年经常性收入达 1 亿美元的 Gamma 创始人：打造智能体运行框架

简要描述Brief Description

Gamma 的联合创始人兼首席产品官探讨了智能体运行框架、创意工作流、模型选择、实验、营销自动化，以及垂直领域 AI 企业的发展机遇。对话还介绍了 Gamma 的演示文稿生成流程，并分享了家具购物和专业协作方面的实用案例。Gamma's co-founder and chief product officer discusses agent harnesses, creative workflows, model selection, experimentation, marketing automation and opportunities for vertical AI businesses. The conversation includes Gamma's presentation pipeline and practical examples from furniture shopping and professional collaboration.

目录Table of Contents

• 引言
  Introduction
• 了解测试框架
  Understanding the harness
• 生成并检查演示文稿
  Generating and inspecting a presentation
• 垂直领域机遇与创始人优势
  Vertical opportunities and founder advantages
• 创意品味与 AI 嗅觉
  Creative taste and AI smell
• 评估和用户信号
  Evaluation and user signals
• 用于专业工作的个人智能体
  Personal agents for professional work
• 有效使用模型和上下文
  Using models and context effectively
• 实验与有益的摩擦
  Experiments and useful friction
• 营销工程师与个性化定制
  Marketing engineers and personalization
• 图像模型与垂直领域产品
  Image models and vertical products
• 细分市场与创始人动机
  Niches and founder motivation

引言Introduction

今天我们邀请到了 Gamma 的联合创始人兼首席产品官 John Nurona，一起聊聊智能体框架及其带来的各种机遇，还有垂直领域的智能体框架。我已经迫不及待想深入探讨了。首先，我想特别感谢 Claude 背后的团队 Anthropic 赞助本期节目。John，非常感谢你来参加我们的播客。听完这期节目，大家会学到什么？Today we're talking with John Nurona, co-founder and chief product officer of Gamma, and we're going to talk about agent harnesses, all the opportunities around it, vertical agent harnesses. I can't wait to dive in, and I want to give a special shout out and thank you to Anthropic, the team behind Claude, for sponsoring today's episode. John, thank you so much for coming on the podcast. By the end of this episode, what are people going to learn?

谢谢你邀请我，Greg。今天能来到这里，我真的非常兴奋。我们将深入探讨如何在实践中真正构建一个智能体，包括揭开智能体运行框架的神秘面纱，以及它在 Gamma 是什么样子。Gamma 是一个创意型智能体运行框架，也是一个用于视觉传达的智能体。Thanks for having me, Greg. I'm really excited to be here today. We're going to dive deep on what it means to actually build an agent in practice, including demystifying what is an agent harness and what that looks like for us at Gamma, which is a creative agent harness, an agent for visual communication.

而且不只是你很谦虚。Gamma，我是说，它增长得非常快。我想它的 ARR 已经达到 1 亿美元了。你们刚刚重新推出了这个平台，发布了一个重磅版本。而且你们正在学到很多东西。我希望从这期播客中获得的，其实就是一些关于智能体工具框架机遇的小见解、你们实时学到的东西，以及其他存在的机会、创业点子之类的。那么，John，你能保证到这期节目结束时，我能学到其中一些东西吗？当然。来吧。让我们实现它。And not just you're you're being humble. Gamma is I mean it's it's grown extremely quick. I think it's hit 100 million ARR. You've just relaunched the platform. Huge release. And there's a lot you're learning. And what I hope to get out of this podcast is really just little insights around opportunities in agent harnesses, what you're learning in real time, and other opportunities that exist, startup ideas and stuff like that. So, John, can you commit that by the end of this episode I'm going to learn some of those things? I'm in. Let's do it. Let's make it happen.

那么，我想先从这里开始说明，你可以看到什么是智能体框架，因为我觉得很多人刷 X 的时候会看到“智能体框架”这个词，却不太明白它到底是什么。你能清楚地解释一下它是什么吗？So, I want to start off by like back here you can see what is an agent harness because I think a lot of people, you know, scroll X, they see agent harness, but they don't really understand what it is. So, can you clearly explain what it is?

了解测试框架Understanding the harness

当然。我们可以把它拆解成两个词：agent 和 harness。我也会用 Gamma 的发展历程来说明这两者。先给大家介绍一下背景：Gamma 主要是一款用于制作演示文稿的工具，也能制作文档、网站、社交媒体图片等类似内容。事实上，我们已经做了好几年，远在 agent 成为热门概念之前。我们于 2020 年创立公司，到了 2023 年才真正起飞。早期，我们只是单独调用 LLM 来完成一些任务，比如写一张幻灯片，或为一套演示文稿制作大纲；之后，我们再非常手动地把这些调用串联起来，做成预设工作流，用来完成诸如生成一份 10 页演示文稿之类的任务。过去一两年里发生的重大变革，就是 agent 的崛起：如今最强大的 AI 产品，不再是手动串联这些单独的 LLM 调用，而是运行一个更高层级的 LLM 循环，让它完成某项工作、检查结果、使用工具，并通过一系列调用最终产出内容。Absolutely. Let's break it down into the two words, agent and harness. And I'll use maybe Gamma's journey to illustrate both. So, for everyone's context, Gamma is a tool primarily for making presentations, also documents, websites, social graphics, that sort of thing. We've actually been around for several years, long before agents were ever a thing. We started the company in 2020, really took off in 2023, and early on we were just making these one-off LLM calls to do things like write a single slide or make the outline for a deck when we very manually strung them together into these preset workflows to do things like generate a presentation with 10 slides. The big revolution that's happened in the last year or two is this agent revolution where the most powerful AI products now are not stringing together these individual LLM calls by hand, but they're actually having one higher level loop of an LLM run that will do some kind of work, inspect it, use tools, and then turn that into something over a whole series of executed calls.

因此，我们实际上在这一年里经历了整个过程，重建我们的产品，让它成为一个智能体。而要把你的产品构建成智能体，就需要构建一个运行框架。这个运行框架包含智能体用来在其环境中执行任务、检查自己的工作，并向用户交付完成成果的所有工具，其中也包括询问用户的实际偏好。最著名的智能体运行框架可能就是 Claude Code 这样的产品。Claude Code 是一个用于编程的运行框架。它会教智能体如何遍历你系统中的所有文件，查看里面有什么；教它如何运行测试；甚至会通过特定规则来引导它，比如在某些情况下进入计划模式、执行完成后运行这些测试，或者在 GitHub 上完成这些操作。And so we actually went through this whole journey over the course of this year of rebuilding our product to be an agent. And building your product to be an agent involves building a harness. So the harness is all the tools that agent uses to execute things in its environment, inspect its own work, and deliver something finished to the user, including soliciting from the user what their actual preferences are. So probably the most famous agent harness is something like cloud code. Cloud code is a harness for programming. So cloud code teaches the agent how to go through all the files on your system and see what's there. teaches it how to like run tests and even prompts it with certain rules like go into plan mode in certain cases or run these tests after completing execution or do this stuff on GitHub.

在我们的案例中，Gamma 不是编程代理，而是创意代理。因此，它需要理解你的艺术风格和创作偏好，帮助你构思故事，而我们也非常重视视觉设计。我们会生成图像并进行排版，所以需要了解它们的效果，并确保呈现得恰到好处。我们的代理框架会把这些能力都转化为代理可用的工具。我可以具体讲讲这套框架是什么样的。In our case, Gamma is not a coding agent. It's a creative agent. So, it needs to do things like understand your art style and creative preferences, help you craft a story, and we care a lot about visual design. We're doing things like generating images and laying them out. So, we need to know how they look and make sure they look right. So, our harness takes all of those things and turns them into tools that an agent can use. And I can talk through what that look like.

嗯，我是说，麻烦你这么做。而且我也挺想知道你是同意还是不同意我的看法。我一直在说，智能体运行框架就是新一代 GPT 说唱歌手，或者说新一代说唱歌手，对吧？这说法只是个不错的推文，还是说确实有几分道理？Yeah, I mean, please do. And I think also I'm I'm curious if you agree or disagree with me. I've been saying agent harnesses are the new GPT rappers or new rappers, right? is that is that just a good tweet or is it is there some truth to it?

我觉得这话有道理。虽然我一开始总是反驳把 GPT 包装层说成这个概念的说法，因为它多少暗示了 90% 的工作都由 LLM 完成，外面只套了一层薄薄的封装，负责来回传递请求。我想，harness 或许说明了，实际上大部分工作都在这层封装上。你会发现，这些 LLM 的核心能力其实相当相似。不是说它们完全没有差异，但总的来说，把它们变成产品的所有奥妙，都在于围绕它们构建 harness。就连 OpenAI 和 Anthropic 这样的模型公司，也投入了大量工作来开发 harness。所以它们自己也在对这些模型进行封装，以实现不同的能力。I think there's truth to it. Although I always disputed this concept with the GPT wrapper originally because it sort of implies that 90% of the work is the LLM and then there's just this thin layer on top which is the wrapping back and forth. I think maybe what harnesses illustrate is that actually the wrapping is most of the work. that you actually have all these LLMs which are quite similar in their core capabilities. Not to say they don't have some variation but generally all the magic of turning them into a product is building the harness around them. And even the model companies like OpenAI and Anthropic are putting so much of their work into harness development as well. So they are themselves doing wrapping of these things to achieve different capabilities.

生成并检查演示文稿Generating and inspecting a presentation

你能不能用大概30秒带我过一遍：有人在 Gamma 上输入提示词后，会发生什么？实际来说，它会调用多少个模型？因为我觉得很多人都不太清楚这到底是怎么运作的。Could you walk me through what happens like in 30 seconds after someone types a prompt on gamma? Like how many models does it touch like realistically you know because I think a lot of people are not sure h how that how that actually works.

我来带你了解一下。我们来看一个示例提示词。Gamma 的一个非常典型的用例是：有人粘贴一堆杂乱的笔记、文档或 PDF，然后说，帮我把这些内容整理成一份 20 页的演示文稿。你输入这个提示词后，幕后其实会发生很多事情。首先，我们会在云端启动一整台计算机，这就是所谓的智能体沙箱。过去几个月里，OpenAI 的那些智能体著名地从这类沙箱中逃了出来。我们会启动整个沙箱，让智能体获取代表用户提示词的所有文件、用户上传的文件，以及关于用户的其他背景信息。然后，我们会提示智能体查看这些文件和用户的提示词，并执行特定任务。对我们来说，最先要做的事情通常之一，就是弄清楚还需要向用户询问哪些信息。Let me walk you through it. Let's go through a sample prop. So very typical use case for gamma is someone will paste in a bunch of messy notes or a document or PDF and say help me turn this into a 20 slide presentation. So the moment you put that prompt in a whole bunch of things actually happen under the hood. The first thing is we actually spin up an entire computer in the cloud. This is what they call an agent sandbox. This is what the opening eye ones broke out of famously in the last few months. So we actually spin up this entire sandbox where the agent gets all these files representing the user's prompt, the file they brought in, also other context about them. we then prompt the agent to go look through all that file and their prompt and run certain tasks. And for us, actually one of the first tasks is usually to figure out what other information we need from the user.

