Why experts writing AI evals is creating the fastest-growing companies in history | Brendan Foody

Brendan Foody: The wealthiest companies in the world are willing to spend whatever it takes to improve model capabilities.

Brendan Foody：世界上最富有的公司愿意不惜一切代价来提高模型能力。

Lenny Rachitsky: We are entering the era of evals.

Lenny Rachitsky：我们正在进入评估时代。

Brendan Foody: We started working with all of the top AI labs.

Brendan Foody：我们开始与所有顶级人工智能实验室合作。

What the labs need is labor marketplace.

实验室需要的是劳动力市场。

They actually need extraordinary professionals that can measure model capabilities.

他们实际上需要能够衡量模型能力的非凡专业人士。

Lenny Rachitsky: They found this pocket, maybe the biggest business opportunity in history.

Lenny Rachitsky：他们发现了这个口袋，也许是历史上最大的商机。

Brendan Foody: We grew from 1 to 400 million in revenue run rate in 16 months, fastest ascent in history.

Brendan Foody：我们的收入运行率在 16 个月内从 1 亿增长到 4 亿，这是历史上最快的增长速度。

Brendan Foody: The market is bound by the amount of things where humans can do something that models can't.

Brendan Foody：市场受到人类可以做而模型无法做的事情数量的限制。

The lab's primary bottleneck to improve models is how they can effectively have some way of measuring what success looks like for the model.

实验室改进模型的主要瓶颈是如何有效地通过某种方法来衡量模型的成功程度。

Lenny Rachitsky: There's this tweet that you retweeted.

Lenny Rachitsky：您转发了这条推文。

"If you really think about it, we were put on Earth to create reinforcement learning training data for labs."

“如果你认真思考一下，我们被派到地球上是为了为实验室创建强化学习训练数据。”

Brendan Foody: It's highly likely that the entire economy will become an aural environment machine, building out all of these worlds and contexts.

Brendan Foody：整个经济很可能会成为一台听觉环境机器，构建出所有这些世界和环境。

And I think the narrative in AI over the last three years has almost entirely been one of job displacement, but very few companies and people have talked about this new category of jobs that's being created.

我认为过去三年人工智能领域的叙述几乎完全是关于工作岗位的取代，但很少有公司和人们谈论正在创造的这一新类别的工作岗位。

Lenny Rachitsky: I talked to a lot of people about what should I be studying?

Lenny Rachitsky：我和很多人讨论过我应该学习什么？

Where should I be getting better?

我应该在哪些方面变得更好？

Brendan Foody: How can they leverage this technology to do so much more?

Brendan Foody：他们如何利用这项技术做更多事情？

We'll give people interviews where we say, "Use whatever tools are available to build a website and let's see what product you're able to build in an hour."

我们会给人们采访时说：“使用任何可用的工具来构建网站，让我们看看你在一小时内能够构建什么产品。”

Lenny Rachitsky: Today, my guest is Brendan Foody, CEO and co-founder of Mercor.

Lenny Rachitsky：今天，我的嘉宾是 Mercor 首席执行官兼联合创始人 Brendan Foody。

Mercor is the fastest-growing company in history to go from 1 to $500 million in revenue.

Mercor 是历史上增长最快的公司，收入从 1 美元增至 5 亿美元。

They did this in 17 months, less than a year and a half.

他们用了 17 个月（不到一年半）的时间完成了这件事。

Brendan is also the youngest unicorn founder ever.

布伦丹也是有史以来最年轻的独角兽创始人。

They just raised $100 million at $2 billion valuation.

他们刚刚以 20 亿美元的估值筹集了 1 亿美元。

Mercor, if you haven't heard of them, helps AI labs and AI companies hire experts to help them train their models using AI.

Mercor（如果您还没有听说过）可以帮助人工智能实验室和人工智能公司聘请专家来帮助他们使用人工智能训练模型。

They've never had a customer churn, their net retention is over 1,600%, and they're on a nine-figure revenue run rate.

他们从未出现过客户流失情况，净保留率超过 1,600%，并且收入运行率高达九位数。

Lenny Rachitsky: In our conversation, we talk about the increasing value and importance of evals, the landscape of AI training companies like Mercor, and why they've become so important and valuable, how Brendan discovered this opportunity, his insights on what product market fit looks like, the core tenets he's instilled within his organization that have allowed him to build the fastest growing company in history, what people writing evals for labs are actually doing day to day, which skills and jobs are going to last the longest with the rise of AI, why he doesn't think we'll see AGI or superintelligence anytime soon, and so much more.

Lenny Rachitsky：在我们的谈话中，我们讨论了评估日益增长的价值和重要性、像 Mercor 这样的人工智能培训公司的前景，以及为什么它们变得如此重要和有价值，Brendan 如何发现这个机会，他对产品市场契合度的见解，他在组织中灌输的核心原则，使他能够建立历史上发展最快的公司，为实验室编写评估的人们每天实际上在做什么，哪些技能和工作将在人工智能的兴起，为什么他认为我们不会很快看到通用人工智能或超级智能，等等。

This episode is incredible.

这一集令人难以置信。

You need to hear this.

你需要听听这个。

Lenny Rachitsky: If you enjoy this podcast, don't forget to subscribe and follow it in your favorite podcasting app or YouTube.

Lenny Rachitsky：如果您喜欢这个播客，请不要忘记订阅并在您最喜欢的播客应用程序或 YouTube 中关注它。

It helps tremendously.

它有很大帮助。

Also, if you become an annual subscriber of my newsletter, you get 15 incredible products for free for one year, including Lovable, Replit, Bolt, N8N, Linear, Superhuman, Descript, Whisperflow, Gamma, Perplexity, Warp, Granola, Magic Patterns, Raycast, ChatPRD, and Mobbin.

此外，如果您成为我的时事通讯的年度订阅者，您可以在一年内免费获得 15 种令人难以置信的产品，包括 Lovable、Replit、Bolt、N8N、Linear、Super human、Descript、Whisperflow、Gamma、Perplexity、Warp、Granola、Magic Patterns、Raycast、ChatPRD 和 Mobbin。

Check it out at lennysnewsletter.com and click Product Pass.

请访问 lennysnewsletter.com 查看并单击 Product Pass。

With that, I bring you Brendan Foody.

接下来，我为您带来布伦丹·美食。

Lenny Rachitsky: This episode is brought to you by WorkOS.

Lenny Rachitsky：本集由 WorkOS 为您带来。

If you're building a SaaS app, at some point your customers will start asking for enterprise features like SAML authentication and SCIM provisioning.

如果您正在构建 SaaS 应用程序，那么您的客户有时会开始要求提供 SAML 身份验证和 SCIM 配置等企业功能。

That's where WorkOS comes in, making it fast and painless to add enterprise features to your app.

这就是 WorkOS 的用武之地，它可以快速、轻松地将企业功能添加到您的应用程序中。

Their APIs are easy to understand so that you can ship quickly and get back to building other features.

他们的 API 易于理解，因此您可以快速交付并继续构建其他功能。

Today, hundreds of companies are already powered by WorkOS, including ones you probably know, like Vercel, Webflow, and Loom.

如今，数百家公司已经由 WorkOS 提供支持，其中包括您可能认识的 Vercel、Webflow 和 Loom 等公司。

Lenny Rachitsky: WorkOS also recently acquired Warrant, the fine fine-grained authorization service.

Lenny Rachitsky：WorkOS 最近还收购了 Warrant，这是一项细粒度的授权服务。

Warrant's product is based on a groundbreaking authorization system called Zanzibar, which was originally designed for Google to power Google Docs and YouTube.

Warrant 的产品基于名为 Zanzibar 的突破性授权系统，该系统最初是为 Google 设计的，旨在为 Google Docs 和 YouTube 提供支持。

This enables fast authorization checks at enormous scale while maintaining a flexible model that can be adapted to even the most complex use cases.

这可以实现大规模的快速授权检查，同时保持灵活的模型，甚至可以适应最复杂的用例。

If you're currently looking to build role-based access control or other enterprise features like single sign-on, SCIM, or user management, you should consider WorkOS.

如果您当前正在寻求构建基于角色的访问控制或其他企业功能（例如单点登录、SCIM 或用户管理），您应该考虑 WorkOS。

It's a drop-in replacement for auth zero and supports up to one million monthly active users for free.

它是 auth Zero 的直接替代品，并免费支持多达 100 万的每月活跃用户。

Check it out at workos.com to learn more.

请访问 workos.com 以了解更多信息。

That's workos.com.

那是workos.com。

Lenny Rachitsky: You fell in love with building products for a reason, but sometimes the day-to-day reality is a little different than you imagined.

Lenny Rachitsky：您爱上构建产品是有原因的，但有时日常现实与您想象的有点不同。

Instead of dreaming up big ideas, talking to customers, and crafting a strategy, you're drowning in spreadsheets and roadmap updates and you're spending your days basically putting out fires.

您没有提出伟大的想法、与客户交谈并制定策略，而是淹没在电子表格和路线图更新中，并且每天都在忙着救火。

A better way is possible.

更好的方法是可能的。

Lenny Rachitsky: Introducing Jira Product Discovery, the new prioritization and roadmapping tool built for product teams by Atlassian.

Lenny Rachitsky：介绍 Jira Product Discovery，这是 Atlassian 为产品团队构建的新优先级划分和路线图工具。

With Jira Product Discovery, you can gather all your product ideas and insights in one place and prioritize confidently, finally replacing those endless spreadsheets.

借助 Jira Product Discovery，您可以将所有产品创意和见解收集到一处，并自信地确定优先顺序，最终取代那些无休无止的电子表格。

Create and share custom product roadmaps with any stakeholder in seconds.

在几秒钟内创建并与任何利益相关者共享自定义产品路线图。

And it's all built on Jira, where your engineering team's already working so true collaboration is finally possible.

这一切都建立在 Jira 上，您的工程团队已经在 J​​ira 上工作，因此真正的协作终于成为可能。

Great products are built by great teams, not just engineers.

伟大的产品是由伟大的团队打造的，而不仅仅是工程师。

Sales, support, leadership, even Greg from finance, anyone that you want can contribute ideas, feedback, and insights in Jira Product Discovery for free.

销售、支持、领导，甚至财务部门的 Greg，任何您想要的人都可以在 Jira Product Discovery 中免费贡献想法、反馈和见解。

No catch.

没有捕获。

And it's only $10 a month for you.

而且每月只需 10 美元。

Say goodbye to your spreadsheets and the never-ending alignment efforts.

告别电子表格和无休止的对齐工作。

The old way of doing product management is over.

旧的产品管理方式已经结束。

Rediscover what's possible with Jira Product Discovery.

重新发现 Jira Product Discovery 的可能性。

Try it for free at atlassian.com/lenny.

请访问 atlassian.com/lenny 免费试用。

That's atlassian.com/lenny.

这是 atlassian.com/lenny。

Lenny Rachitsky: Brendan, thank you so much for being here.

Lenny Rachitsky：布伦丹，非常感谢你来到这里。

Welcome to the podcast.

欢迎来到播客。

Brendan Foody: Thank you so much for having me, Lenny.

布兰登·福迪：非常感谢你邀请我，莱尼。

I'm a huge fan, and so excited to have a conversation.

我是一个超级粉丝，很高兴能进行对话。

Lenny Rachitsky: I'm really excited to have this conversation as well.

Lenny Rachitsky：我也很高兴能进行这次对话。

I'm a huge fan of yours.

我是你的超级粉丝。

I'm excited for more people to learn about you and what you're building.

我很高兴有更多的人了解您和您正在构建的东西。

Lenny Rachitsky: I want to start with a tweet that you have pinned at the top of your Twitter feed right now, and here's the tweet.

Lenny Rachitsky：我想从您现在固定在 Twitter 源顶部的一条推文开始，这是该推文。

"We are now working with six out of the Magnificent 7, all of the top five AI labs, most of the AI application layer companies.

“我们现在正在与 Magnificent 7 中的 6 家合作，所有排名前 5 的人工智能实验室，以及大多数人工智能应用层公司。

One trend is common across every customer.

每个客户都有一个共同的趋势。

We are entering the era of evals."

我们正在进入评估时代。”

Lenny Rachitsky: The reason this caught my attention is that's one of the most recurring trends on this podcast, people talking about the increasing value of learning how to do evals well and the value of evals for companies.

Lenny Rachitsky：这引起我注意的原因是，这是这个播客中最常见的趋势之一，人们谈论学习如何做好评估的价值不断增加以及评估对公司的价值。

It feels like still people don't know what the hell this is what we're talking about, why this is so important.

感觉人们仍然不知道我们在谈论的到底是什么，为什么这如此重要。

Talk about just what you think people are still missing, what they need to know, what this era of evals means.

谈谈你认为人们仍然缺少什么，他们需要知道什么，这个评估时代意味着什么。

Brendan Foody: If the model is the product, then the eval is the product requirement document.

Brendan Foody：如果模型是产品，那么评估就是产品需求文档。

And the way that researchers' day-to-day looks is that they'll run dozens of experiments where they'll make small improvements on an eval set.

研究人员每天都会进行数十次实验，在评估集上进行微小的改进。

And reinforcement learning is becoming so effective that once they have an eval, they can help climb it.

强化学习变得如此有效，一旦他们进行了评估，他们就可以帮助攀爬它。

If you look at just how fast people were able to saturate Olympiad Math once they focused on it, how fast we're even saturating SWE-bench once we focus on it.

如果你看看人们一旦专注于奥数就能以多快的速度饱和，那么一旦我们专注于它，我们甚至能以多快的速度使 SWE-bench 饱和。

And so in many ways, the barrier to applying agents the entire economy to automate every workflow is how do we measure success?

因此，在很多方面，在整个经济中应用代理来实现每个工作流程自动化的障碍是我们如何衡量成功？

How do we eval it?

我们如何评价它？

And write the PRDs for everything that we want agents to do, which Mercor is obviously a huge part of doing.

并为我们希望代理做的所有事情编写 PRD，这显然是 Mercor 的一个重要部分。

Lenny Rachitsky: So people hearing this, they're like, "Oh, yeah.

Lenny Rachitsky：所以人们听到这个，他们会说，“哦，是的。

Okay, shit.

