AI Interview Wiki
主题综述

专才角色坍缩为通才构建者 · The Collapse of Specialist Roles into the Generalist Builder

主题综述

更新日志

主流共识

同一个起点:过去十五年的"碎片化"被认为是要被逆转的历史偶然,不是自然秩序。 Ryo Lu 把这条时间线讲得最直白:

"Over the last, I don't know, 15 years or so, I think the art of making software fragmented a lot and then we kind of split into different roles. Each role kind of used their own tool, used their own artifact. They think in their own kind of words and lingo... But with AI, with Cursor, things kind of flip again."
「在过去的,我不知道,15 年左右,软件制作的艺术变得非常分散,然后我们分成了不同的角色。每个角色都使用自己的工具,使用自己的工件。他们用自己的语言和行话思考……但是有了 AI,有了 Cursor,情况又会发生转变。」
Ryo Lu · Ryo Lu (Cursor): AI Turns Designers to Developers

他把碎片化归因于成本与流程优化,而非产品本身的需要——因此是可逆的:

"Now it's like you kind of break everything down. You try to optimize them with like processes and cost optimizations and people's become like... boxed into little areas and problem sets when the whole thing is actually all together. And that causes a lot of problems, I think."
「而现在你却把所有东西都分解了。你试图用流程和成本优化来优化它们,人们变得像……被框在小区域和问题集中一样,而实际上整个东西是联系在一起的。我认为这导致了很多问题。」
Ryo Lu · Ryo Lu (Cursor): AI Turns Designers to Developers

边界模糊被普遍认为"在 AI 之前就开始了",AI 只是加速器。 Figma 的 Yuhki Yamashita 说得几乎是同一个意思(此篇逐字稿仅中文):

「这种界限的模糊甚至在人工智能出现之前就已经发生了,比如设计师学习如何编码,或者产品经理学习一些设计……很多这种情况的发生是因为人们正在被推到更高的层次。工程师不再需要了解底层发生的所有事情。他们可以使用一堆强大的库和 API,并且仍然可以非常高效。设计师们也很少再进行像素级的操作……因此,你们都在朝着一个相似的领域前进,在那里你可以专注于最具战略意义的事情。」(逐字稿仅中文)
— Yuhki Yamashita · 20Product: Figma CPO on How Figma Builds Products

大家共享一个机制描述:人从"生产制品"上移一层。 工程师这一侧,Legora 的 Jacob Lauritzen 和 Claude Code 的 Boris Cherny 描述几乎重合。Jacob:

"The job of an engineer is changing from typing a bunch of code to sort of one layer above it, which is, what does the system look like? ... you sort of have engineers thinking about the higher level one abstraction above."
「工程师的工作正在从输入大量代码转变为更高一个层次,即系统看起来是什么样的?……你希望工程师思考更高层次的抽象。」
Jacob Lauritzen · 20Product: Inside Legora's Tech Stack

Boris 把上移后剩下的"有趣部分"点名成了跨学科的活:

"The fun part is figuring out what to build and coming up with this. It's talking to users. It's thinking about these big systems. It's thinking about the future. It's collaborating with, you know, other people on the team... And that's what I get to do more of now."
「有趣的部分是弄清楚要构建什么,并提出这个想法。是与用户交谈。是思考这些大型系统。是思考未来。是与团队中的其他人协作。这就是我现在可以做更多的事情。」
Boris Cherny · Head of Claude Code: What happens after coding is solved

这个"抬升"第一次有了刻度——顺带有了天花板。 八人里唯一给这个抬升标出刻度的是 Julie Zhuo(前 Facebook 设计 VP,现 Sundial CEO):AI 把人从 10 百分位抬到 60-70 百分位,但她先自己压了上限——到不了顶部(「现在，我不会以博士或最高的 1％、10％的水平来做。」):

