如何帮助人们借助人工智能蓬勃发展

今天在《AI每日简报》中，我们聊聊如何借助AI帮助人们更好地成长。

Today on the AI Daily Brief, how to help people thrive with AI.

《AI每日简报》是一档每天更新的播客和视频节目，聚焦AI领域最重要的新闻和讨论。

The AI Daily Brief is a daily podcast and video about the most important news and discussions in AI.

好了，各位朋友。

Alright, friends.

在正式开始之前，先说几件事。

Quick announcements before we dive in.

首先，感谢今天节目的赞助商：Robots and Pencils、Blitzy、Section 和 Airtable。

First of all, thank you to today's sponsors, Robots and Pencils, Blitzy, Section, and Airtable.

想收听无广告版本的节目？请前往 patreon.com/aiddlibrief，或者在 Apple 播客上订阅。

To get an ad free version of the show, go to patreon.com/aiddlibrief, you can subscribe on Apple Podcasts.

如果想了解如何赞助本节目，请发送邮件到 sponsors@aideiallybrief.ai。

To learn more about sponsoring the show, send us a note at sponsors@aideallybrief.ai.

这周的核心主题，一直都是模型。

The big theme of this week has been models.

模型，模型，还是模型。

Models, models, and more models.

但就算世界上有再多模型，也帮不了人们学会如何从AI中获得价值。

And yet, all the models in the world aren't going to help people learn how to get value out of AI.

没错，模型的改进能修复以前模型的缺陷，开辟新机会。

Yes, model improvements can deal with fail cases from previous models and open up new opportunities.

但如果没人教大家怎么用，这一切都白搭。

But if people aren't supported in learning how to use them, it's kind of all for naught.

这正好和今天赞助商Section最新的AI能力报告发现一致。

And that certainly seems to be what today's sponsor section found with their most recent AI proficiency report.

这份报告讲的故事，很多在大公司上班的朋友肯定都耳熟能详。

The story the report tells is one that will be very familiar for many of you guys who work inside big companies.

他们总结的第一个关键发现是：智能代理已经来了，但人们还没准备好用。

Their first key finding they summed up: Agents are here, agentic readiness is not.

虽然他们调查的员工中有69%表示公司已经对AI代理采取了某些行动，但真正会在工作中使用智能代理工具的只有16%，不到10%的人能用自己的话准确定义什么是AI代理。

While 69% of workers they surveyed reported that their organization had taken some action on AI agents, only 16% actually use an agentic tool at work, and less than 10 can define an AI agent in their own words.

这其实不难理解，毕竟只有30%的员工在拥有AI代理的公司里接受过相关培训。

This isn't surprising when you find out that only 30% of employees at organizations with AI agents have actually received agentic training.

这项研究虽然是最新的一份，但绝不是唯一一份在讲这个问题的报告。

Now this study is the latest to show this sort of detail, but is far from the only one out there telling this story.

今天我们要聊的，是些实际的方法和例子，看看怎么帮人更好地用好AI。

Where we're going to end today is some ideas and examples of how to help people thrive more with AI.

不过在那之前，既然这期是个周末长文深度思考型节目，我想念一段大卫·布鲁克斯最近在《大西洋月刊》上写的长文《将在AI时代茁壮成长的人》里的几段。

But before we do that, since this is a weekend big thinklong reads type of episode, I actually want to read some excerpts of this recent long form piece in The Atlantic by David Brooks called The People Who Will Thrive in the AI Age.

布鲁克斯认为，真正让人脱颖而出的，不是智商高低，而是你对脑力劳动的态度。

Brooks argues that what will differentiate people is not how smart they are, but instead, their relationship to mental effort.

布鲁克斯写道：还记得当初人们说AI会抢走我们的工作，让人类无事可做吗？

Brooks writes: Remember when AI was going to take away our jobs and leave humans with nothing to do?

可到目前为止，这事儿压根没发生。

So far, that doesn't seem to be happening.

ActiveTrack的研究人员分析了上万名员工的数字行为，发现一旦人们开始用AI，工作反而变得更忙碌了，而不是更轻松。

Researchers from ActiveTrack analyzed the digital activity of more than 10,000 workers and found that when people adopted AI, their work life became more intense, not less.

这些早期使用者在邮件、即时消息和聊天应用上花费的时间翻了一倍多。

The time that these early adopters spent on email, messaging, and chat apps more than doubled.

他们使用商业软件的频率上升了94%。

Their use of business software rose by 94%.

加州大学伯克利分校哈斯商学院的研究人员发现，当人们使用AI后，开始接手过去外包出去的任务，因为编程和工程这类工作变得更容易上手了。

Researchers from UC Berkeley's Haas School of Business found that when using AI, workers started taking on tasks that they had previously outsourced because activities such as coding and engineering became easier to do.

他们利用晚上、周末、排队等候的空档，甚至任何能挤出的零碎时间埋头工作，而AI正好帮上了忙。

They squeezed in workbursts in the evening, on weekends, in waiting rooms, and wherever else they had a spare moment, and AI was handy.