好吧，妈的。

I got to really pay attention to this eval stuff." Any advice about learning how to do this well?

我必须真正关注这些评估的事情。”关于学习如何做好这件事有什么建议吗？

What companies that are doing this well are doing differently?

哪些公司做得很好，但他们的做法却有所不同？

Help people get better at this thing.

帮助人们在这件事上做得更好。

Brendan Foody: Yeah.

布兰登·福迪：是的。

I think that for enterprises especially, the core way to think about it is how can they build a test or systematic way to measure how well AI automates their core value chain?

我认为，特别是对于企业来说，思考这个问题的核心方式是如何建立一个测试或系统的方法来衡量人工智能对其核心价值链的自动化程度如何？

So if it's an architecture firm that's producing these architecture diagrams of what they provide to their end customer, how can they effectively measure that?

因此，如果是一家架构公司正在制作这些向最终客户提供的架构图，那么他们如何有效地衡量这一点？

And each company has its own value chain or maybe a handful of them if it's a multi-product company.

每家公司都有自己的价值链，如果是多产品公司，则可能有少数价值链。

And just thinking about how they measure that is the prerequisite to really effectively applying AI throughout their entire business.

只要思考他们如何衡量这一点，就是在整个业务中真正有效地应用人工智能的先决条件。

Lenny Rachitsky: I saw you talking about this on the No Priors podcast with Sarah and Elad, and I don't know if it was after this or before this, but Sarah tweeted, "Evals equals your new marketing." What does that mean?

Lenny Rachitsky：我看到你在 No Priors 播客上与 Sarah 和 Elad 谈论这个问题，我不知道是在这之后还是之前，但 Sarah 发推文说，“Evals 等于你的新营销。”这意味着什么？

What do you think she's saying there?

你觉得她在说什么？

Brendan Foody: Yeah.

布兰登·福迪：是的。

Well, it ties to what I said earlier about how if the model is the product, evals are the PRD, but also subsequently the sales collateral, right?

嗯，这与我之前所说的有关，如果模型是产品，评估是 PRD，但随后也是销售抵押品，对吗？

Because evals are what you give to researchers to show them what they should be building and going on, but they're also the way that you demonstrate the efficacy of capabilities.

因为评估是您向研究人员提供的内容，以向他们展示他们应该构建和继续进行的内容，但它们也是您展示功能有效性的方式。

Brendan Foody: And historically, everyone's been pointing to these academic evals of PhD level reasoning with GPQA, Humanity's Last Exam, or Olympiad Math, but now it's moving towards the capabilities that people practically care about of how do we get models to automate the way that we build a software platform or automate the way that we do an investment banking analysis.

Brendan Foody：从历史上看，每个人都在关注 GPQA、人类期末考试或奥林匹克数学对博士水平推理的学术评估，但现在人们真正关心的是我们如何获得模型来自动化我们构建软件平台的方式或自动化我们进行投资银行分析的方式。

And I think labs as well as application layer companies will increasingly use evals to demonstrate the capabilities of their models and their products.

我认为实验室以及应用层公司将越来越多地使用评估来展示其模型和产品的功能。

Lenny Rachitsky: Okay.

莱尼·拉奇茨基：好的。

So let's build on this and zoom out a little bit and talk about the landscape of the market that you're in.

因此，让我们在此基础上稍微缩小一下范围，讨论一下您所在的市场格局。

And I was just reflecting on this as I was preparing for this conversation.

当我准备这次谈话时，我正在思考这一点。

If you think about the companies growing faster than any company's ever grown in history, there's essentially three buckets.

如果你想想这些公司的增长速度比历史上任何一家公司的增长速度都快，那么基本上可以分为三个阶段。

There's the foundational model companies, there's vibe coding apps, Cursor and Loveable and Bolt and Replit and all these here, and then there's data labeling data companies like you.

这里有基础模型公司、vibe 编码应用程序、Cursor 和 Loveable、Bolt 和 Replit 等等，还有像您这样的数据标签数据公司。

So I've had the CEO of Handshake on the podcast.

所以我在播客上邀请了 Handshake 的首席执行官。

I have the CEO of Scale coming on.

Scale 的首席执行官即将登场。

There's also Surge.

还有浪涌。

There's you guys.

有你们。

Help us just understand the landscape of what this is all about because I think people don't really know what the hell's going on and see all these companies growing like crazy.

帮助我们了解这一切的情况，因为我认为人们并不真正知道到底发生了什么，并看到所有这些公司疯狂增长。

Brendan Foody: Yeah, I'll give a little bit of the origin story, incorporate in that and how it frames the landscape.

Brendan Foody：是的，我会介绍一些起源故事，并将其融入其中以及它如何构建景观。

Because when we started the company, I met my co-founders when we were 14 years old.

因为当我们创办公司时，我在 14 岁时遇到了我的联合创始人。

We started the company together when we were 19 initially, in January 2023, initially hiring people internationally, matching them with our friends and automating all the processes of how we did that.

2023 年 1 月，我们 19 岁的时候一起创办了这家公司，最初在国际上招聘人员，将他们与我们的朋友匹配，并使我们的所有流程自动化。

So similar to how a human would review a resume, conduct an interview, and decided to hire.

与人类审查简历、进行面试并决定聘用的方式非常相似。

We automated all of those processes with LLMs, bootstrap the company to a million dollar revenue run rate before we dropped out of college.

我们通过法学硕士实现了所有这些流程的自动化，在我们从大学退学之前将公司的收入运行率提升到了 100 万美元。

Brendan Foody: And then a handful of other things happened, but we met OpenAI and we saw that there was this enormous transition in the human data market where it was moving away from this crowdsourcing problem of how do you find low and medium skilled people that can write barely grammatically correct sentences for early versions of LLMs and moving towards this sourcing and vetting problem.

Brendan Foody：然后发生了一些其他事情，但我们遇到了 OpenAI，我们看到人类数据市场发生了巨大的转变，它正在摆脱如何找到能够为早期版本的法学硕士编写几乎语法正确的句子的中低技能人员的众包问题，并转向这个采购和审查问题。

How do we source and assess the best professionals, the experienced?

我们如何寻找和评估最优秀的专业人士、经验丰富的人？

Think software engineers, the investment bankers and doctors and lawyers that can actually help to evaluate and interpret all of the capabilities that people want models to have.

想想软件工程师、投资银行家、医生和律师，他们实际上可以帮助评估和解释人们希望模型具备的所有功能。

Brendan Foody: So from there, we start working with all of the top AI labs.

Brendan Foody：从那时起，我们开始与所有顶级人工智能实验室合作。

We grew from 1 to 400 million in revenue run rate in 16 months, and it's been an extraordinary journey and super exciting.

我们的收入运行率在 16 个月内从 1 亿增长到 4 亿，这是一次非凡的旅程，非常令人兴奋。

Lenny Rachitsky: Okay.

莱尼·拉奇茨基：好的。

First of all, that is out of control.

首先，这是失控的。

I don't know if people understand.

不知道大家是否明白。

I think this is the first time you're sharing that number.

我想这是你第一次分享这个数字。

I know we're recording this, you'll have announced it by now, but 1 to 400 million in revenue in 16 months.

我知道我们正在记录这一点，你现在已经宣布了，但是 16 个月内的收入为 1 到 4 亿。

Brendan Foody: Exactly.

布伦丹·福迪：没错。

So fastest ascent in history, which is an exciting statistic we're very proud of.

这是历史上最快的上升速度，这是一个令人兴奋的统计数据，我们对此感到非常自豪。

Lenny Rachitsky: Okay.

莱尼·拉奇茨基：好的。

So something big is happening here.

所以这里正在发生一些大事。

Why is this so valuable?

为什么这如此有价值？

What is going on here?

这是怎么回事？

So it's just to try to summarize what you guys do simply is you help hire people for labs to help them train their models, and you help them find not just generalist labor, but experts, helping them with very specific gaps in the model's knowledge.

因此，我只想总结一下你们所做的事情，就是帮助雇用实验室人员来帮助他们训练模型，帮助他们找到的不仅是通才，还有专家，帮助他们解决模型知识中非常具体的差距。

Brendan Foody: Yeah, precisely.

布伦丹·福迪：是的，没错。

And so it really ties to your first question around the era of evals that's framing all of this, which is that the lab's primary bottleneck to being able to improve models is how they can effectively have some way of measuring what success looks like for the model, both to use it as the eval for the tests that they're measuring their progress against, as well as the verifiers in an RL environment to then reward the model, improve capabilities, et cetera.

因此，这确实与你围绕评估时代提出的第一个问题有关，该时代构成了所有这一切，即实验室能够改进模型的主要瓶颈是他们如何有效地通过某种方式来衡量模型的成功程度，既将其用作他们衡量进度的测试的评估，也将其用作 RL 环境中的验证者，然后奖励模型，提高能力等等。

And they need this across every domain for every capability that models don't know how to use.

他们需要跨每个领域的模型不知道如何使用的每项功能。

And the wealthiest companies in the world are willing to spend whatever it takes to improve model capabilities where Mercor is sitting at the forefront and the primary bottleneck.

世界上最富有的公司愿意不惜一切代价来提高模型能力，而 Mercor 则处于领先地位和主要瓶颈。

Lenny Rachitsky: Okay, what are these people actually doing?

Lenny Rachitsky：好的，这些人实际上在做什么？

So what's an example of a kind of person that is sought after?

那么，什么是受追捧的人的例子呢？

And then what are they doing sitting there at the computer?

那么他们坐在电脑前做什么呢？

Brendan Foody: Effectively, the market is bound by the amount of things where humans can do something that models can't.

Brendan Foody：实际上，市场受到人类可以做而模型无法做的事情数量的限制。

So I'll make that very concrete.

所以我会说得非常具体。

Say you have a model that you want to write a red line for a contract in the way that a lawyer would, and it makes a handful of mistakes, misses a bunch of key points in doing so.

假设你有一个模型，你想像律师那样为合同写一条红线，但它犯了一些错误，遗漏了一些关键点。

What you could do is have a lawyer create a rubric similar to how a professor might create a rubric to create a deliverable for what are the things we want the model to be able to do?

您可以做的是让律师创建一个类似于教授创建标题的方式来创建我们希望模型能够做的事情的可交付成果？

Brendan Foody: So it can effectively score that, right?

Brendan Foody：所以它可以有效地得分，对吧？

Plus however much of it identifies this or XYZ key point.

另外，它的大部分内容都确定了这个或 XYZ 关键点。

And that's really the foundation to measuring what does progress look like for models?

这真的是衡量模型进展情况的基础吗？

Is this model achieving the capabilities that these professionals want?

该模型是否实现了这些专业人士想要的功能？

As well as how do we use this as training data to reward and to reinforce a lot of the capabilities that people want models to achieve.

以及我们如何使用它作为训练数据来奖励和强化人们希望模型实现的许多功能。

Lenny Rachitsky: Okay, so they're essentially writing evals just to connect it back to original conversation.

Lenny Rachitsky：好的，所以他们本质上是在编写评估，只是为了将其与原始对话联系起来。

Brendan Foody: Exactly.

布伦丹·福迪：没错。

Well, that's an interesting thing is everyone talks about RL environment.

嗯，有趣的是每个人都在谈论 RL 环境。

I feel like the two hot button things are like RL environments and evals, but one thing like Andrej Karpathy's tweeted out about a bunch is there's not actually a nuance.

我觉得这两个热门按钮就像 RL 环境和评估，但像 Andrej Karpathy 在 Twitter 上发布的一堆内容一样，实际上没有细微差别。

It's in the data type.

它在数据类型中。

It's more just a different semantic way of describing what it's being used for.

它更多地只是描述其用途的不同语义方式。

But ultimately, it's just some stasis point for how do you measure what good looks like?

但归根结底，这只是你如何衡量什么是好的东西的一个停滞点？

And you can use that either as the benchmark to the sales collateral, as Sarah was saying, to say, here is why are models the best model in the world and here's the capabilities that we've been working towards, or you can use it on the post-training side to reward certain model trajectories and achieve those capabilities.

你可以将其用作销售抵押品的基准，正如莎拉所说，这就是为什么模型是世界上最好的模型，这就是我们一直在努力实现的功能，或者你可以在训练后使用它来奖励某些模型轨迹并实现这些功能。

Lenny Rachitsky: Okay.

莱尼·拉奇茨基：好的。

So say this lawyer, this person is writing, "Here's what a great red line contract looks like and here's the rubric of what excellent is." Then are they also providing data, like actual examples of red line documents as a part of that?

这位律师说，这个人正在写，“这就是一份伟大的红线合同的样子，这就是优秀的标准。”那么他们是否也提供数据，例如红线文档的实际示例作为其中的一部分？

Brendan Foody: They may.

布伦丹·福迪：可能会。

The data landscape historically has included two kinds of data.

历史上，数据格局包括两种数据。

The first is supervised fine-tuning data, which is input/output.

第一个是有监督的微调数据，即输入/输出。

When people think about fine-tuning in the historical sense, that's what it is.

当人们从历史意义上思考微调时，就是这样。

The second is RLHF where the model will generate a couple of examples.

第二个是 RLHF，模型将生成几个示例。

We'll choose which is the most popular example.

我们将选择最流行的示例。

Brendan Foody: What everyone is generally moving towards is reinforcement learning from AI feedback instead of human feedback where you have instead the human defined some sort of success criteria, some way to measure that.

Brendan Foody：每个人普遍倾向于从人工智能反馈而不是人类反馈中进行强化学习，而是由人类定义某种成功标准，以及某种衡量标准的方法。

And examples in code, it could be a unit test.

代码中的示例，它可能是单元测试。

We can scalably measure success and other domains that could be a rubric.

我们可以大规模地衡量成功和其他可能成为标题的领域。

And then you use that to incentivize model capabilities.

然后你用它来激励模型的能力。

And it's far more scalable and data-efficient, and so that's why a lot of the broader trend in the market across the board is moving towards RLHF to both eval models as well as improved capabilities.

而且它的可扩展性和数据效率要高得多，这就是为什么市场上的许多更广泛的趋势正在全面转向 RLHF，以评估模型和改进的功能。

Lenny Rachitsky: I had one of the co-founders of Anthropic on.

