从逆袭到领先：Anthropic 产品主管 Diane Penn 谈崛起与产品战略

后来我们开始明确自己的定位：我们真正如何看待世界，如何思考 AI，如何将其更贴近公众。但当时是一种非常自下而上的文化。所以整个经历都是自下而上的。我看到工程师、设计师无偿投入时间。所以我喜欢用这个例子来说明早期是什么样的，但文化和价值观，我认为，从那些早期以来一直保持不变。

Like we then started to identify ourselves as what we actually think the world, how to think about AI, how to bring that closer to the public. But it was a very bottom-up culture. And so that entire experience was very bottom-up. I see engineers, I see designers donating time to work on it. And so I like to always use that as an example of what the early days were like, but the culture and the values have very much, I think, stayed the same since those early days.

本期节目由本季赞助商 work OS 呈现。OpenAI、Anthropic、Cursor、Vercel、Replit、Sierra、Clay 以及数百家其他成功公司有什么共同点？它们都使用 work OS。如果你正在为企业构建产品，你肯定体验过集成单点登录、SCIM、RBAC、审计日志等大型企业所需功能的痛苦。work OS 将这些交易障碍转化为即插即用的 API，提供一个专为 B2B SaaS 打造的现代开发者平台。说实话，我投资的每一家开始向上扩张的初创公司最终都会与 work OS 合作。这是因为他们是最棒的。无论你是种子期初创公司试图拿下第一个企业客户，还是全球扩张的独角兽，work OS 都是实现企业就绪和解除增长障碍的最快路径。它本质上就是企业功能版的 Stripe。请访问 workos.com 开始使用，或者直接加入他们的 Slack，那里有真正的工程师等待回答你的问题。work OS 让你通过令人愉悦的 API、全面的文档和流畅的开发者体验更快地构建。立即访问 workos.com，让你的应用企业就绪。

This episode is brought to you by our season's presenting sponsor work OS. What do OpenAI, Anthropic, Cursor, Vercel, Replit, Sierra, Clay, and hundreds of other winning companies all have in common? They are all powered by work OS. If you're building a product for the enterprise, you've felt the pain of integrating single sign-on, SCIM, RBAC, audit logs, and other features required by large companies. Work OS turns those deal blockers into drop-in APIs with a modern developer platform built specifically for B2B SaaS. Literally, every startup that I'm an investor in that starts to expand upmarket ends up working with work OS. And that's because they are the best. Whether you are a seed-stage startup trying to land your first enterprise customer or a unicorn expanding globally, work OS is the fastest path to becoming enterprise ready and unblocking growth. It's essentially Stripe for enterprise features. Visit workos.com to get started or just hit up their Slack where they have actual engineers waiting to answer your questions. Work OS allows you to build faster with delightful APIs, comprehensive docs, and a smooth developer experience. Go to workos.com to make your app enterprise ready today.

当你回想 Anthropic 从当初那个试图与 OpenAI 这个巨头竞争的小实验室发展到今天，还有哪些重大的转折点？有哪些时刻让你觉得‘哇，那真的改变了局面’？

What are some of the other big inflection moments as you think about Anthropic going from just this lab that's trying to compete with this juggernaut of OpenAI at that point to what it is today? What are some moments that stick out as 'wow, that really changed things'?

肯定是我们在训练和测试 Opus 3 的时候。那时公司还不到 200 人，我们非常清楚我们需要也想要创建一个前沿模型。这对于我们触达用户、消费者以及展示研究成果的能力至关重要。我们也在寻找答案：为什么人们应该选择 Claude？那是早期我们被问到的核心问题。Opus 3 是在 2024 年 3 月初发布的。但在那之前，有几个月的时间，来自推理、研究、微调、预训练等各个团队的成员在不同节点集结起来，朝着共同的目标努力，我认为每个参与其中的人都非常自豪。我记得作为 PM，还有研究负责人，我们都在……那是 12 月左右，所以我们都在各自的父母家里，看到每个人的背景都是自己童年的房间，大家都在非常努力地思考：我们到底在训练模型做什么？它是否以正确的方式呈现？所以，我认为这在建立信任方面非常有力。当时很多研究负责人现在都在领导强化学习、角色工作和对齐工作。这种基础信任也帮助我们现在在产品和研究领域与所有生产模型合作良好，因为我们在早期就已经并肩作战了。然后，我们识别出编码很重要。对吧？在 2023 年我刚加入时，没有人会把 Anthropic、Claude 和编码放在同一个句子里。当时的竞品模型如 GPT-4 虽然也用于编码，但那只是众多用例之一。我观察到的一个现象是，人们开始不仅使用这些模型进行代码自动补全，而是实际编写长篇幅的代码。这成为我们训练 Opus 3 以更好应对这一任务的机会。从训练角度来看，这只是相对较小的改变，但它帮助我们在早期竞争中脱颖而出，吸引了许多早期的 Claude 爱好者和开发者，因为我们提供了他们当时认为不可能的价值。

Definitely when we were training and testing Opus 3, I think that was the moment when the company — we were less than 200 people still at that point — and it was very clear that we needed and wanted to create a frontier model. That was very important in terms of our ability to reach users, consumers, and to showcase our research. And we were looking for ways also for why should somebody choose Claude? That was a core question we were getting asked in the early days. I think with Opus 3, it launched early March 2024. But there were many months of various teams across inference, research, fine-tuning, pre-training that rallied at different points towards a common goal, and I think everybody that was involved was really proud. I remember being the PM, the research leads, myself, we were all... this was around December, so we were all at home in our various parents' homes, seeing everybody's background of their childhood room, and everybody was working really hard to figure out what we are training the model for. Is it showing up the right way? So I think that was really powerful in terms of building a lot of trust. A lot of our research leads from that time are now leading reinforcement learning, leading our character work, alignment work. So that foundational trust also helped us work well now with any of our production models across product and research because we worked so much in the trenches together in the early days. And then there were things like identifying that coding was important. Right? In 2023 when I started, nobody said Anthropic and Claude and coding in the same sentence. Competitor models like GPT-4 at the time were used a bit for coding, but it was one of many use cases. One thing I saw was people starting to use these models not just for code autocomplete but actually writing long form code. And that became an opportunity for us to train Opus 3 to be better at it. It ended up being a relatively smaller change from a training perspective, but it helped us differentiate in the early days competitively for users and actually bring a lot of the very early Claude enthusiasts and developers because we were providing a value they didn't really think was possible at the time.

你谈到 Opus 3 时感觉那是很久以前的事了。很难想象那是一个重大的转折点，听到它内部是一个重要的里程碑非常有趣。几乎感觉你们建立了一种信心，‘哇，我们真的能发布一个前沿模型’，而如果和今天拥有的模型相比，它现在其实不怎么样。我一直想到的是 Opus 45，有趣的是，一年后同样是在寒假期间，每个人都能在家编程。那是另一个重大里程碑吗？

It's so interesting you talk about Opus 3 like that's so long ago. It's hard to think that was a big inflection, and it's really interesting to hear that was internally a big milestone. It almost feels like the confidence you all built that wow, we could really ship a frontier model, which is now today not so great if you compare it to what we've got today. What I always think about is Opus 45, which was, interestingly, a year later also during winter break when everyone was home able to code. Was that another big milestone?

是的。Opus 45 绝对是另一个重要时刻。我认为 Opus 45 的神奇之处在于，我们不仅有了一个模型，还有一个载体，那就是像 Claude Code 这样出色的产品体验。我们在团队中常说，你需要前沿产品才能拥有前沿模型，让人们感受到前沿模型的魔力。我认为在那之前的几个月里，我们已经感受到了 Claude Code 的魔力。但模型智能达到一个水平，在非常广泛的层面上，用户能够在新的用例中体验前沿智能，并以智能体方式端到端地运行任务——我认为这才是转折点。这其实是两者共同作用的结果。没有 Claude Code 这样的产品，Opus 45 就不会有那个时刻；而没有 Opus 45，Claude Code 的采用也不会如此迅速。

Yeah. Opus 45 was definitely another large moment. I think what was magical about Opus 45 is we now not just had a model but a vehicle, which is a great product experience like Claude Code. One thing we say a lot on the team is you need frontier products in order to have frontier models and for people to feel the magic of frontier models. I think we felt the magic of Claude Code for many months before that. But the fact that the model got to a level of intelligence where at a very broad level users can experience frontier intelligence in new use cases, allow it to run things end to end in an agent manner — I think that was the inflection. It was actually both. Opus 45 wouldn't have had that moment without a product like Claude Code, and Claude Code wouldn't have had that type of accelerated adoption without Opus 45.

所以，顺着这个思路，Dario 很有趣：如果你回顾他的所有预测，他就会说，‘好吧，编码问题将在一年内 100% 解决，类似这样的话。’他不断提到我们将如何写代码，AI 将完成我们所有的代码。我记得当时所有人都说，‘不可能。这太复杂了。AI 怎么可能真正擅长人类做的这种非常复杂的事情？不，很长一段时间内这还得靠人类。’他完全正确。他还经常谈到我们现在所处的指数阶段。他现在就是这样描述的。我们正处在指数曲线上。我记得不久前，新模型发布时，每个人都说，‘好了，我们完了。没有提升空间了。它在趋于平稳。结束了。没有增长空间了。’而现在情况恰恰相反。

So kind of speaking on this thread, Dario, interestingly if you look back at all his predictions, he's just like, 'Okay, coding is going to be solved 100% in like a year, something like that.' He kept talking about how we're going to do code, like AI is going to do all our code. And I remember everyone being like, 'There's no way. This is way too complicated. How is AI ever going to get really good at this very complex thing that humans do? No, this is going to be humans for a long time.' He was completely right. Something else that he talks a lot about is this exponential we're now on. That's the way he describes it now. We're on the exponential curve. I remember not long ago, new models were being released and everybody was like, 'Okay, we're done. There's no more upside. It's plateauing. It's over. There's no more room to grow.' And now it's the opposite.

现在我们身处其中，就像指数曲线一样。我们正处于指数增长的阶段，这意味着每一次进步都是巨大的飞跃，因为我们正处于那种曲棍球棒式的部分。身处这个 AI 快速进步、解锁了如此多东西的疯狂历史时刻，感觉如何？人们应该如何为 AI 带来的日益加速的进步做准备？

Now we're inside, like if you think about the curve of the exponential. We're like inside of the exponential now, which by definition means every improvement is a massive jump because we're like on that hockey stick part. What's it like just being on the inside of this crazy historic moment when AI is improving so fast, so much is being unlocked? Uh, what is it like and how should people prepare for the coming acceleration of more and more improvement from AI?

我在团队中喜欢说的一件事是，当互联网从新奇事物转变为每个人都能使用的东西时，我们大多数人都还没有活跃地工作。感觉就像现在正在经历类似的过程——我认为类比是有帮助的。这个类比有几层含义：第一，适应能力变得非常重要。我们有评估体系。在安全方面，有安全测试和红队演练。在能力和产品方面，有新的原型和产品，比如 Claude Code Tag 等。但很难预测确切的时间点或具体的模型。所以，当你面对新信息时，如何做出更好的决策，而不是坚持原来的计划？这种灵活性非常重要。我认为另一点是，如何真正从第一性原理出发，推理下一步是什么？那又怎么样？我们如何投资新产品？如何向用户解释差异？所以，身处指数增长中的很多体验就是那种节奏，理解你如何运作并做出更好的决策，然后运用第一性原理思维去做一些事情——比如我们原本计划几个月后才做的，但现在模型已经可以实现了，于是我们把它推向用户。比如 co-work skills tag。这是一个非常积极的自我强化循环。我认为很大一部分也在于彼此信任，确保我们做出正确的决策，带领大家前进。有些团队可能更快地感受到指数增长。那么，我们如何有风度地带领组织、成长中的公司和团队一起前进？

One thing I like to say on the team is most of us weren't actively working yet when the internet transitioned from this novelty to something that everyone can use. And it feels like that's just taking humans—I think analogies are helpful. So the analogy of that is a couple of things. Number one, adaptability becomes very important. We have evals. On the safety side, safety testing, red teaming. On the capabilities and product side, new prototypes and products like Claude Code Tag and others. But it's very hard to predict the exact moment or the exact model. So the adaptability of when you're faced with new information, how do you then make better decisions versus keeping the same plan? That agility is really important. I think another piece is, with that, how do you actually think very first principles and reason through what's next? What's the so what? How do we invest in new products? How do we invest in explaining the differences to users? So a lot of the experiences of being in that exponential is that pace, understanding how you operate and make better decisions, and then applying that first principles thinking to do something that we might have pulled up a plan we were expecting a few months from now, but now the model can actually do it, and work on it, and actually bring that to user. So this is things like co-work skills tag. It's a very positive self-reinforcing loop. And I think a big part of it also is just having the trust in each other, making sure we're thinking through the right decision making, bringing folks along. Some teams might see the exponential, feel it faster than others. So how do we have the grace to bring the organization, the growing organization and company along on that?

