripe是家精干的公司，我们还有很多任务。所以对于边际支付方式，到底值不值得投入？

And so, you know, what is the the challenge? The challenge is integrating with any new payment method, and it takes, like, two months for a couple engineers. It's not the end of the world, but Stripe's a pretty lean company, and we got a lot of stuff to do. And so, you know, for the marginal payment method, is it worth it? Yes or no?

退一步想，这本质上是在做什么？其实就是研究Stripe的代码库，理解本地支付模块(LPM)的运作机制，按照LPM集成指南把两者对接起来。LPM团队第一次花了两周，但他们最近用大语言模型(LLM)仅用两周就成功上线了泛欧支付方式。我估计再过一个月，这个时间能缩短到一两天。

Well, when you step back, what are you doing? You're really, like, looking at Stripe's code base. You're looking at, like, how the LPM works and, like, the integration guide for the LPM, and you're kinda, like, hooking the two up. And so the LPM team, it took them two weeks for the first one, but they just launched a new pan European payment method in two weeks using an LLM to, like, build that integration. And I think they'll probably, you know, have it down to a day or two within within a month.

这是个很好的例子——本质上应该是机器对机器的对话，而且双方都有完善的文档支持。LLM能完成大部分工作。我们还大量使用AI编程助手，大约65%-70%的工程师日常都在用。不过具体成效我还很难量化。

And so that's just a good example of, like, it's kinda should be just like a machine talking to a machine, and there's pretty good documentation on both sides of the house. And so the LLM can make it can make it pretty far. We also use a lot of AI coding assistants. And, yeah, I would say, like, 65, 70% of engineers use them on the day to day. I have a really hard time understanding impact.

比如，我不太清楚该看哪些统计数据。实际上我...

Like, I don't quite know what statistic to look at. I I don't actually

我认为不是代码行数的问题。

think It's not lines of code.

我不认为是代码行数的问题，因为...是的。上周我收到了三份不同的文档，我知道这些文档至少部分是由大语言模型生成的。是文档，不是代码。在每种情况下，我都回去找当事人说，其实我只想看看你输入到ChatGPT里的要点，而不是那八页文档，因为我很难理解那八页的内容。听起来不错，但我不确定它是否与现实有联系。

It's I don't think it's lines of code because Yeah. I have in the last week been sent three different documents that I know were, like, at least partially written by an LLM. Documents, not code. And in all cases, I went back to the individual, and I said, like, I actually just wanna see the bullets that you put in chat GPT or whatever, not the not the eight page document because I have a really hard time reasoning about the eight page document. And sounds good, but I'm not quite sure it's, like, connected to reality.

我对代码行数也有同样的感觉。就像我不想要更多页的文档一样，我也不想要更多的代码行数。是的。所以我们正在关注这一点。

And I feel the same way about lines of code. Like, I don't really want more lines of code just like I don't want more pages of docs. Yeah. Yeah. So we're watching that.

另外，许多这类编码模型的成本实际上相当可观。嗯。因此在我们规划明年工作时，我们一直在思考如何在这方面提高效率——显然这些工具很有价值，我们当然希望人们使用AI编码工具。但要确保我们能为业务带来正确的回报。其中一部分是控制成本，另一部分是更清晰地评估影响和价值。

And then also the cost of a lot of these coding models is actually, like, pretty nontrivial. Mhmm. And so as we're planning forward to next year, we're reasoning a bunch about like, where can we get somewhat more efficiency there given like, obviously, it's valuable and we want people to be using AI coding tools for sure. But we wanna make sure that that we're getting the right returns for the business. And some of that is managing costs and some of that is getting a clearer read on impact and value.

你觉得社会契约正在如何改变？就像刚才说的，直接发我要点就行。对吧？在大语言模型出现前我也可以只发要点，但你们非要我写这份备忘录。对吧？

How do you feel the social social contract is changing? Like, mentioned, just send me the bullet points. Right? It's like, I could have sent you the bullet points before LLMs, but you are making me write this memo. Right?

我感觉在很多组织里存在一种表演性质的行为。部分就像在重要会议上穿西装一样，这是一种方式，像是说'嘿，我在认真做事'

I feel like in a lot of organizations, there's like a performative. And part of it is like, you know, wearing a suit to an important meeting. It's like, a way, like, hey. Doing

穿西装啊。我听说有人确实这么做。

it. Wear suits. I hear some people do.

是的。你知道，我这么做是为了表示对你的尊重。

Yeah. You know, I'm doing it to show you respect.

对。对。

Yeah. Yeah.

没错。同样的道理，我本可以直接发给你这个要点，所以我们这次会议可以穿短裤进行了。

Yeah. In the same way, I could have just sent you this bullet point, so we're gonna have this meeting in shorts.

对。对。对。你

Yeah. Yeah. Yeah. Do you

觉得现在有了AI，就像是，好吧。如果你要用AI处理，直接发我要点就行，我们某种程度上正在突破，像是

feel like with AI now, it's like, okay. If you're gonna do it with AI, just send me the bullet points, and we're kinda, like, breaking through in a way of, like

这真的很有意思。好的。我之前没想过这个，但这是我的工作假设。告诉我是否合理。我关心的是该领域的专家，他们确实是该领域的专家。

So that's really interesting. Okay. So I hadn't thought about this before, but here's my working hypothesis. Tell me if it tracks. What I care about is that the expert in the area like, they're an expert in the area.

否则他们就不会发文档给我了，对吧？该领域的专家已经深思熟虑过。那么写作是什么？真正的写作。

Otherwise, they wouldn't be sending me a doc. Right? The expert in the area has thought deeply. And what is writing? Like, actual writing.

打字。随便吧。这不重要。但不用大语言模型写作会迫使你深入思考。

Typing. Whatever. It doesn't matter. But, like, writing not with an LLM force you to do. It forces you to think deeply.