Lenny Rachitsky：我邀请了 Anthropic 的联合创始人之一。

He said exactly the same thing.

他也说了同样的话。

That's what they've done at Anthropic, is move towards AI-driven reinforcement learning.

这就是他们在 Anthropic 所做的，朝着人工智能驱动的强化学习迈进。

Lenny Rachitsky: So essentially, if I can understand this correctly, I'm the lay person here trying to understand this on behalf of the audience.

Lenny Rachitsky：所以本质上，如果我能正确理解这一点，我就是这里的外行人，试图代表观众理解这一点。

So essentially a lawyer is like, "Here's what correct looks like for redlining," and then it's AI is just on its own almost, just like, "I'm going to try to get this.

所以本质上，律师会说，“这就是红线的正确样子”，然后人工智能就几乎靠自己了，就像，“我会尽力得到这个。”

I'm going to try to improve on this and I know if I'm heading the right direction based on this eval/rubric I've been given."

我将尝试改进这一点，并且我知道根据我得到的评估/标题我是否正在朝着正确的方向前进。”

Brendan Foody: Exactly.

布伦丹·福迪：没错。

Applying all of the criteria of what good looks like similar to how the TA might apply the professor's criteria of does the student's response meet this criteria or this criteria plus however many planes, et cetera.

应用所有关于什么是好的标准，类似于助教如何应用教授的标准，即学生的回答是否满足这个标准，或者这个标准加上多少个飞机，等等。

Lenny Rachitsky: Awesome, okay.

莱尼·拉奇茨基：太棒了，好吧。

Let me shift to talking about the broader labor market here.

让我谈谈更广泛的劳动力市场。

So there's two parts to this question as we talk about this.

因此，当我们讨论这个问题时，这个问题分为两个部分。

One is just how long will we need to do this?

一是我们需要多长时间才能做到这一点？

You guys grew so incredibly fast.

你们成长得如此之快。

Is there a point of like, "Okay, we don't need humans.

是否有这样的观点：“好吧，我们不需要人类。

We're tapped out." So let's start there and then I'll ask a broader question.

我们已经筋疲力尽了。”让我们从这里开始，然后我会问一个更广泛的问题。

Brendan Foody: So the key question is how long there's going to be things in the economy that humans can do that AI can't do?

Brendan Foody：所以关键问题是，经济中人类能做而人工智能做不到的事情还能持续多久？

And I think there's certainly a bucket of people that say we're going to have superintelligence within three years and humans won't play a role in the economy.

我认为肯定有很多人说我们将在三年内拥有超级智能，而人类将不会在经济中发挥作用。

And that's one school of thought.

这是一种思想流派。

Brendan Foody: Our perspective is very different.

Brendan Foody：我们的观点非常不同。

Our perspective is that these models are extraordinary and automating a lot of things very quickly, but there's a lot of things that they're horrible at.

我们的观点是，这些模型非常出色，可以非常快速地实现很多事情的自动化，但也有很多事情它们很糟糕。

Even still, it can't schedule time on my calendar.

即使如此，它仍然无法在我的日历上安排时间。

It can't draft emails for me.

它无法为我起草电子邮件。

It can't use basic tools.

它无法使用基本工具。

And we need evals for everything.

我们需要对一切进行评估。

For everything that the models can't do, we need evals for the tool use, evals for the long horizon reasoning.

对于模型无法做到的一切，我们需要对工具使用进行评估，对长期推理进行评估。

Brendan Foody: Imagine in 10 years when we want models to be able to go out and build a startup for 30 days.

Brendan Foody：想象一下，10 年后，我们希望模特能够在 30 天的时间里出去创办一家初创公司。

We need evals for that to effectively reward it.

我们需要对此进行评估以有效地奖励它。

And I think that that road to improving models will last for as long as there is anything in the economy that humans can do which models can't and be a huge portion of what the future of work looks like.

我认为，只要经济中存在人类可以做到而模型无法做到的事情，并且在未来工作中占据很大一部分，那么改进模型的道路就会持续下去。

And so our mission is creating the future of work, and I think that this is a really exciting industry and giving us a glimpse into the direction that everything is headed towards.

因此，我们的使命是创造工作的未来，我认为这是一个非常令人兴奋的行业，让我们一睹一切的发展方向。

Lenny Rachitsky: There's this tweet that you retweeted that I want to ask you about.

Lenny Rachitsky：您转发了这条推文，我想问您相关问题。

"If you really think about it, we were put on Earth to create reinforcement learning training data for labs."

“如果你认真思考一下，我们被派到地球上是为了为实验室创建强化学习训练数据。”

Lenny Rachitsky: What does that mean to you?

莱尼·拉奇茨基：这对你来说意味着什么？

What is this person implying?

这个人到底在暗示什么？

And it's basically what you're saying is we're just helping train models.

基本上你所说的是我们只是在帮助训练模型。

Brendan Foody: It speaks to conversations I've had with a lot of researchers and executives at top labs, which is that it's highly likely that the entire economy will become an aural environment machine, building out all of these worlds and contexts for us to then have rubrics or other kinds of verifiers.

Brendan Foody：这涉及到我与顶级实验室的许多研究人员和高管进行的对话，即整个经济很可能成为一个听觉环境机器，为我们构建所有这些世界和环境，然后拥有规则或其他类型的验证器。

And that is really exciting in so many ways.

这在很多方面都非常令人兴奋。

Brendan Foody: Because I think let's draw an analog to other revolutions where when we had the industrial revolution, everyone was freaking out about losing their jobs, but there was this whole new class of jobs of how do we build the machines?

布伦丹·福迪（Brendan Foody）：因为我认为让我们与其他革命进行类比，当我们进行工业革命时，每个人都为失去工作而感到害怕，但是我们如何制造机器却出现了全新的工作类别？

How do we have knowledge work?

我们如何进行知识工作？

How do we create everything new?

我们如何创造一切新事物？

And I think that the narrative in AI over the last three years has almost entirely been one of job displacement, right?

我认为过去三年人工智能领域的叙述几乎完全是工作岗位流失，对吗？

Sure, there's ChatGPT is growing fast and it's very cool that everyone loves using it, but from an economic standpoint, people talking a lot about job displacement.

当然，ChatGPT 正在快速发展，每个人都喜欢使用它，这很酷，但从经济角度来看，人们谈论很多关于工作取代的问题。

But very few companies and people have talked about this new category of jobs that's being created and what that's going to mean and how people can prepare and upskill for that.

但很少有公司和个人谈论正在创造的这一新类别的工作、这意味着什么以及人们如何为此做好准备和提高技能。

And I think that the most exciting thing possible is creating that future of how do humans fit into the economy and how will that evolve over time?

我认为最令人兴奋的事情是创造人类如何融入经济的未来以及随着时间的推移将如何演变？

Lenny Rachitsky: I talk to a lot of people about just what should I be studying?

Lenny Rachitsky：我和很多人讨论过我应该学习什么？

Where should I be getting better?

我应该在哪些方面变得更好？

People in school right now are just like, "What is even going to be valuable in the future?" You're at the center of a lot of just what jobs are most in demand, how hiring is evolving.

现在学校里的人都会想，“未来什么是有价值的？”您处于最需要什么工作以及招聘如何演变的中心。

So let me just ask you a very concrete question.

让我问你一个非常具体的问题。

What jobs do you think will remain in the future/what skills are still worth investing in for younger people, especially?

您认为未来哪些工作仍将保留/哪些技能仍然值得对年轻人特别是投资？

Brendan Foody: In terms of jobs, I would respond with a category of things that have very elastic demand are going to be super exciting.

Brendan Foody：就工作而言，我会回答一类需求非常有弹性的事情，这些事情将非常令人兴奋。

Because when we make people 10 times more productive, we'll build 10 times, if not 100 times as much software as an example.

因为当我们让人们的生产力提高 10 倍时，我们将构建 10 倍（甚至 100 倍）的软件。

And so I think the product managers that can now do so much more are going to be extremely well-positioned.

因此，我认为现在可以做更多事情的产品经理将处于非常有利的位置。

And so far as the skills, I think it's people that can leverage AI to do whatever their day-to-day workflows are.

就技能而言，我认为人们可以利用人工智能来完成日常工作流程。

Brendan Foody: I have had a couple conversations with teachers where they get my thoughts on how they should be assessing their students because we originally started out curating all of these AI interviews and assessments for people and have thought about this immensely.

Brendan Foody：我与老师进行了几次对话，他们了解了我对如何评估学生的想法，因为我们最初开始为人们策划所有这些人工智能面试和评估，并且对此进行了深入的思考。

And what we realized is that you don't want to fight against them using the models.

我们意识到你不想用模型来对抗他们。

It's similar to when the calculator came out, you don't want to give people all of this arithmetic work of how do you get them to do it and not use the calculator.

这类似于当计算器问世时，你不想给人们所有这些算术工作，如何让他们做而不使用计算器。

You want to tell them, "Use the tools and let's see what you can do."

您想告诉他们，“使用这些工具，让我们看看您能做什么。”

Brendan Foody: And so we'll give people interviews where we say, "Use ChatGPT and Kodak.

Brendan Foody：所以我们会给人们采访时说，“使用 ChatGPT 和柯达。

Use Claude code.

使用克劳德代码。

Use whatever tool cursor and whatever tools are available to build a website and let's see what product you're able to build in an hour." And so I think that I give that an example in so far as talent assessment because I think it pertains also to the skills that people should be honing in on of how can they leverage this technology to do so much more in whatever industry or vertical they're operating in.

使用任何工具光标和任何可用的工具来构建网站，让我们看看您在一小时内能够构建什么产品。”因此，我认为我在人才评估方面举了一个例子，因为我认为它也涉及人们应该磨练的技能，即他们如何利用这项技术在他们所从事的任何行业或垂直领域做更多的事情。

Lenny Rachitsky: When you talk about elastic, being elastic, is it generalists being good at just a bunch of different things, or what do you say?

Lenny Rachitsky：当你谈到弹性时，弹性是指擅长一堆不同事物的通才，还是你说什么？

What do you mean when you think elastic?

当您想到弹性时，您是什么意思？

Brendan Foody: So I more mean how much capacity for demand there is in that industry.

Brendan Foody：所以我更指的是该行业的需求能力有多大。

So I'll give a couple of examples.

我将举几个例子。

In accounting, I think realistically we only need so much accounting in the world.

在会计方面，我认为实际上我们在世界上只需要这么多会计。

Maybe there's areas where we can do more and that'll be good, but it doesn't feel like the world needs 100 times more accounting.

也许在某些领域我们可以做得更多，那会很好，但感觉世界并不需要增加 100 倍的会计工作。

Brendan Foody: On the other hand, in software development, I think we can ship 100 times more features for our products, move 100 times faster, build so much more.

Brendan Foody：另一方面，在软件开发方面，我认为我们可以为我们的产品提供 100 倍以上的功能，移动速度提高 100 倍，构建更多产品。

There's just it feels like there's unlimited demand for the industry.

感觉这个行业的需求是无限的。

And I think Mark Andreessen tweeted about this recently, that software is the most elastic industry of all where when we increase productivity, there's so much more that will be built.

我认为马克·安德森（Mark Andreessen）最近在推特上谈到了这一点，软件是所有行业中最具弹性的行业，当我们提高生产力时，将会构建更多的东西。

And it's definitely characteristic of a lot of other domains as well.

这绝对也是许多其他领域的特征。

And so I would focus on those domains where if we make everyone 10 times more productive, that'll increase demand, not reduce it.

因此，我将重点关注那些如果我们使每个人的生产力提高 10 倍，就会增加而不是减少需求的领域。

Lenny Rachitsky: Okay.

莱尼·拉奇茨基：好的。

So you're in the bucket of learn to code, still useful as a skill.

所以你正在学习编码，作为一项技能仍然有用。

You take computer science.

你学的是计算机科学。

And so in terms of elastic categories of jobs, sounds like engineering, product management is in that bucket.

因此，就弹性工作类别而言，听起来像是工程、产品管理就属于这一类。

Great.

伟大的。

A lot of people listening to this are PMs.

很多听这个的人都是PM。

What else, like design users?

还有什么，比如设计用户？

I don't know.

我不知道。

What else do you feel is in that bucket from what you've seen?

根据你所看到的，你觉得桶里还有什么？

Brendan Foody: Yeah, I think that there's a lot of things where the whole value chain of building companies has a lot of these variable costs, even large portions of operations or consulting.

Brendan Foody：是的，我认为建筑公司的整个价值链在很多方面都有很多可变成本，甚至很大一部分是运营或咨询。

Imagine if we could have 10 times as many McKinsey consultants, what would be possible in so far as the research we could do, the analysis, et cetera.

想象一下，如果我们可以拥有 10 倍的麦肯锡顾问，那么我们可以进行的研究、分析等等将会是怎样的。

But I think the companies and people that are going to succeed are those that lean into this narrative of abundance of how do we do so much more rather than fighting back against it of how do we try to stop displacement.

但我认为，那些会成功的公司和个人，都倾向于这种丰富的叙述，即我们如何做更多的事情，而不是反击我们如何试图阻止流离失所。

Lenny Rachitsky: So along those lines, I think about your second bucket, which is the people that will be most successful.

Lenny Rachitsky：沿着这些思路，我想到了你的第二个目标，即那些将最成功的人。

It's not like a specific skill, but it's being good with AI, using AI to become better at what you're already doing.

这不像是一种特定的技能，而是善于利用人工智能，利用人工智能来更好地完成你已经在做的事情。

This reminds me of Elon's whole thing with Neuralink, which I don't know if this is how we put it, but the way I've always heard it is you wanted to build Neuralink because in the future when AGI and superintelligence is around, we need a way to compete and the best way to compete is plug our brains into a superintelligence so we have a chance.

这让我想起了 Elon 的 Neuralink 的整个事情，我不知道这是否是我们的说法，但我一直听到的方式是你想要构建 Neuralink，因为在未来，当 AGI 和超级智能出现时，我们需要一种竞争的方式，而最好的竞争方式就是将我们的大脑插入超级智能，这样我们就有机会。

And it feels like that's what AI is.

感觉这就是人工智能。

Getting good at AI tools is essentially is having this super superpower.