所以我在听的是，你几乎不知道每个模型发布后能实现什么。因此，重点是要随着事情的出现而保持适应。正如你所说，产品本身必须跟上、必须追赶可能实现的东西。同样，就像模型能做到很多，但人们可能不知道如何去做，或者做不到。所以产品让事情变得简单，甚至只是告诉你“你可以做这件事”，这似乎是重要的一部分。大致是你在描述的吗？

So what I'm hearing here is you almost don't know what will be possible with every model release. And so the important things to focus on is being adaptable as things emerge. To your point, the product itself has to stay up to, has to catch up to what is possible. To your point again, just like it can do so much, but people may not understand how to do it and may not be able to do it. So the product making it easy and even just telling you 'here's something you could do' feels like an important part. Is that roughly what you're describing?

我想是的。我认为原始的缩放定律论文中有一些非常有趣的图表。大家都很熟悉缩放损失的概念——随着你增加算力和数据，所谓的损失（即下一个词预测的损失）会下降。这是一个非常平滑的线性曲线，模型随着规模扩大而变得更智能。但该论文中真正有趣的还有这些非常不同的涌现能力图表。例如，随着你增加更多数据并用更多算力训练模型，你实际上会看到这些不连续的涌现能力跳跃。模型从无法计算 1+1 变成能够可靠地计算。所以这些涌现能力，这种可预测性并非每个人都知道确切的时间点。你需要评估体系来评估它。这实际上一直是这项技术运作的一部分，也是让安全之类的事情变得更难的原因，因为除非你有评估体系，除非你有测试系统，否则这些跳跃可能真的会发生而你却不知道。

I think so. I think there's some really interesting graphs in the original scaling law papers. And I think folks are very familiar with the scaling loss in the lens of, as you add more compute and data, what's called loss—aka the loss from next token prediction—goes down. And it's a very smooth linear curve of the models get more intelligent as you scale them up. What's actually also interesting in that paper is there are these very different emerging capability graphs. So for example, as you add more data and train the models with more compute, you essentially see these actually discontinuous emerging capabilities jump. So the models go from 1+1 being a thing that it can't calculate to a thing that it can reliably calculate. And so these emerging capabilities, this nature of predictability is not necessarily everyone knows the exact moment. You need the evals to be able to assess that. That has actually always been a part of how this technology works, and also what makes things like safety harder, because unless you have the eval, unless you have the systems to test, these jumps might actually happen and you don't know.

嗯，这太有趣了，你可能开发出了这个能做连你自己都不知道的事情的 AI 大脑。所以工作的部分就是去发现，哇，我们刚刚变得非常擅长这个，我们可以用它做什么？

M, that's so interesting that you may have developed this AI brain that can do something you're not even aware of. And so part of the job is just uncovering wow, we just got really good at this thing, what can we do with that?

我认为有产品悬置和用户悬置。用我们产品管理的语言来说，即使在今天的模型上，我认为我们还有很多可以在当前的 Opuses 上探索的，当然还有 Fable，比如 temple。而这种发现实际上是 Anthropic 早期 DNA 的另一部分，我认为它也将继续成为我们在产品、实验室和跨研究中运作的重要部分。

I think there's like product overhang and user overhang. To put it in our PM language, even on today's models, I think there's a lot that we could be exploring on our current Opuses, and definitely with Fable, for example, temple. And that discovery is actually another part of what's been in the early days of Anthropic's DNA, and I think is also continuing to be a big part of how we operate in product, in labs, and across research.

这让我想起 YC 总裁 Gary Tan 说过的话。我不确定他的头衔。他有一个有趣的观点：如果你现在愿意每年在 token 上花费 10 万美元，你就在体验 2028 年人们的生活方式，因为到那时它会变得非常便宜，每个人都可以这样工作。但如果你这么做，现在就有一个 alpha 机会去活在未来，疯狂地花费 token。所以人们有很大的机会去了解未来是什么样子，并且更快地构建。对这个“token 最大化”的想法和价值有什么看法？

This makes me think about something Gary Tan's been talking about, president of YC. I don't know what his title is. He had this interesting point that if you're willing to spend $100,000 a year right now in tokens, you are living the way somebody in 2028 is going to live, because by then it'll be really cheap. Everyone can work this way. But if you do, there's this alpha opportunity right now to just live in the future, go crazy on token spend. So there's a big opportunity for people to learn what the future's like and also just build much faster. Thoughts on this idea of and the value of token maxing, let's call it.

是的，我更倾向于从产品角度出发。token 花费更像是输入，而真正的输出是你所描述的实验。我认为如果我们将目标围绕实验来设定，那可能是更好的结果框架，因此可能有不同的方式来实现那个结果。我要说在内部，一些最有创造力的思考者、最好的原型设计师确实会花费大量时间与 Claude 以及我们拥有的每一个新版本的研究模型互动。所以你必须使用模型才能想出好的、伟大的、更好的想法，这是无可替代的。当技术发展如此之快时，不接触技术就很难制定完美的策略。同时，我认为我们还有其他可以做的事情。我们经常做的一件事就是在 Anthropic 内部公开工作。在早期，当我们的产品界面较少时，有一个 Slack 频道，几乎整个公司都在测试早期版本的 Claude 并尝试不同的用例。人们不叫它们用例，但你可能让它编辑一篇文章，或者想出发送这封邮件的正确方式。它们都是不同的用例，但我们都是公开工作的。

Yeah, I think I take more of an almost product lens. It's almost like token spend is more the input, and really the output is what you described of experimentation. And I think if we were orienting goals around experimentation, that might be the better framing of the outcomes, and therefore there might be different ways of achieving that outcome. I will say internally, some of the most creative thinkers, the best prototypers do spend a lot of time with Claude with every new version of a research model that we have. And so there is something around you have to be using the models to then come up with good, then great, then better ideas, and there's no substitute for that. It's very hard to come up with a perfect strategy without touching the technology when it's moving this quickly. At the same time, I think there's other things that we could be doing. One thing that we do a lot is actually working in public at internally within Anthropic. And so in the early days when we had less product surfaces, there was a Slack channel where everyone—almost the entire company—was testing early versions of Claude and trying different use cases. People were not calling them use cases, but you might be asking it to edit an essay or to come up with the right way to send this email. They were all different use cases, but we all worked in public.

然后你会神奇地看到，不同用户或团队中的不同成员提出了一个想法，然后其他人尝试这个想法的不同变体。在大约 10 个左右的请求内，就会产生一些神奇的东西或潜在的新用例。我认为这不仅仅是个体自己摸索如何使用这项技术。我认为我们在实验时可以做得更多，真正带来那种集体发现。实验并不总是个人的运动。

And then what you would see magically is different users or folks on the team coming up with an idea and then other people trying different variations of that idea. Within maybe 10 or so requests, there was something magical or potentially a new use case that emerges. I think there's a lot in not just individuals figuring out by themselves how to use this technology. I think we could be doing more to actually bring that communal discovery when we do experimentation. Experimentation is not always necessarily an individual sport.

这太有趣了。是的。这种我们不确定它的能力或我们能做什么的想法，需要不断的试探和人们尝试各种东西，听到其他人在尝试什么，从而找出可能性。多么有趣的技术。或者就像，‘好了，我发现它可以做这个。你会用它做什么呢？’

It's so interesting. Yeah. This idea that we're not sure what this is capable of or what we could do with it, and it takes all this poking around and people trying things, hearing what other people are trying to figure out what's possible. Such an interesting technology. Or just like, 'Okay, here's what I figured out it could do this thing. What are you gonna do with that?'

我认为在大的主题上，我们知道模型可以写很好的文章或长篇文章，但那些你能实际解决的具体痛点，并将其提升到用户可以使用的层面，我认为这更多是基于探索或实验的。

I think at a broad theme we know that the models can write great essays or long-form writing, but individual pain points of what you can actually solve with that and bring it to a user level that people can use, I think is something that is more exploration or experimentation based.

顺这个话题，你负责实验室团队的产品，这非常酷。我们曾在播客中邀请过 Ben Mann 和 Mike Griger，他们都在实验室工作，谈论过实验室。什么是实验室？实验室推出了什么？很多人听说过这些，但实验室是如何运作的，使得它们能够在甚至核心 Anthropic 产品团队之外创造出如此创新的想法？实验室的宗旨很多方面是识别并抓住那些可能不在核心路线图中的非连续大赌注，弄清楚是否存在‘那个东西’，以及它的 10 倍、100 倍、1000 倍是什么。例如，像 Claude Code 这样的东西。我想…

Following this thread, you oversee product for the labs team, which is extremely cool. We've had Ben Mann on the podcast, Mike Griger, who both work on labs now, to talk about labs. What is labs? What's come out of labs? Many people have heard of these things, and how do they work that enables them to create such innovative ideas outside of even the core Anthropic product team? The thesis of labs in many ways is identifying and pulling the thread on discontinuous large bets that might not be in the core roadmap, and figuring out if there is a 'there there', and also what is the 10x, 100x, a thousandx of the 'there there'. So for example, things like Claude Code. I think...

我听说过像 Claude Code、Skills，以及最近的 Claude Design、MCP 之类的东西。我们在团队中真正强调的是，尤其是现在有很多东西可以构建。那么，进行非连续赌注意味着什么？我认为我们今年采取的一种方法是，你可以对主题或领域持有非常坚定的意见，而对具体的原型则持有较弱的态度。所以这是一种实验文化。很多是自下而上的，比如团队中的工程师们非常自我驱动去测试不同的想法。有时我们有一个论点，但它可能暂时行不通，那么我们会在一到两代模型后重新审视它。这些原型最终只是帮助我们学习，即使它没有立即发布，也是有价值的。我认为这允许实验室的孵化和章程真正加速，并为 Anthropic 更广泛地预见未来。

I've heard of things like Claude Code, things like Skills, and most recently Claude Design, MCP. The thing that we really try to emphasize within the teams is especially right now there are so many things that could be built. What does it mean then to have a discontinuous bet? I think one approach that we're taking this year is you can be very strongly held opinion about the theme or the area, and then more weakly held about the exact prototype. And so there is a culture of experimentation. There's a lot of bottom-up, like engineers on the team are very self-driven to test out different ideas. Sometimes we have a thesis and it might not work yet, and so we then might revisit it in one to two model generations. This idea of these prototypes that actually end up just helping us learn is also valuable, even if it doesn't lead to something immediately shipping. I think that allows the incubation and the charter of labs to really accelerate and see around corners more broadly for Anthropic.

想到在已经非常创新、创造力和疯狂发布产品的 Anthropic 内部再设立一个实验室，这很有趣，说明在 Anthropic 内部创建实验室团队仍有价值。是什么让实验室如此成功？因为你列举了所有这些产品，而且感觉 Anthropic 发布的东西几乎都是最大的胜利。我肯定还有很多我现在没想到的。在一个更大的公司内部创建一个成功的实验室，核心是什么？

It's so funny to think about a labs within an Anthropic which is already so innovative and creative and just shipping like crazy, that there's value to still creating a labs team within Anthropic. What enables labs to work as well as it has? Because you listed all these products, and it's like, what else has Anthropic shipped? It feels like all the biggest wins almost. I'm sure there are many that I'm not thinking about right now. What's kind of core to creating a successful labs within a larger company?

我认为团队文化，与 Anthropic 广泛的文化类似，非常有价值。我认为 Ben 设定了令人难以置信的愿景，并推动人们思考想法的 10 倍、100 倍。我们的团队，实验室内的小组，都很小。有时这些想法从一个工程师开始。我认为有时当非常大的团队追求非常模糊的大想法时，你实际上会因此变慢。所以我认为是文化。我们实际上也选择那些想要做从 0 到 1 实验的人，这并不容易。我们有很多赌注最终被拒绝或关闭，也许将来会重新审视。但这很难。当你全心全意投入，作为一个赌注的创始人，而它还不成功时，这很难。所以我认为是选择那种性格，那些对从 0 到 1 非常热情和深入的人。

I think that team culture, similar to broadly at Anthropic, is very valuable. I think Ben sets an incredible vision and pushes people to think about the 10x, 100x of the idea. Our teams, the pods within labs, are small. Sometimes these ideas start with one engineer. I think sometimes when there are really large teams pursuing very ambiguous large ideas, you end up actually being slowed down because of that. So I think it's culture. We actually also select for folks who want to do that zero-to-one experimentation, and it's not easy. There's a lot of bets that we end up turning down or turning off, and maybe we revisit them in the future. But that's hard. That's hard when you pour your heart and soul, you're acting as a founder for a bet, and it's not working yet. So I think it's like selecting for that type of personality, folks who are really passionate and deep about the zero-to-one.