他们还做了更多并行工作，比如同时管理多个机器人，多件事一起干。

They also did a lot more multitasking supervising a bunch of bots, doing things simultaneously.

研究揭示的普遍趋势是，很多人并没有用AI省下的时间减少工作，而是拿来接更多新任务。

The general pattern that the research points to is that many people don't use the time they save using AI to do less they use the time to take on new tasks.

AI还改变了员工和上司对每天该完成多少工作的预期。

AI also seems to shift workers' expectations and their bosses' expectations about how much they should accomplish in a day.

每一小时都变得更满，也更让人焦头烂额。

Every hour feels more crowded but also more frazzled.

Active Track 的研究发现，人们用于专注、不受打扰工作的时长减少了 9%。

The Active Track researchers found that the time people spent on focused, uninterrupted work fell by 9%.

这种精神状态甚至有个名字，叫 AI 脑疲劳。

There's even a name for this mental state AI brain fry.

现在，先暂时放下 Brooke 的文章，这种感觉其实到处都在蔓延。

Now, taking a pause from Brooke's piece for a minute, there is a lot of this feeling going around.

Midjerdy 的创始人大卫·霍尔斯最近发推说：我的朋友们都觉得用最新的代码模型特别有成就感，但也特别累。

Midjerdy founder David Holes recently tweeted: My friends are all feeling extremely productive and also extremely drained with the latest coding models.

这让我觉得哪里不太对劲，但同时好像也藏着巨大的机会？

This makes me feel like something is wrong, and also that there might be a big opportunity?

大家平时有什么办法能让自己感觉好一点吗？

Does anyone have any strategies they use to make it feel better day to day?

这个话题我也聊过很多次了。

This is also something I've talked about a lot.

几个月前，我在一期节目里提过‘无限待办清单’这个概念——就是永远做不完的工作列表，总有一件事儿在等着你。

A couple months ago, in an episode, I introduced the idea of the infinite backlogbasically, this never ending list of work that ensures that there is always a next thing to do.

在AI出现之前，虽然任务清单永远做不完，但至少还能找到一些合理的暂停点。

Now in the pre AI world, while the list was never ending, there were reasonable stopping points on that list.

AI和智能代理的出现改变了这一切——既然能用代理复制自己，人就总觉得工作不该有停歇的时候。

What changed with AI and agents specifically is that now that you can effectively duplicate yourself through agents, it feels as though there should never be any downtime in work.

代理不需要周末，也不需要睡觉。

Agents don't need weekends, they don't need sleep.

那它们是不是该一直处理那个无穷无尽的任务清单？

So can't they be taking on that infinite backlog constantly?

但现实是，限制并没有消失，只是从‘我们能做多少’变成了‘我们能管理多少’。

Of course, in reality, the limits have just shifted from how much we can do to how much planning and oversight we can support.

无论如何，回到布鲁克的观点，他写道：在新兴的AI时代，一个核心原则是——当智能变得充裕，意志力才更显珍贵。

In any case, back to Brooks, he writes: A guiding principle of the emerging AI age is this: When intelligence is plentiful, volition is valuable.

真正能带来改变的人，不是那些追求放松、被动用AI少干活的人。

The people who are going to make a difference are not the ones who seek relaxation and passively use AI to work less.

而是那些主动拥抱AI、不断打磨自己思维能力、力求做得更多的那群人。

They are the ones who will seek improvement and actively wrestle with AI to develop their own mental capabilities and accomplish more.

换句话说，决定人与人之间差异的，不是聪明与否，而是他们对待脑力劳动的态度。

In other words, what will differentiate people is not how smart they are, but their relationship to mental effort.

现在有些人心理学上称之为‘高认知需求’。

Right now, some people have what psychologists call a high need for cognition.

他们喜欢动脑筋，享受深度思考。

They enjoy thinking hard.

这类人喜欢玩难度高的游戏，也爱读内容厚重的书。

These are the people who enjoy playing difficult games and reading dense books.

另一端则是‘认知吝啬鬼’，他们觉得费脑筋很不舒服，能躲就躲。

On the other end of the spectrum, there are the cognitive misers the people who find it unpleasant to think hard and take any opportunity not to do it.

中间那部分人属于中等认知需求。

In the middle are the people who have a medium need for cognition.

他们只有在特别在意的事情上才会投入精力，但并不真正享受思考的过程。

They will put in the effort when they really care about something, but they don't intrinsically enjoy it.

认知需求和智力有关联，但并不是一回事。

Need for cognition correlates with intelligence, but is not the same thing.

我们身边都有很多特别聪明却不愿意下苦功的人。

We all know a lot of really smart people who don't like to work hard.

这促使布鲁克斯开始归纳出几种不同类型的人，他们与人工智能互动的方式会大不相同。

And this leads Brooks to start to identify a number of different archetypes for people who will have different experiences with AI.