擅长人工智能工具本质上就是拥有这种超能力。

Brendan Foody: Figuring out how to leverage them and incorporate it will definitely be of paramount importance.

Brendan Foody：弄清楚如何利用它们并将其纳入其中绝对是至关重要的。

Lenny Rachitsky: It just comes back to this almost cliche quote now.

Lenny Rachitsky：现在又回到了这句几乎是陈词滥调的引言。

It's, "AI won't replace you.

它是，“人工智能不会取代你。

People that are really good with AI will replace you."

真正擅长人工智能的人将会取代你。”

Brendan Foody: I think it's totally spot on.

Brendan Foody：我认为这是完全正确的。

And I've definitely seen this at the enterprise level as well where there's certain enterprises we talk to that are almost fearful not wanting to engage, not wanting to eval their businesses because that'll provide the evidence that their value chain is being automated.

我在企业层面也确实看到了这一点，我们与之交谈的某些企业几乎害怕不想参与，不想评估他们的业务，因为这将提供他们的价值链正在自动化的证据。

And there's others that...

还有其他一些...

Literally some of the most recognized sophisticated Fortune 500 businesses that have this mentality and there's others that are leaning into it of if we have the ability to do 10 or 100 times more, what will that mean and how do we lean into that future?

从字面上看，一些最受认可的复杂的财富 500 强企业都具有这种心态，还有其他一些企业也倾向于这种心态：如果我们有能力做得再多 10 倍或 100 倍，这意味着什么以及我们如何走向未来？

Because there's so many things that are going to change over the next 10 years, and I think those are the kinds of businesses that are going to be successful.

因为未来十年将会发生很多事情，我认为这些都是能够成功的企业。

Lenny Rachitsky: Let's talk about labor markets more broadly.

Lenny Rachitsky：让我们更广泛地讨论劳动力市场。

You guys, so it's interesting though.

你们，所以这很有趣。

You started not feeding people to AI labs, not training models.

你开始不把人们送到人工智能实验室，也不训练模型。

It was just like help people find jobs, help companies hire, and then you're like, "Oh wow, this whole opportunity." You have this really interesting view on the future of just labor markets and hiring.

这就像帮助人们找到工作，帮助公司招聘，然后你会想，“哦，哇，这整个机会。”您对劳动力市场和招聘的未来有非常有趣的看法。

Talk about that.

谈谈那个。

Brendan Foody: Yeah, it's interesting.

布伦丹·福迪：是的，这很有趣。

I remember when we started the company, as I mentioned, we were 19, and just had this gut intuition that it felt so wildly inefficient that labor markets are so disaggregated.

我记得当我们创办公司时，正如我提到的，我们 19 岁，直觉告诉我们，劳动力市场如此分散，效率极低。

And what I mean by that is when we would hire someone internationally, they would apply to a dozen jobs.

我的意思是，当我们在国际上雇用某人时，他们会申请十几个工作。

When we as a company in the Bay Area were considering candidates, we would consider a fraction of a percent of candidates that were available in the market.

当我们作为湾区的一家公司考虑候选人时，我们会考虑市场上可用候选人的一小部分。

And the reason for that is that there was this matching problem that everyone's solving manually where they'll manually review resumes, they'll manually conduct interviews, and manually decide who to hire.

原因是每个人都手动解决这个匹配问题，他们将手动审查简历，手动进行面试，并手动决定雇用谁。

But when we're able to automate that matching problem at the cost of software, it makes way for this global unified labor market that every candidate applies to and every company hires from facilitating a perfect flow of information in the economy.

但是，当我们能够以软件为代价实现匹配问题的自动化时，就会为全球统一的劳动力市场让路，每个候选人都可以申请该市场，每个公司都可以通过促进经济中信息的完美流动来进行招聘。

Brendan Foody: And I think that that future is undoubtedly what we're heading towards, but what we've realized over time is that the nature of work is also changing dramatically.

Brendan Foody：我认为，未来无疑是我们正在走向的，但随着时间的推移，我们意识到工作的性质也在发生巨大的变化。

And part of building that future over a 10-year time horizon is creating that future of work and all of the more tactical things we do and building these incredible data sets across evals and RL environments for our customers.

在 10 年的时间范围内构建未来的一部分是创造未来的工作和我们所做的所有更具战术性的事情，并为我们的客户跨评估和 RL 环境构建这些令人难以置信的数据集。

Lenny Rachitsky: What I've seen in how hiring has changed, I'm doing research on this with a partner, Gnome, it's so much easier to apply for companies that everyone's just applying now, to hundreds of companies.

Lenny Rachitsky：我所看到的招聘方式发生了怎样的变化，我正在与合作伙伴 Gnome 一起对此进行研究，现在每个人都在向数百家公司申请，现在申请公司要容易得多。

AI is just making it easy to adjust their resumes and cover letters and make it feel like, "Oh, I applied to more of course very specifically, but it was one of 100 places." And then on the flip side, hiring managers are getting flooded with applications and so now they need AI to filter.

人工智能只是让他们可以轻松调整简历和求职信，让人感觉“哦，我当然非常具体地申请了更多职位，但它只是 100 个职位之一。”另一方面，招聘经理收到了大量的申请，所以现在他们需要人工智能来过滤。

So even if we didn't want to get to this place, we're almost being pushed into this direction of so much volume on both sides.

因此，即使我们不想到达这个地方，我们也几乎被推向两侧都有如此大体积的方向。

We need something really smart at filtering and helping us hire and select, and this is exactly what you guys have been building for a long time.

我们需要一些真正智能的东西来过滤并帮助我们招聘和选择，而这正是你们长期以来一直在构建的东西。

Brendan Foody: Precisely, yeah.

布伦丹·福迪：没错，是的。

And the fascinating thing a lot of people ask, do we think about ourselves as a labor marketplace or do we think about ourselves as a data company?

许多人问的一个有趣的问题是，我们是否将自己视为劳动力市场，还是将自己视为一家数据公司？

And I think that the reason it's an interesting question is our realization on from what the labs need is that they actually need a labor marketplace.

我认为这是一个有趣的问题，因为我们从实验室的需求中认识到他们实际上需要一个劳动力市场。

They actually need these exceptionally high caliber people.

他们实际上需要这些非常高素质的人才。

And of course we'll layer on some project management and some software platform associated with it.

当然，我们将分层一些项目管理和一些与之相关的软件平台。

But the really core thing that they want is how do they find these extraordinary professionals across all of these different domains that can measure model capabilities and work to build that future work together?

但他们想要的真正核心的事情是如何找到跨所有这些不同领域的杰出专业人员，可以衡量模型能力并共同构建未来的工作？

Lenny Rachitsky: This episode is brought to you by Enterpret.

Lenny Rachitsky：本集由 Enterpret 为您带来。

Enterpret is a customer intelligence platform used by our leading CXN product orgs like Canva, Notion, Perplexity, Strava, Hinge, and Linear to leverage the voice of the customer and build best in class products.

Enterpret 是一个客户智能平台，我们领先的 CXN 产品组织（例如 Canva、Notion、Perplexity、Strava、Hinge 和 Linear）使用它来利用客户的声音并构建一流的产品。

Enterpret unifies all customer conversations in real time, from Gong recordings to Zendesk tickets to Twitter threads, and makes it available for your team for analysis and for action.

Enterpret 实时统一所有客户对话，从 Kong 录音到 Zendesk 票证再到 Twitter 线程，并可供您的团队进行分析和采取行动。

Lenny Rachitsky: What makes Enterpret unique is its ability to build and update a customer-specific knowledge graph that provides the most granular and accurate categorization of all customer feedback and connects that customer feedback to critical metrics like revenue and CSAT.

Lenny Rachitsky：Enterpret 的独特之处在于它能够构建和更新特定于客户的知识图，该知识图提供所有客户反馈的最精细和准确的分类，并将客户反馈与收入和 CSAT 等关键指标联系起来。

If modernizing your voice of customer program to a generational upgrade is a 2025 priority, customer-centric industry leaders like Canva, Notion, Perplexity, and Linear, reach out to the team at enterpret.com/lenny.

如果将客户之声计划现代化至代际升级是 2025 年的首要任务，那么以客户为中心的行业领导者（例如 Canva、Notion、Perplexity 和 Linear）请通过 enterpret.com/lenny 与团队联系。

That's E-N-T-E-R-P-R-E-T.com/lenny.

那是 E-N-T-E-R-P-R-E-T.com/lenny。

Lenny Rachitsky: Going back to just how this all works and what you guys do for models, I was talking to a friend who had an ankle sprain or his foot was hurting and he got an x-ray and he fed the x-ray into ChatGPT and then asked him, "Give me this specific x-ray." And it's like, "Okay, sure." And then it gave him, "Here's what you have." And he was talking to me, he's like, "What is out there on the internet that trained this model to know this stuff?" And I was like, "No, it's actually somebody sitting there helping the model understand this.

Lenny Rachitsky：回到这一切是如何运作的以及你们为模特所做的事情，我正在和一位脚踝扭伤或脚受伤的朋友交谈，他做了 X 光检查，他将 X 光检查输入 ChatGPT，然后问他，“给我这张特定的 X 光检查。”就像，“好吧，当然。”然后它给了他：“这就是你所拥有的。”他跟我说话时说，“互联网上有什么东西可以训练这个模型来了解这些东西？”我当时想，“不，实际上是有人坐在那里帮助模型理解这一点。

Once they recognize, it doesn't fully understand this.

一旦他们认识到，它并没有完全理解这一点。

Humans are actually helping them learn these things."

人类实际上正在帮助他们学习这些东西。”

Brendan Foody: Exactly.

布伦丹·福迪：没错。

Well, so the way it works, at least what most people's understanding is there's a lot of complexity in how the models work, is that pre-training gets a lot of the knowledge into the model of what are all the different things that see into the world.

好吧，它的工作方式，至少大多数人的理解是，模型的工作方式非常复杂，预训练将大量知识带入模型中，了解世界上所有不同的事物。

And then post-training and reinforcement learning is for all of the reasoning of what are the pieces of knowledge that are accurate, what are inaccurate, and what to prioritize at any given time to make a decision.

然后，训练后和强化学习用于推理哪些知识是准确的，哪些是不准确的，以及在任何给定时间要优先考虑哪些知识以做出决策。

And so behind that, there would've been radiologists that worked on the post-training data set to create some stasis point for here's the diagnosis and rewards and penalties associated with it.

因此，在此之后，放射科医生会处理训练后数据集，为诊断以及与之相关的奖励和惩罚创建一些停滞点。

And it's really the quality of those people that went into the quality of the decision and recommendation that ChatGPT ultimately made.

这些人员的素质确实影响了 ChatGPT 最终做出的决策和建议的质量。

Lenny Rachitsky: So let's actually follow that, right, because that's really interesting and I don't know how many people understand it.

Lenny Rachitsky：所以让我们实际上遵循这一点，对吧，因为这真的很有趣，我不知道有多少人理解它。

I understand it.

我明白了。

So the work that you do and these experts do is post-training.

因此，您和这些专家所做的工作是培训后。

It's not feeding data into the model that it's trained on.

它不会将数据输入到训练的模型中。

It's, "We have this model GPT-5.

它是，“我们有这个型号的 GPT-5。

Now here's all the things that's missing.

现在这里是所有缺少的东西。

Let's add to it."

我们来添加一下吧。”

Brendan Foody: Exactly, yeah.

布伦丹·福迪：没错，是的。

It's really unlocking, allowing the model to focus on all the right tokens, from pre-training all the right things in model context, up weighting the effective reasoning chains to enable the models to reason better in a more generalized way.

它确实是解锁的，允许模型专注于所有正确的标记，从在模型上下文中预训练所有正确的事物，增加有效推理链的权重，使模型能够以更通用的方式更好地推理。

Lenny Rachitsky: What's the scale of people just working on the stuff.

Lenny Rachitsky：从事这些工作的人员规模有多大？

It's like thousands, tens of thousands, hundreds of thousands?

好像是几千、几万、几十万？

Brendan Foody: Tens of thousands at any given time, hundreds of thousands more generally.

Brendan Foody：任何时候都有数以万计的人，更普遍的是数十万。

It's huge.

它是巨大的。

And the most exciting thing is that it's growing really quickly.

最令人兴奋的是它的增长速度非常快。

I think that to your question also about the competitive landscape, historically there were all these crowdsourcing companies that would get these super high volumes of low-skilled people.

我认为，对于你关于竞争格局的问题，历史上所有这些众包公司都会获得大量的低技能人才。

I think Scale and Surge were the primary companies that pioneered that industry.

我认为 Scale 和 Surge 是开创该行业的主要公司。

And then in this transition to higher-skilled labor, what people realized is that actually you can go a lot further with just getting higher caliber people even in smaller amounts initially, and now subsequently scaling that back up once they're able to meet the quality bar.

然后，在向高技能劳动力过渡的过程中，人们意识到，实际上，只要最初获得少量高素质人才，即使数量较少，现在只要他们能够达到质量标准，就可以扩大规模，从而走得更远。

Brendan Foody: And I think that there's a bunch of companies that after our success and very rapid revenue growth that started early last year have chased after that, which makes sense.

Brendan Foody：我认为，在我们的成功以及去年年初开始的快速收入增长之后，有很多公司都在追随我们，这是有道理的。

And seeing that the market was changing very quickly, we were taking off, and trying to pursue a similar thesis on the market.

看到市场变化非常快，我们开始起飞，并试图在市场上追求类似的论点。

Lenny Rachitsky: It's interesting.

莱尼·拉奇茨基：这很有趣。

There's always been these companies, AlphaSights and GLG, that did this before AI or is paid to connect to an expert and ask them questions about stuff.

AlphaSights 和 GLG 等公司在人工智能出现之前就已经这样做了，或者付费联系专家并向他们询问相关问题。

And essentially, okay, it turns out this is really useful for models.

本质上，好吧，事实证明这对于模型来说确实很有用。

We don't need the person in the middle.

我们不需要中间的人。

Brendan Foody: Exactly, yeah.

布伦丹·福迪：没错，是的。

Well, but one core difference is that AlphaSights would generally be a one-off call versus a lot of our work is really hiring people for projects of how do they work on something for a longer period of time.