因此，我们通常会弹出一个表单，问几个问题，比如：这份演示文稿是给谁看的？是学校里的教学演示吗？是给成年人的？还是用于你的工作？我们还经常会调用各种不同的连接器。我们已经集成了 Notion、Slack 或 Linear 之类的工具，可以从中提取所有相关背景信息。就在最初的 10 秒左右，我们可能已经调用了三四种不同的模型，包括昂贵的和便宜的。稍后我们也会聊聊如何管理成本；目前，我们只是获取一些信息并提几个问题。我们还围绕这些内容构建了一个界面，并加入了一些生成式 UI，让你实际看到有哪些选项。接下来，我们会为你的演示文稿编写大纲。and so typically we will actually pop up a form and ask a few questions. questions like who's this actually for? You know, is this like an educational presentation for school? Is this for adults? Is this for your work thing? We will actually call out often to different connectors. So we've built integrations into tools like notion or slack or linear where we can actually pull in all this relevant context. Just already in that first maybe 10 seconds we've now made calls using probably three or four different models including expensive and cheap ones. We'll talk about actually how we manage costs as well just to sort of source information and ask a few questions. We've rendered a UI around it with some generative UI to actually show you what those choices are. Then we're going to actually write an outline for your deck.

这通常还需要多调用几次 LLM，我们会为你设计一个模板。所以我们有一个模板库，里面有各种不同的视觉风格。我们会用视觉模型，试着找出哪种风格最适合你。我们还会用一些很酷的搜索和嵌入技术，浏览成千上万个模板，找到合适的那个；如果你说“好是好，不过我想要红色、蓝色，或者其他什么颜色”，我们甚至还会为你定制。然后我们会再次把结果展示给你，让你编辑和检查。我们甚至还没开始制作演示文稿呢，就已经可能调用了大约 30 到 40 次 LLM，使用四五种不同模型来完成不同任务。现在，你就要点击那个生成按钮了，而我们会在你眼前为你构建一份演示文稿。That's usually several more LLM calls and we will style a template for you. So we have a library of templates with different visual styles. We're going to use vision models, try to figure out which one's a good fit for you. we're going to do some cool search embedding things to sort of look at the space of thousands of templates and find the right one and even customize it if you said, "Yeah, but I want it red or blue or whatever it is." Then we're going to once again show that to you and let you edit it and inspect it. We haven't even started making the deck yet. and already we've probably made something like 30 or 40 different LLM calls to maybe four or five different models doing different tasks. now you're going to hit that generate button and we are going to build up a presentation for you before your eyes.

而且，完成这件事实际上涉及很多环节。我们得把文字写到幻灯片上，还得安排好所有内容。就我们而言，我们还会生成大量图像，无论是装饰性的还是内容性的。我们经常会制作图表、信息图或数据图表。所以，实际上，这些工作现在都需要调用更多不同的模型，而且通常是并行调用。性能方面有很大的考量，因为如果我们逐项处理，速度会相当慢，你可能还没等我们完成就已经离开了。因此，我们做了很多工作来让整个流程尽可能快。大约一分钟后，你就会看到整套演示文稿在你面前逐步呈现并生成出来。到这时，我们大概已经进行了近 100 次不同的模型调用。And there's actually a lot of pieces involved in doing that. We have to write the words on the slide. We have to lay out all the things. in our case, we generate a lot of imagery as well, whether decorative or more substantive. We're often making diagrams or infographics or charts. So actually all of those are now more calls that are going out to different models often in parallel. There's a huge performance consideration because if we did these one by one, it's going to be quite slow and you're going to basically leave before we ever finish. So we do a lot of work to make it all pretty speedy. And then in the space of about a minute, you're going to actually see that whole presentation stream in front of you and generate. And now we've made probably close to 100 different model calls.

我想说，整个过程中可能用到了10到15种不同的 AI 模型。而且我们还没结束，因为你大概不会满足于第一稿。事实上，我们也不会满足于第一稿。生成第一版演示文稿后，我们会让 AI 视觉模型逐一检查每一张幻灯片，找出它能发现的问题：文字没对齐、图片里的人物头部被裁掉、对比度奇怪，或者内边距不合适。简单的演示文稿里居然会出这么多问题，真令人惊讶；人一眼就能看出来。有时人也未必看得出来，因为这需要设计师的眼光。不是每个人都能做到，但我们可以训练模型来发现这些问题。所以，我们会进行一整轮视觉质量检查。And I would say probably 10 to 15 different AI models have been used somewhere along the way. And we're still not done because you probably won't stick with the first draft. in fact, we don't stick with the first draft. After we generate that first presentation, we're now going to have a AI vision model look at every single slide and look for problems that it can find. Misaligned text. The person's head got cut off in the image. there's weird contrast. Something's not padded right. It's actually amazing how many things can go wrong in something like a simple presentation that a human eye can detect right away. or sometimes a human eye can't detect because it takes a designer's eye. So, not everyone can, but we can train models that can catch that. So, we go through this whole round of visual QA.

实际上，我们会根据 QA 中发现的问题进行多轮修复，然后重新处理。经过所有这些修复之后，我们才会最终把结果交付给你。到这一步，我们可能已经用过 20 种不同的模型，前前后后进行了大约 200 次调用，才为你交付看起来像是一次性完成的结果。We actually will do multiple rounds of fixing things based on what we find in that QA and then redo them. and then after all those rounds of fixing, we will finally serve it up to you. And now at this point, we've probably used 20 different models making like maybe 200 different calls along the way to deliver what looks to you like one finished output.

垂直领域机遇与创始人优势Vertical opportunities and founder advantages

所以，你刚才跟我描述的，其实是某个特定细分领域的一套工作流程。就这个案例来说，嗯，就是创意内容产出。没错。你们最初从演示文稿起步，现在已经转向了，嗯，本质上就是一套创意工具套件。So what you've just described to me is like a workflow in a particular niche. In this case, it's, you know, creative output. Totally. You started with presentations. You've now moved, you know, it's essentially a creative suite.

是啊。你觉得在哪些领域有机会打造智能体框架？你知道，如果你明天要创办一家公司，你觉得打造垂直领域智能体框架的机会在哪里？你觉得有哪些尚未充分探索的细分领域？Yep. you know, where do you see opportunities to create agent harnesses outside? You know, if you were starting a company tomorrow, where do you see an opportunity to create a vertical agent harness? What's what's an example of a niche that you know you see underexplored?

这是个挺有意思的问题。我的意思是，你知道的，单凭我的经验来看，我觉得当我们发现某个问题和普通编程截然不同时，自己打造一套独特的智能体框架就会很有价值。就我们来说，区别在于这件事非常主观。过程中有很多地方需要做判断，不同的人对什么才算好可能会有分歧，这意味着你希望有人类参与其中。另外，使用它的人也确实在意输出的具体细节。所以我觉得，可以把这两个特征视为打造自己的智能体框架很合适的场景。随便举几个例子吧，我知道很多人都在用 Claude Code 做游戏，而我非常看好这样一种想法：做一款属于自己的电子游戏，几乎就像是自己用氛围编程把它做出来一样。That's an interesting one. I mean, you know, just leaning on my experience, I think where we see a lot of value of making our own distinct agent harness is when that problem looks very different from just plain old coding. In our case, what makes it different is that it's very subjective. There's a lot of judgment calls to make along the way where two different people will disagree about what good looks like, which means you want to have the human in the loop. And also where the person using it actually cares about the specifics of the output. And so I would think about maybe those two traits as something that go into like, wow, these are a good case for making your own agent harness. So just thinking aloud, some examples are like I know a lot of people are making games with cloud code and I'm really bullish on this idea of like the personal video game where you almost like vibe code it yourself.

编程智能体已经很擅长制作电子游戏了，但我觉得它们往往不太擅长真正收集玩家的意见，了解游戏应该往哪个方向发展、如何演变。你想想，像 Roblox 或 Minecraft 这样的游戏里，有很多东西都是玩家自己塑造的。而且通常是一些非技术人员，他们自己并不会编程。我在想，能不能为此搭建一种运行框架，让整个框架实际上都在一个 3D 世界之类的环境里运行。coding agents are already really good at making video games, but I think they're often not good at actually sourcing the players opinions on where it should go and where it should evolve. And if you think about, you know, things like Roblox or Minecraft, there's all these things where people are actually shaping the world for themselves. And usually it's like nontechnical people who are not coders themselves. I wonder if there's a harness to be built there where in fact the whole harness runs in like a 3D world or something.

我觉得大家真的应该拓宽思路，尽量——你知道——总的来说，我的经营理念就是要与众不同。试着做出和别人做的东西完全不一样的产品。对我们来说，这一直是个挑战，因为我们努力做到与众不同，可后来别人就会——你知道——模仿我们，或者做出同样的东西。于是我们就得继续寻找越来越不同的方向。我觉得——你知道——这不只是针对你们。现在任何在软件领域做产品的人都是如此。没错。一旦你偶然找到了成功的秘诀。完全是这样。I think people should really think broadly and try to you know in general my philosophy of business is try to be different. Try to do something that looks nothing like what everyone else is making. And for us that's been a challenge because we try to be different but then others you know copy us or do the same thing. And we have to keep looking for more and more different. I think that's, you know, it's not even just for you guys. It's anyone who's building anything in software today. Yeah. As soon as you catch lightning in a bottle. Totally.

大家都在模仿你。但我觉得，现实情况是，他们模仿的是六个月前——你知道，三个月前、六个月前的你，对吧？所以，作为创始人，你得想清楚：好吧，我鲜明的观点是什么？我对世界将走向何方有什么判断？我该如何打造一个产品工作室，让自己每两周都能从零到一地开发一个新功能，探索这个拼图中的一块新内容？people are copying you. But I think the thing that people I think the reality is they're copying where you were six, you know, 3 months ago, 6 months ago, right? So you as the founder have to be like, okay, where do I what's my strong point of view? What's my thesis on where the world is going? And how do I essentially create like a product studio where you know every two weeks you know I'm building a feature 0ero to one exploring a ter you know a new piece of this puzzle.