"But if I was, you know, at the zero or 10th percentile, it can certainly get me even today, very quickly up to like the 60th, 70th in terms of what the state of the art is. I think that unlocks so many doors. The main thing that I felt so excited about and this is something I tell my team all the time is we need to dissolve the boundaries of these traditional roles. In the past, again, we would have a traditional team, engineers, product manager, designer, researcher, data scientist. And I think now the teams can look more like, well, it's just two people."
「但是如果我处于 0 或第 10 个百分位，它肯定可以在今天非常迅速地将我提升到 60％、70％的水平，就目前的技术水平而言。我认为这打开了很多扇门。我感到非常兴奋的主要事情，也是我一直告诉我的团队的是，我们需要打破这些传统角色的界限。过去，我们有一个传统的团队，工程师、产品经理、设计师、研究员、数据科学家。我认为现在团队可以看起来更像是，嗯，只有两个人。」
Julie Zhuo · From managing people to managing AI (Julie Zhuo)

这个 60-70 分位的刻度对后面所有分歧都关键:它同时支撑"人人变通才"(底座是真的)和"专才不消失"(顶部那截没被吃掉)。

还有一个反复出现的共同信念:上移后,区分人的不是产出量,而是"品味/观点"——AI 给基线,人给取舍。 这一点 Ryo 和 Jacob 用了几乎一样的意象("AI slop"/"grayness"):

"What is good is like the LLM, because it has seen a lot of things, it can do the baseline really fast and really good. Then the thing on top is taste or like your self-selection of what is good... There needs to be something for the human to specify what is good, what is right, how I want to do it. If you don't put in that opinion, it will just produce AI slop."
「LLM 的优点是,因为它已经看到了很多东西,它可以非常快速和出色地完成基线工作。然后最重要的是品味,或者说你对什么是好的自我选择。……需要人为地指定什么是好的,什么是正确的。如果你不加入这种观点,它只会产生人工智能垃圾。」
Ryo Lu · Ryo Lu (Cursor): AI Turns Designers to Developers
"If you don't have taste, then you let AI slop, converge to sort of grayness and everything looks the same and everything's just like, you need to have taste to have sort of an opinion and a stance in the world."
「如果你没有品味,你会让人工智能的混乱趋于灰暗,一切看起来都一样……你需要品味才能在这个世界上拥有观点和立场。」
Jacob Lauritzen · 20Product: Inside Legora's Tech Stack

这个信念在新的三篇里不但复现,还升级成了"瓶颈"语言——Akshay Nathan(OpenAI 核心产品工程负责人,ChatGPT Work 的操盘者)把品味与"想法"并列为唯二稀缺物:

"Bottlenecks becomes sort of like ideas and taste, I guess. I think because anyone can build now, I think it really is the era of bottoms-up ambition. And because there's so much to be built, you're always going to be bottlenecked by the amount of ideas and the amount of things that you're doing at any given time."
「瓶颈变得有点像想法和品味，我想。我认为因为任何人现在都可以构建，所以这确实是自下而上的雄心时代。而且因为有那么多东西要构建，你始终会受到想法数量和同时进行的事情数量的瓶颈限制。」
Akshay Nathan · OpenAI's Vision for the AI Super App (Akshay Nathan)

最后一条新共识是范围外推:坍缩不止于产品/设计/工程三件套。 Akshay 把同一句式推广到整个知识工作——金融、营销、研究同样"突然能构建":

"And so when I look at knowledge work, I think about people are no longer siloed by their roles. They're no longer siloed by maybe the background or training that they have. Like no matter what function you're in, you can suddenly build things. You can suddenly get access to data that you otherwise might not be able to interpret, et cetera."
「所以当我看待知识工作时，我考虑到人们不再被他们的角色所孤立。他们不再被他们的背景或培训孤立。不管你在什么职能中，你都能突然构建东西。你能突然获得你以前可能无法解读的数据，等等。」
Akshay Nathan · OpenAI's Vision for the AI Super App (Akshay Nathan)

分歧在哪

主题定义里那句"坍缩、融合成一类端到端独立构建软件的通才构建者",在这八个人嘴里其实是两种不同的未来。表面上都在说"边界模糊",但对"坍缩到一个人 vs. 保留专才、去分化到什么程度、这样做是否高效"分歧很硬。

一、"一个人做全部" vs. "别让他做全部——机会成本太高"