嗯，但一个核心区别是 AlphaSights 通常是一次性的，而我们的很多工作实际上是雇用人员来完成项目，了解他们如何在较长时间内从事某项工作。

And so that's, I think, one of the reasons that some of the traditional expert networks have struggled to get into this.

我认为，这就是一些传统专家网络难以涉足这一领域的原因之一。

And also how do you retain those people and think about all the incentives where it actually looks more similar in some ways to one of the traditional labor marketplaces of an Uber or DoorDash, just with much higher-skilled talent that's treated exceptionally well?

另外，你如何留住这些人，并考虑所有的激励措施，在某些方面，它实际上看起来更类似于 Uber 或 DoorDash 的传统劳动力市场之一，只是拥有高技能人才，而且待遇特别好？

Lenny Rachitsky: It's such a good opportunity for me to learn so much about this, so I'm going to ask questions.

Lenny Rachitsky：这对我来说是一个很好的机会来了解很多这方面的知识，所以我要问问题。

Lenny Rachitsky: It's so interesting to me.

Lenny Rachitsky：这对我来说很有趣。

How much of the experts are focused on specific concrete knowledge versus personality and softer skills?

有多少专家关注具体的具体知识而不是个性和软技能？

How much of it's like, "Here's how you do an exam.

其中有多少就像，“这就是你如何进行考试。

Here's how you do an x-ray"?

这是如何进行 X 光检查的？”

Brendan Foody: It depends on the lab.

Brendan Foody：这取决于实验室。

It's a lot of both.

两者都有很多。

I think that previously it might've been more softer skills, but now a lot of the labs are focused on their business models of what are the economically valuable capabilities that drive revenue and leaning a lot into these professional domains.

我认为以前可能是更软的技能，但现在很多实验室都专注于他们的商业模式，即哪些具有经济价值的能力可以推动收入，并大力投入这些专业领域。

But I think the creative side is also still really important to everyone.

但我认为创意方面对每个人来说仍然非常重要。

And so we're seeing a meaningful amount of both.

因此，我们看到这两者的数量都相当大。

We hired all the people from the Harvard Lampoon a couple of months ago, their comedy club, to help with making models funnier.

几个月前，我们聘请了哈佛讽刺俱乐部喜剧俱乐部的所有人员来帮助让模型变得更有趣。

And so do all sorts of stuff like that, hiring Emmy award-winning screenwriters and everything across the board on creative capabilities that you'd look for.

诸如此类的事情也是如此，聘请艾美奖获奖编剧以及您想要的所有创意能力。

Lenny Rachitsky: That is amazing.

莱尼·拉奇茨基：太棒了。

What a cool story.

多么酷的故事啊。

I'm excited for this to kick in.

我对此感到很兴奋。

How fast do these things turn around?

这些事情的转变速度有多快？

Say you hired this team, how fast are we going to see the impact potential?

假设您雇用了这个团队，我们要多快才能看到潜在的影响力？

Is like months?

是像几个月吗？

Is it years?

是几年了吗？

Brendan Foody: Well, so it depends, because some models or some labs will release iteratively where they'll just improve the model behind the scenes.

Brendan Foody：嗯，这要看情况，因为一些模型或一些实验室会迭代发布，他们只是在幕后改进模型。

Lenny Rachitsky: Without announcing a new model?

Lenny Rachitsky：没有宣布新型号？

Brendan Foody: Exactly.

布伦丹·福迪：没错。

Every couple of weeks versus others do these big releases.

每隔几周就会与其他人进行这些重大发布。

And so it depends a lot.

所以这取决于很多。

We're behind all of them, but we move really fast.

我们落后于所有这些，但我们进展得很快。

It would be a customer gives us a request of we need these award-winning screenwriters, and within 24 hours we'll turn around the experts.

如果客户向我们提出要求，我们需要这些获奖编剧，我们会在24小时内请专家。

And then there's also this really interesting dynamic where in a set of 100 people that we hire, oftentimes the top 10% of people will drive majority of the model improvement.

还有一个非常有趣的动态，在我们雇用的 100 名员工中，通常前 10% 的人将推动模型的大部分改进。

It's like a company.

这就像一家公司。

If you have 100-person company, oftentimes the top 10% of the company will drive majority of the impact.

如果您的公司有 100 人，通常公司前 10% 的员工将产生大部分影响。

And what that means is that when we're able to build proprietary advantages in identifying who are those top 10% of people, both in so far as how do we have them on our platform but also identify and match them effectively, it creates so much value for customers that it's difficult to compete against.

这意味着，当我们能够在识别谁是前 10% 的人方面建立专有优势时，无论是如何将他们放在我们的平台上，还是有效地识别和匹配他们，它都会为客户创造如此多的价值，以至于难以与之竞争。

Brendan Foody: And so it really does tie back to the founding thesis of the company, which is how do we find these extraordinary people and identify them so that we can reliably deliver these top 10% or top 10X experiences for our customers.

Brendan Foody：因此，这确实与公司的创立理念相关，即我们如何找到这些杰出人才并识别他们，以便我们能够可靠地为客户提供前 10% 或前 10X 的体验。

Lenny Rachitsky: So on that, so is the idea, you hire Jane.

Lenny Rachitsky：那么，这个想法就是这样，你雇佣简。

She's incredible at coding and she now works for Anthropic and that's her full-time job doing this?

她的编码能力令人难以置信，现在为 Anthropic 工作，这就是她的全职工作吗？

Or is this a part-time thing?

或者这是一个兼职的事情？

Is this a project thing mostly?

这主要是一个项目吗？

Brendan Foody: It would sometimes be part-time.

Brendan Foody：有时是兼职。

Sometimes it would be full-time.

有时它会是全职的。

I would say most often it's part-time where it's like someone might work at a thing company where they're underemployed, maybe one of the ones that's moving slower where they have an extra 20 hours a week and then they're able to do this on the side or whatever the equivalent is across a bunch of different industries.

我想说最常见的是兼职，就像某人可能在一家未充分就业的事物公司工作，也许是一家工作节奏较慢的公司，他们每周有额外的 20 个小时，然后他们就可以兼职做这件事，或者在许多不同的行业中从事类似的工作。

But we also do a lot of 40 hour a week roles as well.

但我们也承担很多每周 40 小时的工作。

Lenny Rachitsky: And how much are they making?

Lenny Rachitsky：他们赚多少钱？

Is it meaningful enough for a AI engineer to spend time on this?

对于一个AI工程师来说，花时间在这上面有意义吗？

Brendan Foody: Yeah, very meaningful.

Brendan Foody：是的，非常有意义。

So our median pay rate in the marketplace is $95 an hour, but it can flex up, well up into $500 an hour based on the depth of someone's expertise.

因此，我们市场上的工资中位数是每小时 95 美元，但根据某人的专业知识深度，它可以灵活地提高到每小时 500 美元。

And one thing that highlights this difference relative to a lot of the crowdsourcing companies is if you look at the economics of the crowdsourcing companies, oftentimes they would pay $30 an hour to town as the average.

与许多众包公司相比，突出这种差异的一件事是，如果你看看众包公司的经济情况，他们通常会平均支付 30 美元每小时的费用。

And so think about the people that you can hire, the undergrads for $30 now versus the Goldman bankers, the McKinsey analysts, the Fang software engineers.

因此，想想你可以雇佣的人，现在 30 美元的本科生与高盛银行家、麦肯锡分析师、方软件工程师。

And ultimately it comes down to what are the capabilities that labs want their models to have?

最终归结为实验室希望他们的模型具备哪些功能？

And it much more falls in the latter bucket than the former one.

与前者相比，它更多地属于后者。

Lenny Rachitsky: I know there's only so much you can talk about with this stuff, but so Anthropic, Claude has been so good at coding so much better historically than other models.

Lenny Rachitsky：我知道你能谈论的东西只有这么多，但是克劳德如此人性化，在编码方面一直比其他模型好得多。

I also use it for writing, giving feedback on writing.

我也用它来写作，提供写作反馈。

What is it that allowed them to get so good at this and continue to be so good at this?

是什么让他们能够如此擅长并继续如此擅长？

Brendan Foody: Well, I can't go too much into detail about customer work, but I think that it's this trend of reinforcement learning and being very thoughtful about defining the right rewards that we're releasing across the board.

Brendan Foody：嗯，我无法详细介绍客户工作，但我认为这是强化学习的趋势，并且在定义我们全面发布的正确奖励时非常深思熟虑。

And how we could mitigate reward hacking, set up the right rewards, that's super impactful.

我们如何减少奖励黑客行为，设置正确的奖励，这是非常有影响力的。

Lenny Rachitsky: Evals.

莱尼·拉奇茨基：埃瓦尔斯。

Again, evals is all you need.

同样，evals 就是您所需要的。

Brendan Foody: One of my favorite quotes from customers is that, "Models are only as good as their evals," which has always held true.

Brendan Foody：我最喜欢的客户名言之一是，“模型的好坏取决于他们的评估”，这句话一直都是正确的。

Lenny Rachitsky: I think Greg Brockman tweeted this once.

Lenny Rachitsky：我想 Greg Brockman 曾经发过这样的推文。

"Evals are all you need."

“评估就是您所需要的。”

Lenny Rachitsky: Let's talk about Mercor a little bit more.

Lenny Rachitsky：让我们再多谈谈 Mercor。

One of the maybe, not even maybe, I believe the data tells us it's the fastest growing company in history.

其中一个可能，甚至不是可能，我相信数据告诉我们它是历史上增长最快的公司。

Lenny Rachitsky: I want to understand what you did to make this happen.

Lenny Rachitsky：我想了解您为实现这一目标做了什么。

So let me just ask, what do you think are some of the core tenets of how you built Mercor that most contributed to being this successful?

那么我想问一下，您认为您创建 Mercor 的核心原则是什么，对取得如此成功最有贡献？

Brendan Foody: I think the most important thing is looking at the leading indicators in fast-moving markets.

Brendan Foody：我认为最重要的是关注快速变化市场的领先指标。

I remember when I used to think...

我记得以前我以为...

Everyone in venture talks about the why now, and I used to think about the why now of how from a product standpoint, less from a market standpoint of now we can automate the way that we review resumes or the way that we conduct interviews, et cetera.

风险投资中的每个人都在谈论“为什么现在”，而我曾经从产品的角度思考“为什么现在”，而不是从市场的角度来看，现在我们可以自动化我们审查简历的方式或我们进行面试的方式等等。

But ultimately there is this legacy market that's has all these incumbents and it's relatively stagnant.

但最终，这个传统市场拥有所有这些现有企业，而且相对停滞不前。

But what matters a ton is actually figuring out what are the new markets, the new pockets of demand that are changing very quickly where the wealthiest customers in the world are willing to pay whatever it takes to improve model capabilities, and how do we focus on the leading indicators of those markets to make sure that we have the best solution for the flagship customers in the market and optimize everything around that.

但最重要的是，实际上要弄清楚什么是新市场、正在快速变化的新需求领域，世界上最富有的客户愿意不惜一切代价来提高模型能力，以及我们如何关注这些市场的领先指标，以确保我们为市场中的旗舰客户提供最佳解决方案，并优化围绕该解决方案的一切。

Brendan Foody: And that's what I found has been most impactful in building the business.

Brendan Foody：我发现这对业务建设最有影响力。

I think maybe that's one thing is leading indicators in markets.

我认为也许市场领先指标是一回事。

If I had to choose another, it's customer obsession.

如果我必须选择另一个，那就是客户至上。

We have had for the last...

我们已经有最后...

We're starting to have a couple of product managers help out with go-to-market, but for the last year and a half of the business, we've had no one in sales and marketing.

我们开始聘请几名产品经理来帮助进入市场，但在过去一年半的业务中，我们没有销售和营销人员。

And so we're immature from a sales and marketing standpoint because we focused 100% of company resources on how do we build great products and experiences for our customers.

因此，从销售和营销的角度来看，我们还不成熟，因为我们将 100% 的公司资源集中在如何为客户打造出色的产品和体验上。

Just getting word of mouth, the people that have worked with us at other businesses want to keep working with us and leaning into creating those great experiences.

只要得到口碑，在其他企业与我们合作过的人就希望继续与我们合作，并致力于创造这些美好的体验。

And so that's where I spend all my time.

这就是我所有时间都花在的地方。

And I think that some founders can get caught up in how do they get really good at marketing before they've figured out the thing that really drives a lot of customer love and creates the six-star experiences that you're used to building.

我认为，一些创始人可能会陷入如何真正擅长营销的问题，然后他们才弄清楚真正能激发大量客户喜爱并创造您习惯于建立的六星级体验的事情。

Lenny Rachitsky: I'm going to go back to that first point, which is like, okay, you found this pocket, maybe the biggest business opportunity in history.

Lenny Rachitsky：我要回到第一点，那就是，好吧，你发现了这个口袋，也许是历史上最大的商机。

How did you first find...

你最初是怎么发现...

What was that moment of, "Wait, this could be really big"?

“等等，这可能真的很大”的那一刻是什么？

Brendan Foody: So there's some crazy stories here.

Brendan Foody：这里有一些疯狂的故事。

I remember we started the company as I mentioned in January 2023.

我记得正如我在 2023 年 1 月提到的那样，我们创办了这家公司。

And then in August 2023 when I was still in college, one of our customers introduced us to the co-founders of xAI over a Zoom call saying how we had these really smart Indian software engineers that were great at math and coding.

然后在 2023 年 8 月，当我还在上大学时，我们的一位客户通过 Zoom 电话向我们介绍了 xAI 的联合创始人，并介绍了我们如何拥有这些非常聪明、擅长数学和编码的印度软件工程师。

So we met them and we explained how the software engineers we had were really good at math and coding because they weren't distracted by all the humanities.

因此，我们会见了他们，并向他们解释了我们的软件工程师如何真正擅长数学和编码，因为他们没有被所有人文学科分心。

They didn't have to study history and English and all these other things, and they loved it.

他们不必学习历史、英语和所有其他东西，而且他们喜欢它。

So they had us in two days later to the Tesla office and we met the entire xAI co-founding team except for Elon, while I was still a college student.