那么你领导研究团队的产品。你和 Anthropic 的研究人员一起工作。很多人对什么是研究、研究人员做什么有所了解。我认为很多人并不完全理解这些在 AI 实验室中非常有价值的人。我的想法是，我想帮助人们理解，也帮我理解：研究人员整天到底在做什么？我设想他们有改进模型的假设，他们找数据，调整算法，检查调整训练方式，然后测试，看效果如何，不断迭代，不断寻找改进模型的方法。大致正确吗？帮我们理解研究人员整天在做什么。

So you lead product for the research team. You work with the researchers at Anthropic. A lot of people kind of get a sense of what research is, what researchers do. I think a lot of people don't totally understand these very valuable people at all the AI labs. The way I think about it, and I want to help people understand, help me understand: just what are researchers doing all day? What I imagine is they have a hypothesis for how to improve the model, they find data, they tweak some algorithms, they check adjust how it's trained, and they test it, see how it did, keep iterating, and keep trying to find ways to improve the model. Is that roughly right? Help us understand what researchers are doing all day.

这真的……我认为这可能是更日常工作的一部分。我认为关于研究人员和研究机构，比如在 Anthropic，还有一个更广泛的未来愿景。例如，我认为在公司成立之初，研究人员就在讨论如何让 Claude 使用电脑，如何让 AI 浏览屏幕，对吧？所以研究人员身上有很多非常像创始人的能量，我是这么描述的，或者说是非常大胆和雄心勃勃的研究人员。我们在 Anthropic 有很多这样的人。所以有一层是关于这项技术可以走向何方的愿景。然后在循环的另一边，既然这项技术（或 Claude）已经在人们手中，我们如何让它今天变得更好？所以这是中长期，以及很多精力思考未来视角，同时在当下和短期内，我们可以改进哪些方面？我认为你很好地描述了我们如何在不同版本的 Claude 上进行迭代改进。我的团队与研究人员合作的方式是非常紧密地融入这些循环，尤其是在对用户有重大影响的领域。

That's really... I think that's a lot of maybe the more day-to-day. I think one piece around researchers and research organizations like at Anthropic is there's also a vision of the future more broadly. For example, I think even at the founding of the company, researchers were talking about how do we get Claude to use a computer, how do we get AI to navigate a screen, right? So there's a lot of actually very founder-like energy, is how I describe it, within researchers, or really bold and ambitious researchers. We have a ton of those at Anthropic. So there's one layer of vision of what this technology can go. Then I think on this other side of the loop, there's also now that this technology (or Claude) is in people's hands, how do we make it better today? So it's a medium and long term, and a lot of energy thinking about that lens of the future and also in the immediate and short term, what are the improvement areas we can make? I think you're describing a really good sense of how we make iterative improvements on different versions of Claude. The way my team works with researchers is kind of being very integrated and embedded in those loops, particularly areas where there's a lot of impact on users.

所以这包括视觉、计算机使用、编码、智能体编码、工具使用、测试时算力等直接影响用户的方面，然后找出如何将用户反馈引入并落地到对研究人员可理解且可操作的层面。我认为第二部分实际上是工作的重要组成部分，有时也是最难的部分。例如，我们可能会收到关于 cla.ai 的反馈：Claude 产生了幻觉。这非常模糊。如果你把它交给研究人员，说‘请修复 Claude 的幻觉问题’，这并不具有操作性。因此，团队部分工作是理解用户给出该反馈的轨迹，并且这是经过同意的。所以我们分析，Claude 在那时应调用哪些工具？根据其当前知识，它是否调用了正确的工具？它查看了正确的文档，但查看了错误的事实。第一种情况是工具使用失败；第二种情况可能是搜索、知识或搜索综合方面的失败，或者可能与对齐有关。因此，我们要将这种详细程度带给研究人员，判断这是否是一个足够大的问题，制定评估指标，然后描述我们如何改进——这些就是可操作性层面的内容，也是研究人员的日常语言。我们努力贴近如何以可操作的方式将用户反馈引入核心模型训练和研究开发循环。

So this is things like vision, computer use, coding, agent coding, tool use, test time compute, things where there's a direct user impact and then figuring out what are the ways to bring the user feedback and ground it in a level that is understandable for users and also actionable for researchers. And I think that's the second piece is actually a big part of the job and sometimes a hard part of the job. So for example, we might get feedback on cla.ai. Claude hallucinated. It's very vague. If you bring that to a researcher and you say, 'Please fix Claude from being hallucinated.' It's not very actionable. And so part of the time of the team is understanding, okay, what's the trajectory of why that user gave that feedback? And it's consented. And so we look at okay what should Claude have called tools in that moment or from its current knowledge or it called the right tools, it looked at the right document but it looked at the wrong facts. In the first case that would have been a failure on tool use. On the second case it would have been a failure on let's say search or knowledge and search synthesis or it could be something around alignment. And so bring that level of detail to researchers coming up with like is this a big enough problem figure out things like evals to then describe how we've improved it like those are the levels of actionability and it's the day-to-day language of the researchers. And so we try to stay very close to how to bring that in an actionable manner between users to the core model training and the research development loop.

我前几天和别人聊到，感觉现在 AI 研究领域是如果你想在人生中非常成功就该去的地方。从你的角度，成为一名真正成功的研究人员需要什么？你知道，不是每个人都能进入这个领域，不是每个人的大脑都适合这种方式，但假设人们说‘嘿，我想探索这条职业道路’。从你看到的，要成功需要什么？研究人员通常是做研究还是产品经理与研究人员合作？两者都做。但研究人员，比如产品经理与研究人员合作也会非常成功。但感觉每家公司都在竞相挖走所有顶尖研究人员。所以，我知道你不是 AI 研究人员，但就从你看到的，要在这条职业道路上成功需要什么？

I was talking to someone the other day about how it feels like research AI research is the place to be now if you want to be very successful in life. What does it take to become a really successful researcher from what you can tell? You know, not everyone can get in, not everyone's brain is going to work this way, but just say people are like hey I want to explore this career path. From what you've seen, what does it take to make it there? Researchers generally are research and product managers working with research or both? Let's do both. But the researchers like you know PMs working with researchers also going to be very successful. But it feels like everyone's trying to poach all the top researchers across every company. So just I know you're not an AI researcher, but just from what you've seen, just like what does it take to make it in that career path?

是的，我认为在 Anthropic，许多最成功的研究人员和研究领导者都是那些对问题有很强第一性原理思维的人。他们非常擅长推理问题。他们对自己的研究领域充满热情，并对其前景有大胆的描述，而且他们实际上非常关注细节。所以我们的领导层、首席科学家、微调和强化学习负责人实际上非常接近训练运行，会查看训练运行的评估结果和底层数据。所以保持贴近并乐于深入细节，我认为是优秀研究人员和培养品味的表现。另一部分是随着时间的推移，他们有能力进行大格局思考并非常有雄心。就像 Dario 说的，我们可以改变软件工程。沿着这个方向，你会学到很多，必须在想法上仰望星空，才能成为成功的研究人员。

Yeah, I think a lot of the most successful researchers and research leadership at Anthropic are folks who are really strong first principles thinkers about problems. Like they reason through problems really well. Who are just passionate about their research area and have a bold description of what that could look like and then who are actually close to the details. And so our leadership, our chief scientists, our heads of fine-tuning and RL folks are actually really close to the training runs and actually look at things like how the training run is eval looking at the underlying data. So actually staying really close and be excited to be in the details I think have been a sign of really strong researchers and developing taste. And I think another piece is just their ability to think big over time and be very ambitious. Like Dario said we can transform software engineering. And going in that direction you learn so much, you have to shoot for the stars in many ways across your ideas in order to be a successful researcher.

我非常喜欢这个‘要更雄心勃勃’的梗，现在经常出现，这很难，说起来容易，但真正要做到‘你能想多大’很难，而这就是 AI 现在释放的潜力。就是要更雄心勃勃。

I love just this meme of just be more ambitious comes up so often now which is so hard like it's it's easy to say that it's hard to actually just like how big can you and how that's so much of what AI now unlocks. Just be more ambitious.

是的。我认为需要彻底思考一两次，然后在该领域保持固执，而在具体方法上则更灵活。这是我们常常挑战自己的问题。但技术发展如此之快，你如何确保你正在构建的东西实际上是向前兼容的？因此，做大格局思考也是核心产品开发循环的一部分。我经常问团队的一个问题，或者我在构建产品时的思考方式是：假设 Claude 8 出现了，用户行为发生了什么变化，那对你今天的构建意味着什么？它是否能向前兼容那种体验？所以，将其落地，我认为雄心勃勃是非常广泛的，所以要尝试用一些方式来描述它。

Yeah. I think it's thinking through it once or twice to the end and then being stubborn about the area and maybe more loose around the exact approach. It is a question we challenge ourselves with. But the technology is moving so quickly and how do you make sure what you're building is actually forward compatible? So it's also part of the core product development loop to think bigger. One thing I ask the team frequently or how I think about when we're building a product is: let's say Claude 8 comes around, what changes in what users do, and then what does that mean for how you're building today? Is it going to be forward compatible to that experience? So just grounding it, I think being ambitious is very broad and so trying to ground it in some ways of describing that.

而且，是的，一切都要朝着一个连贯且有意义的方向发展，而不是在一个完全不同的方向上雄心勃勃。说到雄心勃勃和 Claude 8，Fable Mythos 最近似乎达到了一个全新的转折点。以前，你有一个很棒的模型，发布它。‘嘿，大家好，欢迎。Opus 45 出来了，每个人都可以使用。’Mythos 则走了一个非常不同的方向。它被封锁了。有很多审查，很多关于它能力的担忧。所有公司都必须确保它不会入侵他们的所有系统。现在感觉每个模型，因为它们不断变得更好，都会受到更多审查，并且会有更多关于谁可以使用它们的限制，这感觉是一件大事。你怎么看？这如何改变你的运作方式？

And also yeah everything heading in a direction that all is cohesive and makes sense versus just ambitious in a completely different direction. Speaking of ambition and Claude 8, Fable Mythos recently feels like hit this very new kind of tipping point with models where it used to be you have an awesome model, release it. Hey everyone, welcome. Opus 45 is out, everyone can use it. Mythos went in a very different direction. It got blocked. There was a lot of scrutiny, a lot of concern about what it was capable of. All the companies had to go make sure it wasn't going to hack into all their systems. And it feels like now every model because they continue to get better will now have a lot more scrutiny and there will be more restrictions on who can use them which feels like a big deal. How do you think about that? How does that change the way you operate?

我可能把政策和出口管制方面留给负责这些的人。我认为产品问题以及我们内部如何与这些模型互动，正如你提到的，随着前沿模型能力增强，安全保障、红队测试、测试和预发布流程也需要快速发展和适应。一个例子是，在 Fable 模型之前，我们没有那么强的备用用户体验和系统，因为我们的目标是确保这项技术带来不对称的利益，并最小化其负面或严重风险。所以我们构建了备用系统，以便用户仍然可以从 Opus 4 获得出色的响应。因此，我认为在改进安全系统的同时，如何继续开发和提供出色的用户体验？我认为双方都有更多可以做的。所以你会看到我们在未来几周和几个月内不断创新和改进我们现在所谓的‘模型安全保障包’。

I'm going to maybe leave the policy and the export control side to folks that own that and work on that. I think the product question and how we interact with these internally is I think as you mentioned as frontier models become more capable the safeguards and the ways of red teaming and testing and the pre-release process also needs to evolve and adapt quickly to address that. And so one example is you know before Fable models we didn't have as strong of let's say fallback UX's and systems because our goal was to make sure that there is asymmetrical benefit for this technology and to minimize the downside or a severe risk of it. And so we ended up building fallback systems so that users will still get a great response from Opus 4. And so I think there's a piece around as we evolve and improve safety systems, how do we continue to develop and deliver great user experiences? I think there's more that we can do on both sides. And so you'll see us innovating, improving on what we call now the model safeguards package more and more in the coming weeks and months.

这里真正有趣且出乎意料的是，它为 Anthropic 创造了一个非常有趣的优势：你可以使用最新技术，而这种情况将出现在每个实验室。每个人都在不断改进，这就在实验室内部形成了一种不公平优势——你可以接触到别人还无法接触到的最好的东西，而这些都超出了你的掌控。你更希望每个人都能使用它。所以，这是一个非常有趣的新反馈循环即将开始：如此先进的模型只有某些公司才能访问，而这将是所有这些限制的一个全新的、意想不到的二阶效应。我们的目标是开发这些系统和模型，使其尽可能包容。我认为我们的目标是避免这种情况发生在通用技术上，并使其更易获取。我认为这是我们目前的最高优先事项之一，以减少我们正在看到的情况。

What's really interesting and unexpected here is that it creates this interesting advantage for Anthropic where you have access to the latest stuff, and this is going to happen at every lab. Everyone's going to keep improving, and it creates an unfair advantage within the labs to have access to the best stuff that other people can't yet, outside of your control. You'd prefer everyone uses it. So it's a really interesting new feedback loop that's going to start where models that are so advanced are only accessible to certain companies, and that's going to be a whole new unexpected second order effect of all these restrictions. Our goal is to develop these systems and models to be as inclusive as possible. I think our goal is to not have that happen for general-purpose general-use technologies and to make it more accessible. I think this is one of our top priorities right now to reduce what we're seeing there.

