OpenAI Tara Seshan：AI 第三纪元与持久共事者

Lenny Rachitsky

如果把 AI 产品的第一纪元想成聊天，第二纪元是和智能体一起干活，那可能很快到来的第三纪元是：你怎么和一个能跟你一起把事做完的持久共事者协作。

If you think about the first era of AI products as chat, the second era as working with agents, that third era that might come soon is how you work with a persistent coworker who is able to get things done with you.

Tara Seshan

有一种「悬空能力」：AI 已经能做到什么，和我们实际拿它在做什么之间，还有一大截。

There’s this idea of the overhang of what AI is capable of and what we’re actually doing with it.

Tara Seshan

很难预知未来会出现什么。为模型「现在」的能力建产品会失败；为你以为一年后会到的能力建也会失败。两种同样错。唯一可行的建法是盯着两到三个月。

It’s so hard to understand what is going to emerge in the future. You fail if you build for where the models are now. You fail if you build for where you think the models will be in a year. Both outcomes are equally wrong. The only way to build is two to three months out.

Lenny Rachitsky

作为产品经理，你觉得自己最需要适应和调整的是什么？

What do you most have to adapt to and adjust in how you operate as a PM in this world?

Tara Seshan

多产和经验主义，远比学院派或纯理论重要。与其写很长的推理文档，不如尽快到达能试用、能跟用户测的东西。

Being prolific and empirical is way more important than being academic or theoretical. Rather than writing out some long reasoning doc, it’s: how do I get to something I can try out and test with users as fast as possible.

Tara Seshan

感觉我们不仅能更有野心，几乎还必须更有野心——这对很多人并不自然。抬高别人的野心、提醒他们这里什么是可能的，是产品管理角色很大一块。

It feels like not only are we able to be more ambitious, we almost need to be more ambitious, which is not natural for a lot of people. Elevating others’ ambitions or reminding them of what’s possible here is a huge part of the product management role.

Lenny Rachitsky

今天嘉宾是 Tara Seshan。她在 OpenAI 负责 Codex 和 ChatGPT Work 的产品。我认为这是面向知识工作者增长最快、也最重要的 AI 产品之一。她和工程负责人 Andrew Ambrosino 搭档——Andrew 最近也上过这档播客。来 OpenAI 之前，Tara 在 Stripe 待了大约六年，是最早的五位产品经理之一；多年被评为整个 Stripe 组织顶尖的三人之一。她还在 Watershed 带过产品，做过创始人，是 Thiel Fellow，更重要的是，她是我几年前 Lenny’s Newsletter Fellows 三位入选者之一。我非常兴奋看到她现在这个角色。下面请出 Tara Seshan。

Today my guest is Tara Seshan. Tara leads product for both Codex and ChatGPT Work at OpenAI. I believe this is the fastest-growing and arguably most important AI product for knowledge workers today. Tara works alongside Andrew Ambrosino, who was a recent podcast guest — he’s her eng manager. Prior to OpenAI, Tara spent six years at Stripe where she joined as one of the first five product managers, and for many of those years she was named one of the top three Stripes across the entire organization. She also led product at Watershed, was a founder and a Thiel Fellow, and most importantly of all, Tara was one of the three Lenny’s Newsletter Fellows. I am so excited to see Tara in this new incredibly important and impactful role. With that, I bring you Tara Seshan.

Lenny Rachitsky

Tara，非常感谢你来，欢迎上播客。

Tara, thank you so much for being here and welcome to the podcast.

Tara Seshan

谢谢 Lenny。我很高兴来，见到你真好。

Thank you, Lenny. I’m so glad to be here. It’s so nice to see you.

Lenny Rachitsky

我更高兴。你在 OpenAI 大概一年了——在多数地方这很短，按 AI 时间却像一辈子。你加入时大概有过对前沿实验室的想象。实际在 OpenAI 工作，最让你意外的是什么？最好好坏都说。

I’m even more glad. You’ve been at OpenAI for just about a year now, which in most places would be a very short amount of time. In AI time, that’s like a lifetime. I imagine when you joined you had a sense of what a frontier lab would be like. What’s most surprised you about what it’s actually like to work at OpenAI — ideally both good and bad?

Tara Seshan

很多地方反而很熟悉：我以前也待过高增长、高人才密度、高强度、超规模扩张的公司。同事很强、紧迫感很高等，都似曾相识。真正最意外的是：我以前待过的公司几乎都是创始人主导；OpenAI 则是「人人都像创始人」——尤其在自己负责的领域，每个人某种程度上都是创始人。相对我以前的地方，自上而下的方向极少。

So many things about working at OpenAI felt familiar because I’d worked at other high-growth, high-talent, high-intensity, hyperscaling places. Things like “my colleagues are so awesome” or “the urgency is really high” felt familiar. What felt most surprising is that every company I’d worked for before had been founder-led — and OpenAI is actually founders-led, meaning everyone inside the company, especially in their area, is in essence kind of a founder. The level of top-down direction at OpenAI is extremely limited relative to places I’ve worked prior.

Tara Seshan

刚到公司时这既让人愉悦——我自己有过创业经历，心想「我还能继续觉得自己是这块产品或这个团队的创始人」——而且我和市场之间的距离极薄。在大公司有时会觉得被用户或市场需求隔开；在 OpenAI 完全不是。你像创始人一样，为产品做产品市场契合需要做的一切都会做。

When I first got there that was delightful — I’d come from a founding journey and thought, yes, I can continue to feel like the founder of this product area or team — and the distance between me and the market is very thin. At a larger company you sometimes feel insulated from what users want or what the market demands; at OpenAI you do not at all. You are doing everything it takes to get product-market fit for your product, akin to how a founder might.

Tara Seshan

更意外的一面是：我以为会有一箱 OpenAI 的「秘密战略」可学，像以前公司有支付圣经、有一套支付与运营的思考方式。结果 OpenAI 真的很 open——关于世界该什么样、产品该怎么建、模型该怎么运作的想法，很快就会变成公开产品或公开叙事的一部分。这既是正向惊喜，也彻底改了我的运作方式。

The more surprising side is I came in expecting a treasure trove of OpenAI secret strategy I could understand, akin to past companies where you find the payments bible and how we think about payments and operations. Actually OpenAI is open — every thought about how the world should look, how product should be built, or how the model should operate very quickly becomes part of the public product or public messaging. That was incredibly positively surprising and a change in my operating mode.

Lenny Rachitsky

所以你是说没有一间密室，里面跑着 AGI、挂着写满答案的总计划？

So you’re telling us there’s not like the secret room with the AGI running there with the master plan that has all the answers?

Tara Seshan

至少我肯定不在那间房里。真正激励我的是：OpenAI 做的很多东西立刻变成用户能在产品里摸到、感到的东西，这个循环比我见过的任何地方都快。

Or at least I’m not in that room for sure. The piece that is really inspiring is that so much of what OpenAI does immediately becomes something users can touch and feel in the product, and that cycle is faster than anywhere else I have seen.

Lenny Rachitsky

你在很多地方做过 PM，也是资深 PM 负责人。在这个新世界里，你会丢掉什么？

You’ve been a PM at a lot of different places, a longtime PM leader. What do you lose in this new world?

Tara Seshan

市场更静态、更慢时，你有机会做宏大战略式的工作，因为更可预测，至少能看清各块拼图。比如支付市场虽有动态，但是成熟市场：你可以推理「我下这个注，对手可能下那个注」，从第一性原理严谨推演下一步。那种市场甚至要求你这么做——赢家会比别人想得更严；不够严就会显得粗心，因为很多决策本可预见。

When a market is more static or slow-moving, you have the chance to do grand strategy-esque work because it’s more predictable, or you can at least understand all the pieces. Payments is dynamic to some extent, but it’s also an established market — you can say if I take this bet, my competitor might take that other bet, or reason from first principles rigorously through the next actions. That market mandates that winners think more rigorously than everybody else. If you aren’t thinking rigorously, it shows up as carelessness, because a lot of those decisions could have been predicted.

Tara Seshan

但这个市场很难预知未来会出现什么：非常涌现、变化极快、极度动态，而且最重要的是必须贴着研究。所以多产、更经验主义，远比学院派或理论重要。我以前很多公司非常学院、理论；从写像博士论文一样的长推理文档、规划很长一段，切换到「怎样尽快到能试用、能跟用户测的东西」，一开始很别扭。

But in this market it’s so hard to understand what will emerge. It’s very emergent, fast-changing, dynamic — and most importantly, very important to stay tied to the research. Being prolific and more empirical is way more important than being academic or theoretical. Lots of past companies I’ve worked at were very academic and theoretical. Switching from writing a long reasoning doc almost like a PhD thesis for the next stretch of time, to “how do I get to something I can try and test with users as fast as possible,” felt very jarring at first.

Tara Seshan

我会想：我是不是没尽职？不够深思？不该把一切想得很严吗？其实你必须去试、尽量多学。你真正需要的思考，是把核心假设钉得尽可能尖。假设定义最重要——用 Shishir Mehrotra 的话说，就是 eigenquestion：那个最该测的具体最重要的事。其它你编出来的宏大战略都不相关。

I was like, am I not doing my due diligence? Am I not being thoughtful enough? Shouldn’t I be thinking through all of this with a ton of rigor? But actually you’ve got to try stuff and learn as much as possible. The thinking you need is being as pointed as possible about your core hypothesis. That hypothesis definition is the most important thing — to use Shishir Mehrotra’s phrase, the eigenquestion: that specific most important thing to test. Everything else, any other grand strategy you concoct, is not relevant.

Lenny Rachitsky

我想多听一点：世界变了很多，但这块反而更重要了——PM 角色里没变、甚至更关键的是什么？人们该更盯紧什么？

I’d love to hear more about that, because that’s really interesting as almost: here’s the piece of the PM role that is not changing. So much is changing, the world is changing, but there’s still this piece that is even more important. Speak more to what specifically you think people need to focus more on.

Tara Seshan

PM 角色外围有很多陷阱：盯执行按时、写各种文档和演示等等。但核心一直是：关于你的产品，最本质的问题是什么？什么决定产品成或不成？你怎么测？怎么看结果？怎么喂回假设再跑一轮？那才真是 PM 的工作。

There are so many trappings around the PM role — running execution on time, writing all these specific docs and presentations, etc. But the core has always been: what is the most essential question you need to ask about your product? What will determine whether your product works or doesn’t? How do you test that? How do you look at the results? And how do you feed that back into a loop of refining your hypothesis and running it again? That truly has always been the PM job.

Tara Seshan

这当然包括理解用户、市场、你在建的技术，把三者拧成尽可能锋利的假设，再把测试做得又快又有效。这不仅没变，而且成了公司里最重要的能力。工程经理、工程师、数据科学家、设计师都在往「问题定义 + 测试循环」上靠：我们到底在做什么、怎么知道它在生效。对 PM 来说很好，因为 PM 一直盯这个；其它职位附带物掉了很多，这点每次都必须做对。

That involves understanding users, the market, and the actual technology you’re building — pulling those three together into the sharpest hypothesis you can, then making the test as fast and effective as possible. That has not only not changed — it’s become the most important thing at the company to be able to do. EMs are thinking this way, engineers, data scientists, designers — everyone has moved to focus on this problem-definition and testing loop: what are we actually doing and how do we know if it’s working. From a PM standpoint that’s great, because PMs have always been focused on getting that right. Many other trappings of the job have fallen away, and that remains the key thing to get right every time.

Lenny Rachitsky

你提到循环。最近几周 Twitter 上「loops」很火，知识工作也一直在谈。我理解的循环大概是：告诉 AI 成功长什么样，让它去建、去摸索，直到达到成功。你怎么看循环从软件工程扩到产品管理、再到整个知识工作？这会成真吗？

You mentioned this idea of a loop — loops were so hot a few weeks ago on Twitter, and it continues for knowledge work broadly. The way I understand a loop: AI, here’s what success looks like — go off and build and figure it out until you achieve success. How do you think about loops expanding from software engineering to product management and all knowledge work? Do you think that’s going to be a thing?