两天后，他们带我们去了 Tesla 办公室，我们见到了除 Elon 之外的整个 xAI 联合创始人团队，当时我还是一名大学生。

And xAI was just getting started at that point and they were super excited about our focus on the quality of the experts.

xAI 那时才刚刚起步，他们对我们对专家质量的关注感到非常兴奋。

Brendan Foody: And so while they were still doing pre-training, they weren't ready for human data at the time and we didn't start working with them at that point.

Brendan Foody：因此，当他们仍在进行预训练时，他们当时还没有准备好接受人类数据，我们当时也没有开始与他们合作。

We just knew from that point forward before we even dropped out that the market was about to change radically and we needed to be at the frontier of that.

从那时起，我们就知道，在我们退出之前，市场即将发生根本性的变化，我们需要处于这一变化的前沿。

And so then fast-forward a few months, one of the crowdsourcing players came to us and actually used our platform to hire over 1,000 people where this is very interesting experience because we started getting flooded with support tickets about how those people weren't getting paid.

几个月后，一位众包参与者来到我们这里，实际上使用我们的平台雇佣了 1,000 多人，这是非常有趣的经历，因为我们开始收到大量关于这些人如何没有获得报酬的支持票。

And we obviously felt horrible because we had referred them to this opportunity.

我们显然感到很糟糕，因为我们向他们推荐了这个机会。

It was this reputable company.

就是这家信誉良好的公司。

And we realized that a lot of the incumbents were resting on their laurels with respect to what was needed in the experiences they were creating for talent in their marketplaces to help improve models.

我们意识到，许多现任者在为市场人才创造经验以帮助改进模型方面满足于自己的成就。

And there was this opportunity to work directly with the labs in a way that kept the dignity of the experts in the marketplace, paid them extremely well, and cut out the middlemen.

并且有机会直接与实验室合作，以保持专家在市场上的尊严、支付非常丰厚的报酬并消除中间商的方式。

Brendan Foody: And so we started doing that in May of last year, and then the rest is history.

Brendan Foody：所以我们从去年五月开始这样做，然后剩下的就成为历史了。

Lenny Rachitsky: Wow, okay.

莱尼·拉奇茨基：哇，好吧。

Hundreds of millions of dollars in revenue since.

此后收入达数亿美元。

So what I'm hearing here is you were very open to looking for poll.

所以我在这里听到的是，您非常愿意寻求民意调查。

You saw some poll, you explored it.

你看到了一些民意调查，你探索了它。

And then once you saw that there was something really meaningful there, you just went deep on making that an incredible experience as amazing as possible.

然后，一旦你看到那里有一些真正有意义的东西，你就会深入研究，使其成为尽可能令人难以置信的体验。

Brendan Foody: Exactly.

布伦丹·福迪：没错。

I think if I had to distill it into advice for founders, one thing I've realized is that I spent a lot of time trying forced product-market fit.

我认为，如果我必须将其提炼成对创始人的建议，我意识到的一件事是我花了很多时间尝试强制产品市场契合。

And in some ways you should be persistent.

在某些方面你应该坚持不懈。

You should have these theses that you have conviction about how the world will change.

你应该有这些论文，表明你对世界将如何改变有信心。

But sometimes you just need to sheer it from the market and know that it's there, the poll, to know the right places to focus.

但有时你只需要从市场中剔除出来，并知道它就在那里，民意调查，就可以知道应该关注的正确位置。

Because if it's difficult to sell, if it's extremely difficult to sell the marginal customer, you're not going to be able to grow a huge business.

因为如果很难销售，如果很难销售边际客户，那么你就无法发展庞大的业务。

What you actually need to find is the customer that's surprisingly easy to sell into where you're going to be able to grow with them.

你真正需要找到的是那些非常容易推销的客户，并且你能够与他们一起成长。

You know that it's a large pain point.

你知道这是一个很大的痛点。

And so it's some combination of being stubborn with respect to your thesis around how the world will change, but also very open-minded with respect to exactly what form that takes and how the market's developing and how your company will fit into it.

因此，既要固执地坚持自己关于世界将如何变化的论点，又要对具体采取什么形式、市场如何发展以及您的公司如何融入市场持非常开放的态度。

Lenny Rachitsky: That's an amazing insight.

Lenny Rachitsky：这是一个令人惊奇的见解。

In the moments you described, felt like it was a combination of this xAI meeting feeling like, "Oh wow, they really, really want this thing that we have.

在你描述的那些时刻，感觉就像是这次 xAI 会议的组合，感觉就像，“哇哦，他们真的、真的想要我们拥有的这个东西。

We're now doing an amazing job," and then it's 1,000 people hiring in the platform.

我们现在做得非常出色”，然后平台上就招聘了 1,000 名员工。

Was that those two moments that are like, "Wow"?

那两个时刻是不是让人发出“哇”的感觉？

Brendan Foody: Exactly.

布伦丹·福迪：没错。

And those happened, keep in mind, while we were a seed company, right?

请记住，当我们还是一家种子公司时，这些事情就发生了，对吗？

Well, so the first one was before we even raised any seed funding, we were totally bootstrapped because we bootstrapped the company to a million dollar revenue run rate and have always remained super capital-efficient.

嗯，所以第一个是在我们筹集任何种子资金之前，我们完全是自力更生的，因为我们将公司的收入运行率提高到了 100 万美元，并且始终保持超级资本效率。

We've never burned money.

我们从来没有烧过钱。

We were lifetime profitable.

我们终生盈利。

And then we raised our seed round in September from General Catalyst, and it was the other experience after we raised our seed round where we really knew that there was an enormous amount of demand in this market where we saw the volume and we saw that the incumbents were sleeping with respect to how the market was changing and the kinds of people that were needed to make that change happen.

然后我们在 9 月份从 General Catalyst 筹集了种子轮资金，这是我们筹集种子轮资金后的另一次经历，我们真正知道这个市场有巨大的需求，我们看到了数量，并且我们看到现有企业对市场如何变化以及实现这种变化所需的人员类型漠不关心。

Lenny Rachitsky: It's one thing to see this opportunity and start to execute on it.

Lenny Rachitsky：看到这个机会并开始执行它是一回事。

It's another to actually succeed at this scale and consistently win.

在这种规模上真正取得成功并持续获胜是另一回事。

You guys have very specific values within the business.

你们在行业内有着非常具体的价值观。

Talk about those.

谈谈那些。

It feels like that's a big part of your success too.

感觉这也是你成功的重要组成部分。

Brendan Foody: It totally is.

布伦丹·福迪：完全是这样。

So I'll give the three and maybe a brief story associated with each of them.

因此，我将介绍这三个人，也许还有一个与他们每个人相关的简短故事。

Brendan Foody: So the first one is having a can-do attitude, which everyone gives me a little bit of a hard time for because it's a funny saying, but we've always set these ridiculously ambitious goals, and then somehow the trajectory of the company forms around those goals.

Brendan Foody：所以第一个就是要有一种“我能做到”的态度，每个人都让我有点为难，因为这是一个有趣的说法，但我们总是设定这些荒谬的雄心勃勃的目标，然后以某种方式围绕这些目标形成公司的轨迹。

Where I remember when we were talking to Benchmark before they led our Series A, we were at 1.5 million in run rate.

我记得当 Benchmark 领导我们的 A 轮融资之前我们与他们交谈时，我们的运行率是 150 万。

And I said we'd be at 50 million in run rate by the end of the year.

我说过，到今年年底，我们的运行率将达到 5000 万。

And they said we were absolutely insane, right, as anyone would.

他们说我们绝对疯了，对吧，就像任何人都会的那样。

And plus or minus two weeks, we hit it.

前后两周，我们就成功了。

And then we've now well blown past the tracking to 500 million in run rate, which was initially our goal for this year.

现在，我们的运行率已经远远超过了 5 亿次，这是我们今年最初的目标。

So setting these incredibly ambitious goals with respect to the revenue scale of the business, the caliber of experiences for talent, all those dimensions is super important to first have a can-do attitude.

因此，在业务收入规模、人才经验水平等方面设定这些令人难以置信的雄心勃勃的目标，所有这些维度对于首先拥有进取的态度非常重要。

Brendan Foody: The second thing is really high standards, which is who we hire and what we expect of them.

Brendan Foody：第二件事是真正的高标准，即我们雇用的人员以及我们对他们的期望。

We have an incredibly high hiring bar where we hire tons of former founders, people that have incredible experiences.

我们的招聘门槛非常高，我们雇佣了大量的前创始人，他们拥有令人难以置信的经验。

We just hired or partnered with Sundeep Jain who joined us as president.

我们刚刚聘请了 Sundeep Jain 或与其合作，他加入我们担任总裁。

He was previously the chief product officer and chief technology officer at Uber and joined our relatively small in the grand scheme of things company to help scale up all the processes where Uber is of course the largest labor marketplace in the world.

他此前曾担任 Uber 的首席产品官和首席技术官，并加入了我们这家规模相对较小的公司，帮助扩大所有流程，而 Uber 当然是世界上最大的劳动力市场。

So super high standards is of paramount importance.

所以超高的标准至关重要。

Brendan Foody: And then the third one that we really lean on significantly is intensity.

Brendan Foody：我们真正依赖的第三个因素是强度。

And that if you look at the early cultures of the legendary companies, thinking of Meta or Google, they have these incredible, intense early-stage cultures of people just moving heaven and earth and doing whatever it takes to push the frontier of model capabilities.

如果你看看传奇公司的早期文化，比如 Meta 或 Google，你会发现他们拥有令人难以置信的、强烈的早期文化，人们只是在天地间移动，不惜一切代价推动模型能力的前沿。

And so still very much output-oriented of what do people achieve rather than input-oriented of the specific hours they work, but recognizing that it takes a lot to build a legendary business, and that's ultimately what we're optimizing for.

因此，人们仍然以产出为导向，而不是以他们工作的具体时间为导向，但要认识到建立传奇企业需要付出很多努力，而这最终就是我们要优化的目标。

Lenny Rachitsky: I could see why this works.

Lenny Rachitsky：我明白为什么这有效。

Can-do attitude plus high standards plus intensity, I could see how that leads to success.

积极进取的态度加上高标准和强度，我可以看到这如何带来成功。

There's a lot of talk these days about this 6-9-9 culture, working six days a week, 9:00 AM to 9:00 PM.

如今，关于这种 6-9-9 文化（每周工作六天，上午 9:00 到晚上 9:00）的讨论很多。

A lot of people are like, "Why?

很多人会说：“为什么？

That's terrible.

太可怕了。

Why would you make people do that?" But at the same time, I'm just constantly hearing this from the most successful AI companies.

你为什么要让人这么做？”但与此同时，我不断从最成功的人工智能公司那里听到这样的说法。

This is just the way it is to be successful.

这就是成功的方式。

Things are moving so fast.

事情发展得太快了。

This is an opportunity you'll never see again.

这是一个你再也不会看到的机会。

Just talk about your thoughts on that.

谈谈你对此的看法吧。

Brendan Foody: Yeah.

布兰登·福迪：是的。

Well, to clarify, we've never mandated hours.

好吧，澄清一下，我们从未规定过工作时间。

It's more been a byproduct of people that care a lot where we care a lot about the trajectory of the business.

它更多的是那些非常关心我们非常关心业务轨迹的人们的副产品。

And so a lot of people come into the office and stay late.

因此很多人来到办公室并熬夜。

But if they need to leave early and get dinner with their kids or travel on the weekend, of course that's totally fine.

但如果他们需要早点离开并与孩子一起吃晚饭或周末旅行，那当然完全没问题。

And for us, it's much more about finding people who have a lot of ownership and are really bought in, less so about the specific hours in the office, even though we found that oftentimes it's the people that are most bought in, not always, but oftentimes it's the people that are most bought in and that burn the midnight oil with us.

对我们来说，更重要的是找到那些拥有大量所有权并且真正被接纳的人，而不是在办公室的特定时间，尽管我们发现通常是最被接纳的人，并非总是如此，但通常是那些最被接纳并与我们一起熬夜的人。

Lenny Rachitsky: When you say high standards, is there something you could share that gives us an example of what you mean there?

Lenny Rachitsky：当您说到高标准时，您是否可以分享一些东西，让我们举例说明您的意思？

Because a lot of people think they have high standards and they don't.

因为很多人认为自己有很高的标准，但事实并非如此。

Brendan Foody: If you are very patient, there's always some trade-off between speed and quality when hiring.

Brendan Foody：如果您非常有耐心，那么在招聘时总是需要在速度和质量之间进行权衡。

And I remember especially for our first 10 people, we were just so patient and disciplined about finding some of the best people in the world.

我记得特别是对于我们的前 10 名员工，我们非常耐心和自律地寻找世界上最优秀的人才。

Half of them are...

其中一半是...

Our second employee, Sid, as an example, our second employee in the US, Sid was previously the head of growth at Scale who joined us when we were a seed stage company.

以我们的第二位员工 Sid 为例，我们在美国的第二位员工 Sid 之前是 Scale 的增长主管，他在我们还是一家种子期公司时加入了我们。

Daniel who joined us was previously scaled to consumer apps to over 100,000 users and all sorts of just extraordinary backgrounds of our first 10 hires.

加入我们的 Daniel 之前曾负责消费类应用程序，拥有超过 100,000 名用户，并且是我们前 10 名员工的各种非凡背景。

And I think that that initial talent density shaped so much of what the rest of the org looks like as you scale it up.

我认为，当你扩大规模时，最初的人才密度在很大程度上决定了组织其他部分的面貌。

Lenny Rachitsky: I know you also have this perspective that people talk about waiting to hire, to hire really slowly, but it's actually not necessarily the right advice.

Lenny Rachitsky：我知道您也有这样的观点，即人们谈论等待招聘，慢慢地招聘，但这实际上不一定是正确的建议。

Talk about that.

谈谈那个。

Brendan Foody: It's painful because it's a double-edged sword.

Brendan Foody：这很痛苦，因为它是一把双刃剑。

On one hand, I'm thrilled that our first 10 people are so phenomenal and I think that that has paid dividends for the business.

一方面，我很高兴我们的前 10 名员工如此出色，我认为这为公司带来了红利。

But on the other hand, I think that companies do get to the point where you just need to hire really fast.