最尖锐的对撞在 Cursor/Andreessen 一侧与 Legora 一侧之间。

Andreessen 把三个角色描述成一场"墨西哥僵局",结论是最有才的人会"三样全能",而且这正是最有价值的方向:

"There's like a Mexican standoff happening between those three roles. Every coder now believes they can also be a product manager and a designer because they have AI. Every product manager thinks they can be a coder and a designer, and then every designer knows they can be a product manager and a coder. They're actually all kind of correct."
「这三个角色之间就像墨西哥僵局一样。现在每个程序员都认为他们也可以成为产品经理和设计师,因为他们拥有人工智能。每个产品经理都认为他们可以成为程序员和设计师,然后每个设计师都知道他们可以成为产品经理和程序员。实际上他们都或多或少是正确的。」
Marc Andreessen · Marc Andreessen: The real AI boom hasn't even started yet
"The talented people in any of those roles become superpowered and they become good at doing all three of those things. And then those people become incredibly valuable because then those are people who can actually build and design new products right from scratch, which is the most valuable thing."
「任何这些角色中的有才能的人都会变得超级强大,并且他们会擅长做所有这三件事。然后这些人变得非常有价值,因为这些人实际上可以从头开始构建和设计新产品,这是最有价值的事情。」
Marc Andreessen · Marc Andreessen: The real AI boom hasn't even started yet

Jacob 把同一句"产品和工程正在融合成一个人"直接点名,然后反对——不是说做不到,而是说不划算:

"A lot of people are saying that product and engineering are converging. It's becoming one thing. It's like one person can do the product work and build the system and ship it and everything. And I think for some companies that's true. I think for companies where you really need PMs, it's not true, or it can be true, but it's inefficient."
「很多人认为产品和工程正在融合。它正在变成一件事情。就像一个人可以做产品工作,同时建立系统并发布它。我认为对于某些公司来说,确实如此。对于那些真的需要 PM 的公司来说不是,或者可能是,但效率很低。」
Jacob Lauritzen · 20Product: Inside Legora's Tech Stack
"You don't want your product people to do engineering because the opportunity cost of that is really high... if your PMs are coding a lot, if they're spending 50% of their time coding, we're missing out on so much product work."
「你不希望你的产品人员去做工程工作,因为机会成本非常高……如果你的 PM 花费 50% 的时间在编码上,我们将失去太多产品工作。」
Jacob Lauritzen · 20Product: Inside Legora's Tech Stack

这条裂缝现在第一次有了"两边都动过手"的对照。 Julie Zhuo 站到了 Jacob 的对面——同样讲效率,结论相反:Sundial 直接砍掉了 PM 角色,理由是专职角色会诱发"默认外包",砍掉之后所有权才回到工程师身上:

"So, we have eliminated more roles. For example, we thought we would need a bunch of product managers. It's turned out that actually if you don't have a product manager, …when you have like a designer or a product manager and let's say I'm an engineer, Then when I have a problem like, oh, I need to figure out the product definition, my default will be, well, I've got these people and that's kind of their job description, so I'm just going to delegate that to them. …It allowed everyone to be like, oh, wait, we don't have a product manager on the team. Okay, so communication's up to me. Figuring out how we get greatest value to users is something that is now strictly in my charter."
「所以，我们已经取消了更多的角色。例如，我们认为我们需要很多产品经理。事实证明，如果没有产品经理，……当你有一个设计师或产品经理，假设我是一名工程师，那么当我遇到问题时，比如，我需要弄清楚产品的定义，我的第一反应是，我有人可以做这件事，这在某种程度上是他们的工作职责，所以我会直接把这件事委托给他们。……这让每个人都觉得，等等，我们团队里没有产品经理。好的，所以沟通的责任就落在我身上了。弄清楚如何为用户创造最大价值，现在完全是我的职责。」
Julie Zhuo · From managing people to managing AI (Julie Zhuo)

而且她诚实报告了代价曲线——短期确实变慢(这正是 Jacob 说的效率损失),但她认为是回本的投资,并给专才留了按需召入的口子:

"You have to invest a little bit in the beginning, so people are not as comfortable. They have to learn. So, initially, things take longer. …But in the long run, that investment really pays off because now you have a lot more people who are, again, a little more well-rounded and can take on many more pieces just on their own. And then in specific scenarios, oh, this is super front-end heavy, sure, let's still bring in somebody who is more specialized in that particular skill."
「你必须在开始时投入一些，所以人们不会那么舒服。他们需要学习。所以，最初，事情会花费更长的时间。……但从长远来看，这项投资真的会得到回报，因为现在你有更多的人，他们更加全面，可以独自完成更多的工作。然后在特定的情况下，哦，这是超级前端重的，当然，我们仍然可以找一个更擅长特定技能的人。」
Julie Zhuo · From managing people to managing AI (Julie Zhuo)

注意口径差异:Andreessen 讲的是能力上的"能"(你能变全能,所以你就该变全能,那是最稀缺、最不可替代的组合);Jacob 讲的是分工上的"该不该"(能是能,但把稀缺的产品脑力浪费在写代码上是错配)。前者要"坍缩成一个人",后者要"上移各自的一层,但保持分工"。这是本主题里最真实的裂缝——被"边界模糊"这层共同措辞盖住了。Julie 的加入把口径再推进一格:从"能不能/该不该"到"试过了"——她承认了 Jacob 说的短期效率损失,但主张它是投资而非错配。两人可能都对:Jacob 自己留了活口("对于那些真的需要 PM 的公司来说不是"),Julie 的公司是几乎全员工程师的小型创业公司。这场对撞的真实变量恐怕不是谁对谁错,而是公司规模与产品复杂度——但至此,裂缝两侧都有了一线实操者,不再是"愿景家 vs 一线"。

二、坍缩成"通才一个人",还是坍缩成"通才 + 专才的更好搭配"

即使在同意"融合"的人内部,落点也不一样。

Ryo 的落点是"软件构建者"这个单一身份,专长退化为个人色彩,而不是职能墙:

"I think people will always have their strengths or their unique special skill or some spike... But I think of all of these people as just like their software builders."
「我认为每个人都有自己的优势、独特的技能或特长……但我认为所有这些人就像软件构建者一样。」
Ryo Lu · Ryo Lu (Cursor): AI Turns Designers to Developers

Figma 的 Yuhki 明确设想的却是"通用产品构建角色 + 一群把专长做到更深的专家"两层结构,而且他强调"往两个极端推"反而是造好产品的必要张力(此篇逐字稿仅中文):

「由于边界的模糊,很容易想象它们可能会更多地融入到一种通用的产品构建角色中,同时总会有一些专家可以更深入、更高水平地发挥他们在各自领域的技能。但即便如此,我还是会说,总有一群人推动不同极端,这是非常有益的……我认为这种紧张关系是推动产品进步的原因。」(逐字稿仅中文)
— Yuhki Yamashita · 20Product: Figma CPO on How Figma Builds Products

Andreessen 的落点是"T 型/侧 E 型"——深在一处、横跨若干处,而且明确说这不是把人抹平,而是让稀有组合更值钱:

"If you think about the T configuration, the top of the T is like how many individual domains are you familiar enough with to be able to use the AI tools... and then this part of the T is how deep can you go in at least one of those domains so that you really, really deeply know what you're doing. If you're super deep on coding and you can use AI to do design and you can use AI to do product management, that's your T right there."
「如果你思考一下 T 型结构,T 字的顶部就像是你熟悉多少个独立的领域……然后 T 字的这一部分代表你至少在一个领域能深入到什么程度。如果你非常精通编码,又能用 AI 做设计,还能用 AI 做产品管理,这就是你的 T 字型技能。」
Marc Andreessen · Marc Andreessen: The real AI boom hasn't even started yet

Akshay 用的是同一个 T 字母,但把它平民化了——Andreessen 说的是"有才能的人"变超强,Akshay 说的是"每个人"都会 T 型,而且当场拿自己作证(他自认没有视觉品味,照样靠迭代出设计):