Tara Seshan

我确实觉得，工作的未来会越来越像掌舵而不是划桨：会有智能体帮你做大量划桨，你的角色越来越是把船指对方向。而且掌舵可能会越来越高层——以前是写这行代码、按 tab；现在在指挥更完整的东西；也许到目标层，甚至更高。抽象层会上移，但最终仍要有人判断：我们指向哪？有了反馈和新数据，下一步要把这东西带到哪？

I do think the future of work will look more like steering than rowing — there’ll be agents you work with that do a lot of the rowing, and your role increasingly becomes steering the ship and pointing it in the right direction. That steering might grow higher and higher level. Steering used to be “I wrote this line of code, press tab”; now I’m directing something more comprehensive; maybe to the goal level, maybe even higher. Steering will continue to go up layers of abstraction, but ultimately it’s still on a person to find which direction we’re pointing this in, and given feedback and additional data, where do I want to take this next.

Tara Seshan

掌舵一部分看数据告诉你什么，但很多是做有观点的判断。我们有时低估直觉，甚至低估一种正向的决定论：我希望未来长这样——不是因为反面策略不可行，而是我想把世界往我推的方向推。至少现在，这仍需要人给出意见。循环很好，让智能体在越来越大的循环里替你划桨很好，但你仍然真的需要掌舵。工作也会变成：你和其他人一起，共同掌舵一群你们一起用的智能体。把队友拉进你和智能体之间——它划桨、你掌舵——的互动，也非常有价值。

Some of steering is certainly what the data tells you, but a lot is making an opinionated call. We sometimes underrate the power of intuition or even positive determinism about what we want the future to be — picturing “I would like the product to look this way,” not because the converse isn’t equally viable, but because I would like the world to look like the direction I’m pushing. That will always remain an opinion, at least right now, required from a person. Loops are awesome — running agents in increasingly larger loops where they do more of the rowing is great — but right now you really still need to steer. Work will also look like steering with other people over a group of agents you work with together. Bringing teammates into that interaction between you and the agent — it’s rowing, you’re steering — feels incredibly valuable.

Lenny Rachitsky

有意思。一个想法是：如果人人工具一样，拉开差距的是人本身，否则大家会做出一样的东西。几乎不公平的优势是人脑。

That’s such an interesting way of describing it. One thought: if everybody has access to the same tools, what separates you is the human — the person. Otherwise we’re all just going to be building the same thing. Everyone could be asking how do we win, what do we do. The almost unfair advantage is the human brain.

Tara Seshan

这让我想起时尚。确实有人人都能穿、能干活的功能服装；但你穿什么，很多是你想对个性或对外呈现作什么陈述。它之所以有力，往往因为它和别人的表达形成对照——这件衬衫有陈述，正因为和大家或某群人不同，或声明你属于哪一组。我们建的很多产品同样有观点、有艺术性。Patrick Collison 有句很好的话：软件不像房地产，不是投钱就出价值；它更像拍电影——你可以砸很多钱进一部电影，但不保证片子成功或好。里面有作者陈述、有观点和艺术性。那依赖于你或你的团队对产品有有意思的话要说。

It reminds me a lot of fashion. There are functional clothes everybody can wear that get the job done, but so much about what you wear — how I think about what I wear — is about what statement I want to make about my individuality or how I want to reflect to the world. A lot of what makes that compelling is how it contrasts with other people’s expression — the shirt makes a statement only because it’s different from what everybody else is doing, or different from some cohort, or makes a statement about group membership. A lot of the products we build feel similarly opinionated and artistic. Patrick Collison has this really nice statement about software: software is not like real estate — you don’t put money in and get value out. It’s a little more like filmmaking — you can put a lot of money into a film, but that doesn’t guarantee the film is successful or good. There is some auteur statement or opinionation and artistry. That relies on you or your team having something interesting to say about your product.

Lenny Rachitsky

Marty Cagan 常说：你对产品或功能的那个点子，很少就是最终那个东西；你要走一整段过程才搞清它到底该是什么。你好像也在说：人必须走完那过程，才明白它真正是什么、人真正要什么——绝不会是「好，明白了，从一开始就去建这个」。

There’s something Marty Cagan is big on: when you have an idea for a product or feature, rarely is that idea the thing that ends up being. There’s this whole process you go through to figure out what the hell it actually should be. It feels like that’s what you’re saying — you need to go through that process as a human to understand what it really is and what people actually want. It’s never “okay, got it, go build this thing — I got it from the beginning.”

Tara Seshan

当然。而且那些循环越来越快。所以形成直觉、拿到形成直觉所需的信息，再和人与智能体一起落地——这才是关键。

Yeah, for sure. And those loops are moving faster and faster. Your ability to form those intuitions, get the information you need to form them, and then use that with people and agents to put it into action is the key.

Lenny Rachitsky

你觉得知识工作者下一轮工作方式的转变是什么？你不仅能用到别人还没有的最先进工具，还处在全世界最「AI 化」的人群里。接下来三到六个月，你们内部已经在用、你觉得会变成大家更常态的方式是什么？

I’m curious what you think the next shift will be in how we work broadly as knowledge workers. Not only do you have access to the most advanced tools some other people don’t yet — you also work around the most AI-forward people in the world. How are people working internally that you think will become a more normal way we all work using these AI tools in the next three to six months?

Tara Seshan

两方面。一是继续在更高抽象层和智能体协作：让智能体更独立地替你做更多，你进来掌舵，再让它继续「煮」——在更高阶抽象上提供细节。人们越来越把智能体想成持久的、像队友、像同事：像我和团队里某人协作——他们做一大块，我给输入，他们再做，我们按不同节奏同步，看彼此进行中的工作，给更多反馈。这种同事模型是方向，也是和智能体协作更自然的界面；内部已经大量看到。

There’s two aspects. One is continuing to work with agents at higher and higher levels of abstraction — letting the agent do more and more for you independently, coming in and providing that steering, then letting the agent continue to cook and provide details at higher orders of abstraction. People are increasingly thinking about agents that are persistent, that feel like teammates, like coworkers — where you work with them the way I might work with someone on my team: they do a whole bunch of work, I provide input, they do work again, we sync at different cadences, look at each other’s in-progress work, and provide more feedback. That coworker model is the way things are going — a much more natural interface for working with agents — and we already see a lot of that internally.

Tara Seshan

二是：我目前和智能体大多是一对一——我和我的智能体，也许它拉了子智能体做任务，但基本是我和它；这可能和同事各自跟各自的智能体脱节。有段时间内部大家只是在 Slack 互传 Codex 线程截图：「我想让你看看我怎么得到这个数字——截图。」那也不是最自然的一起协作方式。智能体做的工作越多，我们难道不该和智能体一起、一起把事做完吗？最自然的界面是什么？这些是我们在想的。

Second: a lot of my work with agents thus far has been one-on-one — I work with my agent, maybe it spawned sub-agents, but it’s me and my agent, potentially divorced from what colleagues are doing with theirs. There was a time when everyone internally was just sending Codex threads — screenshots of Codex threads — to each other on Slack: “I wanted to share how I got to this number — here’s a screenshot.” That’s also not quite the most natural way to collaborate. As more work gets done with our agents, shouldn’t we be able to get work done with our agents together? What’s the most natural interface to make that happen? Those are things we’re thinking about.

Lenny Rachitsky

我脑补的是聊天。比如这是 Tara 的智能体、这是我的；她做了分析，我说：Lenny 的智能体，去核对这合不合法、是否接得上我对世界的看法。

Chat is what I’m picturing. Here’s Tara’s agent, here’s my agent. She did some analysis — hey, my agent, Lenny’s agent, go check make sure this is legit and connects to the way I think about the world.

Tara Seshan

理想中，工作像多人游戏：我们一起把事做完，掌舵各自的智能体，而智能体继续承担越来越多战术性的划桨。

Ideally work feels like a multiplayer game where all of us together are getting stuff done, steering our agents as our agents continue to take care of more and more of those rowing tactical tasks.

Lenny Rachitsky

有意思的是信任和意识在慢慢推进：好，再干久一点，你能扛更多。大家怕快速起飞——模型突然太聪明就麻烦了。感觉我们更像慢速起飞，在迭代；并不真的慢，但不是突然冒出一个 300 IQ 的 AI。

It’s interesting how it’s been this slow progression of trust and awareness that this can be how we work — okay, go work for longer, you can take on more. There’s talk of slow takeoff vs fast takeoff, and everyone’s afraid of a fast AI takeoff where it’s way too smart and we’re in big trouble. It feels like we’re on the slow takeoff scenario, which is good — slowly iterating. Doesn’t feel that slow, but we’re not some 300 IQ AI.

Tara Seshan

模型确实极其聪明。让我们能和智能体一起工作、让智能体扛更高阶抽象的，当然包括智力、长程任务表现、能在任务上待多久；但也有很多「肉土豆」战术条件。本地跑的智能体很方便，因为能碰你机器上的数据；要让云端智能体成功，得建大量云基础设施。还有系统接入：智能体怎么跟装了你数据的第三方系统说话。就像你雇了同事却锁在房里、不给 Google Docs、Slack 和公司数据库——不会有用。同样，孤立的云智能体也不会高效。所以让智能体有用、实现这些未来，智力侧当然重要，很多却是数据访问、云基础设施和可靠性这些听起来平淡、但对最终效果几乎同样关键的事。

The models are incredibly smart. A lot of what has enabled us to work with agents together or have them take higher-order abstraction things is certainly about intelligence — their ability to perform long-running tasks and how long they can stay on tasks — but there are also very meat-and-potatoes tactical things. Agents working locally are really convenient because they have access to all the data on your machine. To make an agent successful in the cloud, there’s a ton of cloud infrastructure you have to build. And access to your systems — how can agents talk to all these third-party systems that have your data. A colleague you hire and lock in a room without Google Docs and Slack and the company database would not be that useful. Similarly, a cloud agent that is similarly isolated will not be that effective. A huge part of making these agents useful and achieving some of these futures is on the intelligence side, but a lot is also really tactical — data access, cloud infrastructure, and reliability pieces that feel much more prosaic than the broader intelligence questions, but matter in some ways just as much for end effectiveness.

Lenny Rachitsky

这碰到本播客常出现的词：野心。你也常想这个。感觉我们不仅因为 AI 工具能更有野心，几乎还必须更有野心——这对很多人并不自然。容易的事人人都能轻松做；难的事也变容易了。现在拉开人和公司差距的，就是能有多野心。谈谈「需要野心」在你这儿唤起什么。

This touches on something else that has been coming up a bunch on this podcast — ambition. I know you think a lot about this too. It feels like not only are we able to be more ambitious because of these AI tools, we almost need to be more ambitious, which is not natural for a lot of people, because everybody can now do all these easy things really easily. The easy stuff is super easy. The hard stuff is easy. The thing that separates people and companies now is just how ambitious they can be. Talk about what comes up when I talk about the need and emergence of this need for ambition.

Tara Seshan

我们看到用 AI 工具最有效的人，不只是用它自动化机械任务，而是用它扩大自己能做的事的集合。以前那种「独角兽」是：产品感觉很强，碰巧还会工程，也许还会设计——因为他们压扁了职能间的翻译层，能自己很快做出或想出东西并跑起来，再和团队协作。我发现最有说服力、也是我尽量和最成功的同事在做的，就是扩大「可能性范围」，把脑子里的东西更多地变成现实——以前全才做得到的事，我们现在多少都有了这种超能力。

The people we see most effective with AI tools don’t simply use it to automate rote tasks but use it to expand the set of things they are capable of doing. Back in the day, before all this AI stuff, the unicorn was a really thoughtful product-sense person who also happened to be an engineer who may also have been a designer — because they flattened the layers of translation between functions and could build or ideate something quickly themselves, get it up and running, then work with a team. The most compelling thing I’ve found — that I try to do and I’ve seen some of my most successful colleagues do — is really expand the set of things that are within their range of possibilities so they can start realizing more of what’s in their head into reality, the way someone who was previously a jack of all trades could. We kind of all have that superpower now.