它迫使你构建推理逻辑。我不知道你们怎么样，但我每次写文档时都会反复阅读50遍，思考逻辑是否自洽？有没有遗漏的陷阱？思路是否正确？别人会怎么看待这个？

It forces you to structure your reasoning. I don't know about you guys, but when I read a doc when I write a doc, I've, like, read the doc, like, 50 times and thought about, like, does this logic track? Are there gotchas I'm not considering? Like, is that the right train of thought? Like, how might somebody else look at this?

并不是说我把文档写得华而不实而要点本可以更好，而是这种细致、深思熟虑、艰苦耗时的文档构建过程，迫使我从第一性原理出发进行合理推理。而大语言模型恰恰相反——随便扔几个要点，无需追溯原理，听起来不错大家就喜欢，我认为这极其危险。

And it's not like that I wrote the doc to be performative and the bullets would have been better. It's that the careful, thoughtful, arduous, time consuming construction of the doc forced me to appropriately reason from first principles. And I think LLMs do the opposite. Like, oh, you just throw in the bullets. You don't have to raise reverse principles, and it sounds good and people like but I think that's extremely dangerous.

我觉得确实如此，但感觉你依然没有用AI生成文档。因为你属于那种把写作当作思考过程的人

I think that is true, but I feel like you still are not generating documents with AI. So because you are the type of person that uses the writing as the thinking

是的。

Yeah.

你依然会经历这个思考过程

You still go through the process

没错。

Yeah.

与那些使用AI的人相比，他们本来就不会在撰写长篇文档上投入太多思考。对我来说，这才是关键。同样地，我们昨天在另一次访谈中讨论过代码问题，那就像是光标和这些工具带来的老虎机效应。但归根结底，你必须合并PR。所以你得想出对业务有意义的方案。

Versus the people that use the AI still wouldn't have put that much thought into writing the long document anyway. I think to me, that's really the thing. Same way, we were talking about this for code yesterday in another interview, which was like the slot machine effect of cursor and these tools. But at the end of the day, you gotta merge the PR. So you gotta come up with something that makes sense for the business.

没错。这类文档其实也是同样的道理，对吧？重点不在于形式，你只需要设计出正确的ACP方案。我不在乎它是10个要点还是10页纸。

Yep. Like with these documents, it's kind of the same. Right? It doesn't like, you just need to come up with the right ACP design. I don't care if it's 10 bullet points or like 10 pages.

是的。

Yes.

在我看来，现在情况正在变化，因为人们更常阅读摘要了。这就好比——既然你会总结我的内容，那我何必写长篇大论？直接写简短的就行了。我没有标准答案，只是觉得这种现象很有趣，就像你直接说'给我要点就行'。

To me, I think things are changing now because also people read more summaries. So it's like, well, if you summarize my thing, then why should I write a long thing? I should just write a short one. I I don't have an answer. Just interesting to see how, you know, you're like, just send me the bullet points.

对。我最主要的是...虽然我在意很多方面，但有个绝对底线：如果内容生成用到了大语言模型，请注明来源。因为我最反感的情况是，当我认真读了两页以为是人写的文档后，突然发现烦人的空格+双破折号+空格格式...

Yeah. The primary thing that I well, there's many things I care about, but like a very concrete nonnegotiable is if an LLM was used in the generation of this content, please cite the LLM. Because my least favorite thing is to be, like, two pages into what I think is a thoughtful doc and find the annoying, like, space, double dash, space, and then

作为双破折号的使用者，我觉得...我的意思是，早在LLM毁掉这种格式前我就这么写了。

As a double dash guy, I feel like I mean, I was writing it before LLMs ruined it.

不，让我炸毛的不是那个。只有这件事会让我炸毛。但你懂我意思吧？我确实认为我们都需要对LLM保持警惕。

No. That's not what tips me out. That's the only thing that tips me out. But you you know what I mean? Like, I I I do think we all need to be careful about LLMs.

此外，我认为这里还存在一个社会行为层面的问题——我们无法关闭自己的大脑。我的意思是，实际上我在思考：大语言模型是否让深度思考能力在当今世界变得更加重要？这种能力包括质疑、指导机器该做什么，而非亲自执行任务，因为语言模型本身就能完成任务。如果你看待语言模型时会想‘哦，这意味着我不需要深度思考了，因为它们会替我完成’，这种想法很自然，因为它们确实能产生诱人的输出。

And then I think there's like a societal behavioral thing here too, which is like, we can't turn our brains off. I mean, I actually think that, like, do LLMs make all the more important in the world? The ability to think and reason deeply. To, like, question, to, like, tell the machine what to do more so than like to do and execute the task because the LM can do and execute the task. And so if you're looking at LMs and you're like, oh, that means I don't have to think deeply because they're just gonna do it for me, which is very natural because they do produce enticing output.

我只是觉得这对社会来说存在风险。我们已经看到，在社交媒体环境下成长的人普遍存在注意力问题——我指的不是博取关注的注意力，而是保持任务专注度的能力。同样地，在工作中过度依赖大语言模型的人，可能会在深度思考方面投入不足。

I just think it's like I think it's like risky for for society. I mean, I think we've seen how, like, people who grew up on social media have, a lot of issues with attention. I think people who I don't mean attention trying to get attention. I mean attention, like staying focused on a task. I think in the same way, people who overgrow up in their work life on LLMs risk under investing in-depth.

我认为在当下这个充满大语言模型的世界里，这种情况尤其危险——虽然表面可能看不出来，但深度思考才是最重要的。

And I think that's particularly dangerous in a world where with LLMs actually, it may not appear this way in the moment, but, like, depth is the most important thing.

我稍微有点不同意见，可能比你们更倾向于接纳‘思维惰性’。

I would push back a little bit in terms of I think I'm maybe a bit more sloth friendly than you guys.

是吗？

Are you?