我觉得这完全没错。我认为，对于任何正在考虑构建自己的智能体运行框架的人来说，最重要的问题是：我的优势会是什么？优势有几种不同的类型。其中一种就是你刚才说的：你对世界有独到的看法，并且正朝着这个方向更快地前进，而这并不是其他人关注的方向。我认为另一种优势是你拥有某种专有数据。你掌握着其他人没有的信息，也许是因为你已经有了一个成功的平台之类的。我觉得还有一种完全不同的优势：你做的不只是软件，因为软件领域正变得非常拥挤，进入这个市场也令人望而生畏。但你知道，还有硬件、实体服务，还有各种将 AI 以有趣方式融入其中的现实业务可以做。实体服务具体指什么？你能详细说说吗？I think that's totally right and I think for anyone thinking about building their own agent harness the really important question to ask is what's my edge going to be? And there's a couple different types of edge you can have. One type of edge is what you just said which is you have a distinct vision of the world and you are just moving faster towards it and it's not the one that others are focused on. I think another edge is you have some kind of proprietary data. You've got some information that other people don't maybe because you already have a successful platform or something. I think a totally different edge too is that it's not just software because software is getting very crowded and it's a scary market to be in but you know there's there's hardware there's like physical services there's all sorts of real life businesses to be made that incorporate AI in interesting ways how do you know physical services like can you elaborate on what you mean

是啊，我的意思是，先说清楚，我是搞软件的，所以对这些一窍不通。但如果我不是搞软件的，我会花很多心思琢磨，怎么把 AI 融入某种人们正在经历的现实生活体验中，让其中确实发生某种人与人之间的价值交换——比如有人给你按摩，或者一家餐厅给你上菜，诸如此类。我觉得有太多这样的实体业务还没有真正借助 AI 重塑。我不确定它们是不是都需要各自的 harness，不过也许有些确实需要。比如我看到另一个 harness 开发得很多的领域，就是法律科技。我猜医疗科技也是一个很大的领域，里面有很多有意思的数据封锁和监管问题。对，我觉得法律就是那种显而易见很适合 AI 的领域。yeah I mean to be clear I'm a software guy so I don't know anything about this but if I were not a software guy I'd be thinking a lot about how can I integrate AI into some kind of real life experience that people are having where there's actual some kind of human exchange of value whether it's like you know a person giving you a massage or a restaurant that's serving you a meal or whatever it is I think there are so many times of like real businesses that are not quite reinvented with AI yet I don't know that they all need their own harness although perhaps some of them do like another category where I'm seeing a lot of harness development going on is like legal tech for example I suspect medical tech is like a big one that has a lot of interesting like data blocking and regulatory stuff going on. Yeah, I think legal was one of those where it was like obviously this makes sense.

会计显然说得通。我想我会考虑的是：如果是今天，我会不会去做一个法律科技领域的 AI 智能体框架？大概不会。对。只是因为感觉这个领域太拥挤了。完全同意。所以，我喜欢你知道，当你谈到服务型企业，甚至一般来说那些无聊的企业时，比如给暖通空调公司用的智能体框架，或者……accounting obviously makes sense. I think the thing I would be thinking about is like would I be building a legal tech a AI agent harness today? Probably not. Yeah. just because it feels so crowded. Totally. So, I like when you know when you're talking about service businesses, even boring businesses in general, like agent harnesses for HVAC companies or,

你知道，如果有人想在某个细分领域创办一家智能体工具框架初创公司，该如何着手设计呢？也就是说，你需要找到关键洞见和细分领域。关于什么是智能体工具框架，有一个非常简单的答案：它就是一组智能体用来完成任务的工具。所以我认为，应该问的是：有哪些独特的工具是我已经拥有的，或者可以开发出来的，而且目前还没有出现在市面上？我觉得，作为创始人，理想情况下你应该总是问自己：基于我的人生经历、我所处的地方、我认识的人，等等，我最适合做什么？you know, if someone wants to build an agent harness startup, h in a niche, how do they approach designing it in the sense like you need the key insight, you need the niche? One really simple answer about what an agent harness is an agent harness is just a bunch of tools that an agent uses to complete a task. And so I think the question to ask is what are the unique tools that either I already have or I can build that are not already out there. And ideally I think as a founder you should always ask the question of what am I uniquely suited to do based on my life experience where I am in the world the people I know that sort of thing.

创意品味与 AI 气味Creative taste and AI smell

还有，嗯，你怎么看 AI 味儿？是啊，完全同意。你知道，我觉得一个好的智能体运行框架会尽可能消除这种 AI 味儿。而且我知道你对此有很强烈的看法。and h how do you think about AI smell? Yeah, totally. you know, because I think a good a good agent harness is going to kill that smell as much as possible. And I know you have a strong opinion on that.

我确实是。我今年一直在用“AI味”这个词。我一直在说给它除味、把它喷掉之类的。也许有些人只是还不知道这个说法。“AI味”是我们用来指这种现象的：人们用 AI 创作作品时，你往往能看出来他们用了 AI。我觉得这种感觉在文字里最明显，比如有人用 Claude 给你写邮件，你就会觉得：“你平时根本不用这个词，而且你说话也不是这个样子的。”现在我在 AI 图片里也经常能闻到这种味道。我在旧金山走来走去，每家咖啡馆、每家餐馆都有一张海报，我一眼就能看出来：“这就是 Nano Banana 做的，那是 GPT Image 做的。”我整天都在做这些，所以老远就能认出来。我觉得随着 AI 在整个经济领域和普通人的生活中越来越普及，越来越多的人也开始认出这种“AI味”了。I do. It's I have been using this term AI smell all year. I've been talking about deodorizing, spraying it off, all that sort of thing. Maybe for people don't just don't know the term. AI smell is what we use to refer to this fact that when people make work with AI, you can often tell they used AI. I think it shows up most in writing where when someone writes you an email maybe with claude you're just like you don't normally use that word and that's not what you sound like. and I notic it a lot with AI images now where I just like walk around in San Francisco and every cafe every restaurant has a poster where I can just smell like that's nano banana that's GPT image I do this stuff all day so I can like spot them a mile away and I think more and more people are starting to recognize that AI smell as it diffuses just through the economy and normal people.

所以对我们 Gamma 来说，这其实是这次发布的一个重大目标。作为早期的 AI 工具之一，我们某种程度上也算是成也成功、败也成功。那时候还没人用 Gamma，只要你用了，做出来的东西就会很酷、很吸睛。但后来大家都开始用它，一切看起来都像是 Gamma 做的，人们也不想被认出来是在用 AI。所以我们刚刚完成的这次发布，目标可以说就是迅速拓展我们的创意版图。说到底，AI 的痕迹还是会存在，但可以有一千种不同的 AI 味道，而不是千篇一律的单一风格。我们可以帮人们进入不同的小风格圈子。顺便说一句，这也是另一种形式的 harness。我们需要创建自己的自动化系统，不断切换不同的创意风格。So for us at Gamma, it was actually a huge goal of this launch. We had kind of been a victim of our own success as one of the early AI tools where when nobody used Gamma, if you used it, your stuff would look really cool and stand out. But once everybody started using it, everything started to look like Gamma and people didn't want to be recognized as using AI. And so kind of our whole goal of this launch we just pulled off was we need to rapidly expand our sort of creative landscape. the truth is there's still going to be AI smells, but there can be a thousand AI smells instead of one monolithic one. And we can help people jump to different little pockets. And we needed, this is another form of a harness, by the way. We needed to create our own automated system for constantly cycling through different creative looks.

我们其实是用时尚来打比方的。时尚也有一种“气味”。你会觉得：“我认得你穿的那件H&M衬衫，我也有一件一样的。”你知道，在时尚界，人们会不断变换不同的造型，这里面有一种循环变化的动态。现在我们喜欢穿特别宽大的牛仔裤，但过不了多久，紧身牛仔裤又会流行起来。所以我们意识到，必须顺应这个循环，为人们提供适应它的工具。We actually use the analogy of fashion. Fashion has a smell, too. You're like, I know that H&M shirt you're wearing. Like, I have the same one. and you know, with fashion, there's kind of this dynamic where people are cycling through different looks. You know, right now we're into like really big wide jeans, but soon it's going to be skinny jeans again. And so we realized we had to actually ride that cycle and give people the tools to adapt to it.

评估与用户信号Evaluation and user signals

所以，这个有评测吗？看起来挺通用的？你知道，评测是个很有意思的话题，因为我觉得评测特别适用于客观的事。其实，如果你不介意的话，能不能讲讲什么是评测？So is there an eval for this looks generic? You know, eval is such an interesting topic because I think eval work really well for things that are objective. Actually, break, if you don't mind, break down what an eval is.

对，完全没错。Eval（评测）可以理解成一种软件测试：你提出一个预先设定好的问题，然后用一种方式来评判结果。举个我们在 Gamma 领域里的简单例子：当用户的需求有歧义时，我们通常希望智能体能提出问题。比如用户说“做一份关于老虎的演示文稿”，智能体就应该问：“这是给孩子看的，还是给成年人看的？”因为针对这个主题，可以做出风格截然不同的老虎演示文稿。所以，我们通常会用 eval 来测试新的提示词或模型。一般还会给它们打分，也可以把一系列 eval 汇总成一个基准测试，然后说：“哦，结果发现，模型 B 的得分比模型 A 高了 3%。”每次实验室推出新模型，他们都会说：“哦，Opus 7 在这 15 项 eval 上表现都特别出色。”Yeah, totally. So an eval just means a you could think of it almost like a software test. It's a hard-coded question that you ask and a way of judging it. So a simple example in our world of gamma is we often want the agent to ask questions when you do something ambiguous. So we'll give it nel like when the user says make a deck about tigers the agent should ask something like is this for kids or for adults because there's very different kinds of tiger presentations you can make about that. and so we typically use evals to test out new prompts or new models. And usually you score them and you can put a bunch of evals together into a benchmark where you say, "Oh, it turns out model B scores 3% higher than model A." Every time the labs come out with a new model, they're like, "Oh, Opus 7 crushes like these 15 evals."

对我们来说，在一个极具创意的领域，难点在于评测那些主观、以品味为导向的事情，比如“这个看起来酷吗？”更不用说那些非常个人化的问题了，比如“这像我吗？”“这听起来像是我会说的话吗？”对此并不存在单一的客观评测标准。所以，我们尝试创建一些评测，作为这些问题的替代指标。比如，它是否遵循指令？举例来说，如果你提供了品牌指南，它是否会遵循？这是一个客观问题，我们确实会对自己进行测试。但我们从未把自己局限于评测。我认为，这其实也揭示了一个有趣的观点：你可以如何在这个领域竞争。归根结底，创始人能为自己的领域带来的是品味。他们有自己对美的判断。他们确实在与用户交流，而用户也有自己的品味。The problem for us in a very creative domain is that eval subjective tasteoriented things like, "Does this look cool?" and even more so very personal ones like is this me? Does this sound like something I would say? There's no one objective eval for this. And so, you know, we've tried to create some evals that are proxies for this. Like, does it follow instructions? Does it, for example, if you provide your brand guideline, does it stick to it? That's an objective question that we do test ourselves on. But we've never limited ourselves to just eval. And I think this is actually interesting insight too of how you can compete in this world. Ultimately, what founders can bring to their domain is taste. they have their own sense of what looks good. They're actually talking to their users and those users have one.

而且，我觉得你所处的领域越难验证，比如时尚，就越有空间容纳这些独特的视角。我觉得在服装设计方面，应该有一百种不同的束带设计。我想你也知道，Nike 的设计应该和 Adidas 的不一样，因为它们对品味有不同的理解。另外，对我们来说，由于我们的用户群非常多元，用户各不相同，所以我们实际上更依赖 A/B 测试。我们会以很大的规模实际运行两个模型或两组提示词，然后在广泛的用户群中衡量哪一个表现更好，因为这能反映主观差异，也能帮助我们弄清楚哪项指标最重要。And the less verifiable a domain you're in, like say fashion, the more I think there's room for these unique perspectives. I think when it comes to clothing design, there should be 100 different harnesses. I think you know Nike should have a different one than like Adidas because they have a different perspective on taste. Also, for us, because we have a super diverse user base who are all different, we actually rely a lot more on AB testing. We will actually run two models or two prompts at very big scale and measure across that wide user base what's actually performing better because that captures subjectivity and what's what's the metric that matters most there.