是的，有道理。我想你希望尽可能多的客户使用这个东西。

Yeah, that makes sense. I would imagine you'd want as many customers using this thing as possible.

本期节目由 Mercury 赞助，这是革命性的银行服务，深受超过 30 万企业家的喜爱，现在还有 Command 功能。我使用 Mercury 已经超过 6 年了，从未想过离开。Mercury 本质上就是由产品人员而非银行家打造的银行服务。他们让开发票、转账、为团队成员设置虚拟卡变得如此简单，甚至可以说有趣。你的银行有 API、终端原生 CLI 或支持 AI 的 MCP 服务器吗？我认为没有。就在最近，他们推出了 Command，这是直接内置在 Mercury 中的对话式界面，充当你的财务运营者。我一直在用 Command 转账、了解我在哪些类别上花费最多、分析我的现金流，就在今天我还用它查了过去一年从某个特定赞助商那里赚了多少钱。我只需问：“过去一年我从 X 那里赚了多少钱？”10 秒后我就得到了答案。这太酷了。访问 mercury.com 了解更多，几分钟内在线申请。Mercury 是一家金融科技公司，不是 FDIC 投保银行。银行服务由 Choice Financial Group 和 Column N.A. 提供，均为 FDIC 成员。

This episode is brought to you by Mercury, radically different banking, loved by over 300,000 entrepreneurs and now with Command. I've been a customer of Mercury's for over 6 years. I have never once thought about leaving. Mercury is basically what happens when banking is built by product people, not by bankers. They make it so easy, dare I say fun, to send invoices, move money around, set up virtual cards for folks on my team. Does your bank have an API, a terminal native CLI or an AI-ready MCP server? I don't think so. And just recently, they launched Command, a conversational interface built directly into Mercury, which acts as your financial operator. I've been using Command to transfer money around, to figure out what categories I've been spending the most money in, analyze my cash flows, and just today I used it to find out how much I've made from a specific sponsor over the past year. I just ask, "How much have I made from X over the past year?" 10 seconds later, I have an answer. It is so freaking cool. Visit mercury.com to learn more and apply online in minutes. Mercury is a fintech company, not an FDIC insured bank. Banking services provided through Choice Financial Group and Column N.A., members FDIC.

我想谈谈产品角色如何变化，以及在这个 AI 已成为我们生活核心部分的新世界中，哪些人做得好。当你招聘 PM、产品人员时，当你寻找在当今世界表现出色的人时，你注意到了什么？你更看重什么？你发现哪些特质变得越来越重要，哪些正在变得不那么重要？

I want to talk a little bit about how the product role is changing and who is doing well in this new world now that AI is such a core part of our life. When you're hiring PMs, product people, when you're looking at people that do well in today's world, what are some things that you notice? What are you looking for more? What's trending up in what you find is important and what's trending down?

实际上，在我的团队中，我们三年来没有改变招聘流程。所以我们真正寻找的、评估的通才特质，比如 PM、研究产品经理，其实一直没变。我认为最重要的特质是第一性原理思维。这指的是不要用你之前在消费产品或 B2B SaaS 中使用的模式匹配，而是真正去弄明白在这个时刻、为这个用户群体、用这项技术，用户价值是什么。

We actually on my team have not changed our hiring loop for three years now. So what we actually look for and the traits and how we evaluate generalists like PMs, research product managers have actually been the same. I think some of those traits number one is first principles thinking. And this is really rather than pattern matching what you used to do in, let's say, consumer product or B2B SaaS, but actually figuring out in this moment for this user group with this technology what is the user value.

能举个例子吗？很多人听到“第一性原理思维”就会说：“我知道，我很擅长这个。”那么，真正展示出出色第一性原理思维的例子是什么呢？

Is there an example? A lot of people hear "first principles thinking" and they're like, "Yes, I know about it, I'm good at this." What is an example of someone having really demonstrated really good first principles thinking?

我认为一个例子是：你可能会认为产品经理负责产品策略并交付用户价值，日常表现为编写 PRD 或产品愿景文档。对于我的团队（研究产品经理）来说，驱动用户价值的方式是找到正确的用户反馈和评估（evals），对吗？这可以成为用户需求的人格化体现。所以我们确实会写一些产品文档和 PRD，但我们在团队里有一句话：评估就是新的 PRD。因为要交付用户价值，关键不在于过去一二十年人们写的那种具体文档，而是一种新的工作方式。所以第一性原理思维会是：让我弄清楚为了实现目标我应该做什么，而不是说这里有一堆我做过的活动，所以我将继续做下去。

I think one example is: you think of a product manager as owning product strategy and delivering user value, but demonstrating it day-to-day by writing a PRD or a product vision doc. For my team, as research product managers, the way to drive user value is to figure out the right user feedback and evals, right? That then can be a personification of that user need. So we do write some product documents and PRDs, but we actually have a saying on the team: evals are the new PRDs. Because in order to deliver that user value, it's not that exact artifact that people used to write in the last one to two decades; it's a new way of working. So the first principles thinking would be: let me figure out what is the thing I should do to achieve my goals, rather than here is a set of activities that I've done and therefore I will continue to do.

所以以前的做法是：有一个想法，创建 PRD，与人讨论，对齐计划，设计，构建，发布，看看效果，迭代。我现在听到的是：现在，如果有一些关于问题或机会的反馈，第一步是通过评估（eval）来定义工作是什么，而不是通过 PRD。

So the idea here is it used to be: have an idea, create a PRD, talk to people about it, align on the plan, design it, build it, ship it, see how it goes, iterate. What I'm hearing here is: okay, here's some feedback about something that's wrong or an opportunity. Step one is the eval is now how you define what the work is, versus a PRD.

也许第一步是理解用户的痛点。而接触用户痛点的方式也不同，对吧？过去，我们可能会进行用户访谈，如果深入下去，可能让用户带你走一遍他们的用户流程和界面像素。但在这里，你必须像关注像素一样关注词元（tokens）。所以我们团队的一项活动是阅读转录文本，深入理解失败的轨迹，然后判断：这是幻觉吗？Claude 是否过于自信？失败的主题其实有很多细微差别，这让你能够构建对该痛点的持续描述。这本质上可以成为一个新的评估。而且评估的分布：它是否既捕捉到了失败的正面情况，也捕捉到了不应该失败的领域？然后反馈给研究团队，以便我们进行改进并实际衡量质量。当我们有 Opus 5.5 时，这个领域是否在改善？Claude 现在能否识别文档中的正确位置并提取正确的综合信息？所以关键就是可操作性，以及缩短为利益相关者和合作伙伴团队（如研究人员）提供可操作信息的距离。

Maybe step one would be understanding the user pain point. And the way to even access that user pain point is different, right? In the past, we might do a user interview and if you go deep enough, you might have the user walk you through their user flow, the pixels. Here, you have to sweat the tokens as much as you sweat the pixels. So one activity we have on the team is reading the transcripts and understanding the trajectories that failed very deeply, to then say: was this a hallucination? Was Claude being overconfident? So the theme of the failure actually has a lot of nuance, and that allows you to build a sustained description of that pain point. So that could be essentially in a new eval. And the eval on distribution: is it capturing both the positive situations where this is failing and also areas where it should actually not fail? And then bring that back to research so we can make improvements and actually measure the quality. When we have Opus 5.5, is this area improving or not? Is Claude now able to identify the right places in the document and pull the right synthesis out? So it's just the actionability and shortening the distance to actionability for our stakeholders and partner teams like researchers to take action on.

有这样的例子吗——你发现了一个问题或机会，然后编写了评估？在大多数情况下，评估是什么样的？当人们想想象一个评估时，他们脑海中出现的是什么？

Is there an example of something like this where you found an issue or opportunity and then wrote the eval? And what is the eval looking like in most cases? When people want to picture an eval, what do they picture?

我们实际上在 Anthropic 内部开创了这个概念。早期的一个例子是，早期的 Claude 模型不太擅长遵循特定的模式（schema），比如输出 JSON 格式。而这现在是 Claude 成为一个优秀智能体的基础，对吧？如果你不能输出特定格式，你就无法访问 API，无法调用工具等等。

We actually pioneered this concept within Anthropic. So one of the early examples is the early Claude models were not very good at following specific schemas. So things like outputs in JSON, and now that is fundamental to Claude being able to be a good agent. Right? If you can't output a certain format, you don't know how to access APIs, you can't call tools, etc.

那么，现在在 Anthropic，编写评估（eval）是不是已经成了产品经理工作的核心部分？

So is this just a core part of the product management job now at Anthropic, writing evals?

我认为是的。我还跟其他公司的产品经理聊过，我觉得这正成为一项越来越普遍的技能，因为我们构建的很多产品都处于模型、框架和用户场景的交汇处。像评估这样的工具，不仅对研究模型的人有用，对产品团队也有帮助，能带来更好的用户体验，因为你无法改进无法衡量的东西。很多工作仍然非常依赖触感和判断，所以你必须贴近细节。

I think so. I also think it's something I've talked to other PMs at other companies about, and I think it's more and more of the skill set broadly, because a lot of the products we're building are at the intersection of models with harnesses with a set of contexts for a set of users. Having things like eval is a way not just for folks working on models, but generally within product, to get to better user experiences because you can't improve what you can't measure. A lot of this is still very tactile-based, still very judgment-based, so you have to stay close to the details.

而且这个过程非常不具确定性，这也是一个重点。它不会每次都给你同样的答案，所以你必须更宽泛地描述它，不会是完全精确的匹配。所以这是产品工作方式一个非常有趣的变化，而且未来也会这样——评估是其中的重要组成部分。

And also very non-deterministic, which is a big part of this. It's not going to give you the same answer every time, so you have to describe it more broadly. It's not going to be an exact match. So this is a really interesting change in the way product happens and will happen — eval is a big part of this.

你们现在还写 PRD（产品需求文档）吗？还是说仍然有那种一页纸描述问题，或者全是评估？好的，你在摇头。是说还在用吗？

Do you guys still do PRDs? Is there still like a one-pager describing a problem, or is it all evals? Okay, now you're shaking your head. Yes?

我们确实还在用。我认为当问题非常明确时，评估可能是一种简化的方法。但在其他情况下，PRD 仍然非常有价值。PRD 是让一大群人围绕体验和目标的真相来源达成一致的好工具。所以每个模型我们都有 PRD——不一定非给研究员看，而是面向我们不断增长的产品表面、工程团队，以及法律、安全等利益相关方，作为一个真相来源，把我们要实现的目标整合起来，让一大群人朝着同一方向努力。PRD 的另一个价值在于那些更模糊的问题和机会。比如我们还没推出像“计算机使用”这样的功能时，不一定有具体的用户痛点。PRD 中的产品愿景部分有助于探索可能性——即使一项技术尚未准备好为所有人服务，如何让它为某个群体良好运转？你可以探索价值，为用户群体带来连贯的东西。所以我们确实有 PRD。只是现在应用方式有些不同了。

We do. I think when there's a very defined problem, things like eval might be almost a shorthand. But there are other cases where PRDs are really valuable. PRDs are great vehicles for getting a very large group of people aligned on a set of sources of truth about experience and setup goals. So for every model, we do have a PRD — less necessarily for our researchers, but more for our growing product surfaces, for our engineering teams, for our stakeholders like legal and safety, as a source of truth of putting together what we're aiming to achieve so that a big group of people can row in the same direction. The other place where PRDs are valuable is on the more ambiguous problems and opportunities. If we haven't shipped a thing like computer use, we don't necessarily have a set of user-specific pain points. There's value in the product vision portions of a PRD to explore what could be, even if a technology is not yet ready to work for everyone. How do you get it to work well for some group? You can explore the value and bring something that is coherent to a user group. So we do have PRDs. I think the application is a little different now.

好的，太好了。我刚刚采访了 Andrew（他是 OpenAI Codex 应用的负责人），他说你们观点一致：PRD 没死，对特定项目和想法仍然非常有用。很好。好了，我们给 PRD 下了定论——它依然活跃。

Okay, this is great. I just had Andrew (he's the head of the Codex app at OpenAI) and he said you guys are aligned: PRD is not dead, still very useful for specific projects and ideas. Great. Okay, we've closed the book on PRDs are still kicking.