"I think my suspicion is that there's everything, everyone will be like T-shaped in a way and that like AI will enable everyone to become a generalist. Like, you know, things that like I never would be able to like come up with a design before. And like even now, I don't have maybe like the visual taste required, but I can iterate on something with the help of AI. But then people will have a specialty and that's like the straight line in the T or the upward line in the T. And so you can have a specialty that you're interested in and with the help of AI, you can go deeper and become better at it over time, but then you'll also be a generalist."
「我想我的猜测是，届时每个人在某种程度上都会是 T 形的，AI 将使每个人都能成为通才。就像，以前我永远无法想到怎么设计。现在，可能即使我没有所需的视觉品味，但依然可以在 AI 帮助下对某个东西进行迭代。但是人们会有专业领域，这就像 T 的竖线或 T 的上升线。所以你可以有自己感兴趣的专业领域，并在 AI 的帮助下深入并随着时间的推移变得更擅长，但同时你也会是一个通才。」
Akshay Nathan · OpenAI's Vision for the AI Super App (Akshay Nathan)

Julie 的落点比 Ryo 更进一步:builder 不只是身份,应该直接成为头衔——而且她是八人里唯一已经按这个落点改组过公司的:

"But the key thing is they can now use AI to help themselves do a lot of the things that the other folks used to be able to do. So in some ways, we can drop all of these different role distinctions and call ourselves builders. I think that's sort of the most general purpose way of thinking about what we can all be. We can all be builders. We can all be builders. And I love first to get to the world where that's just that's like the title."
「但关键是他们现在可以使用 AI 来帮助自己完成很多其他人过去能够做的事情。所以在某些方面，我们可以放弃所有这些不同的角色区分，称自己为建设者。我认为这是思考我们所有人可以成为什么的最通用的方式。我们都可以成为建设者。我们都可以成为建设者。我很乐意首先进入一个世界，在那里这就像是头衔一样。」
Julie Zhuo · From managing people to managing AI (Julie Zhuo)

不过她随即把"角色"与"技能"拆开——角色区分可以扔,技能上的强弱仍然要认(「我并不是说每个人都必须做所有事情。……但这更多的是关于我们所处的具体环境，而不是关于角色。」)。

Boris 的落点最接近"消解职称": "software engineer" 这个头衔可能直接消失。

"I think in some places the title software engineer is going to start to go away and it's just going to be replaced by builder or maybe it's just everyone's going to be a product manager and everyone codes or something like this."
「我认为在某些地方,'软件工程师'这个头衔将会消失,取而代之的是'构建者',或者每个人都将成为产品经理,并且每个人都进行编码,诸如此类。」
Boris Cherny · Head of Claude Code: What happens after coding is solved

但同一段里 Boris 又说短期内三个职能仍在,重叠约 50%,并非全并成一个:

"I think in the short term, it'll persist. But one thing that we're starting to see is there's maybe a 50% overlap in these roles, where a lot of people are actually just doing the same thing. And some people have specialties."
「我认为短期内会持续存在。但我们开始看到的一个现象是,这些角色之间可能存在 50% 的重叠,很多人实际上在做同样的事情。并且有些人有专长。」
Boris Cherny · Head of Claude Code: What happens after coding is solved

Bret Taylor 给这条谱系添了一个此前没有的叙事方向:通才不是 AI 造出的新物种,而是一直存在、被规模化组织边缘化的老物种——AI 让他们翻身。 他先描述这类人过去的下场:

"And as companies grow, you tend towards more specialization, you know, just because the person who's sort of the Jack or Jill of all trades ends up sort of not fitting in, you know, like there's not really a place for them because, okay, well, you're not really the deepest engineer. You're not really the best designer. You're not really a product manager. If you've been at the company for a while, we'll give you an honorary something to do. …I actually think those people who often exist in early-stage startups are often the people who get sidelined, but actually in a way that actually really harms the company."
「而且随着公司发展，你会趋向于更专业化，因为那种什么都会的人最终会发现自己不适合，因为好像没有他们的位置。好吧，你又不是最厉害的工程师。你真的不是最好的设计师。你真的不是产品经理。如果你在公司待了一段时间，我们会给你一个荣誉职位。……我确实认为那些经常出现在早期创业公司的人，往往是被边缘化的人，但实际上，这种方式确实对公司造成了损害。」
Bret Taylor · Bret Taylor of Sierra on AI agents, outcome-based pricing, and the OpenAI board