你知道，我们有很多不同的指标。举几个例子，我们会关注一些事情，比如生成内容之后，你花多少时间编辑它？你最后会分享或导出它吗？有没有什么信号表明你在投入时间？我们也有付费产品，所以用户会先免费使用我们的产品，如果喜欢，就会升级为付费用户。这算是我最喜欢的信号，因为很难造假。一般来说，只有真正从某个工具中看到了价值，人们才愿意付费。所以，我们会把它当作一个北极星指标。有时候我们甚至会看后续留存情况，比如一周后或一个月后，用户是否还会回来使用产品。这显然需要更长时间才能衡量，但它也是那种很难造假的信号。这周在网上疯传的那张图，我觉得是 A16C 的一张图，说我觉得只有 2% 的人在为 AI 工具付费。你对此怎么看？You know, we have a bunch of different ones. To give you a few examples, we look at things like after you generate, how much time do you spend editing what you made? Do you end up sharing it or exporting it? are there signals that show you're investing time? We also have a premium product, so people will use our product for free and then if they like it, they'll upgrade to paying. That's kind of my favorite signal because it's hard to fake. Generally, people will only pay if they're really seeing value out of a tool. So, we use that one as like a northstar. Sometimes we'll even look at like downstream retention, like are people still coming back to the product a week later or a month later. It's obviously slower to measure, but it's also one of those hard to fake signals. The chart that's going viral this week is I think an A16C chart that says I think only 2% of people are spending on AI tools. What's your opinion on that?

是啊，这不让我意外。我觉得总的来说，订阅服务对那些从事某种有价值工作的人来说，一直算是比较小众的东西。我觉得消费者不喜欢订阅。至少我作为消费者，就不喜欢订阅各种东西。你知道，我还是总想着蹭别人的 Netflix 账号之类的。不，我不会把密码分享给你。你确定吗？要是我跟你解释一下什么是智能体运行框架呢？Yeah, it doesn't surprise me. I think in general subscriptions are always a bit of a niche thing towards people that are doing some kind of valuable work. I don't think consumers like subscription. I, as consumers, don't like subscribing for things. You know, I'm still always looking to jump on somebody's Netflix account or whatever it is. no, I will not share my password with you. Are you sure? What if I explain what an agent harness is?

嗯，不过我觉得工作场景就不一样了。对于工作来说，情况完全不同。你知道，过去大约 10 到 15 年里，软件行业的 SaaS 热潮基本上都是由工作订阅推动的，而且这类订阅还会继续取得成功。我认为，同一张图表也显示，当你把视角聚焦到专业人士身上时，付费使用 AI 的人数正在疯狂增长。所以，这当然是我们最主要的关注点。我认为，大多数人关注 AI 时，主要也在想如何找到这些专业消费者，以及如何打入 B2B 市场——那里的支出模式完全不同。U, but I think it's different for work. I think for work, it's a totally different story where, you know, work subscriptions are what has driven essentially the entire SAS boom the last like 10 to 15 years in software industry and those things will keep being successful. And I think that same chart says that when you actually zoom in to professionals, the numbers are rising like crazy in terms of who's paying for AI. So that's certainly our main focus and I think most people's main focus in AI is how to find those proumers, the professional consumers and how to break into B2B where there's a totally different kind of spending going on.

用于专业工作的个人智能体Personal agents for professional work

如果有人想构建 X 的 gamma，If someone wanted to build the gamma for X,

你会、你会选什么？比如说，如果有人在听这个节目时想：“哇，我挺受 Gamma 正在打造的东西启发的。”对，但我想，就是你刚才描述的那种东西，只不过面向另一个细分领域——你能想到什么吗？what would what would you pick? Like if someone listening to this is like, wow, like you know, I'm inspired by what Gamma is building. Yeah. but I think that like that, you know, basically what you just described, but like for a different niche, anything coming to mind?

我——这真的是件很个人的事，因为你懂的，如果那是我的领域，我早就在做了。我觉得你应该去做那些我一窍不通的事情，不管是宝可梦卡牌、时尚、运动鞋，还是别的什么。找到你熟悉、知道门道的领域，然后把它做出来。那你自己呢？有没有什么你希望它存在、但现在还没有的 harness？如果明天它问世，你就会马上用？I It's such a personal thing because if you know, if it was my thing, I'd be building it already. I think you need to do the thing that I don't know anything about, whether it's like Pokemon cards or fashion or sneakers or whatever it is, like find your domain where you know what's going on and build it. and you personally, is there like a harness that you wish existed that doesn't exist that you'd use tomorrow if it if it came out?

你知道，我对整个个人智能体领域挺感兴趣的。我还没来得及用最新的那些产品，但我觉得，工作协作这块好像还没人真正搞定。我想，如果能为任何东西搭建一个运行框架，我会选 Slack。而且我有点失望，Slack 自己居然还没为 Slack 搭建这样的运行框架。You know, I'm really intrigued by this whole sort of like personal agent space. And I haven't gotten to use all the latest ones, but what I feel nobody's totally cracked is like work collaboration. I think if I could build a harness for anything, it would be Slack. and I'm a little disappointed that Slack themselves have not built the harness for Slack.

在你看来，那会是什么样子？所以，对于公司使用 Slack 或 Microsoft Teams 的人来说，沟通信息简直铺天盖地。当然，消息实在太多了。你总是在努力跟上进度，比如，哪些消息是我真的需要处理的？我见过他们在这方面做一些边边角角的尝试，比如把 30 条 Slack 消息总结成一条，但我想要的是一个真正能替我处理所有事情、并且能了解我的工具。它会学会我会怎么回复，甚至在我回复之前就知道。我希望能打开任何一个有人提到我的 Slack 讨论串，然后就看到一份替我起草好的初稿。它了解你的方式，不是让你回答一份问卷，也不是让你在设置时教它，而是通过观察你的实际行动来了解你。对，通过观察我——我大概已经发了十亿条 Slack 消息。这些就足以让它知道我会怎么回复了，What would that look like in your mind? So, there's just this absolute flood of communication for anybody whose company uses Slack or Microsoft Teams, of course. there's so many messages going through. you're constantly trying to stay on top of like, well, which ones do I actually have to do something about? I've seen them nibbling around the edges of this of like you can sum summarize 30 Slack messages into one, but I want something that just like runs it all truly for me and it learns me. It learns how I would reply before I reply. I want to be able to open any Slack thread someone's pinged me on and have the first draft drafted for me. And it learns you not by answering like a quiz or onboarding it by actually like your action by observing me. Yeah, by observing I have sent approximately a billion Slack messages. That should be all you need to know how I'm going to reply,

对吧？right?

即使是在邮件里，也感觉还没完全开窍。我看到 Gmail 有些小功能，比如说“哦，你知道，没错，我很乐意周四去看牙医”，但我这辈子都不会这么回复。不过我觉得，更重要的是在专业场合真正地展现出真实的我。也许这才是我真正渴望的，而 Gamma 在某种程度上做到了这一点：我想要一个能够学会我的表达风格的智能体。当然，这很难，因为这很私人化，这也是智能体一直没能做到的原因。要让智能体变得贴合个人，真的很难。Even in emails doesn't feel totally cracked. I've seen Gmail has the little things that are like, oh, you know, like, yes, I'd love to go to the dentist on Thursday, which is not how I'd ever reply in my life. But I think there is so much more to just like represent me in a professional setting authentically to me. Like maybe that's what I'm really craving and it's a little bit of what Gamma does is I want an agent that can learn my voice. And of course that's hard because it's personal and this is why the agents haven't done it. It's really hard to make agents personal.

有效地使用模型和上下文Using models and context effectively

所以你知道，在 Gamma，你们是在别人的模型之上构建产品。嗯。对于基于别人的模型创业的创始人，你有什么建议？So you know at Gamma you know you're you're building something on top of other people's models. Mhm. what's your advice for founders building on other people's models in general?

我觉得，基于别人的模型构建产品并不丢人。你知道，我们从一开始就这么做了，经常在 Anthropic、OpenAI 和 Google 的模型之间切换。很多讨论一直都在说：“不行，你得训练自己的模型，得微调模型，因为你需要建立护城河。”我一直觉得，这对早期创业公司来说尤其不现实。首先，训练自己的模型需要投入的资金实在太多；而且，市面上的模型还在不断变强。我很少愿意去和现有的那些模型竞争。不过，随着你一路走得更远，情况确实会有所变化，因为你开始积累真正的用户使用量。说实话，积累真实的用户使用量才是大家应该着重关注的主要战场。I think there's no shame in building other people's models. We you know we have done this from the very start often bouncing between anthropic open AI Google models and so much of the discourse has always been like no you got to train your own stuff you got to fine-tune it because you need the moat. I have always felt that's unrealistic for early stage startups in particular. I mean, first of all, the money involved in training your own models is really hard, but also the models out there keep on getting better. I rarely would want to be competing against the models that are out there. I think though it does change as you get further along in the journey because you start to amass real usage. And I mean, let's be honest, just amassing real usage is the main battle people should be focusing on.

当然可以先用现有模型来获得这类用量。不过，一旦做到这一点，就可以开始做一些非常有意思的事情。实际上，我们在微调模型方面取得了很好的成效，既降低了成本，有时也开始为用户提供更多个性化服务。我们也开始找出一些大模型不擅长的专门任务。就像我之前提到的，整个幻灯片质量检查并判断其是否合格的流程就是一例。实际上，我们用市面上的模型来处理这项任务，效果并不理想。因此，我们正在开展一些自己的研究，为这些专门任务构建模型。By all means, use the models that are out there to get that usage in the first place. Once you do, though, you can start to do really interesting stuff. We've actually had great success fine-tuning models to bring our cost down and in some cases even to start to do more personalization for people. We also have started to identify specialized tasks that you know the big models are bad at. Like I mentioned earlier, this whole process of QAing a slide and telling if it's good. We've actually not had great results using any of the models out there for it. So we are actually doing some of our own research to build models for those specialized tasks.

我认为人们绝对应该以此为目标，但当然也要使用现有的优秀模型。那你怎么看？你知道，你目前对前沿模型或云端模型与开源本地模型相比，有什么看法？I think people should absolutely aspire to that, but by all means use the great models that are out there. What about what about thought you know current what are your current thoughts on frontier models or cloud models versus open-source local?