Tara Seshan

我可以拉出一套设计，做出初始原型，想出定价模型并建模各种情景——可能性集合大幅变宽。这某种程度上让我更能像前面说的电影作者那样实现愿景、达到更高保真。野心不再受限于你自己能执行什么、能沟通什么，可以宽得多。最难的只是扩展思维：能力已剧变，要在不合理的短时间里想清什么是可能的。最好的办法之一是 Patrick Collison 网站上 patrickcollison.com/fast 那类列表——那些在极短时间里完成的不合理地野心勃勃的项目。现在那张列表更惊人：那些项目都出现在这些工具让你几乎瞬间学会建东西、或一问就能总结复杂文本/书籍之前。以前够不到的能力现在够得到——比如让我做一个想法的 CAD 模型。如果那些「快项目」在旧能力下都可能，那有了 AI，我们难道不该看到这类不合理地又快又好的执行指数级变多吗？

I can spin up a set of designs, build an initial prototype, figure out the right pricing model and model out all the scenarios — the set of possibilities has widened dramatically. That means I have the ability to be more of an auteur as I try to get something done and realize my vision maybe to higher fidelity. Your ambitions are no longer limited by what you’re capable of executing yourself or communicating — they can be so much wider. The hardest part is simply expanding your thinking. Capabilities have expanded so dramatically — it’s really expanding your thinking of what’s possible in an unreasonably short time frame. The best way to try that is Patrick Collison’s patrickcollison.com/fast — all these projects that were unreasonably ambitious executed in a really short time period. What’s remarkable now is they all existed before these tools made it possible to learn how to build something almost instantly, or ask one question and summarize a very complicated text or book immediately, or do things previously impossible — can you make me a CAD model of this idea. Capabilities truly beyond my reach are now in reach. If those fast projects were possible before with the capabilities we used to have, shouldn’t we just see an exponential increase of those types of unreasonably quickly and effectively executed things with what AI has given us?

Lenny Rachitsky

正如你说，最难的是记得去试——「哦对，让我看看 Codex 能不能帮我做这个」。这是新习惯。Tyler Cowen 网站上有句话大意是：多数人低估了去找别人说「你做的更野心版本是什么？能不能更快试？能不能按 10 倍规模试？」的影响。现在 PM 特别有效的一件事，就是抬高别人的野心、提醒这里什么是可能的。有人说我们能这样、按这个时间线、或这是第一版时，你的工作之一是说：天花板其实高得多——我们不该更野心吗？能不能更快试？

To your point, the hardest part is just remembering to even try — “oh yeah, let me see if Codex can do this for me.” It’s a new habit. Tyler Cowen has this statement on his site that most people underrate the impact of going to someone else and saying, hey, what’s the more ambitious version of what you’re doing, or couldn’t you try this faster, or at a 10x bigger scale. In some ways, what PMs can do that is incredibly effective now is elevating others’ ambitions or reminding them of what’s possible. When folks say we can get this done this way, or by this timeline, or maybe this is the first version, part of your job now is to elevate everyone’s ambitions and say, actually, isn’t the possibility ceiling meaningfully higher? Shouldn’t we be more ambitious about what we’re attempting? Or couldn’t we try this faster?

Tara Seshan

对，这是很棒的位置——就能建什么、什么可能、以及工作有多兴奋而言。

Yeah, it’s a great place to be in terms of what you can build, what’s possible, and how exciting the job becomes.

Lenny Rachitsky

有意思。我记得 Nick Turley 上过这档——也许是你之前的角色，他现在好像做企业相关。他内部有个 meme：is this maximally accelerated？好像还有 emoji。

That is so interesting. I remember Nick Turley was on the podcast — maybe he had the role before you. I think he’s working on enterprise stuff now. He had this meme internally: is this maximally accelerated? There’s like an emoji inside Slack.

Tara Seshan

是的。「Is this maximally accelerated?」完全是 OpenAI meme。Andrew Ambrosino 和我爱问团队的另一个是：are you mainlining it yet——你是不是整天都在用这个产品把事做完？再加上我们是否尽可能野心——关于你尝试的范围和规模。「是否 maximally accelerated」是我们是否尽可能快；「are you mainlining it yet」是你是否在用、是否把全部品味压到这东西是否管用、是否真是人想要的东西上，并把反馈环收得尽量紧。这三者对我来说是产品开发必须尽量传播的三个 meme。

Yes. “Is this maximally accelerated?” is totally an OpenAI meme. The other OpenAI meme Andrew Ambrosino and I love to ask the team is: are you mainlining it yet — are you using this product all day every day to get your thing done. In combination with are we being as ambitious as possible — about the scope and scale of what you’re trying to do — is this maximally accelerated, are we moving as fast as possible on it; and are you mainlining it yet — are you using it and bringing all your taste to bear on whether this works and is something people really want, and tightening that feedback loop as much as possible. Those to me are the three memes of product development we just have to spread as much as possible now.

Lenny Rachitsky

喜欢。这是新版 dogfooding——你得 mainline 它。从外部看你的团队发推也看得出来：他们有多痴迷产品，不断问能更好什么、现在烦什么、在建什么。很清楚人人都是自己产品的创始人。还有别的内部 meme 吗？

I love that. That’s like the new dogfooding — instead of dogfooding, you got to mainline it. And that shows so deeply in the tweets — mostly how I see your team communicate — how obsessed they are with the product, constantly asking what can we do better, what’s bugging you now, here’s the thing we’re building. It’s very clear, to your earlier point, that everyone is just the founder of their product. Are there any other memes?

Lenny Rachitsky

还有别的内部 meme 吗？那些太有意思了。

Are there any other memes internally? Those are so interesting. Any other — I don’t know.

Tara Seshan

我在想还有没有好的文化 meme。很重要的一个是 feeling the AGI——意识到 AGI 在来。结果可以有很多种、各人想法不同，但把多数人留在这家公司的，是相信 AGI 有益这一使命，并尽一切让它实现：既实现 AGI，又确保对人类有益。

I’m trying to think if there’s other good cultural memes. Certainly a really important one is feeling the AGI — or just being conscious of AGI coming. There are so many outcomes for what it could look like, but a huge part of what puts most people at this company is believing in that mission of AGI being beneficial and trying to do whatever it takes to make that possible — both realization of AGI and ensuring it is beneficial for humanity.

Tara Seshan

建产品时，我脑子里另一个常念叨的是：我们是在为两到三个月后模型会到的地方建吗？为模型现在的能力建会失败；为你以为一年后会到的能力建也会失败。两种同样错。太早错；过度贴合旧模型能力也整盘错。唯一建法是两到三个月，并带着这束光：模型会好得多；模型能力是产品的中心；我建的产品结构不能挡模型的路。我怎么确保这对两三个月后的模型是对的？

In building products, another constant refrain I keep in the back of my mind is: are we building for where the models are going to be in two to three months? You fail if you build for where the models are now. You fail if you build for where you think the models will be in a year. Both outcomes are equally wrong. If you’re too early, you’re wrong. If you build something overly focused on a past model’s capabilities, you’re entirely wrong. The only way to build is two to three months, and having this beam of: models are going to get way better; I need to think about model capability as the center of this product; I need to get out of the way of the model in terms of the product constructs I create. How do I ensure this is right for the model in two to three months’ time?

Lenny Rachitsky

你怎么知道两三个月会长什么样？尤其在指数上，很难。是直觉？研究者会给你感觉吗？怎么运作？

How do you know what two or three months is like? It’s a challenging understanding, especially while we’re on this exponential. Is it just a gut feeling? Do the researchers give you a sense? How does that work?

Tara Seshan

和 research 紧密沟通他们觉得走向哪里极其重要。这些不完全是黑箱——你会知道我们在盯这些事：希望模型在编码的某些具体方面更好，或在写作的某些方面更好。我们确实有聚焦努力去提升特定能力。知道那是什么，并尽量让产品开发贴着 research 的议程和路线图，非常重要。

Communicating really tightly with research on where they think things are going is incredibly important. These things aren’t entirely a black box — you kind of know, hey, we’re focused on these particular things: we’d like models to be better at coding in these specific ways, or better at writing in these specific ways. We certainly have focused efforts on making the model better at specific capabilities. Knowing where that is and ensuring product development is as tied as possible to what research has as its agenda and roadmap is really important.

Lenny Rachitsky

我永远忘不了 Kevin Weil 上播客时——当时他还是首席产品官——说的：这是模型会有的最差水平。听起来简单，但很难真正接受：以后只会更好。现在说起来几乎像陈词滥调，但它是真的，荒谬。

A quote I’ll never forget is when Kevin Weil was on the podcast. He was chief product officer at that time. He said that this is the worst the models will ever be. It sounds so simple, but it’s hard to wrap your head around — this is the worst there will ever be. It’s such a cliché almost now to say that, but it’s true. It’s absurd.

Tara Seshan

对，荒谬。绝对荒谬。

Yeah, it’s absurd. It’s absolutely absurd.

Lenny Rachitsky

好，我想简短聊聊 ChatGPT 这个应用。我现在正开着。我看到 ChatGPT，下拉里有 ChatGPT 和 Codex，还有 Chat 和 Work 的切换。Tara，这都是什么？各自干什么？你觉得以后会怎样——会一直这样，还是你已经在想下一步？

Okay. I want to talk about the ChatGPT app briefly. I have it open right now. Here’s what I see: ChatGPT, and there’s a dropdown with ChatGPT and Codex, and then there’s this toggle — Chat and Work. Tara, what is going on? What are all these things for, and where do you think this goes? Is it going to stay like this? Is there a next step you’re imagining already?

Tara Seshan

我们的北极星是：用户不必在这些选项之间做决定。不该有切换——你走到输入框，打出任务，比如「我想建一个很棒的应用，帮播客嘉宾在节目前做研究」，它就会自动选对 harness、选对模型，帮你把事做完。选择不该压在用户身上，让他们既理解自己要做什么，又理解我们产品的限制和能力。短期这就是我们想去的地方。

Our north star is that users do not need to make decisions between picking all these different options. There is no toggle — you go to the box, you type in your task, like “I’d like to build a really awesome app that helps my podcast guests do research before episodes,” and it will just pick the right harness. It’ll pick the right model for you to get that done. Ideally the choice is not on our users to pick between all these concepts and understand not only what they’re trying to do but the limitations and capabilities of our products. That is certainly where we want to go in the near term.

Tara Seshan

在 ChatGPT 和 Codex 之间选，本质是：你要留在更偏开发的 UI，还是要在 ChatGPT 模式里拿到同样的能力和力量。如果你是 Codex 用户，继续用 Codex——你不会错过什么，尽量多用。如果你是 ChatGPT 用户，想搞清这些新智能体能力，大概该在 ChatGPT 模式。在 ChatGPT 里，若你要对话、要搜索，Chat 模式是对的——你熟悉的聊天模式，模型和能力越来越好。Work 模式底层就是 Codex：我们拿掉了一些编码 UI，不会突然弹出 work tree，但有同样把事做完的能力，比如生成很复杂的财务模型。我们看到——比如公司财务团队——用 Work 模式做出以前要么手工、要么依赖团队里一人深厚专长的事，变成整队都能执行，或抬高时间线、能力、能做到的前沿上的野心。

Picking between ChatGPT and Codex is really a choice: do you want to stay in a more development-oriented UI, or do you want the same power and capabilities in ChatGPT mode. If you’re a Codex user, keep using Codex — you’re not missing out on anything; continue using it as much as possible. If you’re a ChatGPT user who’s like “what are these new agent capabilities,” you should probably be in ChatGPT mode. Inside ChatGPT, if you want conversations, if you want to search, that’s where Chat mode is right — the same chat mode you know and love, with better models and newer capabilities every time. In Work mode, under the covers this is Codex. We’ve removed some of the coding UI — you’re not going to see a work tree pop up all of a sudden — but it is the same power to get things done, for example generate a really complex financial model. We see people — I mentioned our corporate finance team — use Work mode to do incredible things that were previously either manual or required deep expertise from one person on the team, become things the whole team can execute, or just elevate everyone’s ambitions on timeline or capabilities or the frontier of what they can get done.

Lenny Rachitsky

很有帮助。所以大概三种模式：工程模式、聊天模式、知识工作模式；知识工作模式其实是 Codex 在干活，但人可能不知道 Codex 是什么、甚至怕它。Work 模式里有没有不只是 Codex 的东西？有没有额外 harness 微调让它感觉不同，还是同一套换了点 UI？

That’s really helpful. So there’s kind of these three modes currently — engineering mode, chat mode, and do-knowledge-work mode. And the knowledge-work mode is actually Codex doing all that work, but people may not know what Codex is, may be afraid of it. Is there anything in that Work mode that’s not just Codex? Are there additional harness tweaks to make it feel a little different, or is it just the same thing with a little different UI?