因为无论是人类还是AI产生的草率成果，关键在于当我提交文档或发送PR时，我是在为整个内容背书。我不能把责任推卸给大语言模型。

Be just because, like, slob comes from humans and slob comes from AI. Yep. What just matters is, like, when you sign off, when I send you the document or when I send the PR, I am signing off on the whole thing. I I can't abdicate responsibility to the LLM. Yep.

也许大语言模型输出了好内容，也许没有，但最终判断者是我。我是编辑，对吧？所以...

Maybe LLM had good output. Maybe not, but I'm the final judge. I'm the editor. Right? And so, like

其实我觉得我们在这点上意见一致。所以我完全支持在生成过程中使用它。实际上，这真的很酷。它是思维工具，确实是思维工具。

I actually think we're on the same page there. So I am, like, all for using it in the generative process. Actually, it was really cool. It's a tool for thought. It's a tool for thought.

没错。而且它是快速实验和快速迭代的工具。我喜欢看演示或原型，不想看文档，就想看你用任何工具快速做出的东西。实际上，当我们开发可申领沙盒时，永远记得Vercel直接发给我们他们设想实现方式的v0版本。

Yeah. And it's a tool for rapid experimentation and rapid iteration. And I love to look at, like, a demo or a prototype that, like, I don't wanna see a doc on it. I wanna see, like, the quick thing that you spun in whatever tool you use. Actually, when we were working on claimable sandboxes, never forget Vercel sent us basically like the v zero of, like, how they thought it should be implemented.

就是那个UI界面。他们认为用户体验应该是这样的。这比数小时的会议和成堆的设计文档要清晰得多。整天都在搞生成式开发，但在推送PR或发布前，你需要深度认可并盖章批准。

Just like the UI. Like, this is what we think the experience should look like. And it was extremely clarifying, like, much more clarifying than hours of meetings and pages of design docs. And so all day long on the generative, but like you need to like deeply put your stamp of approval before you push the PR, before you publish.

我想深入探讨一下RAG和写代码这两个主要用例。关于内部信息，显然有我们昨天讨论的Glean，还有其他内部代码搜索工具，大家也用Notion。RAG还有意义吗？还在积极开发吗？未来发展方向是什么？

I did wanna double click a little bit on both, you know, the like two two primary use cases on RAG and writing code. Just, I guess on like internal information, there's obviously Glean, which we talked about yesterday and just all the other internal code search tools, guys use Notion as well. Is Rag still relevant? Is that something that's like in active development or what's what's beyond that? What's like So the

我们实际上在大力投入一个叫工具棚的东西，它就像是内部的MCP服务器，可以访问所有Stripe工具。

we've actually been leaning in really hard on, we call it tool shed, but it's like a it's like an internal, like MCP server that basically has access to like all the Stripe tools.

嗯。

Yeah.

我喜欢的是它由中央管理。我们后来停用了之前提到的GoLLM，做了新的开源实现，类似LibriChat的方案，很棒。它能连接所有相同的工具棚MCP服务器，包含你能想到的一切：Slack、Google Drive、Git，还能访问Hubble查看数据目录和所有数据，并能查询数据。我觉得这非常强大。

And, you know, what I like about that is, like and it's managed centrally. Like, you know, there's you know, it's managed centrally and it plugs into we've since killed that GoLLM thing I talked about, and we did a new, like, you know, implementation, like, open source, like, LibriChat situation, which is great. But it, like, hooks up to all those same toolshed, like, you know, MCP servers. And it's got, you know, I mean, everything you would think, like Slack and Google Drive and Git, but also, like, access to Hubble so it can see, like, our data catalog and all the data, and it can query the data and and and. And I think that's been really powerful.

我不认为RAG已死。我认为关键不仅在于这些工具中可获得的信息，还在于能够与所有这些工具交互。对吧？这就像是工具调用。

I don't think RAG is dead. I do think there's an important name of the game around it's not just, like, the information that's available in all those tools. It's also being able to interact with all those tools. Right? It's like the tool calling.

因此我认为它们是共存的，它们确实可以共同存在。虽然Toolshed是集中管理的，但任何人都可以...比如Salesforce团队就能添加Salesforce功能...因为我们不希望因为某个中央团队的阻碍而无法将这些工具暴露给LLM和Stripe的智能体。

And so I think they coexist, and I think they I think they coexist together. Also, while Toolshed is owned centrally, anybody can so you can you know, the the Salesforce team can add Salesforce and and and because we don't wanna be blocked on some central team in order to have those tools exposed to to the LLMs and to the agents for Stripes.

是的，你们需要适当去中心化。从代码层面来说，显然这更贴近我的领域。我觉得这就像是个经济学难题——如何衡量生产力。

Yeah. You wanna decentralize a bit. And then code wise, you know, coming on the code side, obviously, closer to home for me. I think it's also like an economist problem of measuring productivity.

确实。好吧，你现在让我因为没解决这个问题而感到内疚了。

It is. Okay. Well, now you're just making me feel guilty for not having cracked it.

不，不。我是说这是个全球范围内都未解决的难题。

No. No. I mean, it's it's unsolved globally.

所以很难。非常难。

So It's hard. It's hard.

是啊，问题就在这儿。当你核算成本时会发现，哇，这相当高。我在想或许我们可以转向开源模型之类的方案。

Yeah. And that's the thing. Like, you you you're looking at the cost and you're like, oh, it's pretty high. I don't know. Maybe we'd like move to open source models or something.

但是，你并不知道你在中期获得了多少生产力提升，而你雇佣的工程师又相当昂贵。这真的很难衡量。

But, like, you don't know the productivity gains you're off you're you're getting in in in interim and you have pretty expensive engineers. Like, it's it's hard to tell.