他把第一类人称为‘被动受益者’。

The first category he calls productive passengers.

这类人认知需求低，因此会想办法用人工智能来减少自己的工作量。

These are the folks who, as he describes it, have a low need for cognition, and who, because of that, will try to find ways to use AI to do less.

但这并不意味着人工智能对他们没用。

Now, this does not mean that AI won't be valuable for them.

事实上，正因为它让任务变得简单，他们反而能更高效地完成工作。

In fact, it will be valuable for them exactly because it makes tasks easy enough that they can be more productive.

布鲁克斯指出，问题在于，人工智能让事情太容易了，反而可能削弱他们的能力。

The challenge, writes Brooks, is that AI might actually diminish their capabilities because of how easy it makes tasks.

他引用了麻省理工学院媒体实验室的一项研究，发现人们在使用ChatGPT完成任务时，大脑的连接活跃度比不用时下降了多达55%。

He points to research from the MIT Media Lab that found people's brain connectivity declines as much as 55% when they are using ChatGPT compared to when they are not using it to perform similar tasks.

另一项来自可能性科学的研究发现，当人们使用人工智能时，代表认知投入的伽马波活动下降了大约40%。

And another study from Possibility Sciences, which found that gamma wave activity, a sign of cognitive effort, dropped by roughly 40% when people were using AI.

在他看来，这种认知活动的下降会带来可预见的负面影响，

And in his estimation concerningly, this reduction in cognitive activity he thinks will have predictable effects on people's thinking skills I.

E.

E.

会削弱人们的批判性思维能力。

It will make them worse at critical thinking.

布鲁克斯提到的第二类人是那些不太情愿的优化者。

The second category of people that Brooks talks about are the reluctant optimizers.

他把这些人描述为认知需求中等的人，他们明白人工智能可能会掏空自己的能力。

These he describes as people with a medium need for cognition, who understand that AI might hollow them out.

他说，他们会怀着真诚的意图，努力避免自己陷入依赖。

They will resolve earnestly and with good intentions, he says, to not let themselves fall victim.

但在日常生活的繁忙和压力下，他们还是会不知不觉被拉进去，意志动摇，最终过度依赖这些机器人。

But in the crowded and stressful rush of everyday life, they will get sucked in, their resolve will fail, and they'll become over reliant on the bots.

他指出的问题在于人与努力之间的关系。

And the problem he suggests is around the relationship to effort.

他写道：如果你追求的是优化，那你关注的是产出最大化，而不是卓越。

He writes: If you're going for optimization, you're looking to maximize output, not excellence.

根据一家名为GoTo的软件公司所做的调查，43%的员工承认，他们提交过AI生成的内容，尽管他们怀疑这些内容存在错误，整体质量也不高。

In a survey conducted for the software firm GoTo, 43% of workers said they had submitted AI generated content even though they suspected it contained errors and was generally of low quality.

布鲁克斯认为，优化的核心问题在于，它会改变人们对努力本身的态度。

The core problem with optimization, Brooks writes, is that it will change people's attitude towards effort itself.

克里斯·西邦是位于弗吉尼亚州北部的一所小型私立学校里文德尔的校长。

Chris Sibon is the head of school at Rivendell, a small private school in Northern Virginia.

有一天，他给学生们放了一部电影，这部片子由200多位艺术家耗时五年多才完成。

One day, he showed his students a film that took more than 200 artists more than five years to make.

学生们都懵了。

The students were baffled.

为什么要花这么大力气？

Why do that?

有学生说，AI五分钟就能搞定。

As one student put it, AI could have done it in five minutes.

西邦把这称为疏离的工业化。

Sibbon called this the industrialization of detachment.

他指出，布鲁克斯认为，那些曾与艰深文本较劲、在压力下反复修改论点、失败后又重新尝试的学生，收获的远不止知识。

He argued Brooks writes that a student who has wrestled with a hard text, revised an argument under pressure, and failed and tried again is more than informed.

他们变得更扎实了。

He is more solid.

布鲁克斯把第三类人称为精神上的马拉松选手。

The third category Brooks calls the mental marathoners.

事实上，他还用马拉松运动员来做类比。

And in fact, he uses marathon runners as a comparison point.

布鲁克斯写道，汽车完全足以让人跑完26.2英里的距离。

The automobile, Brooks writes, is a perfectly good technology for traveling 26.2 miles.

从实际角度看，根本没必要训练自己跑完这么远。

There is no practical reason that any person should train themselves to run the distance.

但就是有人愿意。

But some people do.

他们愿意付出努力，因为他们想有所成就。

They want to put in the effort because they want to accomplish things.

他们想拓展自己的能力。

They want to expand their capacities.

那些认知需求高的人，思考时就是这样。

High need for cognition people are like this when it comes to thinking.

在《人工智能时代》一书中，布鲁克斯认为，这些精神上的马拉松选手会拼命抵抗AI带来的惰性。

In The Age of AI, Brooks writes, I suspect that the mental marathoners are going to work really hard to resist AI entropy.