Tara Seshan

真的主要在 UI 层。Work 模式和 Codex 模式：你去 Codex 让它生成很棒的财务模型给产品定价，或预测未来六个月收入，Codex 会做得和 Work 模式一样好。差别在于过程中你想在思维链里看到怎样的 UI、暴露多少技术细节。力量是同等的，所以 Codex 用户不切模式也不会吃亏。事实上我们不希望他们切——留在 Codex，做你想在 Codex 做的一切，我们会按你问的事展示合适 UI。真正的北极星是把这些合并，让用户不必做任何这类决定。分开更多是为了尽量在人们所在之处、在他们熟悉的产品和概念上接住他们，确保人人能用上智能体——智能体已经彻底改变了每个开发者的工作方式，我们该对知识工作做同样的事。

It’s really at the UI level. Work mode and Codex mode — if you go to Codex and ask it to generate an amazing financial model to price your product, or predict my revenue for the next six months, Codex will do as good a job as Work mode. It’s really about, whilst it’s doing so, what kind of UI do you want to see in the chain of thought, what kind of technical detail do you want exposed. It’s similarly powerful, so Codex users aren’t missing out on anything by not switching modes. In fact we do not want them to — stay in Codex and do all the stuff you want in Codex, and we will show you the appropriate UI based on the things you asked for. Truly our north star is to merge all these things so users don’t have to make any of these decisions. The separation is really more about how we can meet people where they are as much as possible in terms of the products they use and their familiarity with concepts, and make sure we are enabling everyone to take advantage of working with agents, which has transformed entirely the way every single developer works. We should do the same thing with knowledge work.

Lenny Rachitsky

说得通。东西动得太快：有人说「试试 Codex，会很酷」，然后起飞到一千万月活，接着又有 ChatGPT、又有 Codex，怎么合。过渡期不会立刻显得完美，因为你要边走边调，慢慢把人推向超级应用的愿景。

It makes sense. Because things move so fast — someone’s like let’s try Codex, this is going to be awesome, then it takes off and there’s 10 million monthly active users, and then wait, we’ve got ChatGPT, we’ve got Codex, how do we… It makes sense why these things aren’t going to feel obvious and perfect for a while — you have to adjust as things work and don’t, these transition periods of okay, now let’s get people moving towards this vision of the super app.

Lenny Rachitsky

我想象你工作最难的一块，是平衡这个号称约千亿 MAU、或许是史上最成功消费产品之一的 ChatGPT，和 Codex 这种新东西，以及你们还想试的其它新东西。你怎么想这件事——平衡非常创新、快速移动的团队与产品，和「有十亿人在用，我们不能乱改」？

Okay. I imagine one of the hardest parts of your job is balancing this 100 billion MAU product ChatGPT — maybe the most successful consumer product in history — with Codex, which is this new thing, and other new things you guys want to try. How do you think about balancing these very innovative fast-moving teams and products with “okay, there’s a billion people using this, we can’t change this dramatically”?

Tara Seshan

最有意思的一点是：我们在 ChatGPT 网页和桌面应用上线 Work、把这些合在一起，目标之一就是看着那十亿在用 ChatGPT 的人，把更多智能体能力带给他们。如果第一纪元是聊天，第二纪元显然是和智能体协作——目前主要是编码智能体——我们想带到更多领域，尤其是知识工作；把 Work 的力量交给这十亿聊天用户，是目标的一部分。

One of the most interesting things here is that one of the goals of launching Work in ChatGPT web and in the desktop app and bringing these things together was to look at those billion people using ChatGPT and bring them more and more of the agents’ power. If you think about the first era of AI products as chat, the second era is clearly working with agents — and primarily has been coding agents. We’d like to bring it to more domains, certainly knowledge work, and that is part of the goal of giving all these billion chat users the power of Work.

Tara Seshan

产品挑战是：不仅带给他们，还要自然、易采用，不要变成必须显式做的决定——我们直接帮他们做对的事。怎么去复杂化，让他们不必想 harness 这类对十亿消费者来说很疯的概念。那是主要挑战。然后可能很快到来的第三纪元是：你怎么和一个持久共事者一起把事做完，也许还和别人协作。近期挑战的一部分，是把智能体介绍给可能还没体验过的十亿人——怎样以最容易、最自然、最好用的方式做到。我们还有很多要做。

The product challenge on us is how do we not only bring it to them but make it natural and easy to adopt — make it not a decision they have to explicitly make; we can just help them do the right thing. How do we decomplexify it so they don’t need to think about harnesses, which feel like crazy concepts for a billion consumers to understand. That is primarily the challenge. And of course that third era that might come soon is how do you work with a persistent coworker who is able to get things done with you, maybe collaboratively with other people. Part of this challenge in the near term is we’re introducing agents to a billion people who may not have experienced them yet. How do we do so in the easiest, most natural, and most usable way possible? There’s a lot more for us to do to make that happen.

Tara Seshan

另一课是对比前 AI 时代或过去的产品经验：以前公司 polish 为王——每个 UI 交互、每件小事完全做对，比早发更重要，因为时间对结果没那么大差别。犄角旮旯不抛光、不完全正确，不如不发。这个时代有意思的是：当你有足够确信「这会变革」时，把产品交到用户手里，远好过完美；紧迫感和引入本身极其重要。我们还有很多要做，让它对聊天用户更好用、更易——尤其对不只把它当生产力、也当消费任务用的人——但完成好过完美，我们还有太多要做。

Part of this is also a lesson contrasting the pre-AI era or past product experience: at previous companies polish was king — getting every UI interaction or every little thing completely right was way more important than shipping something early, because time didn’t make as much of a difference in the outcome. If every corner wasn’t perfectly polished and everything wasn’t exactly correct, you might as well not ship it. What’s been really compelling about this era is getting the product in the hands of users when you have so much conviction that hey, it’s transformative — is way better than perfect. That urgency and that introduction of the product is so important. We have a lot to do to make it more usable and easier for chat users, certainly especially for folks who aren’t even using it for productivity but for consumer tasks — but done is better than perfect and we have so much more to do.

Lenny Rachitsky

记得这应用刚上时，很多人吐槽困惑；看团队多快迭代、响应反馈，正是我听到的：先发出去，搞清什么不行、人怎么用，快速迭代。感觉这就是现在的模式。

I remember when this app first launched, there was a lot of comments about the confusion, and seeing how quickly the team iterated and responded to feedback is exactly what I’m hearing — get it out, figure out what’s not working, how people are using it, iterate quickly. Feels like that’s the model now.

Tara Seshan

当然，有些东西你可以在上线前继续迭代、拿反馈；有很多我们总能、也该做得更好。但无论上线前还是后——理想是上线前——尽快迭代、听对信号，是关键。

And of course there are things you can continue to iterate and get that feedback prior to launching, and there’s a lot we can and should always do better. But iterating as quickly as possible and listening to the right signals — regardless of whether that’s pre-launch or post-launch, ideally pre-launch — is the key thing.

Lenny Rachitsky

Twitter 上我明显感到过去几个月有从 Claude Code 到 Codex 的 vibe 转变。以前人人 Claude；最近至少在 Twitter 这个泡泡里——很多技术人在那儿——更往 Codex 倾。除了 Tara 加入、把船稳住之外，过去三到六个月内部有什么可以分享的：我们搞清了这个、切了那个、砍了那个——什么帮 vibe 转过来、帮 Codex 变得像现在这样成功？

Something I’ve noticed on Twitter is there’s definitely been this vibe shift from Claude Code to Codex in the past few months. It used to be everyone was Claude. More recently it feels like people are leaning now towards Codex, at least on Twitter, which is a bubble, but it’s where a lot of tech people are. Curious what’s shifted internally in the past three to six months, other than Tara joining and shaping up the ship. Anything you can share — okay, we figured this out, we shifted this, we cut this — what helped shift the vibes and help Codex become as successful as it is becoming?

Tara Seshan

有句话大概是：开悟前担水砍柴，开悟后担水砍柴。Codex 应用这边，最初把它跑起来、做它的团队超级以用户为中心、迭代环很紧，真的 dogfood——尽可能 mainline 这个应用，把每件事做对。外面和 Twitter 上的人开始意识到在发生什么，用户真正注意到了；但团队一直盯用户、盯迭代。某种程度上，变的是市场在追上来。内部运作方式没变：每个建它的人显然是开发者，在用它做开发，不断修自己的问题，也听公司里别人和用户的问题。

There’s this phrase: before enlightenment, carry water, chop wood; after enlightenment, carry water, chop wood. With the Codex app, the team who initially got it up and running were super user-focused, tight iteration loop, really dogfooded the thing — mainlined the app as much as possible to get everything right. Folks started to realize that was happening externally and on Twitter, and users started to really notice — but the team was always really focused on users, on that iteration. It was merely, to some extent, the market catching up that was the change. That process has not changed internally. Everyone still constantly uses the app. Everyone who’s building it is obviously a developer using it for development, constantly fixing not only their own problems but trying to listen to other people in the company’s problems and user problems.

Tara Seshan

了不起的是运作模式没变。一直是我前面说的：我们是否把使命抬得够高？是否 maximally accelerating 进展？是否尽可能 mainlining？用户和 Twitter 注意到了很好，但那套运作——全部功劳归团队——并没有变。

What’s remarkable is that the mode of operating hasn’t changed. It’s always been the same thing I mentioned earlier: are we elevating our mission sufficiently? Are we maximally accelerating progress? Are we mainlining it as much as possible? It’s great that users and folks on Twitter have noticed, but that operation — full credit to the team — hasn’t changed.

Lenny Rachitsky

这答案有意思的地方是它非常人：是你、Andrew、Tibo、团队痴迷客户和产品。不是 AI 给了答案——是人拉开了差距。

What’s really interesting about this answer is it’s the very human part of it. It’s you, it’s Andrew, it’s Tibo, it’s the team just being obsessed with the customer, the product. It’s not like AI was the answer. It’s the humans that made the difference.

Tara Seshan

团队应得全部功劳。每个人都极其有想法、独立；说到 OpenAI 有很多创始人——那个团队上几乎人人，尤其是桌面团队，都像创始人一样追问、在意每个细节。他们注意到该更好的地方，就很独立地去建、跑起来；若内部测不好——人不用、觉得没用——就迭代，最后再对外发。那个循环，功劳全在团队和个人。

I’ll give the team full credit here. Everyone on the team is incredibly thoughtful and independent. To the point of there being many founders at OpenAI — almost everyone on that team, the desktop team especially, asks like founders and cares about every piece and every detail. When they notice an area that should be better, they go build it very independently, get the thing up and running, and if it doesn’t test well internally — people aren’t using it, people don’t find it useful — they’ll iterate and then finally ship it externally. That loop is full credit to people on the team and individuals for making that happen.

Lenny Rachitsky

你提到角色重叠：工程师在做偏 PM 的事，你大概在发原型、甚至可能发到生产。我也听到很多人问：我现在作为设计师的工作是什么？负责什么、不负责什么？作为营销呢？你注意到这点吗？在怎么处理？有什么想法？

Something you touched on is roles overlapping — engineers are doing PM-ish work, you’re probably shipping prototypes and building, maybe shipping to production. It feels like that also creates a lot of challenges. I hear from a lot of people: what is my job now as a designer? What am I responsible for, what am I not? As a marketer, what am I doing? Is that something you notice, something you’re dealing with — any thoughts along those lines?

Tara Seshan

我一直最喜欢在初创公司工作——有时初创不小心长成大公司——是角色边界极少：一切既是、又不是你的责任，你最终对成功负责。Stripe 非常是这样：工程师、产品经理、设计师之间没什么边界，人人什么都能做。我一直爱那种心态，现在能力终于追上来了。我真正在意的是：必须有人盯着、承担核心问责——产品有没有被人用？是不是人想要的？质量高不高？有没有效？无论那个人是工程师、设计师、PM 还是谁，有人是 DRI；为达那个结果需要做什么，人可以按亲和力和能力捡起来。我喜欢不太在意个人角色边界、人人盯结果发生的团队。

The thing I’ve always liked most about working at startups — sometimes I’ve started at a startup that accidentally grew into a large company, but primarily working at startups — is that there are very few boundaries around your role. Everything and nothing is your responsibility; ultimately you’re accountable for success. Stripe was very much this way — no boundaries around what an engineer vs a PM vs a designer could do; everyone could do anything. I’ve always loved that mentality, and now finally capability is catching up to that. What I really care about is that someone needs to look after or have core accountability for: is this product being used by users? Is it something people want? Is it high quality? Is it effective? Whether that person is an engineer or a designer or a PM or whomever — someone is the DRI — and then whatever work needs to be done to make that possible, people can pick up based on affinity and capability. I like a team that doesn’t really mind what the boundaries are between individual roles, but everyone’s focused on making the outcome happen.