工程师确实昂贵。但和我们所有人一样，工程师们也极度渴望做出他们职业生涯中最出色的工作。所以关键在于，如果人们真正需要某样东西，提供它就具有内在价值。当人们拥有想要的工具，学习想学的知识时，他们会更努力，更有创造力。

Engineers are expensive. Engineers also, like all of us, right, are hyper motivated to do the best work of their lives. And so there's an important component of, like, if the people want it, like, there's inherent value in providing it, right? Like, when people have the tools that they want, are learning the things that they want, like, they work harder. They're more creative.

他们能产出更优质的成果。总之，除了实际产出之外，我认为这些软性的、边缘性的因素都带来了额外价值。还有就是学习曲线的问题。我们经常思考这个问题，比如当你推出一个新的传统机器学习模型，或者现在的人工智能解决方案时。

They produce better output. So, anyway, there's all this, like, sort of soft, fringy stuff that I think is added value above and beyond the actual, you know, production output. And then there's also, like, the learning curve. Right? So, I mean, we think about this a lot, like, when you launch a new traditional ML model or now, like, AI solution, right?

如果你过分关注短期窗口内的结果，很可能会得到错误的负面判断。因为系统还不够完善，参数还没调优，反馈循环尚未建立。它需要时间来获取训练数据以变得更好。

Like, if you overfocus on the results in the immediate short term window, you really risk getting a false negative. Like, it's not good enough yet. It's not tuned yet. It doesn't have the feedback loop yet. You know, hasn't had time to get the training data to get better.

我认为这在AI领域尤其明显。比如使用大语言模型时，GPT-4可能不奏效，但换成GPT-5突然就行了。或者我们使用GPT-4时，其成本可能还难以完全替代特定风险任务中的人力。但转眼间GPT-4o问世，每年能为企业节省300万美元——既因为模型成本降低，更重要的是它能替代更多人力。所以在考虑AI工具的最佳采用策略时，关注当年投资回报率似乎是错误的，应该着眼于两到三年的回报周期——尽管准确预测两三年后的情况本身就很难。

And I think that's, like, kind of particularly true in, you know, AI because you when you're working with LLMs, it's like, well, with, like, GPT four, like, it didn't work, and then you swap in GPT five, and all of a sudden, it does. Or we use, like, a GPT four o, and it was, like, kind of a little bit expensive to justify the humans that it was replacing for a particular risk related task. But then next thing we know, like, o three minutei is out, and it's, like, you know, dollars 3,000,000 a year savings for the business, both because the model is less expensive, but also more importantly, because it replaces more of the humans. And so I I do think that, like, when I think about the optimal adoption of these AI tools, it seems wrong to focus on in year ROI, and it seems right to focus on two year, three year ROI. Now inherently hard to know what two or three years is gonna look like.

但纵观历史，模型性能提升越来越快，成本下降也越来越迅速。这让我总体持乐观态度：我们不应该过分纠结于当年的回报率。

But if we look at the history, models getting much better, much more quickly. Models getting quite a bit cheaper quite quickly. It overall makes me bullish that we shouldn't over obsess around around in year returns.

确实。'当年回报率'这个说法很好，我之前从没这么想过。

Yeah. In year returns. That's a good term I never thought about.

当

When the

经济是这样摊销的。

economy is Amortizing it like that.

数据呢？

What about data?

等等，数据怎么样？

Wait, what about data?

是啊。你知道，就像，工程师的生产力是提高了还是没提高？比如，文本转SQL之类的，对吧？

Yeah. You know, it's like, oh, are the engineers more productive or not? Like, what about yeah. Like, I mean, text to SQL. Right?

这有点像第一代产品。是啊。就像，生产力如何，我是说，你现在可以生成任何图表了，对吧？

It's kinda like the first iteration of this. Yeah. Like, what's the productivity like on, like I mean, you can generate any chart Yes. Now. Right?

但是，这并不意味着

But, like, doesn't mean

是的，没错。我们有个叫Hubert的存在，他其实不是真人，而是一个AI，但名字听起来像个人。这个Hubert本质上是个用自然语言回答业务问题的系统。

it's Yeah. Yeah. So we have this he's not really a guy. He's an AI, but his name's Hubert, which sounds like a guy. So we have this guy called Hubert, which basically is like natural language to ask questions about the business.

另外，我们还有个Sigma助手。用户可以在Stripe数据基础上查询业务信息。这是个更受限的场景，因为Stripe数据就是你的营收数据，结构非常规范，文档也很完善。

By way, we have a Sigma assistant. So like our users can query stuff about their business on top of Stripe data. That's a much more constrained problem because your Stripe data is like your revenue data. It's, like, very well structured. It's very well documented.

这个功能在仪表盘、Sigma和Stripe数据管道里都能用，可以对接任何系统。虽然这个文本转SQL体验不会百分百完美，但已经相当可靠了。而且当你用自然语言描述需求时，我们不仅会生成查询语句，还会用自然语言解释这个查询的作用。

It's available in the dashboard and in Sigma and in Stripe data pipeline. You can hook it up to whatever. And so, you know, a text to SQL experience sitting on top of that. Like, it's not gonna be perfect, but, like, it's pretty airtight. And by the way, like, if you use natural language to describe what you want, we write the query, and then we also tell you in natural language what the query does.

所以即使非技术人员也能验证结果。现在想象一下，Stripe有海量数据表，其数据模型和商业模式都存在大量细微差异。

So even if you're nontechnical, you can validate it. Okay. Now imagine there's a lot of tables at Stripe. There's a lot of Nuance in Stripe data. There's a lot of Nuance in Stripe's business model.