是的。我觉得你知道，云端模型被低估了。再说一次，我觉得开源领域受到了很多追捧和关注，但看看资金流向就知道了：钱全都流向了闭源模型，因为它们确实非常出色。很多工作都已经替你做好了。对我们所有人来说，最棒的是，闭源模型领域竞争非常激烈。我觉得，如果 Anthropic 真就这样一骑绝尘，那情况就会完全不同，因为我们就不会开始在优秀的开源 AAI 模型和优秀的 Gemini 模型之间做选择了。但我感觉这些公司彼此只差几个月的进度，对我们这些应用层开发者来说，这再好不过了。这会让它们始终感受到压力，不断做得更好。Yeah. I think you know cloud models are underrated. Again I think open source gets a lot of hype and excitement but follow the money. It's all going to the closed source models because they're really darn good. So much work has been done for you. And the great thing for all of us is there is intense competition in the world of closed models. I think in a world where you know like Anthropic had just purely run away with it, this would be actually a different story because we wouldn't begin to choose between like great open AAI models, great Gemini models, but I feel like all these companies are just a few months behind each other and for us in the application layer that's the best thing ever. It keeps the heat on them to keep doing better and better.

你曾经说过，你惊讶于自己的 AI 技术栈竟然这么浅。现在还是这样吗？你知道，我会说我们的 AI 技术栈广而不深。我的意思是，我们实际上会用很多不同的模型，但通常用法都相当简单。比如说，通过提示词工程，我们能做到很多；还有，也许你的听众已经听说过“上下文工程”这个词——越来越多的人开始用它来指为模型提供恰当的上下文。我真的觉得，这差不多就是关键所在，而这又回到了构建 harness 这件事上。你能从用户那里获取合适的上下文吗？你能以合适的方式向用户呈现 AI 的输出吗？你能把所有这些提炼成让 LLM 的工作变得非常容易的东西吗？只要让语言模型的工作变得足够容易，你就可以使用任何现成模型，并取得不错的结果。提示词和微调之间有一场大讨论。对。You've once said that you were surprised with how shallow your AI stack was. Is that still true today? You know, I would say that our AI stack is broad but shallow. What I mean by that is we actually use a ton of different models, but we often use them in fairly simple ways. Like we get so far with prompt engineering, with providing and maybe your listeners have heard the term context engineering increasingly of just like providing the right context to the models. I really view that as kind of the whole game here of just and it comes back to harness building. Can you get the right context from the user? Can you express the AI's output to the user in the right way? And can you distill all that into things that make the LLM's job really easy? When you make the LM's job really easy, you can use any offtheshelf model and get good results. There's a big debate prompts or fine-tuning. Yeah.

你在这方面的观点很鲜明。是啊。我觉得对我们来说，98%的价值都来自提示词，因为编写提示词很简单，而且可以移植。You have a strong opinion there. Yeah. I think for us it's 98% prompts is where we get the value out of cuz prompting is easy and it's portable.

所以，顺便说一句，现在已经没人手写提示词了。比如，Claude 会帮你写提示词，所以这事儿特别简单。因此我觉得，大多数问题都能通过提示词解决，包括把一个提示词拆成不同的部分。对我来说，微调更像是在此基础上做优化。比如你想控制成本，或者新模型发布时再挤出额外的 2% 或 3% 质量提升，你们会多快开始用它？还是会先观望？你是怎么考虑这件事的？哦。So like and by the way, nobody's writing prompts by hand anymore. Like Cloud's writing your prompt for you and so it's really easy. and so I think most problems can be solved with a prompt, including breaking up a prompt into different pieces. And fine-tuning to me is more of an optimization on top. When you want to do things like control costs or ek out that extra 2 or 3% of quality when a new model drops, how quickly are you implementing it or are you kind of like waiting and seeing or h how do you think about it? Oh

不，通常一小时之内。如果这是一个我有理由相信会不错的模型。你知道，不是每个模型都值得关注，但我想说，对我们来说，Anthropic 或 OpenAI 推出的任何模型通常都值得关注。我想我提过，我们的测试框架里有各种不同的任务。有些任务我们会用 Fable 或 Astro 级别的模型，有些会用小型的 Haiku、Gemini Flash 级别模型，还有些会用小型的 GPT Luno 模型。所以，每当有属于这一类、能帮助我们补齐帕累托前沿的模型出现，我都想试试看效果如何。通常，由于我们的使用场景非常主观，我们确实会跑一些评测，但我只想先让一部分用户试用。我们很幸运，用户群体很庞大，所以我可以在一小时内开始测试，然后两小时后就拿到效果数据。no, usually within the hour. If it's if it's a model that I have reason to believe is going to be good. So, you know, not every model is interesting, but I would say for us, like any model from Anthropic or OpenAI is interesting, generally if they're dropping it. And I think I mentioned we have all these different jobs in our harness. Some of those jobs we use a like, you know, fable or astro level model for. Some of those jobs we use a little haiku gemini flash level model for. Some we use a little baby GPT luno model for. And so, anytime any model comes in this class that can help us fill out that sort of paro frontier, I want to try it and see how it does. And generally because our use case is so subjective, yes, we'll run some evals, but I just want to run it on a sample of users. And we're very lucky to have a big user base where I can start a test in an hour and get data 2 hours later on how it's doing.

实验与有益的摩擦Experiments and useful friction

所以我学到的一件事是，我是一个名叫 LCA 的设计公司的联合创始人，我们之前也和你们在 Gamma 合作过一个项目。我和你们合作时学到的一件事是，你们会做很多实验。是啊。我很好奇，有没有哪一个实验的结果让你们感到意外？因为感觉你们好像一次会同时做很多很多实验。我们在任何时候都会进行大量实验，可能有 10 到 20 个正在进行，而且我们不断轮换；很多实验不到一天就结束了。so one of the things I learned cuz I'm a co-founder of a design agency called LCA and we actually worked on a project together at Gamma and one of the things I learned working with you guys was you guys run a lot of experiments. Yeah. I'm curious about experiment you maybe one experiment that the results surprised you because you guys are just it feels like you guys are running a lot a lot at a time. We run a huge number of experiments at any given time we probably have 10 or 20 running and we're cycling through many of the experiments last less than a day.

那你给我讲讲实验的流程吧，我觉得这真的很有意思。好，我就讲讲我刚刚开始做的一个实验吧，就在我坐 Uber 过来和你聊天的路上。我刚刚一口气启动了八个不同的实验，因为 Claude Haiku 5.5 昨天发布了。我们特别期待能用上它，因为你知道，我提过，我们有很多任务并不需要那种超强大的模型。它们既简单又便宜，比如从库里挑选一个模板，或者有时候根据某些内容写个提纲之类的。你知道，我们一直期待有新的小模型可供测试。所以今天早上我为它启动了八项不同的测试，等咱们录完这期播客，我就会去查看结果，可能会取消其中四项，看看它们表现如何。我们测试的任务包括在幻灯片上做一些细微修改，或者修改提纲之类的。So walk me through like the process of an experiment because I think it's really interesting. Well, I'll walk you through one that I just started in the Uber ride over to talk to you. I just fired off eight different experiments because Claude Haiku 5.5 came out yesterday. We were really excited to get to use it because, you know, I mentioned that we have a lot of these tasks that don't need that like super fable level model. They're quick and cheap like picking a template from a library or sometimes like writing an outline based on something. And you know, we have been looking forward to new small models we could use to test this. So, this morning I fired off eight different tests for it and right after we get off this podcast, I'm going to check the results and probably cancel four of them and see how they did. and we're testing it for things like making micro edits on a slide or, you know, like revising an outline, whatever it is.

说到我们从这类测试中学到的东西，通常最重要的发现是，我们可以使用成本只有另一个模型一半的模型，却得到完全相同的结果。但有时我们会得出相反的结论：我们会发现，哇，实际上，要完成某项任务，就必须使用那个庞大、强劲而昂贵的模型。有时我们甚至会发现，适度的阻力是有益的。比如，几天前我们刚对产品的一种快速模式做了实验。快速模式的意思就是跳过我之前提到的一些步骤，比如向你展示演示文稿大纲，或者事先向你提问。我们的假设当然是，如果减少流程中的阻力，人们就能更快完成整个流程，也能更快抵达那个神奇的价值时刻。Terms of the kinds of things we learn from these tests, often the great thing that we learn is that we can use a model that is half the cost of another model and get the exact same results. But sometimes we actually get the opposite finding. We find like wow actually the big beefy expensive model is necessary to achieve a certain task. sometimes we even discover things like good friction like we actually just ran an experiment a few days ago of a fast mode of our product. and fast mode just means skip through some of those steps that I told you about earlier like showing you an outline of your presentation or asking you questions up front. And the hypothesis was of course if you remove friction on the process then people will get through it faster and get to that sort of magical value moment more quickly.

我当时真的特别希望这项测试能成功。我想看到它胜出，但实际上它并没有为我们带来胜利。结果发现，我们一路上设置的那些小障碍，比如查看并审阅你的大纲、选择故事的大致走向、回答那些问题，都带来了高得多的满意度评分。虽然速度慢了一些，但那些坚持完成的人对自己的成果投入感更强，完成的内容也更多，而且他们转化为付费用户的比例确实更高。不过我觉得另一个很棒的地方是，测试的答案从来不只是非此即彼。你可以再深入挖掘一点。所以，我们实际上为移动端的新用户推出了一个快速模式，因为你知道，在手机上，你肯定不想为了回答问题而输入一大堆内容。I was I was really rooting for this test. I wanted to see it win but it did not win for us actually. it turns out those speed bumps we had along the way like seeing your outline and reviewing it, you know, choosing the kind of highle arc of your story, answering those questions led to much higher satisfaction ratings. even though it was slower, the people that got through felt more invested in their output and did more and they actually converted to paying at a higher rate. but I think what's also cool is a test is never just one answer or another. You can like dig a little bit deeper. So, we actually shipped a version of fast mode for brand new users on mobile because, you know, on mobile you don't want to be typing a bunch of answers to a question.

营销工程师与个性化定制Marketing engineers and personalization

就在这之前，我刚发布了一篇关于营销工程师的 X 文章。MI just right before this I posted an ex article about marketing engineers. M

我的论点是，营销已经变了。想想《广告狂人》时代，那个时候，你知道，在 50 年代、60 年代和 70 年代，你一年做一个大型广告就行了。后来到了 90 年代和 2000 年代，数字营销人员开始做更多一些实验。接着我们进入了增长黑客时代，每周多做一些实验。我想，到了 2022 年，我会说那是 AI 辅助营销人员的时代，实验做得更多了。你知道，我觉得我们现在正在进入营销工程师的时代：作为营销人员，你每天要运行几十个实验。你怎么做得到？你要用 AI 智能体替你去做。我很想听听你对此的看法。你觉得营销变了吗？你觉得它会朝什么方向发展？挺有意思的，因为我在通用汽车工作之前，曾在一家叫 Optimizely 的公司工作，我就是在那里爱上 A/B 测试的。my thesis is that marketing's changed. If you think about the Mad Men, Madmen era, it was like, you know, you did one big commercial a year and in the ' 50s and ' 60s and ' 70s. Then it became in the '9s and 2000s, the digital marketer, maybe a few more experiments. Then we had the growth hacker era, a few more experiments per week. I think in 2022 I would say AI assisted marketer more experiments and you know I think we're now entering the age of the marketing engineer where you're you're you know as a marketer you're running dozens of experiments per day. How are you able to do that? You're using AI agents to go and do that for you. I'm curious your P POV on that. Do you think that marketing has changed and like where do you think it where do you see it going? It's funny because before I worked at GM, I worked at a company called Optimizely and this is where my love of AB testing came from.