Tara Seshan

另一面是我也真爱 PM 的 craft。Shreyas、Marty Cagan、Shishir 这些人倡导的很多 craft，我觉得很棒。有时问题来自：我那么爱自己领域的 craft，用更流动的团队协作方式做事，会不会丢掉打磨 craft、变得更好的机会？我真没有答案。我们都在一起经历：craft 的某些部分正被模型抽象掉，模型做得可能比个人更有效；你的 craft 从过去那个很具体的任务，变成把它用到产品或学科的另一块。我仍在想：怎么平衡——想成为团队一员、用这些工具、被现在的效率吸引——和我对 craft 的爱。对工程师来说，手写代码的心流曾经很有趣，现在很多人不再整天那样写了。

The converse is I also really love the craft aspects of being a PM. There are so many aspects to PM craft that folks like Shreyas or Marty Cagan or Shishir have really espoused that I think are wonderful. Sometimes maybe some of these questions come from: I so love the craft of my domain — by taking this more fluid approach to teamwork and collaboration to get something done, do I lose out on getting better and polishing my craft? I truly don’t have an answer for that. It’s something we’re all experiencing together — some pieces of our craft are actually getting abstracted by models being able to do it really effectively, maybe better than individuals can. Your craft moves from being able to do that very specific task you did in the past to now applying it to some other part of the product or the discipline. That is still a question I’m thinking about: how do I balance my desire to be part of a team and use these tools and feel so compelled by how effective one can be now, with my love of the craft. It’s really fun handwriting code for an engineer all the time, and one doesn’t really do that anymore.

Lenny Rachitsky

工程角色现在有多不同，简直难以置信。以前你的工作是整天写代码，现在不再是。变得太快了。有人怀念自己手动写代码的心流，对比现在做的事。

Yeah, that’s where I was going to go. It’s just unbelievable how different the engineering role is now. You used to write code all day — that was your job — and that is no longer your job. And that happened so quickly. People mourn the flow state of writing code manually yourself versus what one does now.

Tara Seshan

这是很难的过渡。有人爱，有人不爱——那是另一整套话题。

Yeah. It’s a tough transition. Some people love it, some people don’t. And that’s a whole other topic.

Lenny Rachitsky

顺着说，我常问站在 AI 前沿的人：你觉得人脑未来还会在哪里有价值？长期无法预测——我们还需要人吗？希望需要——但就接下来几年，人脑会在哪里最有价值？

Kind of along those lines, something I’d like to ask people at the frontier of AI is: where do you think human brains will continue to be valuable in the future? It’s impossible to predict long term — will we need humans? Hopefully — but I’d say in the next couple years, where do you think human brains will continue to be most valuable?

Tara Seshan

人会继续最有价值的一块，肯定是作为问责主体：谁最终拥有结果？某种程度上，你可以把一起工作的智能体想成你的下属；最终谁拥有最终产品——质量够不够、是不是你要它做和说的——至少现在，那仍会是人。尤其在强监管、或需要直接人际界面的行业和地方，这很说得通。

I think humans will continue to be most valuable certainly as an entity of accountability — who ultimately owns the outcome here. In some ways you can think of your agent as your report. Ultimately who owns what was the end product — was it high quality, was it the thing you wanted it to do and say — that will certainly remain a person, at least for now. Especially in industries and places that are highly regulated or require a direct human interface, that makes a ton of sense.

Tara Seshan

人脑对表达也很有价值。我前面说软件不像房地产，更像电影：砸钱也不保证出好片；最伟大的电影不一定预算最大。建软件里有艺术性、观点和表达，你会感到有人或一群人的作者身份。选建什么、它感觉如何，仍是非常人的问题。

The human brain is also really valuable for expression. I’d mentioned earlier that software is not like real estate — it is more like a film where you could put money in and a great film does not come out; the greatest films are not the ones with the biggest budgets. There’s a certain artistry and opinionation and expression in building software where you feel there is some authorship by a person or a group of people. That part remains so human — what you choose to build and how it feels feels like such a human question.

Tara Seshan

我也觉得人脑会继续在我们如何彼此关心、彼此联结上有价值。我工作里那一块一直很「人」，而且其实变得比以往更重要：和团队里其他人谈话，一起搞清怎么对某个领域保持热情，怎么一起学、一起干，怎么抬高彼此的野心。那一切仍是非常人的事。我无法预测模型会怎样，但那些块面对我极其人。

I also think the human brain continues to be valuable in how we care for each other and relate to one another. That piece of my work has remained so human and has actually become more important than ever — the part where you talk to other people on your team and collectively figure out how you can be enthusiastic about an area, how you learn and work together, how you elevate each other’s ambitions. All of that feels and remains such a human thing to do. I can’t predict what’ll happen with the models, but those pieces feel to me to be incredibly human.

Lenny Rachitsky

我喜欢这个答案。

I love that answer.

Lenny Rachitsky

你谈到过 AI 能力与我们实际用法之间的悬空。最大缺口之一常常是：好，我到底该拿它干什么？有没有一些你工作中用 AI 的方式，可能启发别人「哇我没想过那样用」？两桶：一是 AI 让你的 PM 工作变最多的地方——现在我靠 AI 干这些；二是最近有没有特别有创意、你觉得「我也该试试」的用法。

There’s this idea you talked about of the overhang of what AI is capable of and what we’re actually doing with it. People are always like, one of the biggest gaps is okay, what should I do with it? I’m curious what are some ways you use AI in your work that may inspire people — oh wow, I didn’t think about using it that way. Two buckets: one is how your PM job has changed most thanks to AI — okay, now I use AI for this stuff; and is there any super interesting creative use of AI recently that you’re like, oh yeah I should try this.

Tara Seshan

我工作中最兴奋的用法之一是：我现在一直在建 Sites。你试过用 Sites 建站吗？

One of the most exciting ways I use AI in work is I actually build sites all the time now. I don’t know if you’ve tried building sites —

Lenny Rachitsky

没有。聊聊 Sites。

I haven’t. Talk about sites.

Tara Seshan

Sites 特别好玩。你可以在 Work 里把站点建成演示产物，但我几乎什么都用站点建。我给团队建过一个游戏站点，大家一起玩——因为站点有数据库。我最近去背包旅行，建了个站点画路线、标海拔，旅行的人还输入各自的食物。又快又有效。Sites 某种程度上实现了 Alan Kay 六十年代提过的可塑个人软件梦想：真正的个人计算机要有个人软件。人们用 Notion 之类用积木配置，但站点就是一句提示：按我说的建我需要的这个工具，它就做了。可分享，能自动更新，能用内部数据建指标仪表盘。比起费力做幻灯片，站点是动态得多的演示表面。

Sites is a really fun, amazing product. You can basically build a site in Work as a presentational artifact, but I also build sites for literally anything. I built a site for the team as a game where we all played together using a site, because sites have a database. I went on a backpacking trip recently and built a site of the route that tracked the elevation of everywhere we were going. Everyone on our trip inputted all their food. Super fast and effective. Sites kind of realized the dream of malleable personal software that Alan Kay flagged in the 60s — the true personal computer is one that has personal software. People with tools like Notion try with blocks to configure what that could be. But with a site, it is literally a prompt. I literally with a prompt say build me this exact tool that I need to get this thing done and it just does it. They’re sharable. They can auto-update. You can use internal data to build a dashboard with lots of metrics. Rather than painstakingly laboring over a slide deck, a site is just a way more dynamic surface for presentation.

Lenny Rachitsky

你怎么用站点？要特别操作吗，还是说「建一个站点」就行？

How do you use a site? Do you have to do anything special, or you tell it create a site?

Tara Seshan

在 Codex 里说：给我建一个给团队玩的黑手党游戏站点——它就会做。

In Codex be like: create a site that is a mafia game for my team, and it will just do it.

Lenny Rachitsky

我脑里是大写的 Site，但大概无所谓。它就知道 Sites 是什么？以前是给你源码、你自己找地方部署——你是说它直接帮你托管？

I’m thinking capital-S Site, but doesn’t matter I imagine. So it just knows what sites are? It used to be here’s some source code, go figure out where to deploy it — and what you’re saying is it just hosts it for you?

Tara Seshan

对，你可以选公开、和团队共享、或仅自己私有。随时能建站点，改变了我日常：以前日常是大量文档、表格之类产物，现在我一直在做站点。

Host — you can choose whether it’s public, with your team, or private to you. The easy reach of building a site all the time has changed what my day-to-day looks like, which often in previous worlds used to look like creating lots of artifacts like docs and sheets. Now I just make sites all the time.

Lenny Rachitsky

可以通过 Work，还是 Codex、Web 都能？

And you could do that through Work, or through all the surfaces — Codex, Work?

Tara Seshan

Work 可以，Codex 可以，Web 可以，移动端也可以。哪都能做。

You can do it through Work. You can do it through Codex. You can do it in the web. You can do it on mobile. You can do it anywhere.

Lenny Rachitsky

好，我刚踢出了一个「建一个关于 Tara Seshan 的站点」。对了，你姓怎么念？我还没问过。

Okay. I just kicked off a create a site about Tara Seshan. Is that how you pronounce your last name by the way? I haven’t asked.

Tara Seshan

Tara Seshan——有点像 station 那个音。

Tara Seshan — like “station.”

Lenny Rachitsky

好，酷。Sites。

S. Okay cool. Sites. Okay.

Lenny Rachitsky

还有没有快速小贴士？Sites 这条就很好，很多人不知道，很有用。

Any other quick tips while we’re on this topic? That was a great tip because I don’t think a lot of people know about sites. That’s very useful.

Tara Seshan

Sites 太棒了。另一件我很爱的是在 Codex 里用 visualize。你用过 /visualize 吗？

Yeah, sites are awesome. The other thing I really love is using visualize in Codex. Have you used /visualize?

Lenny Rachitsky

没有。

No.

Tara Seshan

/visualize 极其兴奋。你可以 /visualize my ChatGPT usage until now 之类，它会拉进你做过的一切，做出很棒的可视化。我无数次想：怎么不仅拉一堆图和数据，还用对我要讲的故事来说易懂有用的方式呈现——visualize 让这变得极其简单。用起来出奇地令人愉悦。

Oh, /visualize is incredibly exciting. You can just do /visualize my ChatGPT usage until now or something like that, and it will pull in all the things you’ve done and create an amazing visualization for it. The number of times I’ve been thinking about how do I not only pull in a bunch of charts and data but present them in a way that is understandable and useful for the story I’m trying to tell has been infinite — and visualize makes that incredibly simple. It is surprisingly delightful to use visualize.

Lenny Rachitsky

这些都是好例子：能力这么多，我们甚至不知道或不理解——而这正是你们的挑战。帮我们知道这些。播客也因此有用——没法全塞进产品里。

These are such good examples of there’s so much power here we don’t even know about or understand — and that’s the challenge you have here. Help us know all these things. That’s why podcasts like this are also useful. Can’t put it all in the product.

Lenny Rachitsky

我要换个方向聊写作。我问过很了解你的 Brie Wolfson 该问你什么。她说该问写作/思考——Tara brief 是标志性的。帮我们理解你的写作、你的 brief 为什么标志，以及对想写得更好的人有什么建议。

I’m going to go in a totally different direction. I want to talk about writing. I asked Brie Wolfson, who knows you well, what to ask you. She said you should ask her about writing slash thinking. A Tara brief is iconic. Help us understand what makes your writing, your briefs iconic, and any tips that might be helpful for people trying to get better at writing documents.

Tara Seshan

我强烈认为工作中有两种写作：一种是写作即思考，一种是写作即汇报。写作即思考，是我写 brief 说明为什么该建某产品、为什么该走某策略，或某个带刺观点。写作即汇报，是总结团队这周状态、发报告，或某次发布/公告的计划。汇报类写作我乐意自动化，尽量用模型简化；思考类写作我永远不会自动化。

I really strongly believe I do two types of writing at work. One is writing as thinking and the other is writing as reporting. Writing as thinking is me writing a brief about why we should build a certain product or take a certain strategy, or maybe a spicy take. Writing as reporting is summarizing the status of what our team has been up to this week and sending a report, or this is our plan for this particular launch or announcement. Writing as reporting I happily automate — I use the models all the time to make that as simple as it can be. But writing as thinking is something I never will automate.