然后给翻身后的这类人开价——单项技能都"相对好找",稀缺的是组合,组合值千倍:

"We might invent one who has taste, but didn't necessarily know CSS. Who has infrastructure ability, meaning that you understand the basics of distributed systems and debugging, and you understand your customer very deeply. With the presence of CodeX, you can produce amazing results. Those people are truly worth a thousand X other people because it's relatively easy to find someone who's a great infrastructure engineer. Not easy, easy, but like relatively. Finding someone with good taste, that's, that's relatively easy. Finding someone who also understands your customers extremely well, like the nuances of the problem they're solving."
「我们可能会创造一个有品味，但不一定懂 CSS 的人。拥有基础设施能力，意味着你了解分布式系统和调试的基础知识，并且你非常了解你的客户。凭借 CodeX 的存在，你可以产生惊人的结果。这些人真的价值千倍于其他人，因为找到一个伟大的基础设施工程师相对容易。不是很容易，而是相对容易。找到一个有品味的人，这相对容易。找到一个非常了解你的客户的人，了解他们正在解决的问题的细微之处。」
Bret Taylor · Bret Taylor of Sierra on AI agents, outcome-based pricing, and the OpenAI board

当 John Collison 说"我们需要发明一种新的工作角色"时,Bret 给出的名字是 "hyper generalists"(「我的意思是，超级通才。」)。注意他的版本天然是精英叙事:值千倍的是稀缺组合的少数人,专才仍是"相对好找"的基线人群——这与 Akshay 的"人人 T 型"共用词汇,却不是同一个社会。

把这八个人排在一起,坍缩的"终态"从最激进到最保守依次是:Boris(头衔消失/人人是会编码的 PM)→ Julie(builder 直接当头衔,且已动手砍掉 PM——唯一进入实施的)→ Ryo(统一身份"软件构建者")→ Akshay(人人 T 型:通才底座+各自纵深)→ Andreessen(个人层面的 T/E 全能人)→ Bret(超级通才是值千倍的少数人,专才仍是基线)→ Yuhki/Jacob(职能仍在、只是各自上移一层,分工保留)。他们说的是同一个方向,但不是同一个目的地。

三、"什么被留给人"其实指向不同的技能

Boris 版的通才是跨学科的好奇心+能想更大的问题——落在工程师身上是"混合型":

"So some of the strongest engineers are hybrid product and infrastructure engineers, or product engineers with really great design sense, and they're able to do design also. Or an engineer that has a really good sense of the business and can use that to figure out what to do next... they're generalists, and they cross over multiple disciplines and can think about the broader problem they're solving, rather than just the engineering part of it."
「一些最强大的工程师是混合型产品和基础设施工程师,或者具有出色设计意识的产品工程师……或者是一位对业务有很好理解的工程师……他们是通才,他们跨越多个学科,能够思考他们正在解决的更广泛的问题,而不仅仅是工程部分。」
Boris Cherny · Head of Claude Code: What happens after coding is solved

Jacob 版留给人的不是"跨学科好奇心",而是一个新的专门工种:meta-engineering(他说这是他们"很快会有的一个明确角色")——把人重新聚焦到"让 agent 有效"这件很窄很硬的事上。这其实和"通才化"方向相反:它是在坍缩之后长出的新专才。

"Kind of the meta-engineering of making agents really effective... How do we make sure that we can enable agents to independently self-improve the system?"
「有点像是使代理真正有效的元工程……我们如何确保能够让代理独立地自我提升系统?」
Jacob Lauritzen · 20Product: Inside Legora's Tech Stack

Andreessen 版留给人的是编排 agent 的能力本身,并且他把这件事往上再推一层——最后连"公司"都可能坍缩成一个人:

"Whatever that job is called, who even knows what it's going to be? But it's going to be incredibly important because the people doing that job are going to be orchestrating the AI."
「不管那份工作叫什么,谁知道它会是什么样子?但它将非常重要,因为做那份工作的人将会协调人工智能。」
Marc Andreessen · Marc Andreessen: The real AI boom hasn't even started yet