Hubert是这款Hubble工具的负责人，你们可以用它在内部查找、探索和查询Stripe数据。目前它还处于早期阶段，每周约有900人使用。我们特意将主要用户群体聚焦在了解该领域的技术人员身上，原因正如你之前提到的——它可能会给出根本性错误的答案，而技术人员更有能力进行验证并提供反馈。在审阅Hubert评估报告时，我发现最有趣的一点是：它在数据发现环节表现最差。

Hubert is the guy that sits on top of this Hubble tool, which you use to find, explore, and query Stripe data that does that internally. And it's early. I mean, we have, like, 900 people who use it a week. We have tried to focus the people who use it mostly on technical folks who know the domain for exactly the reason you were citing earlier, which is it could get the answer fundamentally wrong, and technical folks are gonna be better positioned to validate and provide feedback. One of the most interesting things to me as I was going through the Hubert evals was the place where Hubert did the worst was around data discovery.

也就是说，它很难确定用哪个表和哪个字段来回答当前问题最合适。就我个人使用体验而言，当我知道具体表名时，现在会直接在GareChat界面里指明表名。但更重要的是，我们正在大力推动淘汰低质量数据表，并要求所有者完善高质量数据表的文档。我目前还不确定是否该让大语言模型负责文档工作，所以对于核心数据集，我们暂时还是依靠熟悉该领域的人力来硬性处理。

That is to say, it had a hard time finding which table and which field was best to answer the question at hand. And, I mean, personally, as a user, when I know the table, I actually now just articulate the table in the in the GareChat interface. But more importantly, we're doing a big push right now to deprecate low quality tables and have the owners document high quality tables. I haven't yet figured out if I trust an LLM to do the documentation. So for, like, the canonical datasets, we're kind of brute forcing it with with humans who who know the domain.

我们正在探索但尚未落地的另一个方向是：离线状态下，如果告诉Hubert我在组织中的具体位置，它的表现会明显提升——因为它能知道'我对Lpm数据感兴趣'或'我关注链接数据'，优化结账套件团队的人员通常会查询这些表、查看这些字段、提出这类问题、关注这些指标。虽然这个功能还未上线，但我认为存在一个有趣的问题：我们既可以让人类进行提示词工程或文档编写，也可以直接给大语言模型更多历史上下文——比如'像我这样的人通常喜欢做什么'。这就是我们的下一步计划，我对此非常看好。

The other thing that we are exploring, but we haven't landed yet, is offline, it looks like there's actually really Hubert does much better if you tell it where in the organization I sit because it knows, oh, I'm interested in Lpm data, or I'm interested in link data, I'm interested in OCS data. People who work on the optimized checkout suite tend to query these tables, look at these fields, ask these kinds of questions, look at these types of metrics. Now that's not in production, but I I think there's this interesting question of, like, we can have humans do some, like, prompt engineering or documentation or whatever, but we can also give the LLM more just historic context about what people like me liked to do, basically. So that's that's the next step there. I'm bullish on it.

我不确定是需要两个月还是两个季度才能让所有人都用上。我觉得可能需要一些时间才能足够确信我们不会犯重大错误。不过，对于结构良好、文档完善的数据库，文本转SQL确实很容易。但问题是大多数数据都没有完善的文档。

I think I don't know if it's going to be two months or two quarters before, like, everyone's on it. Like, I think it might take some time to get high enough conviction that we're not going to have an important wrong answer. Text to SQL is really easy, though, with really well structured, well documented. Yeah. It's just most data is not well documented

而且结构也不规范。其实在此之前，我就在数据工程行业工作。我们讨论过DBT五大趋势，你说你没有看法，但我一直认为数据发现是数据目录从业者和语义层从业者的关键胜利或终极胜利。你赞同数据界的这些动向吗？你对现代数据栈有什么调整建议？

and well structured. So immediately before this, I was actually in the data engineering industry. We talked about DBT five trend and you said you didn't have an opinion, but I always thought that data discovery is the important victory or, like, the the the ultimate victory of data catalog people and semantic layer people. Do you agree with those movements in the data world? Do you have any tweaks on the modern data stack that So you

我们正越来越多地转向语义事件基础设施。我认为近实时、高质量、文档完善的数据价值即将飙升，因为我敢肯定九个月后，没人会愿意去查看静态仪表盘并点击操作。他们会希望直接获取洞察，或者让他们的AI代理获取洞察，并且能够实时获取高质量数据。对我们而言，最重要的两个领域是支付和基于用量的计费系统——后者需要非常非常实时，以支持我们讨论过的各种AI商业模式。因此，我们的路径是通过语义事件，在仪表盘中提供近实时的规范数据集（是的，毕竟还有人会用仪表盘）。

we are increasingly moving to, like, semantic events infrastructure. I think the value of near real time, high quality, well documented data is about to skyrocket because I'm pretty sure that nine months from now, no one is gonna wanna go and, like, look at a even, like, static dashboard and click around. They're gonna wanna be fed insights or they're gonna want their agents to be fed insights, and they're gonna wanna be able to just, like, pull real time high quality data. For us, what that looks like, like, the the two most important domains for our users in that regard are payments and this, like, usage based billing, which needs to be very, very real time for all sort of like the AI business model stuff we talked about. And so there, our our path is like semantic events, canonical data sets available in near real time in dashboard, yes, because some people will still use dashboard.

在Sigma中可查询，但也需要类似Stripe数据管道的架构。因为很少有人会孤立地查看Stripe数据，他们希望看到Stripe数据与其他系统的关联。

In Sigma, so, like, queryable, but also in sort of a a Stripe data pipeline type. Right? Because very few people wanna look at Stripe data in isolation. They wanna look at Stripe data connected to the other stuff. Right?

你可以想象直接导入BigQuery之类的平台——用户需要看到数据间的关联。说实话，历史上这些产品的数据源并不统一，这也造成了混乱。所以我们将在未来六个月内对计费和支付流程进行架构重构，之后再逐步扩展。

So, I mean, you could just even imagine, like, pulling into BigQuery or whatever. They wanna see it connected to other stuff. And, you know, historically, honestly, like, it wasn't all the same data feed for all of those products, and that also creates confusion. So we're doing a bit of a of a rearchitecture for that flow starting in the next six months, which is billing and payments. And then I I think we'll I think we'll expand from there.