Tara Seshan

至少对我来说，走完大纲、变成一定程度的散文、删改、继续迭代，是把想法理顺最重要的步骤之一。多数人刷子刷太宽：「我永远不用模型写作」或「我总用模型写作」——两种都错。汇报类应尽量用模型；若你像我一样用写作思考，就不要用它取代你的思考。

I really strongly believe that at least for me the act of going through and outlining something, turning it into some level of prose, cutting and editing it, continuing to iterate on it is one of the most important steps for me to get my ideas in line. Most people paint with a really broad brush — I will never use the models for writing, or I always use the models for writing — and both those broad brushes are wrong. You should use the models as much as possible for writing as reporting, and insofar as you think with writing as I really do and I think a lot of people do, you shouldn’t replace your thinking with it.

Tara Seshan

我过去的 brief：因为我大量用写作思考，会钻进洞里写新产品或新想法的 brief，花大量时间打磨，拿去给人「攻击」想法、挑洞、变强，再找下一个人重复。在 Stripe 我做过无数次——启动新产品线、建议大转向、分析问题并提出路径。Stripe 极度是写作文化，Jeff Weinstein 等人也很爱写 brief、分享；Stripe 是少数 brief 能在公司里「病毒传播」的地方。写作即思考在那里很受珍视，我大部分那种写作是在那里做的。

My briefs in the past — because I write so much as a way of thinking — I will go into a hole, write a brief for a new idea or product, spend a ton of time refining that idea, shop it around with people and have them attack the ideas as much as possible and poke holes, make it stronger, then take it to the next person and do the same thing. At Stripe this is something I did many, many times — whether to kick off a new product area, suggest a big change in direction, or analyze a problem and suggest a path forward. Stripe is incredibly oriented as a writing culture, and there are many people like Jeff Weinstein who are also very into writing and sharing briefs. Stripe is one of the few places where a brief will go viral inside the company. Writing as thinking there is really prized, and that’s where I did the majority of that writing work.

Tara Seshan

在 OpenAI 我仍一直写作即思考，但可分享的产物不再是那种长文档或「证明你想过」的东西。部分因为时代变了：长文档不再是你想清楚的信号——你可以轻易产出表明你没想过的长文档。我个人日常最大、也最别扭的变化之一是：以前我在文档里想，再翻译成演示产物，那就是我想过问题、这是我们要做的、团队往那走的信号；现在我远更偏 mocks 而非 docs、原型而非 docs。如果有人能试用、能交互，更好——我们试了、跑了 A/B、这是结果、所以我认为该往这走——那比文档本身好得多的沟通工具。我仍写大量文档，但那是写给我自己的，不再真的是写给别人看的最佳沟通方式。这大概是这个时代相对上一个时代，我个人经历的最大变化。

At OpenAI I still write as thinking all the time, but the shareable artifact here is not really a long doc or a sort of proof of work in that way — partially because times have changed and a long doc is not a signal that you thought through something, because you can easily produce a long doc that indicates that you haven’t. One of the biggest changes I’ve experienced personally in my day-to-day — a big maybe jarring change — is that I used to think in a document and then do some translation of that into a presentational artifact, and that would be my indication that I thought through a problem and this is what we’re going to do and the team moves in that direction. Now I am way more on mocks not docs, or prototypes not docs. If I have something people can try and interact with, or even better, results where we tried this, we ran an A/B, here’s the results, this is why I think we should go in this direction — that is a way better communication tool than the doc itself. I still write hundreds of docs all the time, but I do it for me. I no longer do it for other people really — that no longer is the best way to talk and communicate. That is probably the biggest change I’ve experienced personally in this era versus the previous era.

Lenny Rachitsky

太有意思了。我喜欢你说的大量拿反馈：听起来显然，但你几乎像「作弊」一样在迭代中拿很多反馈，才能写出标志性 brief，而不是第一次就「好了，这是它」——那样很少成功。

That is so interesting. I really liked your tip of getting tons of feedback on a doc. It sounds obvious, but you can get to an iconic doc or brief by just cheating almost and getting lots of feedback as you’re iterating, to make it stronger and stronger, versus “cool, here it is first time” — and it’s rarely.

Tara Seshan

以前有个经理告诉我：对的做法永远是把文档写到大约 70% 完成，再拿去找你需要买进的人，一起从 70% 到 100%。我现在仍一直这么做，因为很少有真正强的人想碰一个抛光完美、已经完成的想法——完美想法上，他们的新点子会弹开；而有棱角、有毛边、他们也能一起打磨的东西更好。把人拉进过程，文档作为底下的工件，是我发现最好的协作方式之一。

I previously had a manager who told me that the right thing to always do is write a doc to 70% completion and then take it to the people you need buy-in from and get it from 70% to 100%. That still is a thing I do all the time, because very few great people want to interact with a perfectly polished finished idea. A perfectly polished idea — their new ideas just bounce off it — versus something that has more crags and rough edges that they too can polish with you together. Bringing people into the process that way, where a doc is an underlying artifact for that, is one of the best ways to collaborate that I found.

Lenny Rachitsky

你怎么看 AI 脑腐、开始过度依赖 AI？人人都会碰到的挑战。为什么不用这魔法帮忙看东西——然后我们开始丢掉写长文、读长文的能力。你有没有刻意避免的做法？

How do you think about AI brain rot and starting to over-rely on AI? This is just a challenge everybody’s going to have. Why not use this magic to help look at something — and then we start to lose our ability to write, read long documents. Is there anything you do that you are trying to avoid that?

Tara Seshan

「写作即思考」纪律是我日常用来避免思考能力过度萎缩的主要办法之一。汇报类写作我会尽量外包给模型；思考类写作必须自己做。我有个个人信念：如果我要让别人读我的文档，我至少要先自己读过同样次数。开会也一样：如果我要开一群人的会，我需要在会前准备好大家将在会上花费的集体时间量。保持思考锋利时，我先自己写文档，至少投入我期望别人阅读所花的集体时间；我也不太靠模型抛光文笔——我觉得它其实做不好——尤其不会靠它生成第一版。但总结、把内容从一种格式翻译到另一种，我会大量让模型帮忙。

This writing-as-thinking discipline is one of the main pieces I employ day-to-day to make sure I’m not overly atrophying my thinking abilities. I will again outsource all writing-as-reporting as much as possible to the model, but writing-as-thinking I have to do myself. I have this personal belief that if I’m going to make someone read my document, I have to at least read it first that number of times. Or I think about this in meetings too — if I’m going to call a meeting with a set of people, I need to have prepped the collective amount of time people are going to spend in that meeting before the meeting. When it comes to keeping my thinking sharp, I do that writing for the document myself first and make sure I’ve invested like the collective amount of time I expect people to read it at least in writing it and producing it. I don’t really rely on the model either for polishing my prose — which I don’t think it really does — or especially not in generating the first version. But I do of course have the model help me a lot when it’s summarization or translation of content from one format to the other all the time.

Lenny Rachitsky

所以我听到的是：想法、brief、计划你自己写。自己写。不要从 AI 起笔，甚至不要用它改进文笔——保持全是人。

So what I’m hearing is: write the idea, the brief, the plan yourself as a human. Write it yourself. Don’t start with AI. Don’t even use it to improve on the writing. Just keep that all human.

Tara Seshan

对，至少对我：自己开始，自己结束。中间可能用 AI 研究特定要素、塞进数据、去拉数据，或帮忙——

Yeah. At least for me, I start myself and I end myself. I might use AI in the middle to research specific elements or drop in some data or go pull some data or help —

Lenny Rachitsky

反驳一些想法。

Push back on some ideas.

Tara Seshan

对，反驳一些想法；但用写作时自己开头、自己收尾，那不会劣化你的思考。

Yeah, push back on some ideas, but start yourself and end yourself with a piece of writing, and that doesn’t deteriorate your thinking.

Lenny Rachitsky

最后一个关于 Sutter Hill 的问题。你有过很不寻常的职业一步：你是 PM、创始人一类，然后去做了 Sutter Hill Ventures 的 EIR——标志性 VC，孵化过很多了不起的公司，比如 Snowflake，方式很独特。那是怎么回事？你学到了什么？

Okay, one last question I want to ask about Sutter Hill. You had this very unusual career step. You went from PM / founder person and then EIR at Sutter Hill Ventures, which is an iconic VC. A lot of amazing companies came out of Sutter Hill. It has a very unique way of approaching founding where basically they incubate companies — Snowflake as an example. What was that about? What did you learn from that experience?

Tara Seshan

Sutter Hill 是标志性公司，而且故意很难读懂。你去网站上几乎什么都看不到。它不高调，尽量低调、谦逊地运作，却对硅谷一些最标志性的成功负有责任。他们有非常独特的孵化模式，Mike Speiser 等出色合伙人开创并一次次滚出成功。

Sutter Hill is an iconic firm and is intentionally a very illegible firm. If you go to the Sutter Hill website, you will see nothing on the website. It is a firm that doesn’t operate loudly. It tries to operate as under the radar as possible, as modestly as possible, yet is somehow responsible for some of the most iconic successes Silicon Valley has seen. They have this very unusual incubation model which Mike Speiser — one of the amazing partners there — started and has rolled out success after success.

Tara Seshan

对我最标志的是：人们把找到产品市场契合、或建成数百亿美元公司当成黑箱艺术——哦靠运气、靠偶然、靠各种必须凑齐的东西。可 Mike Speiser 做成过多次，所以显然有办法、有路线图、有一套可重复做的事——不只是运气或黑箱。那套 playbook 活在 Sutter Hill 里；他们学会了大量押对注，也学会了日常复利式地建成功公司：企业销售团队怎么建、产品怎么定位、初始创始团队怎么组。Sutter Hill 的招聘极其出色；他们有个内部工具叫 Reticle，地图上有他们接触过的每个人，以及那些人接触过的十个最强的人——这帮他们极其有效。我去 Sutter Hill，是因为我的职业某种程度上一直在问：怎样尽可能多次找到产品市场契合——做创始人、在 Stripe 开新产品、加入 Watershed 这类创业公司。Sutter Hill 搞清了怎么在 B2B 产品上找到产品市场契合，我想向他们学能学到的。

The thing that was most iconic to me about Sutter Hill is that people look at finding product-market fit as a dark art, or building a tens-of-billion-dollar company as a dark art — oh it’s luck, oh it’s chance, oh it’s all these things that must come together. Yet Mike Speiser has done it multiple times. So there’s clearly a way to do it, clearly a roadmap, a set of things one can do to get this repeatably — it’s not just luck, not just a dark art. There is a playbook as it were, and that playbook lives inside Sutter Hill. They’ve figured out how to be right a lot in calling shots and making bets, and they’ve learned how to be right a lot in the daily compounding things one does to create a successful company — whether that’s how you set up your enterprise sales team, how you position your product, how you build the initial founding team. Recruiting at Sutter Hill is unparalleled. They have a secret tool called Reticle where they have a map of everyone they’ve interacted with and the 10 best people those people have interacted with. That helps them be so effective. I went to Sutter Hill because in some way my career has been about how do I try to find product-market fit as many times as possible — whether as a founder, starting new products at Stripe, or joining a startup like Watershed. Sutter Hill is a place where they’ve figured out how to find product-market fit on B2B products, and I wanted to learn what I could from them.

Lenny Rachitsky

你学到了什么？除了「他们知道怎么做」之外，带走的一件事是什么？

What’d you learn? What’s one thing you took away from that experience other than they know how to do it?