我们当时提供 A/B 测试即服务，而我们的很多愿景都围绕个性化展开。这个想法是：你知道，随着修改网站变得越来越容易，我们就会创建一百种细微的变体，让每个人看到的版本都略有不同，并以各种有趣的方式优化转化率。令我意外的是，我想说，这种愿景在互联网上基本没有实现。我觉得网站大多仍然是千篇一律的。大多数移动应用也差不多。除了那些像是高度个性化且大获成功的特定领域。我说的特定领域，是指 TikTok 信息流、Instagram Reels 之类的东西。我们知道，各家公司正投入数十亿美元，想方设法从这些界面上榨取你的每一分钱。我听说麦当劳正在测试个性化定价。你听说过这事吗？We were AB testing as a service and a lot of our vision was about personalization. This idea that you know as it gets easier to change your website, we are going to create a hundred little micro variations of it and everyone will get served a slightly different version and that will optimize conversions in all these interesting ways. Surprisingly to me that vision I would say largely has not come true out on the internet. I would say websites are still mostly one-sizefits-all. so are mobile apps by and large. except for these like specialized areas which are like hyperpersonalized and succeeding. When I say specialized areas, I mean things like your Tik Tok feed, your Instagram reels. Like we know that companies are spending billions of dollars to ek out every last, you know, dollar from you in those surfaces. I heard that McDonald's is testing personalized pricing. You hear about this?

哇，听起来这迟早会引发反弹。这就是我看出这一点的原因。Wow, that sounds like a backlash waiting to happen. That's why I saw it.

是啊，是啊。不过我觉得，麦当劳那个例子或许说明了为什么这种事没有发生。首先，我觉得社会对这类事情确实有种抵触情绪，就像你发现我们俩买了同一张机票，但因为我人在旧金山，他们多收了我五美元。嗯，我觉得一般来说，就连动态加价也是这样。比如昨晚我吃完晚饭回家时遇上了动态加价，我当时就，呃，你知道吧，感觉……就是不太舒服。Yeah. Yeah. But I think maybe that McDonald's example illustrates why this isn't happening. First of all, I think there's a real societal allergy to this kind of thing in the same way that if you found out that we both bought the same airline ticket, but they charged me five bucks more cuz I was in San Francisco. Well, I think in general, like even surge pricing, like I feel like when I had surge pricing last night coming home from dinner and I was like I was like h you know, like it just it's not a good feeling.

是啊，是啊。你知道，我其实觉得，在这个极度追求效率、讲究效果营销的时代，真正奏效的反而是那些坚定不移地投入品牌建设的公司。这种苹果式的理念，就是不惜一切代价保护品牌。所以我觉得，这种趋势会在一定程度上抵消你刚才说的那种大家都放手让营销智能体大展拳脚的情况。话虽如此，我还是认同“营销工程师正在崛起”这个观点。而且我特别喜欢“营销工程师”这个说法，因为听起来就像自相矛盾一样，正如“GTM 工程师”听起来也像自相矛盾。这些本来应该是不同的部门，但我很喜欢看到大家的技能不断融合。其实我们在 Gamma 上就看到了这种情况：越来越多的人用 Gamma，不只是偶尔做一份演示文稿，而是创建一些很酷的自动化流程，比如每当 Salesforce 中有新的交易机会创建时，就根据通话记录自动生成一份推介演示文稿。Yeah. Yeah. And you know, I actually think in this age of hyper efficient, you know, like performance marketing, actually what's working is companies that really doggedly invest in their brand. This like Apple mindset of like just protect the brand at all costs. And so I think that trend is going to counter some of what you're describing of everybody just going wild with their marketing agents. That said, I still buy the thesis that like the marketing engineer is rising. And I just love that term marketing engineer because it sounds like a contradiction, you know, much like GTM engineer sounded like a contradiction. These are supposed to be different departments, but I have loved seeing how everybody's skill sets are just blending. we're seeing this with Gamma actually where people are more and more using Gamma not to just make one presentation here or there, but creating these really cool automations where they're doing things like every time a new deal opportunity is created in Salesforce, let's automatically create a pitch deck based on the call notes that have happened.

你是怎么设置的？How do you set that up?

实际上，我们一直在构建各种连接器，让你可以把更多流程自动化。这样你就能直接接入 Salesforce。不过我想说，目前仍有不少底层配置工作，人们可能会用 NAN、Zappia 之类的工具来触发整个流程并把各部分连接起来。这就是一个很好的例子，说明这仍然是专家的领域。我想，有一些非常聪明的人或机构知道如何帮助企业把这些东西连接起来。我认为，把那些最懂工程技术的前 1% 或 2% 公司所采用的做法，转化成人人都能轻松使用的东西，仍然是一个商业机会。再回到个人智能体这个话题，另一个例子就是目前 Muse 和 Instinct 这类工具的发展情况。We've actually been building kind of connectors where you can automate more of it. so you can just like plug into Salesforce, but I would say there's still a fair amount of plumbing going on where people are using maybe a tool like NAN or Zappia or something to trigger it all together and wire it. that is a good example of a thing where it's still the domain of specialists. I would say there's like these really smart people or agencies that know how to help a business like wire things up. I think there is still a business opportunity in taking these practices that are done by the top 1 or 2% of companies who are really savvy about engineering it and just turning it into a super easy thing that everyone does. Another example of this to go back to the personal agents is like what's happening now with tools like Muse and Instinct.

如果你能把 Linux 电脑配置到合适的位置，还能避免被黑客入侵，那用 OpenClaw 或 Hermes 就能做出一些很棒的东西。说实话，我没跟这个风潮。我当时想，我可是个大忙人，没时间折腾我的 OpenClaw。我打算等几个月，等他们把它弄得简单好用些。结果现在，Muse Instinct 一直给我打电话，因为他们把事情弄简单了。对吧？所以你没有一台开着的 Mac Mini？我还真没买 Mac Mini。作为一名 AI 创业者，我得惭愧地承认这一点。回到你脑海中的营销工程话题，你觉得顶尖 1% 的营销工程师是什么样的？You could do some amazing stuff with OpenClaw or Hermes if you could get your Linux box set up in the right place and avoid getting hacked. Personally, I skipped out on that whole trend. I was like, I'm a busy dude. I don't have time to set up my open claw. I'm going to wait a few months for them to just make it easy for me. And now I got Muse Instinct blowing up my phone because they made it easy for me. Right. So you don't have a Mac Mini running. I never got my Mac Mini. I'm ashamed to say as an AI founder. You know, going back to marketing engineering in your mind, what is a top 1% marketing engineer look like?

嗯，有一点我想到的是，他们确实是营销工程师兼数据科学家。所以他们极度依赖数据，以实际表现为依据。而且现在要做到这一点比以往任何时候都容易。比如，Claude 绝对就是我的数据科学家。所以我想说，现在我们团队里的每个人都随时掌握着非常细致的分析，了解各种失败模式和机会。他们会查看我们网站的流量，找出其中的缺口。你知道，我会看看我们团队里最接近营销工程师的那个人，他总是在思考如何以一个人也难以置信的规模，找到新的方式来做 SEO 和广告创意。你知道，他们已经从独立创作者，变成了管理一支智能体大军的人，而且他们如今承担的角色高度偏向监督。well, one thing that comes to mind is they're really a marketing engineer data scientist. So they are like ultra data driven and grounded in what's performing. And this is something that's gotten easier than ever. Like absolutely Claude is my data scientist. And so like I would say now everybody on our team is constantly armed with like very fine grain analysis of like failure modes and opportunities. They're looking at traffic on our site. They're finding gaps. there, you know, like I look at our, you know, the person we have on our team who's closest to a marketing engineer and he is just constantly thinking of new ways to do SEO ad creative at just an unbelievable scale for one person. You know, they're like they've gone from being like individual creatives to people who are running this army of agents. and they are just in this highle supervisory role.

是啊。我是说，它基本上就是一个部门。它就是个部门。对，对。感觉我们正在走到这样一个阶段：如果你是顶尖的 1% 营销工程师，你的工作范围就不只是……Yeah. I mean, it's basically a department. It's department. Yeah. Yeah. It's like we're getting to the place that if you're a top 1% marketing engineer, your surface area of what you're working on isn't just

嗯，涉及的范围很广。你要关注 SEO、GEO、社交媒体内容、付费广告，基本上还有所有品牌创意相关的东西，你懂的。well, the surface area is huge. You're looking at SEO, GEO, you're looking at social content, you're looking at paid ads, you're looking at pretty much like all creative brand, you know.

当然。我觉得这正是让我感兴趣的地方，因为我认为，那些摸索出如何成为顶尖 1% 营销工程师的人，在让自己的产品获得曝光方面拥有不公平的优势。而其他 98% 的人，如果跟不上市场上正在这么做的人，很快就会变成恐龙。Absolutely. It's I think that's what's interesting to me because I think that the people that figure out how to become a top 1% marketing engineer have an unfair advantage around getting their product seen. Well, and the other 98% are quickly becoming dinosaurs if you can't keep up with the people in the market that are doing this.

我一直有个想法：过去大概一两年里，我们所有人——至少是所有在用这些工具的人——好像都经历了一连串快速晋升。我们从个人贡献者升到经理，再到总监，然后好像又升到了副总裁。现在我们手下有一层又一层的智能体，像金字塔一样。结果发现，这份工作其实挺难的。少数人才会做这份工作是有原因的，因为你会不断被信息轰炸。你得做那么多决定，整个人很容易被这一切压得喘不过气，也摸不着头脑。你需要工具才能理清这一切，也需要有人帮你理清这一切。你需要助理，需要幕僚长，还需要这一切其他的支持。所以我们所有人都得迅速提升技能，学会如何管理上千人；培养这项技能，是我们所有人能做的最重要的事情之一。I've kind of had this idea that all of us over the last maybe like year or two, at least all of us who are using these tools kind of got this rapid series of promotions. We went from individual contributor to manager to director to like VP. We now have these like layers and layers of agents under us in a pyramid. And it turns out that job's actually kind of hard. Like there's a reason that like few people do it because you are sort of constantly barged by information. You're having to make so many decisions. you kind of are like overwhelmed and confused by all of it. You need tools to even make sense of it all. You also need people to help you make sense of it all. you kind of need your assistant and you need your chief of staff and you need all of these things and so all of us are having to rapidly upskill to essentially like how to manage a thousand people and building that skill is one of the most important things we can all do.

此外，我觉得，开发工具帮助普通人成为自己人生的“副总裁”，仍是一个尚未充分探索的领域。回到我关于 Slack 的想法，Also I think building tools to help the average person become a like VP of their own life is still underexplored. Going back to my Slack idea,

我也觉得你对实验总体上很熟悉。感觉如今创办一家初创公司，或者说总体上做生意，似乎更像科学，而不是艺术。是的。I also think you know experiments in general, it feels like building a startup today is or building a business in general is it's like you have is more it feels like almost a little more science than art. Yeah.