我们刚才聊了产品人员需要哪些新技能。你是否还发现，在这个新 AI 世界里，那些做得好的产品经理和产品团队成员，他们身上还有哪些共同模式发生了转变？有没有什么特质是你更看重的？尤其是对于处于职业生涯中期或者更多担任管理/产品领导角色的人？

We've been talking a bit about just what kind of skills are emerging for product people. Is there anything else that you find has shifted in what patterns are common across people that are doing well in this new AI world, in terms of product managers and folks on the product teams? Something you look for more in people? Maybe specifically for folks who might be mid-career or folks who have been more in a managerial, product leadership seat.

我强烈认为：要成为出色的团队经理和与这项技术打交道的产品经理，你自己必须非常亲力亲为，不仅花时间捣鼓，还要真正用这项技术交付产品。也就是说，要深入细节，和你的产品经理、工程师、团队一起“抠 token”。所以我招来的资深产品经理，他们的入职计划和早期职业者完全一样：理解用户、阅读经用户同意的反馈、与客户交谈。我认为关键在于能够理解如何处理问题、什么是好的样子，并且通过非常亲自动手的方式来培养这种能力。如果一个人没有亲身构建过 AI 产品，他很难认同或看出一个好的 AI 产品/功能应该是什么样子。所以我强烈认为，如果你是一名经理，你必须亲力亲为。你必须花一部分时间实际交付产品。你必须设身处地站在团队的角度。我总是设法留出一部分时间，在模型发布时亲自负责一两个工作流，以保持我对模型动向的感知、对模型改进速度的判断，从而帮助团队做出更好的决策。

One thing I feel pretty strongly about is: in order to be good managers of teams and PMs working with this technology, you have to be really hands-on yourself and have spent not just time tinkering but actually shipping with this technology. Again, being in the details and sweating the tokens along with your PMs and your engineers and your teams. So even for folks I hire who have more tenure PM experience, the onboarding plans are exactly the same as somebody who is more early career: it's around understanding users, reading consented user feedback, talking to customers. I think there's something around being able to understand what to do with this, what good looks like, and having developed that in a very hands-on manner. It's not necessarily easy for someone to agree or be able to see what a good or great AI product or AI feature could look like if they haven't experienced building themselves. So I do feel pretty strongly that if you're a manager, you have to be hands-on. You have to spend a portion of your time actually shipping. You have to kind of walk in the shoes of your teams. I always try to carve out a portion of time to actually own one to two work streams when we have models, in order to keep my theory of mind, keep my sense of how the models are moving, how quickly it's improving, so I can help the team make better decisions.

所以，我从这里听到的是：不管你在公司的职级阶梯上处于什么位置，如果你自己不亲手搭建、不真正和 Claude 对话、不真正和 Codex 对话、不去构建东西，你就做不好这份工作。

So, what I'm hearing here is: if you're not, no matter where you are in the ladder of hierarchy at a company, if you're not building yourself, if you're not actually talking to Claude, talking to Codex, building stuff, you're not going to make it.

而且你应该享受和这项技术一起工作的乐趣。我觉得这是另一部分。

And you should have fun working with this technology. I think that's the other piece.

我认为无论处于哪个级别，最能成功的人是那些热爱与 AI 一起工作、积极探索和实验的人，他们不仅花时间做实验，而且真正动手从端到端交付产品并获取用户反馈。我认为这对每个人来说都应该是基础。

I think the folks that would be most successful regardless of their level are people who love working with AI and are exploring and experimenting and carving out the time not just for the experimentation but actually hands-on shipping end to end getting the user feedback. I think has to be fundamental for everyone.

我百分之百明白你的意思。就像我坐在我的新闻通讯和这个播客上谈论这些事，感觉听起来很不错。但每次我真正动手搭建一些东西，摆弄各种小项目时，你就会觉得：“好了，我看到这里发生了什么。”你会获得更多体验，很难准确描述和模型一起工作、构建东西的实际感受，但那完全是另一个世界：“好了，我明白了，这就是他们所说的计算机使用，这就是他们说的这个限制、这个用户体验情况。”

I 100% know what you mean there. Just like me sitting on my newsletter and this podcast just talking about stuff and like yeah that sounds great. Like every time I actually build something and I tinker with all kinds of little projects, you're just like, 'Okay, I see what's happening here.' And you just get so much more, it's like hard to exactly describe what you experience actually working with the models and building stuff, but it's like a whole different world of like, 'Okay, I see. Here's what they're talking about computer use. Here's what they're talking about with this limitation, this UX situation.'

是的。

Yeah.

所以，是的。就是这样，你提出了一个非常有趣的观点——你必须从中找到乐趣，这对很多人来说并不容易，因为他们是被迫使用 AI，或者根本不知道该用它做什么。对于一些人来说：“我不知道，这太烦人了。我只是必须这样做。我不知道为什么，我讨厌这该死的东西。为什么我要和它一起工作？变化太多了。我累了。”对于如何帮助人们找到工作中的乐趣，你有什么建议？

So, yeah. So, it's just like, and you made this really interesting point that you have to have fun with it, which is not easy for a lot of people because they're pushed to use AI or they just don't know exactly what to do with it. For people that are just like, I don't know, it's just so annoying. I just have to do this. I don't know what's so like, I hate this freaking thing. Why do I have to work with this? Things are changing so much. I'm tired. Uh, advice for helping people find that joy in this work.

我想重申我之前说过的一个观点：实验不是一项个人运动。我接触过几乎所有版本的研究模型，以及 20 个版本的 Claude 生产模型，我认为乐趣的一部分来自于看到其他人也发现了使用案例。所以一个想法是，与一个充满热情的人配对，针对你关心的用例，一起探索发现，而不是自己想办法找到完美的用例——因为那可能感觉像工作。与他人一起工作往往感觉像是在享受乐趣。我们还能做些什么来带动其他人？很多时候，在公司内部，如果有一个人非常好奇，分享一个新原型的想法，就会吸引一大批人，他们会说：“哦，我不知道现在 Claude 可以实现这个了。”这就形成了一些良性循环，以及继续享受这项技术的方式。

I think maybe I'll reemphasize something I said earlier around just that experimentation is not an individual sport. Like some of the moments where I think I've touched practically every version of research models across 20 versions of production Claude at this point and I think part of the joy comes from seeing other people discover use cases too. And so maybe one idea here would be pairing with somebody who is excited and seeing what on a use case that you care about and working together versus identifying or trying to figure out the perfect use case yourself because that might feel like work. Working with others feels like joy a lot of the time. And is there more that we could do to bring other people along? That's something like a lot of times internally we have somebody who is very curious and them sharing an idea of a new prototype actually brings a ton more people who are like 'oh I didn't know this could work now with Claude' and so there's just some virtuous cycles here and ways to continue to have joy with this technology.

这点说得太好了。我认为这也是为什么 Twitter 在这方面如此有用——你看到别人分享他们做过的东西，这会激发你自己想出一些小小的创意，而且分享自己做过的事情也很有趣。

That's such a good point. I think that's also why Twitter's so useful for a lot of this is you see other people sharing what they've done and it inspires you to come up with your own little ideas and also it's just like fun to share your own thing that you've done.

所以这是一个很好的观点——找其他人一起玩耍，寻找用例。我还经常听到的是，找到你在生活或工作中想要解决的一个问题，然后打开 Claude Code，告诉它你想做什么，仅凭一个模糊的想法，你就能取得令人难以置信的进展。

So that's a really good point just like find other people to kind of play around with and look for use cases. The thing I've also heard a lot is just find like a problem you want to solve in your life or work and just open up Claude Code tell it here's what I want to do and it's incredible how far you can get just with like a vague idea of a problem you want to solve.

是的。我认为难点在于有太多不同的东西可以尝试。

Yeah. I think it gets hard in that there's so many different things that you could try.

是的。

Yeah.

所以你要么专注于和某个人配对，与对这项技术充满热情的人一起工作，要么找一些你能立即发现价值的事情。无论哪种方式，都能让你更深入，而不是停留在很多事情的表面。我很难跟上那么多原型和产品的步伐，所以我的方法是自己在其中一两个上深入下去。

And so you just like narrowing in on either pairing with someone, working with somebody who has a lot of joy about this technology or figuring out something that you could immediately find value. Like either of them those things allow you to go deeper rather than like more high level about too many things. I find it hard to keep pace with the number of prototypes or products that are out there and so my lens has been how do I go deep in one to two of them myself.

你这么说太有趣了，因为事实正是如此。我们刚刚做了一项调查，是我和同事 Noam 一起做的，询问我的读者们对当前科技和 AI 发展的感受。我们最有趣的发现之一就是，幸福感正如你所说：在几件事上深入，而不是尝试一大堆小事情。找到几件真正能解决好的事，然后深入下去。这是一种源泉，因为人们感到幸福往往是在他们终于找到一种方法，让 AI 真正改善他们的生活，而不是得到一些乱七八糟、支离破碎、半成品的东西。

That's so interesting you say that because that's exactly it. We just had this survey that I ran with my colleague Noam asking my readers just how they're feeling about all the things going on in tech right now and AI. And one of the most interesting takeaways we had was to find that happiness is exactly what you said: go deep in a couple things versus trying to just a ton of little things. Find a couple things to really solve well and then go deep. And that is a source because a lot of the happiness people feel is when they finally unlocked a way for AI to actually make their lives better versus just a couple messed up, broken half-working things.

是的，关键是如何从“打勾完成”变成真正有价值。作为产品人，这是对你时间和精力的产品优先级排序。如果目标是为了快乐而实验，那么你需要哪些输入？但我觉得 Anthropic 的很多秘诀在于文化和自下而上的工作方式，以及公开实验。通过这样做，我们带动了其他人。我认为这最终非常有价值。

Yeah, it's how do you go from this being a check the box, right? And so like us as product people, it's then an exercise of product prioritization of your time and your energy. If the goal is to experiment with joy, then what are the inputs that you need for that? But yeah, I think a lot of the secret sauce of Anthropic is the culture and the bottom-up nature of how people work and this like experimenting in public. And by doing that, it's very much about how to bring other people along. That ends up being, I think, really valuable.

是的，我从各个实验室听到过很多次——没有人确切知道某些东西将如何被使用，很多时候就是尽早发布，看看人们如何使用，看看什么是可能的，然后利用这些信息来构建实际的产品方向。

Yeah, I've heard this so many times from all the labs just like no one's exactly sure how some of this is going to be used and a lot of it is just putting stuff out early, seeing how people use it, seeing what's possible and then using that information to build the actual product to lead in.

是的。是的。

Yeah. Yeah.

我很好奇，沿着这条寻找 AI 帮助工作和生活方式的思路，你作为产品经理，最近有没有用 Claude 做一些有趣的用途？

I'm curious how kind of on this thread of finding ways AI to help you in your work in life. Are there any interesting ways you've been using Claude lately in your work as a PM?

我认为有很多事情，比如 Fable 和 Tag。Tag 还非常早期。我认为这里有一种不同的工作范式：让一个智能体去工作，然后带回产品和体验给你。一个不太近期但我和团队经常讨论的话题是，我认为我们可以更多地用 AI 来改善彼此之间的对话，成为更好的管理者。我觉得这不只是提升我们构建体验的智商，我自己也大量用它来准备如何在关键对话中即时进行更好的交流。我很喜欢那本书，所以我实际上有一个技巧，帮助我判断在当前情况下我是否使用了适当的细节层次，这确实帮助我成为更好的管理者和团队支持者。

I think there's a lot of things with Fable and Tag. I think Tag is in the very early days. I think there's something around how you work in a different paradigm of allowing an agent to go off and work and then bring back product and experiences to you. I think one area that it's not very recent, but one that I bring up a lot with the team and I think we could do more on using AI is just like how to use it to also be more to have better conversations with each other to be better managers. I don't think it's necessarily just about raising the IQ of experiences we build, but also I used it a lot in prepping for how to have better conversations in the moment during crucial conversations. So, I love that book and so I actually have a skill that helps me figure out am I having the right level of detail given the situation at hand and actually helping me be a better manager and better supporter for the team.

对于团队里的管理者，我一直在分享如何用 Claude 成为更好的教练。有时候很难找到合适的措辞，而模型有很多完美的词。我认为它可以从伦理角度提升我们。

For managers on the team, I've been sharing how to use Claude to become a better coach. It's hard sometimes to find the right words, and the models have many perfect words. I think it can augment us from an ET perspective.

哇，这里面有太多有趣的东西了。我先确认一下你在做的事：你建了一个技能，让 Claude 基于《关键对话》这本书构建一个技能，它本身对这本书已经足够了解，你甚至不需要提供内容。然后你用这个技能和 Claude 交流，比如“我马上要和一个同事进行一次非常困难的对话，给我一些建议该怎么应对。”

Oh man, there's so much interesting stuff there. So to understand what you're doing: you built a skill. You just had Claude build a skill pulling in lessons from the book Crucial Conversations, which it knows enough about. You don't have to give it the content. Then you use that skill to talk to Claude: 'Hey, I have this very difficult conversation coming up with a colleague. Give me some tips on how to approach it.'