他们会强烈渴望保持原创性。

They're going to feel a strong desire to be original.

马拉松选手会希望产出真正属于自己的作品，体现独特的自我。

Marathoners are going to want to produce work that feels personal, that reflects their unique self.

他们会想办法利用AI来增强自己的主动性，而不是削弱它。

They're going to want to find ways to use AI to increase their agency rather than diminish it.

到目前为止，这篇文章的基调还算悲观。

Now so far, the essay has been fairly bleak.

但正如布鲁克斯正确指出的：虽然我一直把认知需求当作一种固有的特质，而且意志力确实有遗传基础，但它也非常受环境影响。

But, as Brooks rightly points out: While I've been treating the need for cognition as some sort of ingrained trait, and although willpower has some hereditary basis, it is also extremely sensitive to context.

换句话说，他写道，如果人工智能倾向于削弱人的自主性，人类完全可以改革制度，来帮助重建这种自主性。

In other words, he writes, if AI has a tendency to undermine volition, humans can reform institutions to help build it up.

他思考了教育体系该如何改变，从现在这种侧重死记硬背和功能性产出，转向更注重培养自主性。

He meditates on how the education system might change to shift the orientation from rote memorization, and the types of functional outputs he has now, to instead focus on things like volition.

换句话说，他写道，真正重要的不是智力高低，而是愿意投入脑力马拉松、追求高质量成果的意愿。

In other words, he writes, what really matters is not brainpower but the willingness to run the mental marathons that produce high quality results.

他指出，关键任务是培养人们主动追求认知复杂性的渴望。

The crucial task, he writes, is to cultivate people's desire to seek out cognitive complexity.

他以乐观的语气收尾：如果我们能帮助人们学会渴望更多、追求更多，他们就会愿意付出脑力去完成艰难的事，从而避免即将降临的认知极化。

He ends on an optimistic note: If we can help people learn to want more or hunger more, they'll be willing to undertake the mental effort to do hard things, and we'll avoid the cognitive polarization that is staring us in the face.

如果我们能教育人们清楚而真诚地知道自己真正热爱什么，那么人工智能负责计算和整合，而人类依然决定什么重要、值得探索、该追求怎样的目标，以及最终走向何方。

If we can educate people to be clear and wholehearted about what they truly love, then AI will do the calculating and the synthesizing, but humans will still define what matters, what is worth exploring, what missions we go on, and where we end up.

这样会造就一个机器人无处不在的社会，但人的尊严不仅得以保留，甚至可能得到提升。

That would produce a bot filled society in which human dignity is preserved and, perhaps, even enhanced.

我在这档节目里天天都在讨论AI潜力和现实之间的差距。

I cover the capability gap between AI potential and AI reality every day on this show.

大多数公司还在摸索该怎么起步。

Most companies are still figuring out how to start.

Robots and Pencils 已经开始落地并快速扩张了。

Robots and Pencils is already launching and scaling.

大型企业几周内就能上线具备自主性和生成能力的AI应用。

Agentic and generative AI in production at large enterprises in weeks.

AWS高级合作伙伴的数量在过去一年里翻了一倍多。

AWS advanced tier pattern partner more than doubled in a year.

而且他们正在招人。

And they're hiring.

目前有五十个职位空缺。

50 open roles.

如果你觉得此刻与众不同，想亲身参与其中，而不是旁观，那这绝对值得你关注。

If you're someone who knows this moment is different, who wants to be inside it, not watching it, this is worth a look.

在 Robots and Pencils，最好的点子才能胜出，团队也始终保持着极高的水准。

At Robots and Pencils, the best ideas win, and the team is purposefully kept super high quality.

这里会成为你日后回想起来，觉得这辈子最明智的选择。

This is the kind of place you look back on as the best decision you ever made.

去看看 robotsandpencils.com/careers 吧。

Take a look at robotsandpencils.com/careers.

周末就该用来做有感觉的编程。

Weekends are for vibe coding.

现在把一个热爱的项目做出来，比以往任何时候都容易，赶紧打开你最爱的 vibe 编程工具吧。

It has never been easier to bring a passion project to life, so go ahead and fire up your favorite vibe coding tool.

但周一终究会来，等你反应过来，眼前已是密密麻麻的微服务、70年代留下的 COBOL 老系统，还有条路根本看不到头的工程路线图——比你退休那天还长。

But Monday is coming, and before you know it, you'll be staring down a maze of microservices, a legacy COBOL system from the 1970s, and an engineering roadmap that will exist well past your retirement party.

正因如此，你需要 Blitzy，首个专为大型企业级代码库打造的自主软件开发平台。

That's why you need Blitzy, the first autonomous software development platform designed for enterprise scale codebases.

每轮开发一开始就能部署，让你的开发路线图提速五倍。

Deploy the beginning of every sprint and tackle your roadmap 500% faster.