总会有些数据因各种原因不符合你的‘北极星架构’——我习惯这么称呼理想数据架构。真希望有人能解决这些历史遗留的混乱数据问题。

There's always gonna be a bunch of data that for whatever reason doesn't fit your I like to call it a north star architecture, but like your north star data architecture. Right? And I wish that someone could figure out how to make sense of the old bad data.

嗯。

Mhmm.

所以你不需要完全重构所有东西然后抛弃旧的，对吧？就像，你总会保留一些东西。比如stripe.com网站，在你创建账户之前就已经存在，那是完全不同的数据类型。但我该怎么理解这种情况呢？

So you don't have to, like, rearchitect everything and throw away the old. Right? Like, you'll always have, like, okay, you know, like, there's like the stripe.com website, which happens, like, you know, before you even create an account. It's very different type of data. But like, you know, how should I reason about that?

我该如何管理这种情况？我觉得传统企业经常要处理这类问题。

How should I manage that? And I I think traditional enterprises deal with this a lot.

你是说把网站分析数据转化为注册用户之类的？对，这个问题我想了很多。哦，是吗？这有点像指纹识别技术，虽然让人有点不舒服，但确实可行。

Like converting website analytics to signed up users and all that? Yep. Oh, I've thought about that so much. Oh, yeah? You just need like it's kind of fingerprinting, which is like something that people are kind of uncomfortable with, but you can.

你可以做到的。

You can do it.

是啊，没错。

Yeah. Yeah.

是的。我们有几个问题想问，比如关于自建还是采购的问题。你们开发了很多内部工具这很棒，但市面上也有很多优秀工具。你们如何权衡？显然你们掌握独特的内部信息，但也会与外部供应商合作。我想人们既想知道如何与你们合作，也想了解同行公司如何做这类决策。

Yeah. We we have like a couple questions on like, I think there's a build versus buy question on you're building a lot of internal tooling and that's great, but also there's a lot of great tooling out there. How do you navigate this? Obviously you have a lot of unique internal context, but you also have, work with external vendors. Like people, I guess, want to know how to work with you, but also people in your shoes at peer companies also wanna know how you do this decision.

是的。对我们来说这不是非此即彼的选择，在自建还是采购问题上我们更倾向于'两者兼顾'。而且有些选择是分阶段的。就像我们之前聊过的，当GPT3.5刚出现时，我们觉得Stripe每个员工都应该能用上大语言模型，但当时找不到符合企业级安全标准的供应商，所以我们就自己开发了。

Yeah. I think for us, it's not an either or, it's very much an and when it comes to build versus buy. And some of that and is sequential, right? So you and I were talking earlier about like when GPT 3.5, I think, like, first hit the scene where like, oh, everybody at Stripe needs to have access to LLMs, but like, we don't quite know how to do that in a way that's, like, enterprise grade safe and we feel good about. Like, we don't see a provider there right now, and so we built it.

但现在我们用的是开源方案，比如LibriChat，明白吗？所以我认为这是一个随时间演进的过程。我觉得最困难的事情之一，特别是对那些只盯着自己产品的团队来说——你热爱自己的产品，想不断改进它——就是你可能会陷入很多局部优化。比如我们本可以拿着GoLM说，哦，我们应该想办法让它接入工具库，或者想办法让它做深度研究，我们本可以那样做。

But now we use, like, open source, LibriChat, you know? So I think there's, like, an evolution over time. And one of the things that I think can be really hard, especially for the team who has tunnel vision for the products they own, you you love your product, you wanna make it better over time, is you can get stuck in a lot of hill climbing. Like, we could have taken GoLM and been like, Oh, we should figure out a way to, like, give it access to tool shed. Oh, we should figure out a way to make it do, like, deep research or you know, like, we could have done that.

有时候你需要一个更外部的视角：如果我忽略沉没成本谬误，忽略我对熬夜加班做出来的东西的情感依赖，从第一性原理出发，如果今天要做这件事，我会怎么做？其中部分工作也包括确保人们对自己的能力充满信心，认识到他们可以为公司跨领域贡献很多，从而把他们从必须拥有这个产品的执念中解放出来。我们还做了另一件事，在AI领域尤其如此，因为有太多新东西不断涌现。我们称之为'聚光灯计划'，其实就是对我们想购买的产品发布需求提案。

And sometimes you just need, like, sort of more of an outside in perspective of, Hey, if I ignore the sunk cost fallacy, ignore my emotional connection to the thing that I spent nights and weekends building, first principles, like, if I were to do this today, what would I do? And some of that is also making sure people feel a lot of confidence and conviction in their own abilities and the fact that there's a ton that they can contribute to the company across domains to kinda liberate them from from from needing to own needing to own this product. Another thing that we've done, and this is especially true in the AI space just because there's, like, so much new stuff coming online. We call it the Spotlight program. But basically, like, we put out RFPs for products that we want to buy.

举个最近的例子，大概是一年前做的，是关于评估系统的。我们说：好，这是我们在评估方面遇到的问题，谁来解决？于是我们通过聚光灯计划发布了这个需求提案。我们当然直接接触了很多这类公司，因为它们要么使用Stripe支付，要么投资者与Stripe有关联，所以我们都认识。

So one example that we did recently, I guess it was like a year ago now, was evals. So we were like, okay, here's our problem with evals. Like, who's gonna solve it? And we put out this RFP in the spotlight program. We obviously see a lot of these companies directly because they run on Stripe and or, like, their investors have some affiliation with Stripe, so we know them.

所以她收到了超过二十份关于这个评估系统需求提案的申请。

So she had, like, more than two dozen applicants for this Eval's RFP.

你们根本不可能全部评估完

There's no way you can evaluate all

其实我们确实都评估了。他们都写了不错的单页摘要，我们全部看完后筛选出两家入围公司。

of them. Well, we actually, we did. So the they wrote, like, nice one pagers. We read them all. We narrowed it down to two finalists.