是的，这就像个性化的辅导。我们整天都在做大量的上下文切换，让 Claude 帮我配对、帮我梳理，有时候我最终没有采纳它的建议，但它对头脑风暴非常有帮助。我在想我对反应的思考方式对吗？如何更快地深入、更快地建立信任、更直接。

Yeah. It's almost like individualized personalized coaching. There's so much context switching we do all day, and having Claude help me pair and help me, sometimes I don't end up using its suggestions, but it ends up being very helpful for brainstorming. Am I thinking about reactions in the right way? How do I go a bit deeper faster? Build trust faster, be more direct.

是的，我有很多问题。这太有趣了。人们有一个担忧：他们会开始像 AI 写作那样说话。我知道你没有那样，但这确实是一个担忧。你害怕这种大脑退化吗？我们如此依赖 AI，以至于停止学习和思考，完全与 AI 融合在一起。

Yeah, man. I have so many questions. This is so interesting. One concern people have is that they'll start talking like AI writes. I know you're not doing that, but it's a concern. Do you fear this brain rot and atrophy? We're so reliant on AI that we stop learning and thinking, being overly integrated with AI constantly.

对我来说，思考和写作是紧密相连的。我用 Claude 来增强思考，但确保它不会接管我的思考。根据情况，我可能会先形成自己的观点，再和 Claude 协作，保持自己的感觉和语气。其他事情比如月度业务回顾，我希望能获得标准、清晰的信息。写作部分可以完全交给 Claude，让我更多成为审查者。所以这取决于你用 Claude 做什么，以及把更多任务委托给 Claude 是否存在不对称的价值。

For me, thinking and writing are tied together. I use Claude to augment my thinking, but I make sure it doesn't take over. Depending on the situation, I might develop my own point of view first and then work with Claude, maintaining my sense and tone. Other things like monthly business reviews, I want standard, crystallized information. The writing can be delegated to Claude fully, making me more a reviewer. So it depends on what you're using Claude for and if there is asymmetrical value in delegating more to Claude.

我还听到第一个建议很棒：先思考，有自己的观点，然后把 Claude 当作切磋伙伴来推进想法、对想法提出反驳。

What I'm also hearing the first tip is really great: think first, have a point of view, then use Claude as a sparring partner to evolve the idea, push back on the idea.

是的，这就是对齐研究有价值的地方。你不希望 AI 只是赞同你。你希望它能提升你，达成更好的结果。让 Claude 提出反驳让我变得更好。就像同事一样，我希望在我想法不成熟时有人提出反对意见。

Yeah. This is where alignment research is helpful. You don't want an AI that just agrees with you. You want it to augment and get to a better outcome. Having Claude push back makes me better. Like a co-worker, I want somebody to push back when my ideas are not fully formed.

我想多听一点。我听 Ben Mann 说过 Claude 有一套宪法。直觉上，专注于安全和对齐应该会限制 Claude，但结果恰恰相反：Claude 是最有趣的个性。人们更喜欢和 Claude 聊天。为什么这种对对齐和明确宪法的关注反而让 Claude 变得更好、更有趣？

I want to hear more about that. I've heard from Ben Mann that Claude has a constitution. Intuitively, focusing on safety and alignment should limit Claude, but the opposite happens: Claude is the most interesting personality. People prefer talking to Claude. Why does this focus on alignment and a clear constitution make Claude better and more interesting?

为了让 Claude 尽可能智能和强大，它需要在适当的时候提出反驳。比如，我问一个研究版的 Opus 关于下一版 Claude 的定价问题，它帮助我们得到了更好的结果。AI 不应该只是助手或执行者，而应该弄清楚自己是否在做正确的事，这与知道何时反驳是分不开的。主动性不是按计划做事情，而是知道何时想出新的主意。要让 Claude 更有用，它必须知道何时反驳。这是核心特性。

To make Claude as intelligent and capable as possible, it needs to push back at the right points. For example, I asked a research version of Opus about pricing for the next version of Claude, and it helped come to better outcomes. AI should not just be an assistant or doer, but figure out if it's doing the right thing, which is integrated with knowing when to push back. Proactivity is knowing when to come up with a new idea, not just doing scheduled tasks. For Claude to be more useful, it must know when to push back. That's a core characteristic.

这太有趣了。不那么顺从反而让它更好、更有用。我听过很多人说“AI 告诉我我是对的”，但我希望是相反的。

That is so interesting. It being less compliant makes it better and more useful. I've heard many people say 'AI told me I was right,' but I wish the opposite.

是的，这又回到保护你的思考上。如果你有一个 AI 作为思考伙伴，它不会只是赞同。它会给你增加价值，最终你会得到更好的想法。

Yeah. It comes back to protecting your thinking. If you have an AI that is a thinking partner, it doesn't just agree. It adds to you, and you end up with better ideas.

那应该是英雄目标，而不只是把你的想法改进 10%。是的，我很喜欢这个，因为以前人们常说要思考 10 倍。我记得创始人会这样推动人们：如果我们把这个做到 10 倍会怎样？而我最近常听到的是：我们如何从这个想法实现千倍的增长？什么是这个想法最雄心勃勃的版本？我想回到之前我们谈论经常与 Claude 交流时想到的一点。很明显，当 AI 写了什么东西时，仍然能看出来。有趣的是，它是一个大型语言模型，你会认为它最擅长写作，但奇怪的是，AI 并不擅长写作，总是一眼就能看出是 AI 写的。你觉得我们会到达一个无法分辨这是 AI 写的的地步吗？

That should be the hero goal, not just making your ideas 10% better. Yeah, I love this since it used to be think 10x. I used to be the way you know founders push people like what if we 10x this and I love what I keep hearing is like it's like how do we go thousandx from this idea? What is the most ambitious version of this? I want to come back to something that I was thinking about as we were talking about talking to Claude constantly. It's very clear when AI has written something still. It's funny that it's a large language model. You would think of all things it would be very good at writing and interestingly just no AI is very good at writing it's always very clear this was AI written. Do you think we'll get to a place where we will not know this was AI?

我认为这取决于你通过知道（是否为 AI 所写）想要达到什么目标，以及评估标准是什么。我确实认为我们可以在让 Claude 写得更好方面做得更多。实际上，我的团队和研究方面都有非常积极的努力，致力于让 Claude 写得更好。总的来说，我认为应该清楚一个想法是由你、Lenny 还是我 Diane 主导的。我认为这真的取决于写作的目标。比如像月度业务回顾这样的东西，我其实很乐意让 Claude 从头到尾写出来。呃……

I think it depends on what's the goal that you're looking to achieve by knowing yeah uh what's the eval. I actually do think there's more that we could be doing on making Claude write better. There's actually very active efforts on my team and on the research side about making Claude write better. Just generally I think it should be clear where an idea is being led by you or by you Lenny or me Diane. I think it really depends on what's the goal of that writing. Like for something like a monthly business review, I would actually love to have that end to end be written by Claude. Uh,

并且显然，不让它感觉像是人类写的。你提出的这一点很有意思：我们是否真的更适合知道它是 AI 还是不是？

And obviously and not make it feel like it was written by a human. It's such an interesting point you're making like is it actually better for us to know that it's AI versus not.

是的。但也许更关键的视角是可验证性，或者是谁在验证。

Yeah. But it's also for maybe the lens is more around like verifiability or who's verifying

输出。对，对。比如谁在签字确认。

The output. Right. Right. Like who's signing off.

呃，也许更少关注谁写的，而是谁在验证，谁在签字确认。这比谁写的更重要。

Uh maybe less around who's writing, but who's verifying who's signing off. That becomes more what matters than who's writing it.

你认为为什么 AI 不太擅长写作？我猜它学习了人类所有最好的写作，已经找到了最好的写作方式。而现在我们认识到，好吧，AI 就是这样，它有这些套路。这是核心原因吗？还有别的因素阻止它成为优秀的写作者吗？讽刺的是，作为一个大型语言模型，你本以为它应该非常擅长语言。

Why do you think AI is not great at writing? Like my guess is it has studied all of the best writing in all of humanity. It's figured out here's the best way to write. And now that we've recognized, okay, this is what AI does. It has these tropes. Is that the core of it? Is there something else that's keeping it from being a great writer? Ironically, being a large language model of all things, you think it'd be really great at language.

我认为部分原因是我们需要在训练改进上投入更多，以使 AI 在写作等领域持续强大。另一方面，这项技术有锯齿状的边缘，就像我们之前提到的。所以有时候模型在写作方面不错，但不够智能体式，我们的目标是如何让模型更具智能体式或调用正确的工具。现在这方面有所改进，于是其他领域就成了新的粗糙边缘。我认为我们在写作方面正处于这样一个时刻：我们需要真正专注于优先训练模型，使其在这个领域表现出色，而这也是你提到的一个非常活跃的领域。

I think part of it is also we need to invest more in training improvements to make AI continuously strong on areas like writing. I think it's also like the technology is jagged edged, like we mentioned. So sometimes when the models were good at writing but not agentic, our thesis is how do we make the models more agentic or call the right tools. Now that that's improved a bit then it's well now these other areas actually become more of the rough edges. And so I think we're in one of those moments with writing where we need to actually just focus and prioritize on training the models to be great at this area and that is an active very active area for us that you mentioned.

好的，我很高兴。另外，一旦 AI 变得非常好，以至于我们不知道是谁写的，但正如你所说，有时我们确实想知道它是 AI。这真的很有趣，我从没这么想过。另一个有趣的点是，有个喜剧演员开玩笑说，我们在飞机上，Wi-Fi 断了，我们会抱怨“搞什么鬼？飞机上的 Wi-Fi 坏了，太烂了，你怎么敢？”而当你在一个天空中的管子里像鸟一样飞翔时，你怎么敢抱怨 Wi-Fi 不好用？你的观点是，有如此多的进步和力量，我们无法解决所有问题，无法让一切做到最好。所以基本上 AI 写作一直不是优先事项，而且感觉那里有更多的投资。

Okay. I'm glad I'm glad. And also it's going to be interesting once AI is so good we're like I don't know who wrote that but to your point sometimes we actually want to know that it's AI. That's really interesting. I never thought of it that way. The other interesting part of this is that there's that comedian who was joking that we're like on a plane and the Wi-Fi is down and we're just like, "What the hell? The Wi-Fi is not working on this plane. This sucks. How dare you?" When you're in a tube in the sky flying like a bird and how dare you complain that the Wi-Fi doesn't work. Your point is there's so much advancement and so much power. We can't fix it all. We can't make it all work the best possible. And so basically AI writing has been not the priority and it feels like there's more investment happening there.

是的，我认为语气和个性是一个优先事项。这项技术的进步是一个持续进行的工作，所以我们看到了智能体式行为的飞跃或出现，那是一个新常态，然后其他能力需要继续改进。

Yeah, I think tone and character is a priority. I think this advancement of the technology is a work in progress and so we see a leap or emergence of a jump in agentic behaviors and so that is a new normal and then these other capabilities needs to continue improving

而且我认为一旦我们改进了，比如说写作、语气和个性，我们大概就会说……

And I think once we improve let's say writing and tone and character we probably will say like

我们如何让 Claude 更加积极主动？比如生产力是一个机会，这是人类的本性——我们想让自己变得更好，想让这项技术变得更好。所以是的，我认为我们正在将这种思路应用于 AI，这是正确的做法。我们应该让它变得更好。

how do we have Claude be even more proactive like productivity is an opportunity and that's human nature like we want to make ourselves better. We want to make this technology better. So yeah I think we're applying it to AI which is the right thing. We should be making it better.

我想问你几个问题，这些问题我想问那些站在未来最中心的人。第一个是：你认为未来几年人类大脑在哪些方面会继续保持最大价值？我知道 Anthropic 的使命和愿景是我们将达到 AGI（通用人工智能），达到超级智能。所以未来可能没有价值，但在那之前，随着我们接近那个时间线，你认为人类大脑在哪些方面会继续保持最大价值？

I want to ask you a couple questions I'd like to ask folks working at the very center of the future that is coming. One is where do you think human brains will continue to be most valuable over the years? I know Anthropic's mission and vision is we'll reach a AGI, a superintelligence. So in the future maybe nowhere but before we get there where do you think human brains will continue to be most valuable as we've approached that timeline?