Blitzy的智能代理会分析你的全部代码，规划任务，并自主完成超过八成的工作。

Blitzy's agents ingest your entire codebase, plan the work, and deliver over 80% autonomously.

代码经过验证、端到端测试，品质一流，交付速度却快如计算。

Validated, end to end tested, premium quality code at the speed of compute.

几个月的开发工作，几天就能搞定。

Months of engineering compressed into days.

周末就用 vibe coding 做你的热情项目。

Vibe code your passion projects on the weekend.

周一上班，就把 Blitzy 带上。

Bring Blitzy to work on Monday.

去 blitzy.com 看看，为什么《财富》500强都信赖 Blitzy 来处理关键代码。

See why Fortune 500s trust Blitzy for the code that matters at blitzy.com.

就是 blitzy.com。

That's blitzy.com.

这儿有个残酷的真相。

Here's a harsh truth.

你的公司很可能在AI工具上花了几千甚至几百万美元，但这些工具根本没被充分利用。

Your company is probably spending thousands or millions of dollars on AI tools that are being massively underutilized.

虽然有一半公司都买了AI工具，但只有12%真正用它们创造了商业价值。

Half of companies have AI tools, but only 12% use them for business value.

大多数员工还在用AI来总结会议纪要。

Most employees are still using AI to summarize meeting notes.

如果你负责公司里的AI落地，那你需要Section。

If you're the one responsible for AI adoption at your company, you need Section.

Section是一个帮助你全面推动AI转型的平台。

Section is a platform that helps you manage AI transformation across your entire organization.

它指导员工掌握实际应用场景，追踪谁在用AI创造业务价值，还能让你清楚看到AI在哪儿有效、在哪儿无效。

It coaches employees on real use cases, tracks who's using AI for business impact, and shows you exactly where AI is and isn't creating value.

结果呢？

The result?

你不再只是部署工具，而是真正推动可衡量的AI价值。

You go from rolling out tools to driving measurable AI value.

你的员工不再只做会议总结，而是解决实际的业务问题，而你也能证明投资回报率。

Your employees move from meeting summaries to solving actual business problems, and you can prove the ROI.

别再猜你的AI投入有没有效果了。

Stop guessing if your AI investment is working.

访问 sectionai.com 了解 Section。

Check out Section at sectionai.com.

就是 sectionai.com。

That's sectionai.com.

本集《AI每日简报》由 Hyperagent 赞助播出，你可以管理一支团队协作的智能代理舰队。

This episode of the AI Daily Brief is brought to you by Hyperagent, where you run fleets of agents your team can manage together.

新用户可获得1000美元的推理算力。

New users get $1,000 in inference.

别再等着在电脑上手动触发本地代理或聊天流程了。

Forget local agents and chat workflows waiting on your laptop to be prompted.

Hyperagent 在云端部署全天候运行的智能代理，直接使用你们团队已经在用的工具完成实际工作。

Hyperagent deploys always on agents in the cloud, doing real work across the tools your team already uses.

市场团队的代理能将竞争对手的动态自动生成落地页。

Marketing's agent turns competitor moves into landing pages.

销售团队的代理负责完善线索信息、撰写邮件，并更新客户管理系统。

Sales's agent enriches leads, drafts emails, and updates the CRM.

运营代理自动跟进文书流程，追踪预算使用情况。

Ops agent chases the paperwork and tracks the budget.

每个代理都能访问共享的上下文信息，并严格遵守你们设定的权限和审批规则。

Every agent has access to shared context and follows your rules about scope and approvals.

现在是时候引入那些像同事一样可靠的智能代理了。

It's time you add agents that feel like teammates.

立即在 Hyperagent 上雇佣你的智能团队，由 Airtable 原班人马打造。

Hire yours at Hyperagent, built by the team at Airtable.

前往 hyperagent.com/aideallybrief 领取你的 1000 美元算力额度。

Claim your $1,000 in inference at hyperagent.com/aideallybrief.

所以我更想聊的，不是学校该怎么改——虽然我完全同意，教育的核心目标确实需要彻底转变。

So where I want to take the conversation is not so much about schools and how they can change, although I agree wholeheartedly that the entire core goal of education needs to shift.

我真正关心的是，现在怎么让人和AI建立更好的关系。

What I'm interested in is how to improve people's relationship with AI in the here and now.

布鲁克斯的文章里有一段，提到那些思维耐力强的人是怎么好好利用AI，又不放弃自己的思考主导权的。

Now, there is one section in Brooks' piece where he talks about some of the ways that those mental marathoners use AI well, without surrendering their cognitive agency.

有人总结出一些实用的方法，比如：别让AI替你思考，等你自己有了分析和结论之后，再请它来挑战你的观点。

A couple of the tips and things that people have found include things like: Asking not to produce your thinking, but to challenge it once you've already come up with your own analysis and conclusions.

另一个建议是，明确区分机械性工作和创造性工作。

Another suggestion is to make sharp distinctions between rote work and creative work.