新加入的两位给出第四种答案:留给人的是反馈回路里贴着客户的那一环。 Bret 把通才的核心资产点名为"带着能动性理解客户需求",并把它加码成一种近乎性格的东西("It's like care, period."「而是关心，一直关心。」):

"And I'm hopeful that in a world of AI agents, those generalists who, again, I think the most important part is understanding the customer need with agency, no pun intended, and empowerment can end up more powerful in the Silicon Valley company."
「我希望在人工智能代理的世界里，那些通才，我再次强调，我认为最重要的是理解客户的需求，并具有能动性，这绝不是双关语，并且有能力在硅谷公司中变得更强大。」
Bret Taylor · Bret Taylor of Sierra on AI agents, outcome-based pricing, and the OpenAI board

Akshay 的答案同构——被问到"LLM 为什么带不来新想法"时,他说想法不凭空来,通才的价值在闭环:

"One interesting part of that idea is they're not in a vacuum. They usually come from somewhere. In product development, they're coming from talking to users or reacting to friction that you're seeing or feedback, building on some foundation that you already have planned out before, whatever. And so I think that's where I think there will always be value in these generalists that we talked about. Closing that loop and coming up with those ideas that are grounded in that feedback or talking to users, whatever it is."
「那个想法的一个有趣部分是，它们并不是孤立存在的。它们通常来自某个地方。在产品开发中，它们来自与用户的对话，或对你所看到的摩擦或反馈的反应，基于你以前已经规划好的某些基础。所以我认为这就是我认为总会有价值的那些我们讨论过的通才的地方。完善这个循环，并提出那些基于反馈或与用户对话的想法，无论是什么。」
Akshay Nathan · OpenAI's Vision for the AI Super App (Akshay Nathan)

同一现象("代码变便宜、人上移一层"),四种截然不同的处方:变成好奇的通才(Boris)、变成一种新的窄专才 meta-engineer(Jacob)、变成 agent 的编排者甚至一人公司(Andreessen)、变成贴着客户闭环的高能动性通才(Bret/Akshay)。第四种离 Boris 最近,但重心从"跨学科"挪到了"贴客户"——留给人的不是更宽的技能面,而是反馈回路里最难自动化的那一段。

四、设计这一侧:是"设计师终于能发货",还是"设计被跳过、只剩品味"

Ryo 的叙事是赋能——设计师终于能自己造出活的东西,把过去在 Figma 里"死掉"的稿子变成产品:

"I made mocks and prototypes and I waited for them to happen for like years and they don't happen or like they ended up shipping as like a YouTube video. That's crazy. But, you know, with this new tool, a designer can build."
「我制作模型和原型,然后等待它们实现,等了好几年都没实现,或者最终以 YouTube 视频的形式发布。太疯狂了。但有了这个新工具,设计师就可以构建了。」
Ryo Lu · Ryo Lu (Cursor): AI Turns Designers to Developers

Jacob 的叙事更冷:功能层面的设计阶段可以被跳过,设计只在"设计语言/品味"这个更高层上保留——设计作为一道逐个功能的工序,正在消失:

"Probably some companies will skip the design phase. I think we can skip the design phase on functionality... I do think design still has a place, but it's, you know, one level above the individual features... It's the design language that we choose to have. It's the taste."
「可能一些公司会跳过设计阶段。我认为我们可以在功能上跳过设计阶段……我确实认为设计仍然有一席之地,但它是在单个功能之上的一个层次……它是我们选择的设计语言。它是品味。」
Jacob Lauritzen · 20Product: Inside Legora's Tech Stack

Andreessen 又是第三种:设计师不但没被跳过,反而有了通往 Johnny Ive 的新路——AI 把设计师抬到只想高层问题的位置:

"The job of designer will involve much more of those higher level, more important components... if I'm a 25-year-old designer and I aspire to be Johnny Ive in a decade, all of a sudden I have a new path... because Johnny Ive did everything he did without AI."
「设计师的工作将涉及更多这些更高层次、更重要的组成部分……如果我是一个 25 岁的设计师,渴望在十年内成为 Johnny Ive,突然之间我有了一条新路……因为 Johnny Ive 所做的一切都没有 AI。」
Marc Andreessen · Marc Andreessen: The real AI boom hasn't even started yet