我们开始谈论让 Claude 和模型在判断力方面变得更好，尤其是在过去一年左右。我认为判断力是一个领域，它积累了大量的细微差别和经验，而这些系统还没有像人类那样经历过那么多，所以我认为来之不易的判断力是产品领导者乃至整个领域将继续至关重要的因素。AI 可以构建很多东西。但实验室应该构建哪些东西？这很大程度上需要人类的判断力、持久力和主动性——这些特质超越了通用能力，而是关乎人们如何找到最佳解决方案、如何创造最佳体验的行为和特征。所以我认为这些特质实际上是具体的、会继续重要的特质。我认为还有许多能力和专业领域知识。比如软件工程已经被 AI 彻底改变。还有像生物学、生命科学等领域。这些领域我们才刚刚处于指数增长的起点，而软件工程可能已经在指数曲线上。有些领域我们还没达到。所以你会看到我们推出像 Cloud Science 这样的产品并投资这些领域，因为这些领域能将这项技术带到社会并产生积极效益。所以我认为还有很长的路要走。

We started to talk about making Claude and models better at judgment especially in the last year or so. I think judgment is one and is an area where it's an accumulation of so much nuance and so much experience and these systems haven't experienced as much as humans have and so I think that hard-earned judgment is an area for product leaders and just generally will continue to be really critical. There are so many things AIs can build. Which one are the things that a lab should build right? A lot of that requires human judgment persistence proactivity these are all traits that are beyond just general capabilities but just behaviors and characteristics of people at that level of how do you get to the best solutions how do you create the best experiences so I think those types of traits are actually the tactile traits that I think will continue to be important. I think there is also still a lot of capabilities and subject matter expertise as well. I think software engineering has been really transformed by AI. I think there's areas like biology, life sciences. These are all things that we're just kind of at the foot of the exponential on like maybe software engineering. We're on the exponential on some of these areas. We're not quite there yet. And so I think you're seeing us ship things like cloud science investing in these areas because those are areas that I think bring this technology to society and having a positive benefit for society. So I think there's a lot more to go there.

我还有一个问题想问：作为有孩子的人，你如何看待鼓励他们学习什么？或者你会不会引导他们在这个我们正进入的狂野新世界中获得成功？

Another question I want to ask is, as someone with kids, how do you think about what you are encouraging them to learn? Or do you think you're going to nudge them to be successful in this wild new world that we're entering?

我觉得其实很多品质和我们自己成长过程中培养的是一样的：学习的好奇心、毅力、相信自己内心的声音，然后不断发展这种声音。我有两个孩子，一个四岁，一个八岁。帮助他们发展内心声音是我的责任，比如让他们有主见、敢于表达立场，并鼓励这一点。我认为这些技能在未来很重要，比如拥有自己独特的声音。

I actually think it's a lot of the same traits that you and I probably grew up with, which is curiosity for learning, persistence, believing in your own inner voice, and developing that. Like I have a four-year-old and an eight-year-old. It's on me to help them develop their inner voice, whether that's being opinionated, taking a stance, and encouraging that. I think those skill sets are important in the future, like having their own individual voice.

这太有趣了。这和你之前回答如何避免思维僵化以及过度依赖 AI 时的说法很相关——就是不要过度依赖 AI，先保持自己的观点。你描述的从小培养孩子这种能力的想法真的很重要。我很喜欢这一切如何串联起来：判断力、毅力、自己的观点。

That is so interesting. It's so related to the answer you had when I asked about how to avoid brain rot and overrelying on AI, which is to keep focused on your own point of view before you overly rely on AI. And this idea you're describing of building that in kids is really important. I love how this all connects: judgment, persistence, a point of view of your own.

嗯。

Yeah.

既对孩子，也对成年人。

Both for kids and also adults.

嗯。

Yeah.

哎呀。我想的问题是：什么时候让孩子接触 AI 合适。我有个三岁的孩子，现在还太早，但你是如何引导他们进入这个疯狂世界的？最近参加一个活动，很多父母在讨论如何让孩子接触 AI。有个人有个很有意思的方法：让他们用早期的模型，这样他们还需要稍微费点劲，不能立刻得到所有答案。我觉得很有趣，比如开源的本地模型。

Oh man. Well, the question I'm thinking about is just when to get them on some AI thing. I have a three-year-old, so it's pretty early, but how do you onboard them to this crazy thing? I was at an event recently and a bunch of parents were talking about how they think about AI with their kids. One person had a really interesting approach: keep them on the very early models so they still have to struggle a bit and not get all the answers immediately. I thought that was interesting, like an open-source local model.

不稳定。

Not stable.

是的。

Yeah.

嗯。好奇心是我一直提到的，Ben Mann 对这个问题的回答一直让我印象深刻：好奇心，而且他非常推崇蒙特梭利教育，我们也在鼓励孩子这种方式。所以这里面是有道理的。

Yeah. And curiosity is something I keep mentioning Ben Mann, but his answer to this question has always stuck with me: curiosity, and he's a big fan of Montessori, which is what we're encouraging for our kids. So there's something there.

最后一个类似的问题。这是最近上过播客的 Fiona Fung 建议我问你的。作为一名在 Anthropic 做研究的母亲，身处 AI 这场疯狂风暴的中心，你是如何保持精力、避免精疲力竭的？我们只是从外部看 Anthropic 就已经觉得疯狂了。你在避免倦怠、保持充电和保持理智方面有什么心得？

Maybe a last question along these lines. Something Fiona Fung actually suggested to ask you, who was recently on the podcast. How do you stay recharged and not burn out being in the center of this crazy storm of AI as a mom working in research at Anthropic? We're living through the most unprecedented time working on the outside of Anthropic; it's crazy. What have you learned about avoiding burnout, staying recharged, staying sane?

2024 年全年我们发布了四个模型，或者说四个系列。而今年仅仅第二季度，我们的发布量就超过了去年全年。我很幸运能与我们在研究和产品管理团队中培养和建立的团队一起工作。处理这一切的一个神奇之处在于，这不是一项个人运动。这里有一种彻底的 ownership 感和团队协作精神。有时感觉像一项高强度运动，因为你要做出非常关键的决策，不断有关于用户和训练的新信息，你需要非常迅速地提出建议、做出判断。没有人能独自持久地做到这一点。真正有帮助的是拥有一支出色的团队，他们互相照顾。即使不是某个模型的核心 DRI，在发布前夜他们也会熬夜帮助 DRI，审阅博文、修改、想出更好的演示。这就像是彼此的额外帮手。如果你把所有这些变化都扛在自己肩上，很容易感到孤独，觉得必须自己做所有事。但 Anthropic 的一个神奇之处在于，我们能找到互相帮助的机会，然后更进一步地实现思维融合——我们称之为进入蜂群意识。有一篇文章讲过这个。这让团队能够自我补充。并不是说你休假回来事情会翻三倍，而是你可以放心休假，知道团队能处理好该做的事。我个人也很幸运，我的伴侣非常支持我。今年是我在 AI 领域工作的第六年——先是在亚马逊，然后在 Anthropic——所以他明白我有多热爱这项技术以及它能做什么。这从个人角度也很有帮助。

In 2024, we shipped four models for the whole year, or four series of models. I think we did more than that volume in just Q2 of this year. I've been really lucky with the team that we've grown and built, both the stakeholders on the research side and within our research product management team. One of the magical parts about approaching all of this is that it's not an individual sport. There's a sense of radical ownership and team collaboration. Sometimes it feels like a high-performance sport because you're in very critical decisions, with new information about users and training, and you have to make recommendations and judgments very quickly. Nobody can do that sustainably by themselves. What's really helped is having an incredible team that looks out for each other. Night before a launch, even if they're not the core DRI on that model, they'll stay up and help the DRI, review blog posts, make edits, come up with better demos. It's about being each other's extra hand. If you take all this change on your own shoulders, it's easy to feel alone and like you have to do everything. But one of the magical parts of Anthropic is this ability to figure out opportunities to help each other and then take the next mile of mind-melding—we called it entering the hive mind. There was an article about this. That allows the team to replenish. It's not just that you can take PTO and come back to three times the amount of things to do; it's that you can take PTO and know the team can figure out the right things to do. Personally, I'm really lucky: my partner is very supportive. This is year six of me working in AI—first at Amazon, then Anthropic—so he sees how much I love the technology and what it can do. That really helps from a personal perspective as well.

我很喜欢这些答案之间的关联。所以我听到的是：有其他人、与他人合作、相互依靠、在事情变得疯狂时互相帮助。这和你之前回答如何从工作中找到快乐和乐趣时说的很相似：从他人那里获得灵感，看他们在做什么，一起合作。

I love how many of these answers connect. So what I'm hearing is just having other people, working with other people, relying on each other, helping each other out when things get crazy. That's a similar answer to how to find joy and fun in this work: be inspired by other people, see what they're doing, work together.

嗯。

Yeah.

而且有意思的是，最近 Fiona 上播客时，我问她软件工程领域发生了什么变化。她指出现在比以前孤独多了，因为我们在和智能体而不是其他人一起工作。团队更小了，人们整天在和一堆智能体对话。所以这提醒我们，身边真实的人的力量。

And it's interesting when Fiona was on the podcast recently, I was asking her about what's changed in software engineering. She pointed out it's a lot lonelier now because we're working with agents instead of other humans. Teams are smaller; people have all these fleets they're talking to constantly. So this is a reminder of the power of actual other humans around you.

我们被要求去处理真正重大的事情并做出决策，因为技术给了我们更大的扩展能力。我认为，拥有一些能在工作上和你达到某种程度思维共鸣的人——不是每个细节，而是第一性原理和基本假设——这能让他们支持你、推动你的决策并锤炼你的思考。所以我在组建团队、扩大团队、招聘时非常看重这一点：这个人是在意自己的 ego 和构建一个大组织，还是在意为 Anthropic 和团队的影响力做贡献？我倾向于那些低 ego、团队导向的人。这是可持续性的重要部分。

We're asked to work and make decisions on really big things because we have more scale from the technology. I think having individuals who can have some level of mind-meld with what you work on—how you approach things, not every detail, but the first principles and assumptions—then helps them back you up, push your decision, and sharpen your thinking. So I really try to look for that when building the team, growing the team, hiring: is this person going to care about their own ego and building a big org, or are they going to care about contributing to Anthropic and the impact of the team? I orient towards folks who are low ego and team-oriented. That's a big part of sustainability.

很多问题归根结底还是文化和招聘。我记得那个月每天都有产品上线，还有一个发布日历，大家都在讨论这怎么可能。我听到最多的是，因为每个人都对使命和价值观高度一致，所以人们能很快做出决策。在我们进入激动人心的快速问答环节之前，Dan，你还有什么想分享的吗？还有什么想谈的吗？或者对我们聊过的话题有什么想再强调的吗？

A lot of it always comes down to culture and hiring. I remember that moment when something shipped every day of the month. There was a calendar of launches, and people were talking about how this was possible. What I heard a lot is that because everyone is so aligned around the mission and values, it allows people to make decisions really quickly. Before we get to our very exciting lightning round, is there anything else, Dan, that you wanted to share? Anything else you wanted to touch on? Anything you want to maybe double down on of things we've talked about?

这其实非常有趣，因为我觉得你的问题让我把各个点之间的联系想得更清楚了。我是你的人类版 Claude。我想传达或让大家带走的一点是，一是工作方式，二是这里有很多成长和变化，以及使用这项技术的乐趣。如果你觉得此刻你并没有那么多最初的快乐，那么如果你对这个领域感到兴奋并且想在此工作，你如何找到那些拥有这种快乐的人？我认为发展技能、更新技能有很多方式，比如从第一性原理思考你要解决什么问题。从根本上说，你没有问我这个，但社区里有这样一个问题：当模型已经如此强大、工程师们也积极投入时，我们还需要产品经理吗？我认为那些以用户为中心、深入理解用户试图达成什么、以可操作的方式提炼出来并坚持不懈去做的人——对我来说，这就是产品人员的核心，而且我实际上认为我们需要更多这样的人。我觉得我们正变得过于技术驱动，而要让技术产生影响，你必须深入、必须好奇、必须非常亲力亲为，这些特质我认为也帮助了 Anthropic 在产品开发和模型开发方面，也是文化的一部分，希望这对其他人也有价值。

This was actually really fun because I feel like your questions actually sharpen some of my thinking around how the dots connect. I'm your real human Claude over here. One thing that I really want to convey or have people take away is, one, the ways of working, but also two, that this is a lot of growth and change and having the joy in using this technology. And if you're feeling like in this moment you don't have as much of that feeling of initial joy, how do you find people who do if this is an area that you're excited about and want to work on. I think developing skill sets, replenishing skill sets in many ways, things like thinking from a first principles manner about what you solve. I think fundamentally, you didn't ask me this, but there is this question in the community: do we still need PMs when the models are so capable, when engineers are leaning in? I think the role of people who are user-centric, who go into the details of understanding what users are trying to accomplish, bubbling that up in an actionable manner and doing the relentless work to do that — that to me is the core of a product person, and I actually think we need more of that. I think we are becoming very technology-layered driven, and actually to make that impactful you have to go deep, you have to be curious, you have to be super hands-on, and those are things that I think are also traits that have helped Anthropic from a product development and model development perspective and as part of the culture, and hopefully that's valuable for others as well.