但另一方面，我认为公司确实已经到了只需要快速招聘的地步。

And there's some things where you need a lot of people to do them and you need to recognize that there's going to be some variants associated with hiring, but moving quickly is the priority.

有些事情需要很多人来完成，并且您需要认识到招聘方面会存在一些变体，但快速行动是首要任务。

Brendan Foody: And I think that in some ways, we move too slowly with how we scaled out the team.

Brendan Foody：我认为在某些方面，我们在扩展团队方面进展得太慢了。

And so the benefit is that everyone is extraordinary.

所以好处是每个人都是非凡的。

We have this super high bar and we want to maintain that over time.

我们有这个超高的标准，我们希望随着时间的推移保持这一水平。

But I think the downside is that while the company has grown incredibly quickly, we likely could have grown even faster if we had moved a little bit more quickly with especially ramping from call, like 10 to 100 people.

但我认为缺点是，虽然公司发展得非常快，但如果我们行动得更快一点，尤其是从呼叫人数增加到 10 到 100 人，我们可能会发展得更快。

Lenny Rachitsky: Okay, I was going to ask.

Lenny Rachitsky：好的，我正想问。

So it sounds like the first 10, be very careful, take your time, 10 to 100, maybe speed up a bit.

所以听起来像前 10 个，要非常小心，慢慢来，10 到 100，也许加快一点。

Brendan Foody: Yes, though I wouldn't say it's necessarily 10.

Brendan Foody：是的，不过我不会说一定是 10。

It's determined by the point where you know it's really working.

它取决于您知道它真正起作用的时间点。

And I know that's still not a bright line, but it's like once you know that there's so much more demand than you can handle, that's when you want to step on the gas and optimize for speed in a lot of ways.

我知道这仍然不是一条明确的界限，但就像一旦你知道需求量超出了你的处理能力，那就是你想要加大力度并以多种方式优化速度的时候。

But I think especially until then, it's important to be patient, be disciplined.

但我认为，尤其是在那之前，保持耐心和遵守纪律很重要。

Get the best people is always important, but speed becomes more important once you find the market opportunity, the market vacuum.

获得最优秀的人才总是很重要，但一旦找到市场机会、市场真空，速度就变得更加重要。

Lenny Rachitsky: I know you've started a couple companies in the past, much smaller scale.

Lenny Rachitsky：我知道您过去创办过几家公司，规模要小得多。

In this new role as CEO of this massive hyper growth company, what surprised you most about where you spend the time most or just what the role involves?

在担任这家大型高速增长公司的首席执行官的新职位中，您最惊讶的是什么是您在哪里花费最多时间或该职位涉及哪些内容？

Because a lot of people want to start companies dream about being in your shoes.

因为很多想创办公司的人都梦想着能站在你的立场上。

What are they maybe not understanding about where a lot of your time goes?

他们可能不明白你的大部分时间都花在哪里了？

Brendan Foody: Yeah, it's actually not too surprising.

Brendan Foody：是的，这实际上并不太令人惊讶。

The top two buckets are always working on hiring and time with customers of how do I really deeply understand what customers need and how we can support them?

最重要的两个部分始终致力于招聘和与客户相处的时间，即我如何真正深入了解客户的需求以及我们如何支持他们？

And then how do I build the team and a lot of the processes around that?

然后我如何建立团队以及围绕它的许多流程？

Of course, there's all of the ad hoc things I didn't expect of dealing with the people questions of how do we set up our levels and our comp bands and all of that, which you learn as you scale a business.

当然，在处理人们的问题时，有一些我没有想到的临时事情，比如我们如何设置我们的级别和我们的竞争范围等等，这些都是你在扩展业务时学到的。

But I think that the core places that I spend my time are in line with what I expected as well as what I love doing, which is very fortunate.

但我觉得我花时间的核心地方符合我的期望，也符合我喜欢做的事情，这是非常幸运的。

Lenny Rachitsky: So these two companies you've started in the past, maybe share what they work because they're fun, and then how do they help you be successful in this?

Lenny Rachitsky：所以你过去创办的这两家公司可能会分享他们的工作，因为它们很有趣，然后他们如何帮助你在这方面取得成功？

What's something that they taught you that helped you in your current role?

他们教给你的哪些东西对你目前的职位有帮助？

Brendan Foody: Yeah, so there's been like a dozen, but I'll choose my favorite two.

Brendan Foody：是的，大概有十几个，但我会选择我最喜欢的两个。

So when I was in eighth grade, I started Donut Dynasty where I saw that Safeway Donuts were selling for $5 a dozen, and I was amazed because I felt like as an eighth grader, this was such an incredible deal.

所以当我八年级的时候，我创办了《甜甜圈王朝》，在那里我看到西夫韦甜甜圈的售价为一打 5 美元，我很惊讶，因为我觉得作为一个八年级学生，这是一笔令人难以置信的交易。

And I started to bike down to Safeway, buy Safeway Donuts for $5 a dozen, and then go back to my middle school and then sell them for $2 each, running really good margins of course.

我开始骑自行车去西夫韦，以每打 5 美元的价格购买西夫韦甜甜圈，然后回到我的中学，然后以每个 2 美元的价格出售它们，当然利润非常高。

It sold out super quickly.

它很快就卖完了。

And so then I need to scale up.

所以我需要扩大规模。

So I would pay my mom $20 to drive me in her minivan down to Safeway, buy 10 dozen donuts, go to my middle school, sell them all out.

所以我会付我妈妈 20 美元，让她用她的小货车载我去西夫韦，买 10 打甜甜圈，去我的中学，把它们全部卖掉。

Brendan Foody: And then the school tried to shut me down because I was selling food on school campus, which they didn't like.

Brendan Foody：然后学校试图关闭我，因为我在校园里卖食物，他们不喜欢这样。

So they had me in the principal's office asking me to not do that.

所以他们把我带到校长办公室，要求我不要这样做。

And then I moved my donut stand over 50 feet, so it was off school campus, saying that they could no longer police me.

然后我把甜甜圈摊移到了 50 英尺以外，远离校园，说他们不能再警察我了。

I remember we had competitors pop up where the competitors were charging.

我记得我们在竞争对手充电的地方突然出现了竞争对手。

They bought these Chuck's Donuts, which if anyone in the Bay Area knows, are higher end donuts than Safeway Donuts, but they have a higher cost basis.

他们买了这些查克甜甜圈，如果湾区的人知道的话，这是比西夫韦甜甜圈更高端的甜甜圈，但它们的成本基础更高。

They cost a dollar per.

每张花费一美元。

And so I dropped my prices to $1 for two weeks to run them out of business before I knew what anti-competitive practices were.

因此，在我知道什么是反竞争行为之前，我将价格降至 1 美元，为期两周，让他们停业。

And I'd hire all my friends, paying my friends in donuts because they perceived the donuts as $2 each where they could sell them throughout the school and I could have a lower cost basis on them.

我会雇佣我所有的朋友，用甜甜圈付钱给我的朋友，因为他们认为甜甜圈每个 2 美元，他们可以在整个学校出售它们，而我可以以更低的成本购买它们。

Brendan Foody: So I had all of these fun experiences in selling donuts, and then I could talk more about my high school business as well, which was a more significant scale.

Brendan Foody：所以我在销售甜甜圈方面获得了所有这些有趣的经历，然后我也可以更多地谈论我的高中生意，这是一个更重要的规模。

But I think the takeaway from that was just like you can just do things.

但我认为从中得到的收获就是你可以做事。

So many people have ideas, but the barrier to more companies being built, I think, is just initiative and taking the steps to build the product or experience that customers want and investing the time and the ambition to scale that up.

很多人都有想法，但我认为，建立更多公司的障碍只是主动性和采取步骤来构建客户想要的产品或体验，并投入时间和雄心来扩大规模。

And so I think it was really getting reps of that that enabled me to realize that I should do it later on at a much larger scale.

因此，我认为正是获得了代表，才使我意识到我应该在以后以更大的规模来做这件事。

Lenny Rachitsky: Amazing story.

莱尼·拉奇茨基：精彩的故事。

I love how wholesome that is versus drugs, selling donuts.

我喜欢卖甜甜圈，这与毒品相比是多么有益健康。

Brendan Foody: Then my mom was very worried.

Brendan Foody：然后我妈妈非常担心。

She was like, "Oh, is there any pot of these donuts?" I was like, "No, mom, I assure you these are pure donuts."

她说，“哦，有这些甜甜圈吗？”我当时想，“不，妈妈，我向你保证这些都是纯甜甜圈。”

Lenny Rachitsky: I love that you paid your mom $20 to drive.

Lenny Rachitsky：我喜欢你付给你妈妈 20 美元让她开车。

Brendan Foody: Yeah.

布兰登·福迪：是的。

She was adamant it couldn't be a handout that she was taking her time to drive me, so she needed to make a little bit of money off of it.

她坚持认为她花时间开车送我不可能是施舍，所以她需要从中赚点钱。

We haggled over her title where eventually she wanted to be head of global operations, which we found very entertaining.

我们就她的头衔进行了讨价还价，最终她想担任全球运营主管，我们觉得这很有趣。

Lenny Rachitsky: I hope that's on her LinkedIn.

Lenny Rachitsky：我希望这是在她的 LinkedIn 上。

Brendan Foody: Not yet.

布兰登·福迪：还没有。

Maybe she'll have to add it.

也许她必须添加它。

Lenny Rachitsky: So you said that you've started a dozen companies?

Lenny Rachitsky：所以你说你已经创办了十几家公司？

Brendan Foody: Well, a dozen projects, but I think it was that, and then my AWS company were the two that I scaled up.

Brendan Foody：嗯，有十几个项目，但我认为就是这样，然后我的 AWS 公司就是我扩大规模的两个项目。

Lenny Rachitsky: What's the story behind Mercor as the name?

Lenny Rachitsky：Mercor 这个名字背后有什么故事？

Brendan Foody: Mercor means marketplace in Latin or to buy, sell, trade.

Brendan Foody：Mercor 在拉丁语中的意思是“市场”，意思是购买、出售、交易。

And we want to build the largest marketplace in the world, the marketplace for how everyone finds jobs, and that was really the draw to it.

我们希望建立世界上最大的市场，每个人如何找到工作的市场，这确实是它的吸引力。

Lenny Rachitsky: Okay, maybe a last question.

Lenny Rachitsky：好的，也许是最后一个问题。

This is going back to earlier in discussion because it's something I've been thinking about as we're talking.

这可以追溯到之前的讨论，因为这是我在我们谈话时一直在思考的事情。

There's been this shift from data as the fuel for models, and now it's experts.

这种转变已经从数据作为模型的燃料，变成了专家。

Do you think there's a next step, or is this just will take us to AGI, superintelligence?

你认为还有下一步吗，或者这只是将我们带向通用人工智能（AGI），超级智能？

Brendan Foody: I don't think it's necessarily changing from data to experts.

Brendan Foody：我认为这不一定会从数据转变为专家。

It's more just the paradigm of realizing that labs need this close collaboration with experts to help understand what are the evals that they're building and how can they push the frontier.

这更像是认识到实验室需要与专家密切合作的范例，以帮助了解他们正在构建的评估是什么以及他们如何推动前沿。

But I think it's very clear that evals are evergreen, that so long as we want to improve models, we'll need experts to create evals for them and to create the post-training data for them to learn those capabilities.

但我认为很明显，评估是常青的，只要我们想要改进模型，我们就需要专家为它们创建评估，并为他们创建训练后数据来学习这些功能。

And of course there might be changes in the exact way that people do training with RL or otherwise, but they will always need an eval to measure what does success look like across every domain that they want to build.

当然，人们使用强化学习或其他方式进行训练的确切方式可能会发生变化，但他们始终需要进行评估来衡量他们想要构建的每个领域的成功情况。

Lenny Rachitsky: Okay.

莱尼·拉奇茨基：好的。

So then building on that, a question that comes up a lot these days is, and I know we're talking about fun stuff but I'm getting to serious stuff again, scaling laws and just progression of model intelligence.

因此，在此基础上，这些天经常出现的一个问题是，我知道我们正在谈论有趣的事情，但我又开始讨论严肃的事情，缩放定律和模型智能的进步。

A lot of people are feeling like, "I don't know, it's slowing down.

很多人都感觉，“我不知道，它正在放缓。

We're not going to really get to superintelligence at this rate." What is your sense?

以这样的速度，我们不会真正实现超级智能。”你的感觉是什么？

Brendan Foody: I totally agree with that.

布伦丹·福迪：我完全同意这一点。

I know there's been some executives to big labs that say we'll have superintelligence in three years, but I think the truth is that it's a longer road.

我知道有些大型实验室的高管说我们将在三年内拥有超级智能，但我认为事实是这是一条更长的路。

And that's not to diminish from how extraordinary the models are.

这并不能掩盖这些模型的非凡之处。

I think we'll be able to automate a majority of knowledge work tasks in the next 10 years for sure, but that long road is paved with all of the evals that help to make those capabilities possible.

我认为我们肯定能够在未来 10 年内实现大部分知识工作任务的自动化，但这条漫长的道路是由所有有助于使这些功能成为可能的评估铺平的。

And it's not going to be 10X more pre-training data that gets those capabilities.

并且获得这些功能的预训练数据不会增加 10 倍。

It's much more going to be all of the post-training data sets that are far more data-efficient and thoughtful that help us get there.

更重要的是所有训练后数据集的数据效率更高、考虑更周全，可以帮助我们实现这一目标。

Lenny Rachitsky: David Sachs tweeted this interesting point that the situation we're now is almost the best case scenario where AI is not in this fast takeoff to superintelligence.

Lenny Rachitsky：大卫·萨克斯（David Sachs）在推特上发表了一个有趣的观点，即我们现在的情况几乎是人工智能尚未快速发展为超级智能的最好情况。

There's a lot of competitors keeping each other in check.

有很多竞争对手互相制约。

Models are already very valuable and only getting valuable, more valuable, but there's not just this winner superintelligence taking over the world situation.

模型已经非常有价值，而且只会变得越来越有价值，但不仅仅是这个胜利者超级智能接管了世界局势。

Brendan Foody: Yeah, I think that's true.