也就是说，可以让AI写功能性的邮件，但别让它替你写论文或备忘录。

In other words, to let AI write functional emails, but not to let it write essays or memos.

但我认为，布鲁克斯忽略了最大的机会——简单来说，别只用AI做你本来就会做的事，而要用它来做你根本做不到的事。

I think, though, that Brooks is missing the biggest opportunity here, which is, very simply put, to not just use AI for things you can already do, but to use AI for things you can't do.

布鲁克斯建议，应该对那些过度依赖AI写文章的人加以批评。

Brooks suggests that we shame people who overly rely on AI for writing.

我也不太确定，老兄。

I don't know, man.

我还没让AI替我写邮件呢。

I haven't turned over my email writing to AI.

但咱们真觉得，那些企业沟通里，有多少是真正用心讨论最新会议成果的吗？

But do we really think that most of the corporate communications that were responsible for writing involve within them some paragon of virtue of the effort to discuss the results of the latest meeting?

我觉得我们更该分清不同类型工作的价值，别太纠结于那些AI能替我们搞定的事。

I think we need to better distinguish between the value of different types of work, and not be so concerned in many cases about the work that AI can take off our plates.

更重要的是，那些没被AI搞垮脑子、反而被激发出了新想法的人，往往都明白一点：AI帮我们甩掉琐碎工作固然好，但真正厉害、真正让人兴奋的，是去做以前根本做不到的事。

But more than that, the people who I find whose brains are not atrophying because of AI, but are in fact lighting up with new possibilities, are those who recognize that for as useful as the efficiency side of AI can be in getting that type of rote work off their plate, the real power in fact, the exciting thing is in doing things that weren't possible before.

你知道现在还有啥不容易的吗？

You know what's not easy even right now?

如果你不是程序员，也不太懂技术，怎么去搭建一个能帮你做事的智能代理。

Figuring out how, if you are not a coder by background and not particularly technical how to build an agent that can do things for you.

这事儿得放低姿态，先问AI怎么操作，等它告诉你了，截图下来再问另一个AI：这些词到底啥意思？

Doing so involves a lot of humility of asking AI how to do something, and then when it tells you how, screenshotting what it said and asking another AI what the heck those words mean.

尝试新东西，遇到错误，然后得琢磨这个错误到底是什么意思。

Of trying things, coming up against an error, and then having to figure out what that error means.

感受到亲手推出以前根本不可能完成的东西的成就感，可刚一上线就被别人碰出问题，心里一紧，赶在别人发现前拼命修复。

Of feeling the power of releasing something that you never could have built before, only to have it crumble on the first touch with other people, and to feel the pang in the race as you try to fix it before anyone else shows up.

时间久了，我们熟悉的事就变得简单了。

Over time, the things that we know how to do become easy.

思维的灵活性，就像身体锻炼一样，来自于做那些不舒服、没做过的事。

And mental elasticity, just like physical workouts, comes from doing things that are uncomfortable and that we haven't done before.

重点不是一开始就做得好，而是先做起来，做到好为止，然后重新再来一遍。

The point is to be not good at things, but to do them anyways until we are good at them, and then to run the cycle back all again.

AI并没有改变这一点。

AI hasn't changed that.

但对于善于使用AI的人来说，它改变了他们接下来想尝试的新事物的野心高度。

But for the successful AI users, it's changed the level of ambition around those new things that they might go try next.

我敢打赌，你们当中大多数人都属于那种：在朋友、同事或家人里，唯一一个这样用AI的人。

And I'm sure most of you listening are either: one, the person among your group of friends and colleagues and family who uses AI like that.

或者，是那个正在尝试这样使用AI的人。

Or, alternatively, the person who is trying to use AI like that.

无论这是你本人、你身边的人，还是你正在努力成为的未来自己，那些把AI当作一种能实现前所未有之事的机遇技术、从而拓展自我与能力的人，正是组织重塑所必然依赖的核心支柱。

And whether it's you or people close to you, or the future you that you're working to be, the people who treat AI as this opportunity technology to accomplish things that weren't possible before to stretch themselves, in other words, and stretch their capabilities are, in fact, the key pillar upon which the organizational redevelopment will necessarily be built.

《华尔街日报》的CIO期刊最近发表了一篇文章，主题是AI倡导者。文章指出，AI使用率的提升以及对非使用者疑虑的化解，很大程度上依赖于组织内部这一类人。

The Wall Street Journal's CIO Journal recently wrote an article about AI champions The article argues that a large part of the increase in AI usage and the fighting of skepticism from non users is reliant on this category of people inside organizations.

期刊写道：一线倡导者在推动这些增长中发挥了关键作用。

The journal writes: On the ground champions are playing a key role in those increases.

通过这些项目，员工自愿申请提前使用新工具、参加专项培训，并获得向高层管理者展示的机会。

Through these programs, workers volunteer to receive early access to new tools, special training, and opportunities to present to senior executives.