我的意思是，最终能胜出的，是那些具备实验能力、愿意尝试各种不同做法的人。所以我很喜欢和你聊天，因为我觉得实验能力可以说是你的强项。在 Optimizely，你学到的哪件事是每位 AI 创业者都应该了解的？Now in the sense like you're the ones who are going to win are the ones that have the experimentation muscle and just trying a bunch of different things. That's why like I love talking to you because it feels like that experimentation muscle is I mean that's your like your thing. what's one thing at Optimizely that you learned that every AI founder should you know would be good for them to know?

说得真好。你知道，我想到的一件事是，多年来我帮助其他公司做实验，我觉得这件事很大程度上归结为你有多少次尝试机会。这完全就是个数字游戏，就像销售、约会，或者生活中的其他事情一样。你得真正尝试很多不同的做法，而不是指望只做一次大规模测试，就能神奇地解锁你的业务。我很喜欢你把它称作一种肌肉，因为我确实觉得是这样。首先，因为它是可以训练出来的。你知道，也许有些人与生俱来的这块肌肉比别人更强，但只要你付出努力，它就会增长。我也觉得，你的第一个实验不会成功。完全没错。就像你想变得更健康、锻炼肌肉一样，你必须每天去健身房，付出努力。绝对是。然后……That's a great one. You know, one thing that comes to mind for me, I think from years of like helping other companies do experimentation was that so much of it just comes to like how many atbats you have. it's all just sort of a numbers game, just like sales, just like dating, just like kind of anything else in life. You got to really try a lot of different stuff versus expecting that you'll run one big, you know, test that magically unlocks your business. I like how you called it a muscle because I do think it is. First of all, because it's trainable, you know, yeah, maybe some people are born with more muscle than others, but if you put in the work, that muscle is going to grow. I also think like your first experiment is not going to work. Totally. Just like if you're trying to get fit and you're trying to build muscle, you have to just show up to the gym every day and you have to put in the work. Absolutely. And then

两个月后，你可能就能看到六块腹肌了，对吧？但你不能指望第一天就看到效果。我觉得，做实验最难的地方就在于，你得经历一段悲伤低谷。我觉得很多人最后会想：“算了，我还是凭直觉来吧。”没错。顺便说一句，发社交媒体、写博客之类的事情也是如此，对吧？two months later, then you might see a six-pack, right? But you can't expect it on day one. I think that's the hard part about being an experimentter is like there's like a trough of sorrow that you have to go through. and I think what a lot of people end up doing is they're like, you know what, I'm just going to go with my gut. Absolutely. and by the way, the same is true of things like posting on social media, writing a blog, right?

发送冷邮件进行外联，这整个过程中总会有一段悲伤低谷。我认为，人生中的所有成功都源于熬过某种悲伤低谷。Sending cold email outbound. There's a trough of sorrow in all of it. And I think all success in life comes from swimming through some kind of trough of sorrow.

图像模型与垂直行业产品Image models and vertical products

现在有什么东西是被低估了、你觉得“你知道吗，大家真的聊得不够多”的？我之前和一个朋友聊过语音 AI，感觉这方面的关注度还不够。现在大家都在热议个人智能体。你能想到什么吗？What's what's something that's underhyped right now that you're like, you know what, people are just not talking enough about this? I was talking with a friend of mine about voice AI. that felt like underhyped. right now people are talking about personal agents a lot. Anything that comes to mind?

也许这个回答有点蹩脚，但我其实觉得 AI 图像模型受到的关注还不够。说它蹩脚，是因为我们聊图像模型已经聊了好几年了。从 2022 年 Stable Diffusion 问世以来，人们就一直在用它们做各种很酷的东西。但我觉得有意思的是，随着它们越来越好，人们反而不怎么谈论它们了。随着它们不再频频出错，逐渐成了我们生活中的一部分，我们似乎忘了自己拥有这项不可思议的技术，它几乎能创造任何东西。我觉得，看看人们怎么用 ChatGPT 的图像生成功能和其他一些工具就知道了，来来回回好像就那三四种用法。你知道，他们可能会给自己的小生意做一份菜单，或者给爸爸的业余项目设计一个徽标。Maybe it's a lame answer, but I actually think AI image models are underhyped. It's a lame answer because we've been talking about image models for years. Ever since like stable diffusion came out in 2022, people have been making like cool stuff. But I think ironically as they got good, people stopped talking about them. As they stopped messing up in so many ways and just became sort of like part of the fabric, we sort of like forgot that we have this unbelievable technology that can make anything. And I think if you look at the way people are using like chatgbt images and a few other things, it's kind of just the same three or four use cases over and over. You know, they're they're making like maybe the menu for their small business. They're making like a logo for their dad's side project.

他们现在做的事情可能就那么几样，但这些模型的能力其实非常强大。我们在 Gamma 对此有着非常独特的视角。我们没怎么谈到图像，但我们也会针对不同任务使用大概 10 种不同的图像模型。那里的创新步伐非常快，还有各种尚未探索的能力。对我们来说，可以用它们制作精美的图表或信息图表，放进演示文稿里；制作很酷的小纹理，作为演示文稿的背景；定制图标——这些现在都能实现，只是大多数人还没意识到这一点。我觉得你几乎可以关注任何创意领域，找到运用这些工具的方式。尤其是普通人，根本不知道这些工具能做什么。There's like a handful of things they're doing, but these models are capable of so much. And we have a really cool vantage point to this at Gamma. We didn't talk as much about images, but we use probably like 10 different image models as well for different tasks. And the pace of innovation there has been so much and just this sort of like unexplored capabilities. For us, it's things like making beautiful diagrams or infographics to go in a presentation. Making cool little textures that can be the background of a presentation, making custom icons, like all these things are possible now and most people just don't realize it's there. And I think you can pay attention to almost like any domain that's creative and find ways that you could be using these things. And especially the average person just has no idea.

你知道，我觉得有一件事可能很多 AI 原生用户都在做，但大多数人还没意识到你可以这么做：用 AI 图像模型来搭配衣服、布置家居。比如说，你想在亚马逊买一张新沙发。你完全可以问 ChatGPT：“这张沙发放在我家会是什么样子？”然后再让它帮你找几款别的沙发，也把它们的效果图做出来。很多这类应用都还等着被更深入地产品化。You know, even one that I think is like probably many AI native people are doing but most people don't realize you can do is using AI image models for like clothing and like home decor. You know, like you've had a new couch you want to buy on Amazon. You can actually chat GPT to say what would that look like in my house? and then say like go find me a few others and visualize those two. Like so many of these things are just waiting to be more deeply productized.

你觉得有没有机会在这个领域打造垂直产品，比如专注于室内设计之类的？还是说大模型公司会把这个应用场景全都拿下，所以根本不值得尝试？我觉得垂直应用场景的空间非常大，关键在于了解使用这些产品的人具体有什么需求。就拿家具这个例子来说吧。我最近刚买了张沙发。所以我……恭喜你。Do you think there's an opportunity to build vertical products there like just focused on you know interior design for example or do you think the big model companies are just going to own that use case so it's not even worth trying? I think there is so much room for these vertical use cases and it's all about understanding the specific needs of the people that do them. Like let's just take this furniture one for example. I actually just bought a couch recently. So I Congratulations.

谢谢。真的很舒服。我试了一下这个，觉得特别有意思。比如说，尝试做这件事时我发现，AI 图像模型很擅长向我展示沙发看起来会是什么样子，但它们并没有准确展示沙发的尺寸，对吧？对吧？如果你的墙面尺寸固定，那么一张 68 英寸的沙发和一张 58 英寸的沙发差别可就很大了。所以，你能不能直接在显示画面上标出尺寸单位？这是一个很好的测试框架示例，对吧？模型确实有能力做到很多这类事情。其中一种做法是，你可以想象这样一个流程：把沙发渲染到我的房间里，然后用一些传统编程手段，在图像上画一条横跨沙发的线，用像素来测量它的长度，看看最终尺寸是否正确，再提示模型进行调整，让结果完全准确。Thank you. It's really comfy. And I went through this and it was so interesting like trying to do this and it turns out the AI image models are amazing at showing me what the couch will look like but they're not actually showing me exactly how it's going to measure, right? Right? And there's a huge difference between, let's say, a 68 couch and a 58 in couch if your walls a certain size. So like, could you actually show me the units in the display? That's a great example of a harness, right? The models are actually capable a lot of these things. One way they're capable is you could imagine doing a thing where you render the couch in my room, but then you use some good oldfashioned programming to draw a line across in pixels and measure how long it is and see if it actually came out at the right size and prompt the model to tweak it to get exactly right.

我不觉得那些实验室近期会这么做。或者提示词。嘿，比如说，这张沙发不适合你。是啊，完全同意。不过，Restoration Hardware 还有另一张沙发，我觉得基于这些理由会非常合适。I don't see the labs doing that anytime soon. or prompt. Hey, like this couch isn't the right couch for you. Yeah, totally. But here's another couch from Restoration Hardware that I think would be perfect for these reasons.

完全同意。哥们，我买沙发时发现，有几百个网站在卖沙发，而且你得在每个网站上点一大堆小按钮。最后我逛到了 Google Shopping。体验一点都不愉快。也许已经有人为这事做出了很棒的产品，只是我没发现，但我目前还没看到。我觉得以前有过一些例子。我觉得 Levels IO 好像有个……对，他好像有个室内设计的东西，我觉得是这么叫的。对，对。Totally. And dude, what I learned about couch shopping is there's hundreds of websites that sell couches and you'll have to click through a lot of little buttons on all of them. I ended up on Google Shopping. It was not a pleasant experience. Maybe somebody's already built a great product for this and I missed it, but I haven't seen it yet. I think there's been examples. I think Levels IO has like a Yeah, he has like an interior I think it's called. Yeah. Yeah.

但我觉得还有很大的空间。你懂我的意思吧？比如说，尤其是模型越来越好，图像模型也越来越好。harness 这个概念也越来越成熟。这里还有很多可以做的。我觉得可以肯定地说，如今消费者购买沙发或家具的体验也还没有优化好。完全同意。而这个特定用例很棒的一点是，它属于高客单价消费。完全同意。But I think there's still room. You know what I mean? Like I think especially as the models are getting better, the image models are getting better. the concept of a harness is getting more and more mature. There's just there's more here. I think I think it's safe to say that the also the consumer experience today around couch shopping or furniture shopping isn't optimized. Totally. and what's cool about that specific use case is it's high ticket. Totally.