布伦丹·福迪：是的，我认为这是真的。

I think a lot of the super intelligence fearmongering is probably overrated, but at the same time a lot of people's framing around that is even if there is a 5 to 10% chance of this P-Doom, then we should be careful, which seems logical.

我认为很多关于超级智能的恐吓言论可能被高估了，但同时很多人的框架是，即使这种 P-Doom 的可能性有 5% 到 10%，那么我们也应该小心，这似乎是合乎逻辑的。

But I think that it's going to be an extraordinary 10 years for all of Silicon Valley and all of the world as this technology is able to create abundance and giving everyone better medical treatment, the best access to legal recommendations, and the ability to build great products more than we've ever seen before.

但我认为，对于整个硅谷和全世界来说，这将是一个非凡的十年，因为这项技术能够创造财富，为每个人提供更好的医疗、获得法律建议的最佳途径，以及打造前所未有的伟大产品的能力。

Lenny Rachitsky: And education feels like is transforming.

Lenny Rachitsky：教育感觉正在发生转变。

Brendan Foody: Absolutely, right.

布伦丹·福迪：当然，对。

I even have felt bits of this over the last 10 years where I remember ever...

在过去的十年里，我什至在我记忆中曾经有过这样的感受……

My parents would give me a hard time for not going to classes in college and I'd be like, well, there's way better lectures on YouTube.

我的父母会因为我不去大学上课而为难我，我会想，好吧，YouTube 上有更好的讲座。

Why not just listen there?

为什么不直接在那里听呢？

But I can only imagine as the models get extremely good at conveying information, better than the best professor, what that'll mean and access to all sorts of information to better forward humanity and upskill everyone.

但我只能想象，当模型变得非常擅长传达信息，甚至比最好的教授还要好时，这将意味着什么以及获取各种信息以更好地推进人类发展和提高每个人的技能。

Lenny Rachitsky: So I'll use that as a segue to a final question.

Lenny Rachitsky：所以我将以此作为最后一个问题的延续。

I'm going to take us to AI Corner, which is a recurring segment on the podcast.

我将带我们去人工智能角，这是播客上反复出现的部分。

What's some way that you personally use AI to do better work to help you in life?

您个人使用人工智能来更好地工作以帮助您生活的方式是什么？

Brendan Foody: Well, let's see.

布伦丹·福迪：好吧，让我们看看。

I use it a lot to write documents, as you would expect.

正如您所料，我经常使用它来编写文档。

I also talk to get advice on problems.

我也会就问题寻求建议。

I find it helpful to just reason through almost as a thought partner because, yeah, I don't know.

我发现几乎作为思想伙伴进行推理很有帮助，因为，是的，我不知道。

I find I think better sometimes when I'm talking something through, but I can't talk through everything with colleagues or people around me.

我发现有时当我把事情讲清楚时我会思考得更好，但我无法与同事或周围的人讲所有的事情。

Lenny Rachitsky: And so this is like ChatGPT Voice Mode mostly or something else.

Lenny Rachitsky：所以这主要类似于 ChatGPT 语音模式或其他模式。

Brendan Foody: Yeah, I like ChatGPT Voice Mode a lot.

Brendan Foody：是的，我非常喜欢 ChatGPT 语音模式。

There's stuff-

有东西-

Brendan Foody: ...

布伦丹·福迪：……

or room for improvement, but I am very excited about the future of Voice.

或还有改进的空间，但我对 Voice 的未来感到非常兴奋。

Lenny Rachitsky: Let me show you something I built, actually.

Lenny Rachitsky：实际上，让我向您展示我构建的一些东西。

I wasn't planning to talk about this, but there's this guy, Eric Antonow, who's been recommended by a lot of people to get him on this podcast.

我本来不打算谈论这个，但是有一个人，Eric Antonow，他被很多人推荐让他加入这个播客。

He's this creative product person that's under the radar now.

他是一位现在受到关注的创意产品人士。

He's at Facebook for a long time.

他在 Facebook 工作了很长时间。

He built this project called Pirate GPT, which is you basically put ChatGPT into a stuffed animal to talk to it.

他建立了这个名为 Pirate GPT 的项目，基本上就是将 ChatGPT 放入毛绒动物中并与其交谈。

So built a little wise owl.

于是就造了一只聪明的小猫头鹰。

I don't have it on right now.

我现在还没有戴上

Lenny Rachitsky: But basically you sew in a little speaker right here and you put a little magnet underneath and you can put it on your shoulder and then you just talk to it.

Lenny Rachitsky：但基本上你在这里缝了一个小扬声器，在下面放了一个小磁铁，你可以把它放在你的肩膀上，然后你就可以和它说话。

Brendan Foody: That's so cute.

布伦丹·福迪：太可爱了。

Wow.

哇。

I love it.

我喜欢它。

I'll have to get one of those.

我得买其中之一。

Because I have some of the voice assistants in my apartment, but I really want a ChatGPT voice assistant, so I'm excited for-

因为我的公寓里有一些语音助手，但我真的想要一个 ChatGPT 语音助手，所以我很兴奋 -

Lenny Rachitsky: I was just thinking that.

Lenny Rachitsky：我只是这么想。

Yeah, just come on.

是的，来吧。

Why can't we have a ChatGPT voice just sitting around listening to us all the time.

为什么我们不能有一个 ChatGPT 声音一直坐在旁边听我们讲话。

And you can't on your phone because it goes to sleep and it's like, "Hello, what?"

你也不能打电话，因为它会进入睡眠状态，然后你会问：“喂，什么？”

Lenny Rachitsky: Yeah, so it's what this is trying to be.

Lenny Rachitsky：是的，这就是我们想要的。

Well, there's a kickstarter he started that we'll link to that.

嗯，他启动了一个 kickstarter，我们将链接到该启动器。

You could help out.

你可以帮忙。

That's really easy.: Brendan, is there anything else that you wanted to share or touch on or maybe leave listeners with before we get to a very exciting lighting round?

这真的很简单。：Brendan，在我们开始令人兴奋的灯光回合之前，您还有什么想分享或触及的，或者可能留给听众的吗？

Brendan Foody: Tying to the point around initiative and that you can just do things, I encourage everyone, especially with AI and it being so much easier to build, just take the initiative to go out and build products and talk with customers and take that leap of faith because I think that that is in so many ways, the largest barrier to more innovation, the economy in any way that we can support that.

Brendan Foody：围绕主动性和你可以做的事情这一点，我鼓励每个人，尤其是人工智能，而且它更容易构建，只要主动走出去，构建产品，与客户交谈，并采取信仰的飞跃，因为我认为，从很多方面来说，这都是更多创新的最大障碍，无论我们以何种方式支持经济。

Lenny Rachitsky: Yeah.

莱尼·拉奇茨基：是的。

There's so many people that just, let's not bash the podcast, but just listen to podcasts, read posts, just keep reading and listening and don't do anything with that information.

有这么多人，我们不要抨击播客，而只是听播客，阅读帖子，继续阅读和聆听，不要对这些信息做任何事情。

And there's never been an easier time to actually build stuff and try stuff.

现在是真正构建和尝试东西的更容易的时刻。

Lenny Rachitsky: So definitely take that advice.

Lenny Rachitsky：所以一定要采纳这个建议。

Just you can do things.

只要你能做事。

You can move your donut stand 50 feet and get out of their jurisdiction.

您可以将甜甜圈摊移至 50 英尺并脱离他们的管辖范围。

Lenny Rachitsky: Okay, Brendan, with that, we've reached a very exciting lightning round.

Lenny Rachitsky：好的，Brendan，我们已经进入了非常激动人心的闪电回合。

I've got five questions for you.

我有五个问题要问你。

Are you ready?

你准备好了吗？

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

Lenny Rachitsky：您最常向其他人推荐的两三本书是什么？

Brendan Foody: Let's see.

布伦丹·福迪：让我们看看。

I would say in order, High Output Management is a phenomenal book on running companies.

我想说，《高产出管理》是一本关于经营公司的非凡书籍。

Second is Zero to One, which of course is a classic.

第二部是《从零到一》，这当然是经典。

And then third is Shoe Dog, where I just find it to be a really inspirational story.

第三个是《Shoe Dog》，我发现这是一个非常励志的故事。

Lenny Rachitsky: What is a recent movie or TV show you really enjoyed?

Lenny Rachitsky：您最近真正喜欢的电影或电视节目是什么？

Brendan Foody: I really liked Oppenheimer.

布伦丹·福迪：我真的很喜欢奥本海默。

My favorite TV show of all time is Suits, so I know not recent, but if I had to choose a recent one, probably Oppenheimer.

我一直以来最喜欢的电视节目是《金装律师》，所以我知道不是最近的，但如果我必须选择最近的一部，可能是奥本海默。

Lenny Rachitsky: Very cool.

莱尼·拉奇茨基：非常酷。

Suits, first time someone's mentioned that.

西装，第一次有人提到。

Favorite product you recently discovered that you really love?

您最近发现自己真正喜欢的最喜欢的产品？

Brendan Foody: I love using Codex, like the new version.

Brendan Foody：我喜欢使用 Codex，就像新版本一样。

I know it's sort of new in terms of version.

我知道它的版本有点新。

Yeah, I think it's incredible and just a huge, huge improvement.

是的，我认为这是令人难以置信的，而且是一个巨大的进步。

So yeah.

所以是的。

Lenny Rachitsky: Do you have a life motto that you find yourself coming back to, sharing with folks, finding useful in work or in life?

Lenny Rachitsky：您是否有一个人生座右铭，您发现自己会回想起它，与人们分享，并发现它在工作或生活中有用？

Brendan Foody: I think it's you can just do stuff, what we were talking about earlier.

Brendan Foody：我认为你可以做一些我们之前讨论过的事情。

Take the leap of faith.

带着信念的飞跃。

Lenny Rachitsky: I thought you were going to say can do, which is in your Twitter profile.

Lenny Rachitsky：我以为你会说“可以”，这在你的 Twitter 个人资料中。

Lenny Rachitsky: Two great ones.

Lenny Rachitsky：两个很棒的。

Final question.

最后一个问题。

So we were chatting before this about things that we could talk about and you shared this interesting thing that you haven't shared anywhere else, which is that you're dyslexic.

所以我们之前聊过一些我们可以谈论的事情，你分享了一件你在其他地方没有分享过的有趣的事情，那就是你有阅读障碍。

Why don't you share that with folks?

你为什么不与人们分享呢？

And just how do you get around that having built the fastest-growing company in history?

在建立了历史上增长最快的公司之后，你是如何解决这个问题的呢？

Brendan Foody: I don't hide it at all.

Brendan Foody：我根本不隐瞒这一点。

I think a lot of my colleagues know.

我想很多同事都知道。

And I think on one hand it definitely makes it difficult to go through 1,000 emails a day or read every document that I'm supposed to, but on the other hand, I feel like it helps me to think a little bit differently, to be more creative, and perhaps see that markets are changing that not everyone sees.

我认为，一方面，这确实让我很难每天浏览 1,000 封电子邮件或阅读我应该阅读的每一份文档，但另一方面，我觉得这有助于我以不同的方式思考，变得更有创意，也许会看到市场正在发生变化，但并非每个人都能看到。

And so it's turned out okay so far.

到目前为止结果还不错。

And so I think one thing it's helped me realize from a management standpoint is that we focus much more on how we can leverage people's strengths rather than helping to improve weaknesses, because there's some things that I'm not great at and I'll never be the best in the world at, and there's others that I can hopefully refine and strive to be.

所以我认为，从管理的角度来看，它帮助我认识到一件事，那就是我们更多地关注如何利用人们的优势，而不是帮助改善劣势，因为有些事情我并不擅长，而且我永远不会成为世界上最好的，而其他一些事情我可以希望改进并努力成为。

Lenny Rachitsky: That's such a also recurring theme on this podcast of just focusing on strengths and not focusing over all your focus on weaknesses.

Lenny Rachitsky：这也是这个播客中反复出现的主题，即只关注优势而不是全部关注劣势。

Lenny Rachitsky: Brendan, this was incredible.

莱尼·拉奇茨基：布伦丹，这太不可思议了。

I learned so much.

我学到了很多。

I have a billion more questions, but you got shit to do.

我还有十亿个问题，但你还有很多事要做。

Two final questions.

最后两个问题。

What should people know about what you're doing and roles you're hiring for?

人们应该了解您正在做的事情以及您正在招聘的职位的哪些信息？

And then how can listeners be useful to you?

那么听众如何对你有用呢？

Brendan Foody: Absolutely.

布伦丹·福迪：当然。

We're hiring a ton across the board on our team.

我们正在全面招聘大量员工。

We're hiring strategic project leads on our operations team, software engineers in our engineering team, as well as researchers.

我们正在招聘运营团队的战略项目主管、工程团队的软件工程师以及研究人员。

And so please go to mercor.com and we would love to work with you, and that's the largest way that you can help us.

因此，请访问mercor.com，我们很乐意与您合作，这是您可以帮助我们的最大方式。

Share it with your friends as well.

也与您的朋友分享。

Over half of people in our marketplace come from referrals because we have a platform of people that love us.

我们市场上超过一半的人来自推荐，因为我们有一个由爱我们的人组成的平台。

And so any jobs that you want to apply to or send your friends to, we would love to have you.

因此，如果您想申请或派遣您的朋友去从事任何工作，我们都欢迎您的加入。

Lenny Rachitsky: Brendan, thank you so much for joining me.

Lenny Rachitsky：布伦丹，非常感谢你加入我。

Bye, everyone.: Thank you so much for listening.

再见，大家：非常感谢您的聆听。

If you found this valuable, you can subscribe to the show on Apple Podcasts, Spotify, or your favorite podcast app.

如果您发现这很有价值，您可以在 Apple Podcasts、Spotify 或您最喜欢的播客应用程序上订阅该节目。

Also, please consider giving us a rating or leaving a review as that really helps other listeners find the podcast.

另外，请考虑给我们评分或留下评论，因为这确实有助于其他听众找到播客。

You can find all past episodes or learn more about the show at lennyspodcast.com.

您可以在 lennyspodcast.com 上找到所有过去的剧集或了解有关该节目的更多信息。

See you in the next episode.

下一集见。