Tara Seshan

他们绝对知道怎么做。很意外的一点是：产品市场契合当然重要，但我以前严重低估了产品营销契合——你怎么谈产品、怎么营销它，甚至可以先于真正把产品建出来。它大概该来自对技术的深度理解和企业销售过程的理解拧在一起；那种叙事、定位，在你建产品体验之前，其实是该先测的。你该去 pitch 100 个人，尽量打磨 pitch，把「为什么这东西变革」的营销叙事做对，然后、也只有然后，才承诺「产品形态就是这样」。Mike Speiser 在这门艺术上几乎不可战胜。我以前低估 PMM 工作，觉得它只是职能之间的胶水；后来才意识到，做得极好时，那份工作对公司结果可以是变革性的，甚至可以是让公司成功的那个要素。

They definitely know how to do it. One thing that was very surprising that I learned there is that product-market fit is sure important, but actually I really underrated product marketing fit. The idea that the way you talk about the product and the way you market it can precede actually even building the product. It should probably come from bringing together understanding the technology deeply and then understanding the enterprise sales process — and that product marketing fit, that narrative, that positioning, is actually even before you build a product experience the right thing to test. So you should go pitch 100 people, figure out how to refine that pitch as much as possible, get the marketing narrative of why this thing is transformative right, and then and only then go commit: okay, this is exactly the product shape. Mike Speiser is unbeatable at this art. Previously I’d always kind of underrated PMM work — I was like it’s whatever, it’s the glue between these functions, it’s fine — and then I realized how transformative that work done excellently is to a company’s outcome, and in fact can be the element that makes a company successful.

Lenny Rachitsky

太棒了。我非常同意定位这件事，我们在这档播客谈得很多。

Amazing. I so agree with that — positioning. We talk a lot about that on this podcast.

Lenny Rachitsky

好，我给你看刚才 Sites 边聊边跑出来的东西。看这个。

Okay, I’m going to show you what sites got created real quick. It was running while we were talking. Check this out. Look at this.

Tara Seshan

哦天。

Oh man.

Lenny Rachitsky

我说「做得更酷一点」，它就更酷了。多产品、设计很正经。还有引用——Big conviction. Small teams. Start with the buyer. 你觉得把它当你新官网怎么样？

I was like make it more awesome and it made it more awesome. Multiproduct. Beautiful. Look at this. This is like legit design. Look — you got quotes. Big conviction. Small teams. Start with the buyer. How do you feel about this being your website — your new website?

Tara Seshan

我觉得顶部那张大概十九岁的我的照片很好笑。除此之外我很爱这个站点。

I do think that the picture of me at maybe 19 years old at the top is really funny. But yeah, otherwise I love love the site.

Lenny Rachitsky

好，干得漂亮。

Okay, good job.

Tara Seshan

看起来不错。我想那是我在 Stripe 的工牌照片。

Looking good. I think that was my badge photo from Stripe.

Lenny Rachitsky

哇。太妙了——我已经取消分享了——它还围着你的头建了一小圈装饰，好可爱。Tara——

Oh wow. Amazing. I love that it built — I already unshared it — but I love that it built a whole little thing around your head. So cute. Tara —

Lenny Rachitsky

闪电轮之前，还有什么想分享、想碰的吗？

Is there anything else that you wanted to share? Anything else you want to touch on before we get to our very exciting lightning round?

Tara Seshan

我们在产品建设、尤其是 ChatGPT Work 和新纪元里想得很多的一点是：知识工作和编码其实根本不同。我们学到的一个意外是：编码非常以输出为导向——你让它做编码任务，可以用测试验证做得对不对、好不好；你可以试用，看它通不通。有办法基于输出验证。但知识工作不同：我不能只看最终那份 deck 里的数字说「哦大约 90% 成功」就当真。我真的需要想过程、输入、推理，以及沿途怎么走的。

Yeah, one thing that we’ve been thinking about a lot in product building, especially with ChatGPT Work in this new era, is how knowledge work and coding are actually fundamentally different. One of the surprising things we learned is that coding is so output-oriented that when you ask it to do a coding task, you can verify whether it did the task correctly or well via tests. You can try it out and see if it works. There is a way to validate it based on the output. But knowledge work is different in that I can’t simply look at the deck in the end and see the numbers — oh, it’s like 90% success or whatever in the deck — and actually believe that. I really need to think about the process and the inputs and the reasoning and how it went along the way.

Tara Seshan

所以产品上，我们已做、还要继续做的很多，是继续把产品适配知识工作：更强调让 ChatGPT 成为你的协作者，让你看到所有进行中的工作、引用和输入，带你和模型一起走完到达最终输出的旅程，好让你最终知道：等等，这东西是对的、是好的、是有用的。这当然大量体现在 UX 上，也应体现在推理和思维链上——比如沿途该不该看到更多它如何从数据到达终态的引用？对编码很合用的「线程」表面，对知识工作是否也是看这些东西的对的地方？有很多重大产品问题。所以当我们想把人类协作者带进你的工作，也要想怎样让模型在你们一起把事做完时更像协作者。

In terms of how that looks in the product, a lot of work we have done and have to continue to do is continue to adapt the product to knowledge work, which means way more focus on making ChatGPT your collaborator, allowing you to see all the in-progress work, see its citations and inputs, help you go on the journey with the model to get to that end output such that you know in the end that oh wait, this thing is right, this thing is good, this thing is useful. That shows up certainly in the UX of the product quite a bit, but also should show up in things like the reasoning and the chain of thought — should you see more citations along the way, for example, of how it got to that end state in that data? Is the surface of a thread which is so suited to coding the right place for you to see all of that for knowledge work as well? There’s so many big important product questions. As we think of maybe bringing human collaborators into your work, we also need to think about how we can make the model more of a collaborator with you as you get things done together.

Lenny Rachitsky

这一点太好了。我想象高管会上你在 pitch 计划：很多是帮他们看见「我做了哪些工作才到这里、有哪些步骤」。所以 AI 也需要给你看同样的「工作量证明」；工程则是「我不需要知道所有小架构决定，看起来怎样、测试过不过」。两种模型差别很大。还有上下文：它有没有做你要的事所需的上下文——能不能看你的邮件、能不能看你所有的 Notion 文档。

That is such a good point. I’m imagining an exec meeting where you’re trying to pitch the exec on here’s what the plan is, here’s what I think we should be doing. So much of that is helping them see here’s the work I did to get there, here’s all the steps. So it makes sense that you need the AI to show you that same sort of work that it did — the proof of work essentially — versus engineering where okay, I don’t need to know all of the little architectural decisions you made, just what does it look like, is it passing all the tests. That’s a really good point just how different those two models are. And also there’s the context — does it have the context it needs to do the thing you want it to do — does it know, can it see your email, can it see all your Notion docs.

Tara Seshan

嗯。

Mhm.

Lenny Rachitsky

说得好。我看到你工作的挑战了——这一切是一个产品，很棘手。闪电轮前还有别的吗？

Such a good point. So I see the challenge in your job. I think all this work is one product. Tricky, tricky. Amazing. Anything else before we get to our very exciting lightning round?

Tara Seshan

好，我们开始吧。

Yeah. Let’s jump into it.

Lenny Rachitsky

到了非常激动的闪电轮。我有五个问题。准备好了吗？

With that, we’ve reached our very exciting lightning round. I’ve got five questions for you. Are you ready?

Tara Seshan

好了。

Yes.

Lenny Rachitsky

你最常向别人推荐的两三本书是什么？

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

Tara Seshan

一本我很推荐的是 William Finnegan 的《Barbarian Days》。写一个《纽约客》记者如何爱上冲浪作为热情。我带走的是：人可以深深热爱、投入，把某事当人生目的，即使你并不擅长。它关于爱上冲浪的艺术，以及在明知永远够不到的情况下仍追求卓越与完美。那故事对我如何继续生活非常有说服力、很变革。我真的很爱那本书。

One book I really recommend is Barbarian Days by William Finnegan. It’s about the life of a man who is a New Yorker reporter, but how he fell in love with surfing as his passion. The thing I took away is that one can be deeply passionate and dedicated and have something be your life purpose without you being good at it. It is about the art of falling in love with surfing and his striving for excellence and perfection whilst knowing that he will never reach it. It is such a compelling and transformative story for how I think one should continue to live our lives. I really love that book.

Tara Seshan

另一本我会推荐的是《安娜·卡列尼娜》。我最近在重读经典。我爱它，因为它是有层次的书。我觉得这个新纪元很大一块是我们要改造自己，走上和习惯不同的旅程。我想起十三岁读它时基本只懂情节；十七岁懂欧洲史动态和阶级冲突；三十岁再读，才觉得这是一个关于女人和人类的故事。它提醒我成长是可能的——随着成长，可以用多种透镜看同一件事。我觉得这有点像我们所有人面前职业上的挑战。

Another book I might recommend — I’ve been rereading the classics lately — I really love Anna Karenina, because it’s like a book of layers. A huge part of what we’re going to have to do in this new era is transform ourselves or take ourselves on a journey to do different things than what we were used to. When I think about that book, when I was 13 and I read it, I understood basically the plot. When I read it at like 17, I understood the European history dynamics and class warfare. And when I read it at 30, I was like, oh, this is like a story about a woman and humans. It just reminds me of growth and that it is possible to look at the same thing through multiple different lenses as you continue to grow — which I think is kind of the challenge ahead for all of us as we consider our careers as well.

Lenny Rachitsky

两本都能接到我们现在活着的 AI 时代。我今年早些时候也第一次读了《安娜·卡列尼娜》，很棒——是在「最聪明的人读过什么」书单上看到的。有人剧透了结局，少了点惊喜。我觉得很长；现在在读《权力掮客》，树立了新的「长」标准，读了半辈子了。

It’s interesting on both these — I could connect to AI in the time we’re living in now too. I also recently read Anna Karenina earlier this year. Amazing. I’ve never read it before. I saw it on a book list of here’s what the smartest people in the world have read. It was amazing. Someone gave away the ending which kind of made it less surprising. No spoilers. And I also felt like it was very long — but now I’m reading The Power Broker which set the new precedent for long. I’ve been reading it for half my life at this point.

Tara Seshan

我爱《权力掮客》。另一个强烈推荐：如果有人跟 Simon Hazel 的 Substack——他慢读重要书，做过《战争与和平》，也做过或在做 Hilary Mantel 的《狼厅》——一章一章来。那大概是读《权力掮客》或《战争与和平》甚至《安娜》的唯一办法。

I love The Power Broker. Another thing I highly recommend is if anyone follows Simon Hazel’s Substack where he does a slow read of important books. He did one of War and Peace and he’s doing one of Wolf Hall — or he did one of Wolf Hall, a Hilary Mantel book — take it chapter by chapter. That’s like the only way to read something like The Power Broker or War and Peace or even Anna — chapter by chapter.

Lenny Rachitsky

说到这个，有人跟我说 99% Invisible 有《权力掮客》读书会拆解，大概十三集，每集一两小时，一次几章，还有 Pete Buttigieg、AOC 和在那片区域生活过的人做客，Robert Caro 也上过几次。太棒了。好，继续闪电轮。最近最喜欢的电影或剧——如果有时间看的话？

Speaking of that, there’s a 99% Invisible book club breakdown of The Power Broker — 13 episodes, an hour or two each — they go through a couple chapters at a time and talk about it, special guests like Pete Buttigieg and AOC and folks that lived in that area, and they have Robert Caro come on a couple times. That’s amazing. Hot tip. Okay, we’ll keep going with our very lightning round. Favorite recent movie or TV show you’ve really enjoyed if you’ve had time to watch anything?

Tara Seshan

当然看了《奥德赛》。我觉得是难以置信的电影。我的热看法是它关于 AI——或 Christopher Nolan 对 AI 如何改造社会的看法。我爱它，强烈推荐。他是难以置信的导演，用别的当代导演没做到的方式把艺术性和商业成功桥接起来。我最近还看了黑泽明的《罗生门》——第一次用多人视角讲故事、你到最后也不知道什么是真的；那套电影手法由黑泽明开创。它提醒我在约束下能做到什么：五十年代、黑白、有人扛着摄影机，却如此完美、有品味、创新，是创造力的惊人例子。看那部电影时我想：我的 iPhone 上有他拍那部片时一百倍的力量和工具——我有什么借口不抬高野心、做出更好的东西？

Of course I watched The Odyssey. I found it to be an incredible film. My hot take is that it’s about AI — or Christopher Nolan’s view on how AI transforms society — which I loved, and I highly recommend watching The Odyssey. He is just an incredible director and has bridged artistry and commercial success in a way that I think no other modern director has done. I also recently watched Rashomon, the Akira Kurosawa film that for the first time did that technique of telling a story through multiple people’s perspectives where you never know what was true in the end — that technique in film was pioneered by Kurosawa. It reminds me what one can do under constraints. That film was made in the 50s, it was black and white, there’s a guy holding the camera, and yet it is so perfect and such a tasteful, innovative, amazing example of creativity. What I’m reminded of watching that film is: I have a hundred times the power and tools that he had making that film in my iPhone. And what’s my excuse for not elevating my ambitions and making better stuff?