作为回报，他们被要求向同事推广AI的使用，并在正式会议和非正式交流中解答疑问。

In exchange, they're asked to promote AI adoption to their colleagues and field questions through both formal meetings and informal conversations.

文章举了一个律师事务所的例子：该所员工使用AI的频率显著提升，现在正着手系统化一套针对六十多位倡导者的计划，以更有效地推动AI应用并追踪成效。

They give the example of a law firm who has seen significant increases in the way that their employees use AI, and are now formalizing a program around their 60 some champions on how to promote AI more effectively and track the success of them to do so.

这篇报道的可贵之处，在于它准确识别出了这一关键角色。

Now, what this article gets right is to identify this key role.

但这篇文章忽略了一点，那就是AI倡导者本质上就像是内部的公关人员。

But where it misses a little bit is the idea that AI champions are effectively just internal PR agents.

没错，确实需要有人愿意和同事坦诚地一对一交流，回应质疑和顾虑。

Yes, it is useful to have people who are willing to have frank one on one conversations about AI and answer challenges and skepticisms.

事实胜于雄辩。

The proof is in the pudding.

倡导者真正的价值不在于告诉别人AI有多好，而在于让他们看到，只要愿意尝试，自己到底能做成什么。

And the real value of Champions is not in telling people how good AI is it's in showing them what they could actually be doing if they tried.

我早在2026年之前就预测过，我们会开始看到一种我暂且称为‘内嵌氛围工程师’的角色。

One of my predictions coming into 2026 was that we would start to see a role that I loosely called internally deployed vibecoders.

很明显，现在有个大趋势，就是部署AI的公司会把工程师直接派到业务部门，帮助团队更好地融入这项技术。

Obviously, there is a huge trend towards forward deployed engineers, where companies who are provisioning AI are also placing engineers inside the organization to embed and help those organizations better integrate the technology.

我提出的‘内嵌氛围工程师’，其实就是这类倡导者计划的延伸——那些越来越熟练使用AI新功能，特别是AI代理的编程和构建能力的人，会和业务部门结对合作，帮他们摸索出这些新能力到底能怎么改变工作方式。

And my argument around internally deployed vibecoders was basically an extension of these types of champions programs, where people who are increasingly using the new capabilities of AI and specifically agents including the coding and building capabilities of AI agents, to pair and partner with business functions in ways that could help those business functions start to figure out how that new capability set could actually change how they work.

换句话说，我所说的这些人，不是去帮别人把现有工作提速20%的。

In other words, I argued, these would not be folks who were helping people figure out how to make their current work happen 20% faster.

这些人要做的，是帮业务部门重新思考：不仅怎么做事，甚至包括到底该做什么。

It would be people who would pair to help business functions figure out how to fundamentally change not only how they do what they do, but even in some cases, what they do.

我想用一个实际案例来收尾，那里已经出现了一些类似的做法。

And I wanted to end on a case study of one place where some version of this seems to be happening.

优步的首席技术官普拉文·纳波利最近发推说：优步的智能体AI应用正在迅猛发展，它正在改变我们的开发方式。

Uber's CTO Praveen Napoli recently tweeted: agentic AI adoption is on fire at Uber, and it's changing the way we build.

不只是工程团队，而是整个公司都在发生变化。

Not just in engineering, but across the entire company.

如今，我们99%的工程师都在使用AI工具。

Today, 99% of our engineers use AI tools.

超过70%的代码合并请求都来自本地或云端的智能体。

More than 70% of pull requests are attributed to local or cloud agents.

我们的工程师已经构建了两千多个覆盖整个软件开发生命周期的智能体技能。

And our engineers have built 2,500 plus agent skills across the software development lifecycle.

这些数字很惊人，但也引出了一个更关键的问题：我们该怎么把智能体AI推广到工程部门以外？

Those numbers are exciting, but they led us to a much bigger question: how do we bring agentic AI beyond engineering?

财务、法务、运营、市场、客户服务、人力资源、采购。

Finance, legal, operations, marketing, customer support, HR, procurement.

这些部门依赖的流程非常复杂，往往靠人工操作，细节繁多，还分散在几十个系统里。

These functions run on complex workflows that are often manual, highly nuanced, and spread across dozens of systems.

光看流程图或者文档，根本没法有效自动化。

You can't automate them effectively by looking at process diagrams or documentation.

你得真正搞清楚工作是怎么实际完成的。

You have to understand how the work actually gets done.

所以我们创建了一种叫‘智能协作小组’的东西。

So we created something called agentic pods.

想法很简单：我们挑了大约30位最懂AI的工程师，他们对Uber的系统了如指掌，然后每人配一位来自业务部门的专家。

The idea is simple: we handpicked around 30 of our most AI proficient engineers people with deep knowledge of Uber's systems and paired each of them with a domain expert from a business function.

接着，我们给每个小组只留了两周时间。

Then we gave every pod just two weeks.