最终Brain Trust胜出。我们和他们做了概念验证，他们表现很棒，就继续合作了。我们也喜欢weights and biases、flight、Kubernetes这些基础技术栈。但有些情况下我们也必须自己开发。

Brain Trust ended up winning. We did a POC with They rock. We stuck with them. We also love, like, weights and biases, flight, Kubernetes, like, sort of like sort of like the basic stack you can think of. But there have also been cases where we have had to build.

举个传统机器学习的例子，这也与我们基础模型领域相关，嵌入基本上可以成为特征，我们的特征工程平台。我们曾有一个自研的特征工程平台，它已经很老旧了，处于淘汰边缘。当时我们内部有个团队非常想采用Tekton。

So one example is if you think about traditional ML for a second, also relevant in the world of, like, our foundation models and that embeddings basically, like, can become features, our feature engineering platform. So we had a homegrown feature engineering platform. It was old. It was on its last legs. We had a team internally that really wanted to adopt Tekton.

我们评估过Tekton，这是几年前的事了。当时从延迟和可靠性角度，我们无法理解如何在计费路径上使用它。我们需要达到六个九的可靠性，某些模型的决策必须在几十毫秒内完成。

We evaluated Tekton. This was a couple years ago now. At the time, we couldn't wrap our heads around using it on the charge path just from like a latency and reliability perspective. Like, we've got to be operating at six nines. We got to be like decisioning in, you know, tens of milliseconds for some of these models.

在计费路径上我们确实无法解决这个问题。最终我们与Airbnb合作，内部称之为Shepard的项目（开源后叫Cronon）。我认为两种方案各有适用场景，但我们总是优先考虑采购方案，有时这是显而易见的解决方案。

Like, we just we we can't reason about that on the charge path. We ended up pairing up with Airbnb and and building we call it Shepard internally, but it's open source under the name Cronon. You know, I I think there are there are cases there are cases for both. But we always start with, like, what could we buy? Sometimes those are obvious solutions.

如果没有现成的采购方案，我们会关注新兴初创公司的产品，启动'聚光灯计划'评估。如果实在找不到，才会选择自建。在建设过程中，我们会持续评估：半年或一年后，是继续自建还是改用采购方案——毕竟市场在不断发展。在AI领域我们坚信：未来会有众多模型供应商。

Then we say, oh, there's no obvious buy solution, but, like, maybe there's some new startup thing. Let's run the spotlight program. And if we really come up dry, then we will build. And as we build, we reason about, okay, six months from now, twelve months from now, should we still be building, or should we actually swap in a buy solution because the market has evolved? The other thing that we feel pretty strongly about in the world of AI is there will be many model providers.

模型会越来越多样化，我们不愿把命运押注在单一供应商身上。企业级'选择你的LLM供应商'方案对我们吸引力不大，我们更关注能兼容多种供应商和模型的解决方案，实现灵活切换。

There will be many models, and we do not want to hitch our wagons to just one horse. You know, there's lots of enterprise grade versions of choose your LLM provider. Like, that's of much less interest to us than solutions that sit on top of many different providers and many different models and allow us to swap in and out.

是的。Stripe实验团队在决策时...你们是固定由同一批人负责这类工作，还是会根据评估类型重组团队？

Yeah. And this this the Stripe Experiments team, when you decide something, we need to go Yeah. You have the same people always do this, or do you rebuild this team based on if it's Evals or if it's like

这应该是自下而上的，由具体负责人...

It's probably bottoms up, like, whoever Yeah.

这是全力以赴的时候。实验在很多不同地方进行。比如评估工作，我们就在ML基础设施内部运行，配合Go Chat的重建版本。我们让EP部门的人和负责Go LLM的人搭档。实验性项目有趣的一点在于，其目标是快速验证产品市场匹配度——无论是面向内部用户还是外部用户。

It's it's bottoms up. So experimentation happens in a lot of different places. So like for the evals, we just did it like we ran it within ML infra with like the new rebuild of Go Chat. We paired up someone from EP with the people who had been owning Go LLM. You know, one of the things that's interesting about experimental projects is the goal is to learn quickly whether there's product market fit, whether that's with your internal users or your external users.

但核心目标是达到逃逸速度，让产品成功上线运营。无论是GoChat正式发布取代GoLM，还是代币计费服务用户，或是现在实现的商业代理功能。实验项目团队成立约一年半，样本量相对较小。但我们从这些数据中发现，嵌入式项目——即从产品或基建团队抽调几人，与实验项目组员混编——更容易达到逃逸速度。代币计费就是成功案例。

But the goal is to basically get to escape velocity, like have a product that launches and goes live, whether that's like GoChat GA ing and replacing GoLM or whether that's like token billing, serving our users or agent to commerce now being a thing. And what we found and, you know, this is the experimental projects team has been around about a year and a half. And so these are relatively small samples we're talking about. But what we have found in that sort of anic data is what we call embedded projects, projects where you take a couple people from a product or infrastructure team and a couple people from the experimental projects and group them together are more likely to reach escape velocity. Token billing is a good example.

对吧？比如我们需要计费团队推进项目。如果计费团队从一开始就深度参与，推进就会顺利得多。同理，如果ML基础设施团队真正理解新架构并产生主人翁意识，项目才能长期成功。我不认同有些人把这些团队视为孤立运作的实验室团队。

Right? Like, we need the billing team to take it forward. And if the billing team was, like, core to it from the start, it's much easier for them to take it forward. Same thing, you know, if if the ML infra team, you know, deeply understands the new build and feels ownership over it, it'll be successful in the long run. So I don't think of it you know, some people think of these teams as, like, labs teams off in the side, off in a corner, operating totally in isolation.