所以这里面是有利润空间的，对吧？比如说，我觉得跟 Restoration Hardware 谈个合作也不算离谱，他们可以说：“如果你推荐……我们就给你 5% 的佣金。”so and there's margin, right? Like if for example like I don't think it's crazy to do a deal with Restoration Hardware where they're like we'll give you 5% commission if you recommend

多卖些沙发。是的，绝对是。我觉得，考虑到哪些东西能在与模型公司的竞争中生存下来，还有另一个版本的问题。我们在 Gamma 一直都在面对这个问题。有面向纯消费者的版本，用沙发来打比方，就是“帮我买张沙发”。但还有面向专业消费者的版本，比如：“嘿，我是室内设计师。也许我只是兼职做室内设计，但家里还有两个孩子，这就是我的工作。不过你知道，我是按小时收费的。”这些人需要一款工具，也会主动寻找最好的工具；如果它能帮他们把工作做得更好，他们很乐意为它支付订阅费。此外还有真正的专业人士。比如，Restoration Hardware 本身就有员工在那里工作。他们用的是什么工具？sell more couches. Yeah, absolutely. I think the other version of this when you think about like what will survive competition with model companies and we're certainly facing this question constantly at gamma. There is the pure consumer version which to use the couch analogy is like help me buy a couch. but then there's the like proumer version which is hey I am an interior designer. Maybe I'm a part-time interior designer but I also have like two kids at home and that's what I do but I you know I do it by the hour. Those people need a tool and they will seek out the best and if it helps them do their job better, they will happily pay a subscription for it. Then there's also the true pros. There's restoration hardware itself has people who work there. What tool are they using?

我敢打赌，能做这些事的肯定不只是 ChatGPT。我觉得确实有机会围绕这些领域创办公司。而且我确实认为，在大公司里服务最高端的专业人士，才是真正能打造出经久不衰的产品的地方。对。我觉得这个方向还没有得到充分探索。对，你知道，因为我觉得现在的主流说法是，打造垂直 SaaS 或垂直 AI 产品并没有太多机会。是啊，这挺可惜的，因为我认为机会确实存在。I bet it's not just chat GPT to do all of this. I think there is actually an opportunity to build companies for those. And I do suspect that highest tier of professional the bigger companies is where you can really build something durable. Yeah. And I feel like it's underexplored. Yeah. you know, cuz the narrative I think is there's there's not that much opportunity creating vertical SAS or vertical AI products. Yeah. which is unfortunate because I think that there is

利基市场与创始人的创业动机Niches and founder motivation

而且外面的世界很大。你知道，这是我们从 Gamma 的发展历程中学到的另一件事：我们的用户群非常广泛。想想地球上有多少人要做演示文稿，就知道用户类型有多么多样。我们的收入中，来自美国的不到四分之一。用户遍布各种职业，比如顾问、教师、营销人员、销售人员，还有律师。我以前不知道律师也会做演示文稿。结果他们真的会。你知道，你越能触达一些小众领域就越好。很多时候，大公司其实就是上千个小众领域叠加在一起。所以，如果你能找到一个能跨越多个领域的切入点，就不必一开始就面面俱到；可以先从一个领域入手，再去寻找那些你完全有能力拓展的相邻领域。而且你其实也不知道这会把你带向哪里，对吧？完全没错。就像你当时也没想到，嘿……and it's a big world out there. You know, that's maybe one other thing we've learned from the Gamma journey. our user base is very broads. If you just think about how many people on earth make presentations, it's a lot of different kinds of people. for us, less than a quarter of our revenue is in the United States. it's spread across many different job roles from like consultants to teachers to marketers to sales people to lawyers. I didn't know lawyers make presentations. Turns out they do. and you know, the more you can tap into little niches, you know, like in many ways a big company is just a thousand little niches piled on top of each other. So if you can find something that cuts across and you don't have to start broad, you can start with just one but then look for the adjacencies you can tackle 100%. And you actually don't know where you're where it's going to lead, right? Totally. Like you didn't realize like hey

律师们正在使用的工具是用来制作演示文稿的，然后你最终会，谁知道呢，最终做出一些专门面向律师的东西。没错。接着突然之间，这就成了你业务中非常重要的一部分。是啊，完全如此。所以我总是说，先开始吧，把东西发布出来。lawyers are using are creating presentations and then you end up who knows you end up creating something specific for lawyers. Absolutely. And then all of a sudden that becomes a huge part of your business. Yeah, totally. And that's why I'm always like get started like put something out there.

先把东西推出去。你得试了才知道。这又回到了实验。实验不只是指找一万名用户分成两组来做 A/B 测试。通常只要做个落地页，弄出一个别人能试用的产品就行了，而现在做这些从来没有这么容易过。我知道你不会离开 Gamma，不过假设你明天就离开了 Gamma，你会想创办一家什么样的公司？Put something out there. You got to just find out. That goes back to experimentation. Experimentation doesn't just mean run an AB test with like 10,000 users in each arm. It usually just means put up a landing page, get a product someone can try, which has never been easier. If you I know you're not leaving Gamma, but like let's say you were to leave Gamma tomorrow. What's something what's a company you would start to build?

哦，这问题问得好，老兄。你知道，我就是特别喜欢探索如何激发人们创造力这个领域。我觉得还有其他类型的创造力也挺吸引我的。不过，尽管这个领域已经很拥挤了，我还是特别着迷于氛围编程，也一直在自己实践。还有，今年早些时候我有过这样的经历：我认识不少科技圈的人，也住在旧金山，每次和一个三四个月没见的朋友叙旧，我都会问：“最近在忙什么？”他们就会说：“看看我做的这个应用。”这事儿一下子就风靡开来。我想知道，氛围编程的下一个方向会是什么，或者说这个领域的下一个前沿在哪里？当人们什么都能做出来时，他们会走向何方？所以说，我确实很期待视频模型终于变得好用，也很期待语音模型越来越强。会不会出现某种东西，你打个电话过去，它就能替你做点什么？游戏对我来说也特别有意思。我总觉得电子游戏行业正适合被某种新东西颠覆，但我还没想明白会是什么。Oh, that's a good question, man. You know, it's I just love circling this space of like enabling people's creativity. I think there's other kinds of like creativity that like intrigue me though like as crowded as the space is I'm so intrigued by just like vibe coding as I practice and just the way I had this experience maybe as somebody who knows a lot of other tech people and lives in San Francisco earlier this year where every time I would catch up with a friend I hadn't seen in 3 or 4 months I'd be like what are you up to and they'd be like check out my app I made it sort of took the world by storm I want to know maybe what's the next vibe coding or what's the next frontier here of vibe coding as people can make anything where will that take them and so like within that I'm I'm certainly intrigued by video models finally getting good I'm intrigued by voice models getting good like what's the version of some kind of thing you could call and it would do something for you games are so interesting to me I just think like the video game industry is right for some kind of disruption but I haven't cracked it yet.

我不知道。我觉得在视频方面，我们确实感觉像是迎来了视频领域的氛围编程时刻。是啊。Opus 5.5 显然……他们做出了一些很酷的东西。做出一些很酷的东西。是啊。所以我想问题是……I don't know. I think on the video side it does feel like we're hitting like the vibe coding moment for video. Yeah. Opus 5.5 has is obviously they make some cool stuff. Make some cool stuff. The Yeah. So I think the question is

在一个任何人都能创作视频的世界里，他们能创作哪些类型的视频？我对此也很感兴趣。作为一名创始人，我会想：好吧，在一个任何人都能创作视频的世界里，我能做出哪些创业项目？我完全认同这个方向。这又回到了智能体运行框架。是啊，当然。一切都绕不开智能体。一切都绕不开运行框架，老兄。哥们，非常感谢你来参加播客。有什么话想留给大家吗？说一件事。in a world where anyone can create video, what are the types of videos they can create? And then I'm also interested. It's like as a founder I'm like okay in a world where anyone can create video, what are some startup ideas that I can create? Totally on top of that. And that goes back to agent harnesses. Yeah. Absolutely. Everything goes back to agent. Everything goes back to the harness, baby. dude, thank you so much for coming on the podcast. Is there anything you want to leave people with? One thing.

我总是会鼓励那些考虑创业的人，要知道这是一段漫长、有时也很孤独的旅程。正因如此，你真的必须热爱自己正在做的事情。在我们创办 Gamma 之前，我给自己定下的一条原则是：这个东西必须是我自己也能使用、能拿来“吃自己的狗粮”，并且真正相信的；如果我以后再创办另一家初创公司，这条原则也会继续适用。说实话，这确实排除了很多无聊的生意，不过……I would just always encourage people who are thinking of starting a business to know that it is a long and sometimes lonely journey. And because of that, you really have to love the thing you're working on. one of my kind of rules for myself before we started Gamma and it would stick around if I ever did another startup is it has to be something that I could dog food that I could be a user of and that I could believe in. And to be honest, that actually rules out a whole lot of boring businesses, but

这对我的心灵很有益。我只是需要找到一件我每天都愿意花好几个小时去做的事。所以我总是鼓励大家深入挖掘，想想自己的热情所在——那件只有你自己知道、你愿意每天花 8 个小时去做，而你环顾朋友们却发现他们都不愿意每天花 8 个小时去做的事。把它锁定为你的北极星。我觉得这也是重点：你愿意每天花 8 个小时去做、而别人却不愿意做的那件事。完全没错。因为……it was good for my soul. I just needed to have something that I wanted to spend hours every day in. And so I always encourage people to dig deep and think about like what is your passion and the thing that only you know the thing that you want to spend 8 hours a day on that when you look around your friends don't want to spend 8 hours a day on and zero in on that as your north star. I think that's the important point too around like the thing that you want to spend 8 hours a day on that other people don't. Totally. Cuz like

你可能特别喜欢体育之类的，但很多其他人也喜欢体育，所以这个领域可能竞争非常激烈。我觉得，如果你在挑选创业点子或选择做什么生意，我总是会说：“和细分市场结婚，和产品约会。”我很喜欢这句话。要爱上你想要服务的那群人。you might be really into sports or something but a lot of other people are into sports and that might be very competitive. I think like if you're if you're choosing a startup idea or choosing a business, I always say like marry the niche, date the product. I love that. Like fall in love with who you're trying to serve.

没错，没错。而且我们生活在这样一个时代，你知道，你会不断迭代产品很多次，产品也会变得很不一样。但如果你热爱这个细分领域，这就是为什么我总是鼓励大家在这个领域做媒体、积累粉丝，因为这样之后你就能向他们销售各种不同的产品。Yeah. Yeah. And because we live in an age where you, you know, you're going to be iterating on the product so many times and it's going to look so different, but if you love the niche, and that's why I always encourage people to, build media, build a following in that niche, cuz then you can sell them a bunch of different products.

你已经建立起了自己的受众群。我觉得这完全没错。你知道，杰夫·贝索斯说过一句很相似的话：“对愿景要坚持，对细节要灵活。”嗯，能和杰夫·贝索斯并列，我很高兴。确实如此。谢谢。我很感激你，也谢谢你邀请我来。太棒了。真是美好的时光。You have built an audience. I think that's totally right. You know, Jeff Bezos has a very similar quote where he says, "Be stubborn on the vision, flexible in the details." Well, I'm glad I'm in good company with Jeff Bezos. You sure are. Thank you. I appreciate you and thanks for having me. This is great. Great time.