新增三篇把"设计 = 品味 vs. 手艺"的拆分推得更远,而且开始互相打架。Bret 设想的千倍"产品工程师"是"有品味,但不一定懂 CSS"(引文见分歧二)——与 Jacob"设计只在设计语言/品味层保留"同款,把品味与手艺正式解绑。但 Akshay 的第一人称证词给这条防线捅了一个洞:他承认自己连视觉品味都没有,靠与 AI 来回迭代照样出设计(「现在，可能即使我没有所需的视觉品味，但依然可以在 AI 帮助下对某个东西进行迭代。」,全文见分歧二)。如果迭代能替代品味,那么"AI 之后人还剩品味"这道所有人共用的防线,在设计这一侧已经出现第一个自称的反例。

五、坍缩的单位:一个人,还是一条流程?

前四节全部在"个人技能"的尺度上争论。Bret Taylor 是八人里唯一换了尺度的:他认为 AI 吞掉的原子单位根本不是"某个角色的工作",而是横穿部门的流程——公司吸收不了 AI 红利,是因为出货的其实是自己的组织架构图,没有人对流程负责:

"I think the atomic unit of productivity in AI is a process, not a person. …But the hard part isn't like a person's job. It's actually all the systems and people in between it. And so I think part of the reason why I think it's been slow to get the productivity enhancement is we sort of ship our org charts as companies naturally. That's the natural state. There's not usually a person responsible for that process."
「我认为人工智能生产力的原子单位是一个流程，而不是一个人。……但困难之处不在于某个人的工作。实际上是所有系统和人员的协同。因此，我认为生产力提升缓慢的部分原因是，我们自然而然地把组织结构图也搬过去了。这是一种自然状态。通常没有人负责这个流程。」
Bret Taylor · Bret Taylor of Sierra on AI agents, outcome-based pricing, and the OpenAI board

这与其说是第五种坍缩终态,不如说是对整个提问方式的异议:如果他对,那么"哪些角色会合并成通才"是次要问题,"哪些流程不再需要跨部门交接"才是主要问题——通才构建者只是"流程终于有了单一负责人"的副产品。值得注意的是,他同时也是"超级通才"的命名者:个人尺度(通才翻身)与组织尺度(按流程重组)在他那里是同一件事的两面——八人里只有他把这两个尺度接了起来。

都没说透的

我的看法

以下是判断,把握中等。共识("代码/画稿/PRD 变便宜,人上移一层")比上一版更可靠了——七家公司用不同词汇交叉印证,且 Julie 给出了第一个刻度:10 → 60-70 百分位,顶部够不着。这个刻度本身就近乎裁决:AI 给的是通才底座,不是专才替代。所以在技能层面,我维持"通才底座 + 保留专才"(Yuhki 的两层结构,如今 Akshay 的"人人 T 型"是它的同构表述)是最可能落点的判断,置信度上调。但在组织层面(要不要保留专职角色),我上一版倾向 Jacob("能但不划算"),现在必须降级为未决:Julie 用砍掉 PM 的实操给出了反例,而且她把短期成本("先慢")交代得很清楚,这种自报家丑反而增加可信度。两人可能都对——Jacob 自己说了"对于真的需要 PM 的公司";真实变量大概是规模与产品复杂度,而不是谁对谁错。另外,"品味/taste"被当成不必解释的挡箭牌的嫌疑加深了:第四个人(Akshay)又把"AI 之后人剩什么"答成 taste,但他自己同时承认没有视觉品味也能靠迭代出活——taste 到底是稀缺输入,还是给"让 AI 干活"保留体面的事后归功?回答这个,比预测头衔存废重要。

还想知道什么

取材

继续读
线索
阅读线索：agent 不到 9 nines 可靠，那应该卖给企业还是消费者
线索
阅读线索：如果 coding 是 commodity，AI startup 的护城河长在哪儿
线索
阅读线索：推理成本一年跌 10×，AI app 该按什么定价
线索
摘报
线索
摘报
线索
摘报