Lenny Rachitsky

又回到野心。《奥德赛》我还在抢票，太难了——睡过头，现在一个月都没座。

All comes back to ambition. On The Odyssey, I’m still trying to get tickets. It’s so hard. I slept on it and now it’s like impossible for like a month. There’s no seats anywhere.

Tara Seshan

Kevin Clark 这周稍早给我们弄到晚上十点 Metreon 的票。太好看了。

Kevin Clark got us tickets at 10 PM at the Metreon earlier this week. It was so good.

Lenny Rachitsky

下次叫我，我入。我开着机器人、有人在帮、朋友也在找。希望这集出来时我已看过；没看的话谁有门路请告诉我。我想看 IMAX 全火力 Metreon 那种。好，下一题：最近发现的最爱或有意思的 AI 产品——理想不是 OpenAI 的，但你也可以说 OpenAI。

Next time, call me. I’m in. Oh man, I have bots running on it. I have a person working on it. I have a friend. We’re all trying to find a… Okay. Next question. Favorite or interesting AI product that you’ve recently discovered — ideally not OpenAI product, but you can also go there if you want.

Tara Seshan

当然我最爱的 AI 产品是 ChatGPT，以及在 Codex 里用酷站点和 visualize，太棒了。但 OpenAI 之外，我最爱的是朋友为我做的产品——因为现在人真的能做了。我超级粉 cozy software 运动：为五个朋友做软件工具，大家一起用。我有个朋友 Sebastian，做了个很酷的 AI 应用，把任何东西变成播客丢进一个小播客应用；他还做了个给我们朋友的私密社交网络，叫 GATS。那正是我认为未来该有的：人该做精确满足自己和朋友需求的软件。

Of course my favorite AI product is ChatGPT and using cool sites and visualize stuff in Codex, which is amazing. But outside of OpenAI products, my favorite AI products are products that my friends make for me, because now actually people can do that. I’m such a huge fan of the cozy software movement where you make software tools for like five of your friends and you guys use it together. I have a friend named Sebastian who made a really cool AI app that turns anything into a podcast and puts it in a little podcast app for you. And he also made a really great private social network for our friends called GATS. It is exactly what I think the future should be — people should make software that exactly meets their and their friends’ needs.

Lenny Rachitsky

GATS 是什么缩写？内部笑话吗？

What does GAT stand for? Is that some inside joke?

Tara Seshan

不是——至少如果是内部笑话我也不知道。它像给一小群朋友的私密 Twitter。我在那产品上能学到最有意思的东西。

It is not. Or at least if it is an inside joke, I don’t know it. It’s like private Twitter maybe for a small group of friends. And I learn the most interesting things on that product.

Lenny Rachitsky

像 WhatsApp，但又不是。播客那个有意思——我更想要的版本是直接进你的播客订阅流，想读的东西变成新剧集。

It’s like a WhatsApp, but not. The podcast app is interesting, but I feel like the version I would love is actual podcast in your feed of podcasts and then just new episodes get added of things you want to read or whatever.

Tara Seshan

对，它就是那样——丢进你的 Apple Podcasts feed。

Yeah, that’s what it does. It drops it in your Apple Podcasts feed.

Lenny Rachitsky

太棒了，我想要。帮我订阅。好，还有两题。有没有常在工作或生活里回到的人生座右铭？

Amazing. I want this. Help me subscribe to this. Okay. Amazing. Okay. Two more questions. Do you have a favorite life motto that you find yourself coming back to often in work or in life?

Tara Seshan

我工作中常回到的座右铭其实是托尼·莫里森关于工作的几条。我很快翻一下。是四条，出自她的文章《The Work You Do, the Person You Are》。第一：无论工作是什么，把它做好——不是为老板，是为自己。第二：是你造就工作，不是工作造就你。第三：你真正的生活在家人那里。第四：你不是你做的工作，你是你之为人。

My life motto that I come back to all the time in work is actually Toni Morrison’s takes on work. Let me pull it up really quickly. Okay. It’s four things. It’s from her essay, The Work You Do, the Person You Are. The first one is: whatever the work is, do it well — not for the boss but for yourself. The second is: you make the job; it doesn’t make you. The third is: your real life is with your family. And the fourth is: you are not the work you do; you are the person that you are.

Lenny Rachitsky

起鸡皮疙瘩。太好了。我想这是你 pinned 在 Twitter 资料上的——我记得见过。钉在 Twitter 顶上真好，每次上 Twitter 又看见。最后一题：你当年是 Thiel Fellow。校友群体太夸张了。有没有那段时间好玩的故事——或者你骄傲的其他 Fellow、面试怎样之类？

I got tingles. Wow. So good. And I think that’s what you have pinned to your Twitter profile — yes, because I remember seeing that. So cool. Okay. Final question. You were a Thiel Fellow back in the day. What an alumni group. Holy moly. Any story from that time that might be fun to share — something that’s like oh wow that was crazy? Or any other Thiel Fellow you’re proud of, what was the interview like — anything along those lines?

Tara Seshan

Thiel Fellowship 是我人生的拐点；没有它我不会在这里。对应前面说的：有关键时刻有人抬高你的野心，你照做了，人就变了。那是有人来找我、抬高我的野心、说你可以做这个、不必走你原来那条路的时刻。我永远感激他们能那么做。

The Thiel Fellowship was an inflection point in my life. I wouldn’t be where I am without it. Maybe to the point of there are key moments where you can tell people to elevate their ambitions and they do and that changes them — that was a moment where someone came to me and elevated my ambitions and said no, you can do this, you don’t have to take the path that you were on. Truly I’m eternally grateful for them being able to do that.

Tara Seshan

我现在常一起工作的一位 Thiel Fellow 是 Ari Weinstein——他创办的公司 Sky 被 OpenAI 收购；在那之前他创办过公司、因上一家公司被收购在 Apple 待过一阵。Ari 是我见过最有创造力的思考者之一，真正是 Mac 上能玩出所有酷事的专家。他在 OpenAI 带很多 computer use 的事，已经发出一堆很棒的 computer use 功能。他的创造力、做事的喜悦、对 craft 的爱很激励我。Ari 是酷人。

One of the Thiel Fellows that I get to work with all the time now is Ari Weinstein, who founded a company called Sky that was acquired by OpenAI. Prior to this, he founded and worked at Apple for a while because they acquired his previous company. Ari is just one of the most creative thinkers I’ve ever seen and is truly the expert on what are all the cool things you can do on a Mac. Ari leads a lot of our computer use stuff at OpenAI and he’s shipped a whole bunch of great things for computer use. His creativity and his joy in what he does and his love of his craft really inspires me. Ari’s a cool guy.

Lenny Rachitsky

你想的时候，我解释一下 Thiel Fellowship——错了请纠正。大概是 Peter Thiel 说：人不该去上大学，该去建自己想建的东西；给你十万美元不去上大学，去追随野心。大致对吗？

As you think about it, I’ll explain the Thiel Fellowship for people that don’t know — correct me if I’m wrong. Basically Peter Thiel is like: hey, people shouldn’t go to college. Instead they should just try building something that they want — and you get $100,000 to not do college and instead just go follow your ambition. Is that roughly correct?

Tara Seshan

对，就是那样。当时你和另外 19 个人在一起——每年大约 20 人，因为是 20 under 20。

Yeah, that is exactly right. And you’re with 19 other people at the time. It was like 20 people every year because it’s 20 under 20.

Lenny Rachitsky

办了多少年？还在办吗？

How many years did it go on for? Is it still going?

Tara Seshan

我想还在办，但前四五年大概一直卡在 20 人那个规模。我那一年很疯的一件事是：Fellowship 每年都会做成 CNBC 纪录片。所以我上台 pitch、呈现我要做的想法，全不幸活在 YouTube 上。如果你真想看十九岁的我做尴尬的事，就在那儿。那一批里最惊人、最成功的人之一是 Dylan Field——不仅才华横溢，人也非常善良。能和那些人共事我觉得很幸运。

I think it’s still going, but I think it was constrained at the 20 number for like the first four or five years or something like that. A really crazy thing that happened my year is that every year of the fellowship they decided to make it all a documentary on CNBC. And so my pitch for the fellowship — getting up on stage and presenting the idea I was going to do — all of that is unfortunately live on YouTube. So if you really want to see me as a 19-year-old doing something embarrassing, it’s there. Of course one of the most amazing and successful people who came out of that batch of the fellowship is Dylan Field, who is not only an incredible talent but also a very kind person. Feel very lucky to be able to work with those folks.

Lenny Rachitsky

有意思，想到 Thiel Fellows 时大家脑子里第一个往往是 Dylan。什么品牌啊。Tara，这集太棒了。有什么想推的、想让人去看的？听众怎样对你有用？

Amazing. Yeah, it’s interesting that Dylan’s like the guy I think everyone thinks of when they think of Thiel Fellows. What a brand. Okay, Tara, this was incredible. Is there anything you want to plug? Anything you want to point people to? And how can listeners be useful to you?

Tara Seshan

也许他们该用 ChatGPT 桌面应用，在网页上用 ChatGPT，试试 Work——可惜有个小切换，切过去试。让它做点酷的事，让它建一个关于你的站点开开胃，或做一块你的 ChatGPT 使用可视化。那是很亲密地体验这股力量的酷方式，能做的用例清单是无限的。

Anything I want to plug and point people to? Maybe they should use the ChatGPT desktop app. They should use ChatGPT in the web and try Work. It’s like unfortunately a little toggle. They can toggle over to it and try out Work. Ask it to do some cool thing. Ask it to build a site about you maybe to start, or ask it to make a little visualize block of your ChatGPT usage. It’s a really cool way to start experiencing the power of this stuff very intimately. And the list of use cases they can do from that are infinite.

Lenny Rachitsky

更好的主意：让它建一个站点，告诉你用 Work 能做什么。

Here’s a better idea. Ask it to build a site to tell you what you could do with Work.

Tara Seshan

太好了。那会管用。

Great. That will work.

Lenny Rachitsky

抱歉打断。你还要补什么？

Solve all the problems. Okay. I interrupted you. I apologize. What else were you going to add or say?

Tara Seshan

主要就是：去下载 ChatGPT 应用，在网页上用；更变革的是在移动端试。然后坐一趟长地铁或 Muni；没信号时钻出来，东西已经帮你做完了。那部分超级有魔法——你不用全程开着笔记本晃；你终于有这些东西在云端跑着做真正的工作。

Yeah, my main plug is: go download the ChatGPT app. Go use it on web. Even more transformatively, go try it on mobile. Then take a long subway ride or something like that or a Muni ride. And when you pop out after having no service, the thing is done for you. That’s the part that feels super duper magical. You’re not wandering around with your laptop open the entire time. You’ve finally got these things running in the cloud doing real work.

Lenny Rachitsky

对，最后那点我本来也想提——今天产品上很被低估的一块，我还以为是移动端独有的云能力。

Yeah, that last piece I was going to bring up, but that’s I think a really underappreciated element of the product today on mobile — and I thought that’s just a mobile-only feature, the cloud piece.

Tara Seshan

不，到处都有。

No, it’s everywhere.

Lenny Rachitsky

到处都有。所以在移动应用里可以打开 ChatGPT，切到 Work，让它做事——它不是在本地跑，是在云端，会一直干到完，你再回来聊。听起来简单，力量巨大。Tara，走之前还有吗？

It’s everywhere. Okay. So amazing. So on your mobile app, you can go to ChatGPT, toggle Work, ask it to do some work, and you don’t need to actually have it running locally — it’s running in the cloud. It’ll go keep doing work until it’s done, and then you could chat to it. That feels really simple, but that’s a massively powerful thing. Okay. Anything else, Tara, before we let you go?

Tara Seshan

没有了，就这些。

No, that’s it.

Lenny Rachitsky

好，谢谢。太棒了，谢谢你来做这集。从 Fellowship 那年一路走到现在，真是一段旅程。谢谢大家收听。

Okay. Thanks. This was awesome. Thank you so much for doing this. What a journey since the fellowship back in the day. All right. Well, thanks for being here. Bye, everyone. Thank you so much for listening.

Tara Seshan

荣幸。谢谢。

Such a pleasure. Thank you.