我们确实有些独立项目，比如正在做的代理间支付系统——因为这需要彻底重构产品形态和技术架构。但只要条件允许，我们都会作为联合项目推进，深度嵌入设计合作伙伴，既有意向客户，也联合Stripe内部其他团队。

We do some of the like, we're doing that for agent pay, like agent to agent payments because that that just, like Right. Needs a big rev on, like, what's the product shape? What's the technology shape? But wherever possible, we actually do it as a joint project, very deeply embedded with design partners, like with customers that wanna do it with us, but also with with other teams at Stripe.

关于自建与采购的闭环，我通常的观点是'先买后建'。从执行顺序来看，你的观点显然更加细致——

My typical line on just got just closing the loop on the build versus buy thing is usually buy then build. If you think about the sequencing, I think yours is much more nuanced in terms of like Buy

如果市场有现成方案就采购。如果没有才考虑自建，但必须每季度重新评估采购可能性。

then build if a buy exists. If a buy doesn't exist, you might want to build, but pick up your head every quarter to make sure you can't buy.

确实。我特别强调这点是因为常见相反做法。很多人试图'先建后买'，从第一性原理推导方案。但实际上，外界海量的实验意味着很多人在专攻你的领域（比如开发环境），你完全可以借鉴他们的经验，不必重复造轮子。

Yeah. Yeah. And also I think mostly because I see the opposite. A lot of people try to do the opposite of build, then buy to, to, to like reason things from first principles, but actually, the sheer amount of experimentation in the wider world means that a lot of people are being specialists in your thing, like devals, where you can just benefit from their experience instead of reinventing the wheel.

在加入Stripe之前，我在Coursera工作。你们了解过教育科技领域吗？我在那里工作了八年，加入时公司还不到40人。团队里有很多来自斯坦福大学Andrew和Daphne实验室的绝对天才——他们之前从未工作过。顺便说一句，我可没把自己算进那些'绝对天才'里。

So before Stripe, I was at Coursera. Have you guys ever looked at the ed tech Of Okay. I was there for eight years, and I joined when we were less than 40 people. And it was a lot of absolutely brilliant folks from Andrew and Daphne's lab at Stanford who had never had a job before. And by the way, I do not count myself in the absolutely brilliant folks from Andrew and Daphne's lab.

我当时在东海岸，算不上什么天才，但同样也是初入职场。我们这群毫无工作经验的人，凭着勤奋和决心，硬是搭建了一堆本不该由我们自研的系统。明白吗？我们有自己的实验平台，有自己的分析系统。

I was on the East Coast and not absolutely brilliant, but I had also never had a job before. And a bunch of us never had a job before, but, like, really hardworking, determined people built a bunch of stuff, homegrown, that we shouldn't have. Right? Had our own experimentation platform. We had our own analytics platform.

这就是学习价值。我们自己的机器学习平台。我们收获了很多经验。但问题是：Coursera的核心竞争优势究竟是什么？

This is learning value. Our own machine learning platform. Had our we learned a lot. We learned a lot. But, like, what is Coursera's core competitive advantage?

肯定不是他们的实验平台。要知道那是2014年，实际上当时很多基础设施

It's not their experimentation platform. And, you know, that was 2014. So actually, a bunch of that stuff

都还不存在。

Didn't exist.

确实不存在，但2018到2020年间推倒重来的过程很痛苦。不过彻底重构绝对是正确选择。另一个观察是：当你看到Stripe如何成为AI企业的骨骼系统时，就会立即明白——在支付、结算、税务、收入确认、报表、反欺诈、消费者结账体验、定价和变现框架等方面，这些公司完全选择采购。他们只采购Stripe的服务。

Didn't exist, but it was painful in 2018, 2019, 2020 to rip and replace. And rip and replace was definitely the right thing to do. The other sort of, thing I'll note on that is if you look at how Stripe is the skeletal system for all the AI companies, it very quickly becomes clear that when it comes to payments, billing, tax, revenue recognition, reporting, fraud protection, consumer checkout experience, pricing and monetization frameworks, they are completely buy. Like, they're completely buying Stripe. That's all they're buying.

这里有个有趣现象：前几天我和某家人工智能公司交流，他们用另一家供应商在注册环节拦截机器人。对方抱怨说，让多个第三方分别负责不同环节的反欺诈很麻烦——有的管上游，有的管下游。他们想转用Stripe，虽然我们目前没有这个特定功能，但完全可以开发出来。

And I think there's an interesting thing there where I was talking to an AI company the other day who uses another provider to block bots at the time of sign up. And they said, it's actually really annoying to have multiple third parties doing my fraud protection. Like, one doing it up funnel and the other doing it down funnel. Well, they asked to switch to Stripe. We don't yet have that particular functionality, but we we could build it.

但我认为同样重要的是要思考你能成为什么样的第三方角色，不是对所有事情，而是更专注投入某些领域，这样内部运作会更顺畅，关系网更精简，能获得优惠价格等等。想想云服务提供商，这种情况也经常发生。

But I think there's also it's also important to reason about what is the third party that you can be, not for everything, but like more all in on so that it plays nice internally, that you have fewer relationships, that you have like preferential pricing, etcetera, etcetera. And, you know, when you think about cloud providers, like that often happens a lot as well.

Vercel绝对在这么做。完全正确。试图打包整合。

Vercel is definitely doing that. Totally. Trying to bundle.

是的。那么这就是经济部分。我们刚才在说你更感兴趣的是AI经济的走向。最明显的大问题是：我们是否处于泡沫中？

Yeah. So So this is the economy section. We we we were saying you're you're more interested in sort of the AI economy takes. The obvious big one is, are we in a bubble?

我们是否处于泡沫中？好吧，这取决于你如何定义泡沫。但我想，你知道，一个问题...

Are we in a bubble? Okay. So it depends what you mean by a bubble. But I think, you know, one question

我作为经济学家的经典回答是：视情况而定。

My my classic economist answer is like, it depends.