太棒了，多么鼓舞人心的结尾。老兄，是的。这是我长期以来的观点：构建很容易。困难的部分，正如你所说，是我们应该构建什么，以及我们构建的东西是否正确、是否好、是否值得投入？对我来说，这就是产品经理的工作，也是他们擅长的。

Amazing. What an inspiring way to end it. Oh man. Yeah. And this is something I've been saying for a long time: building is easy. The hard part becomes, as you said, what should we build, and is the thing we have built correct and good and worth leaning into? And to me, that's what PMs do and what PMs are good at.

是的，而且要深入用户的细节。

Yeah. And it's getting into the details of the user.

是的，同理心。好的，很好。产品经理会成功的。好的，产品需求文档没有死。这里有很多重要的教训。Dan，我们这就进入激动人心的快速问答环节。我有五个问题问你，准备好了吗？

Yeah. Empathy. Okay. Great. PMs are going to make it. Okay. PRD is not dead. All kinds of important lessons here. Dan, with that we've reached our very exciting lightning round. I've got five questions for you. Are you ready?

准备好了。

Yep.

第一个问题。你发现自己最常推荐给别人哪两三本书？

First question. What are two or three books that you find yourself recommending most to other people?

一本私人很喜欢的书是《如何培养成年人》。我是妈妈，我经常思考我想在孩子身上灌输什么。那本书很好地描述了：我们不是在培养孩子，而是在培养成年人。这个框架是什么意思？我们想培养孩子的哪些特质？另一本我最近在 Audible 上听的书是埃里克·莱斯的《不可腐蚀》。

One personal one I really like is 'How to Raise an Adult'. So I'm a mom. I think a lot about what are the things I want to instill in my kids. That book is really helpful for describing we're not trying to raise children, we're trying to raise adults. So just the framing of what does that mean and what are the characteristics we want to hone, harness, and foster in our kids. The other book I was listening to on Audible recently is 'Incorruptible' by Eric Ries.

《不可腐蚀》，《不可腐蚀》。是的，是的。

Incorruptible. Incorruptible. Yes. Yes.

是的，他最近刚上过播客。

Yeah. His recent podcast guest.

嗯，是的。

Um, yeah.

而且我觉得如何建立伟大的公司这个问题很重要。我个人一直最着迷于如何让优秀的团队和公司走得更远。看到他对这个问题的框定和重新框定非常有趣。我喜欢其中一些关于用指标衡量文化的例子。如果你只衡量收入，那就会朝那个方向走，但如果你有其他更好的指标，那才是维持你重视的价值观的方式。我一直在思考如何将其落实到团队层面，比如如何更好地阐述我们的规范，我们团队讨论的很多事情。所以我认为这也是一本很好的读物。

And I just think the question of how to build great companies is important. I personally have been most fascinated with how to keep great teams and great companies going further. And it was very interesting to just kind of see his framing and reframing of the question. I loved some of the examples around having metrics around culture. If you only measure revenue, then that's kind of how you're going against, but if you have other better metrics, that's actually the way to sustain the values you care about. I've been trying to think about how to actually bring that to the team level, like how do we better articulate our norms, a lot of the things we talked about on the team. So I think that's also a really good read.

就这样。大家听着，这是你们下一个要看的，就是 Eric Ries 的那期节目。是的。

There you go. That'll be your next watch everyone, as you're listening to this, the Eric Ries episode. Yeah.

那期节目非常好。是的。

Such a good episode. Yeah.

他的书刚出版，《不可腐蚀》。

And his book just came out. 'Incorruptible'.

是的。

Yes.

而且它好像是《纽约时报》畅销书，实际上卖得非常好，我很高兴看到。

And I think it was like a New York Times bestseller. Like it's actually doing incredibly well, which I was really happy to see.

是的，没错。

Yeah, exactly.

下一个问题。最近你非常喜欢的电影或电视剧。Anthropic 的大多数人没时间看剧，但我很好奇你有没有答案。

Next question. Favorite recent movie or TV show you really enjoyed. Most people at Anthropic don't have time to watch things, but I'm curious if you have an answer.

我会说上个月休息的时候，我刷了亚马逊 Prime 的《辐射》。那实际上……它有点……你听说过吗？

I would say during some time off last month, I did get to binge watch 'Fallout' on Amazon Prime. So that was actually... It's kind of... Have you heard of it?

是的，它改编自电子游戏。

Yeah. Yeah, it's based on the video game.

是的，改编自电子游戏。我觉得它非常机智、幽默，同时也超级动作向。所以强烈推荐。

Yes, it's based on the video game. I think it's really witty, it's humorous, it's also super action-oriented. So highly recommend.

好的，下一个问题。你最近有没有发现特别喜欢的产品？

Okay, next question. Do you have a favorite product you recently discovered that you really love?

我觉得 Claude Tag 在产品体验上非常有趣。我们在 Anthropic 内部有这个产品的不同版本，我认为它确实是一个非常强大的工具。

I really do think Claude Tag is very interesting in terms of a product experience. We actually have different versions of this within Anthropic and I think it's actually been a really powerful tool.

是的，我觉得有些人会觉得'这有什么大不了的？'但 Anthropic 的每个人都在夸它，这告诉我这里发生着重要的事情。我正试着在我付费订阅的 Slack 社区里让它跑起来。那会有多酷？

Yeah, it feels like I think some people are like 'what's the big deal?' The fact that everyone at Anthropic is raving about it tells me something important is going on here. And I'm trying to actually get it working within my Slack community that I have for paid newsletter subscribers. How cool would that be?

是的。

Yeah.

是的。我在想，当它不是一家公司，而是一群互不认识的人时，它是如何运作的，以及该如何运作。但我们正在尝试。好的，还有两个问题。你最喜欢的人生格言是什么，工作或生活中你经常回想起的？

Yeah. I'm trying to figure out how it works when it's not a company when it's just a bunch of people that don't know each other and how that might work. But we're trying it out. Okay. Two more questions. Your favorite life motto that you find yourself often coming back to in work or in life.

实际上，我人生的前 10 年是由祖父母抚养长大的，我的父母是在美国的移民大学生和硕士生。我祖父总是说：'无论你走多远，总还有更高的层次。'这是一个很好的说法，尽管是一种相当强烈的方式来描述他的人生哲学。但每当我们遇到新的或前所未有的事情时，我都会回到这句话上。

So I was actually raised by my grandparents for the first 10 years of my life and my parents were immigrant college and master students in the US. And my grandfather always says, 'No matter how far you go, there's always another level,' which is a really good way though, like a pretty intense way of describing his life philosophy. But I go back to that whenever there's something new or unprecedented that we experience.

我觉得，你知道，今年上半年确实有很多这样的情况——我们学到了很多新东西。我在学习，感觉总有一座新的山，一个新的机会。‘还不够好，Dan，我们需要做得更好。我们需要做得更大。’这让我想起了 Ben Man 在他的播客中说的另一句话：现在是最正常的，将来只会变得更奇怪、更疯狂。

And I think, you know, the first half of this year there was definitely a lot of that – a lot of new things that we were learning. I was learning, and it felt like there was always another mountain, another opportunity. 'Not good enough, Dan, we need to go better. We need to go bigger.' That makes me think of another Ben Man line from his podcast: this is the most normal it's ever going to get. It's only going to get weirder and crazier.

是的。是的。不，我们没问题。好了，最后一个问题。我翻了翻你的 LinkedIn。你职业生涯早期在摩根大通做高收益债券交易员。你有一个涂黑了的、一亿美元左右的交易组合。你从那段经历中学到了什么一直影响着你？或者那段时期有没有什么疯狂的故事？那是你生命中四年。

Yeah. Yeah. No, we're good. Okay, final question. I was poking around at your LinkedIn. You were a high yield bond trader at JP Morgan Chase early in your career. You had this redacted hundred million dollar trading portfolio of some kind. What did you learn from that time in your life that has stuck with you, or is there a crazy story from that period? It was four years of your life.

我觉得我实际上学到了很多，现在应用在 Anthropic 和之后的其他工作中。所以当我在摩根大通的时候，交易大厅——你可以想象成《华尔街之狼》那样的场景——非常不同。大多数交易员都坐在终端前，在电脑上做分析。但那里仍然，我认为，男性主导。我是唯一的女性。我是交易台上唯一有这种背景的人。我学到的是，这是一个很好的环境来建立，第一，我的真实自我意识；第二，即使是最初级的人，即使我看起来不同，最好的想法和对最好想法的信念，不管其他所有因素，才是最重要的。所以我带着这种态度工作——我对团队非常坦诚和真实。我努力确保不管人们的级别或资历，如果他们有好主意，我就帮助他们去追求，我自己也这样做。所以把想法提出来，坚定信念，跟进，做那些琐碎的工作来实现它。这些就是我交易中学到的东西。是的，我认为它在很多方面适用于任何工作。

I think I actually learned a lot that I apply here at Anthropic and in other jobs thereafter. So when I was at JP Morgan, the trading floor – you could kind of envision a sort of Wolf of Wall Street – it's very different. Most traders are in front of a terminal, doing analyses on their computers. But it's still, I would say, very male-dominated. I was the only woman. I was the only person with my background on the trading desk. And I learned that it was a very good environment to build, one, my sense of authentic self, and two, that even if I was the most junior person, even if I may look different, the best ideas and having conviction in the best ideas, regardless of all those other factors, is the most important thing. So I think bringing that sense of how I show up at work – I'm pretty vulnerable and authentic with my team. I try to really make sure that regardless of people's levels or tenures, if they have a great idea, I help them pursue it and I do the same. So to put the idea out there, have conviction in it, do the follow-through, do the nitty-gritty work to make it happen. Those were all things I learned from trading. And yeah, I think it applies to any job in many ways.

太棒了。如果人们想关注你，他们可以在哪里找到你？听众如何能帮到你？

That is beautiful. Where can people find you online if they want to follow you, and how can listeners be useful to you?

我在社交媒体上没有太大存在。我觉得找到我工作、我团队工作的最好方式其实是 Anthropic 的博客，以及我们发布新模型、新产品体验的时候。至于对我有用的事情，我认为第一好的是你的反馈。如果你在我们的任何产品界面上点赞或点踩，如果你联系你的销售代表并提供关于模型的反馈，这些反馈会传到我这里。实际上，每一个研究模型，我都很接近地了解用户满意度和反馈。所以给我们反馈，推动 Claude，告诉我们它哪里不足，这些都能帮助我们改进 Claude。另一件事是，如果你的网络中有看起来像我刚才描述的那种人——我正在招聘，团队在成长。我们真的很喜欢热爱这项技术、深具好奇心、第一性原理思考者、敢于质疑假设、并且有修补和黑客精神的人。

I don't have a large presence on social media. I think the best way to find my work, my team's work, is really the Anthropic blog and when we're publishing new models, new product experiences. In terms of being useful to me, I think the best thing number one is your feedback. If you thumbs up or thumbs down on any of our product surfaces, if you contact your salesperson with feedback about the model, it will make its way to me. Actually, with every research model, I get pretty close to understanding favorability and feedback. So giving us that feedback, pushing Claude, telling us where it's falling down, those help us make Claude better. The other thing is if you have folks in your network who seem like this type of profile of person that I just talked about – I'm hiring, the team is growing. We would really love people who love this technology, who are deeply curious, first principles thinkers, who are fearless in questioning assumptions, and who have a tinkering, hackery spirit.

哇，梦想的工作。所以基本上，Anthropic 研究团队的产品经理职位开放。

Wow, dream job. So basically, open PM roles at Anthropic on the research team.

是的。

Yes.

我猜他们是在网站的职业页面上申请吧。

And they apply, I assume, on the website – the careers page.

是的。

Yes.

天啊。好了，开始吧。享受即将收到的简历潮吧。

Holy moly. All right. Here we go. Enjoy the flood of resumes you're about to receive.

谢谢。

Thank you.

Dan，非常感谢你来到这里。

Dan, thank you so much for being here.

非常感谢你邀请我。谢谢你提出非常有帮助、发人深省的问题，甚至帮我连点成线，理解我们如何工作、整个技术如何融合，以及作为产品人置身其中。

Thank you so much for having me. Thank you for really helpful, thought-provoking questions, helping me even connect the dots on how we work, how this whole technology is coming together, and being product people in it.

我真的很感激。但说真的，谢谢你，Dan。好的，好了，再见大家。非常感谢收听。如果你觉得有价值，可以在 Apple Podcast、Spotify 或你最喜欢的播客应用上订阅本节目。同时，请考虑给我们评分或留言评论，这真的能帮助其他听众发现这个播客。你可以在 lennispodcast.com 上找到所有往期节目或了解更多信息。下期再见。

I really appreciate that. But thank you, Dan, for real. Okay. Well, bye everyone. Thank you so much for listening. If you found this valuable, you can subscribe to the show on Apple Podcast, Spotify, or your favorite podcast app. Also, please consider giving us a rating or leaving a review, as that really helps other listeners find the podcast. You can find all past episodes or learn more about the show at lennispodcast.com. See you in the next episode.