头两天：跟着专家实地观察，记录每一个步骤，提问题，建立直观理解。

Days one and two: Shadow the expert: Observe every step, document workflows, ask questions, build intuition.

第三天：根据规模、重复性、业务影响和数据可得性来优先排序机会。

Day three: prioritize opportunities based on scale, repetition, business impact, and data availability.

第四到第五天：和正在做这项工作的人一起开发一个可用的智能代理。

Days four to five: build a working agent alongside the person doing the job.

第六到第九天：让其他做同样工作的人来验证——这个方案能通用吗？

Days six to nine: validate with several others performing the same work does it generalize?

它真的让他们的工作变好了吗？

Does it actually make their job better?

第十天：上线。

Day 10: Ship.

过去两个月里，我们已经在16个不同业务部门运行了16个智能代理小组：资本分配从150个城市各15小时缩短到30分钟；财务进度报告从两天压缩到10分钟；市场网页质量检测从两周减少到50分钟。

In just the past two months, we've run 16 agentic pods across 16 different business functions: Capital allocation across 150 cities, from fifteen hours to thirty minutes Financial pacing reports, from two days to ten minutes Marketing web quality assurance from two weeks to fifty minutes.

支持工作流创建：将9000个手动流程转为自助式自动化。

Support workflow creation 9,000 manual workflows to self-service automation.

他写道，生产效率的提升很显著，但最让我们意外的并不是速度。

The productivity gains, he writes, are impressive, but what surprised us most wasn't the speed.

真正让人惊讶的是，那些被派到陌生领域的工程师，居然这么快就发现了那些一直摆在明面上却没人注意的机会。

It was how quickly engineers embedded in unfamiliar domains uncovered opportunities that had been hiding in plain sight.

最大的收益通常不是来自自动化某个单一任务。

The biggest wins rarely come from automating one task.

而是来自对整个工作流程的重新思考。

They come from rethinking an entire workflow.

一旦你围绕人工智能重新设计流程，往往就能消除交接环节、去掉不必要的审批、替换老旧工具、减少供应商开支，并大幅加快决策速度。

Once you redesign the workflow around AI, you often eliminate handoffs, remove unnecessary approvals, replace legacy tooling, reduce vendor spend, and dramatically accelerate decision making.

自动化的基本单位应该是工作流程，而不是单个任务。

The workflow becomes the unit of automation, not the individual task.

最有影响力的AI能力，往往能跨越团队、部门、职能、工具和系统。

The most impactful agent skills cut across teams, orgs, functions, tools, and systems.

最大的教训是什么？

The biggest lesson?

最有价值的AI机会，通常从外面是看不到的。

The best AI opportunities are rarely visible from the outside.

你得坐在实际做事的人旁边，深入了解每一个痛点，和他们一起打造，而不是替他们做。

You discover them by sitting next to the people doing the work, understanding every friction point, and building with them, not for them.

我们现在正组建一个专门的团队，来进一步推广并深入这项工作。

We're now forming a dedicated team to scale this further and go deeper.

他们会彻底理解工作流程，从头开始重新设计，并利用人工智能从根本上改变企业的运营方式。

They'll deeply understand the work, redesign it from the ground up, and use AI to fundamentally change how the business operates.

我觉得这太棒了，其他团队几乎可以照搬这套做法，很快就能上手。

Now I think this is super cool, and is a type of program that others could imitate almost whole cloth, fairly right away.

但我感兴趣的不只是这两周的成果。

But what I'm interested in is not just the two week results.

我坚信，这类低垂的果实——提升效率的应用，迟早会不断冒出来。

I think inherently, you're going to see these types of low hanging fruit productivity use cases surface.

这当然很好。

And that's great.

组织应该尽快走过这一阶段。

Organizations should get through that as fast as they possibly can.

那么问题就变成了，他们该如何重复利用这些成果？

The question then becomes how they reuse those gains.

我的直觉是，虽然普拉文在这里主要讲的是工程师通过与业务专家紧密合作来决定该做什么的流程，但如果你能把这种工程思维与业务实践之间的互动模式制度化，真正的收益其实不会体现在这两周之内。

And my instinct is that while Praveen here is talking mostly about a flow where the engineer figures out what to do based on their close work with the business expert, I think if you start to institutionalize this sort of interaction pattern between engineering and technical thinking and business performers, the real benefits wouldn't be in the course of those two weeks.

而是在接下来的几个月里，变革的重心会从工程师做那些低垂果实式的优化，转移到业务人员身上——他们受这种自主式工作方式的影响，开始从根本上重新思考整个工作版图。

They'd be over the course of several months, where the main locus of change would shift from the engineers, doing that low hanging optimization, to the business people themselves, who, influenced by the type of agentic working they were now a part of, would start to think differently at core levels about the broader expanse of the work themselves.

换句话说，工程师或许能帮财务进度报告从两天缩短到十分钟，但更多时候，真正决定如何利用剩下的一天二十三小时五十分钟的，是那些被这些自主技术影响、甚至自己动手搭建和使用智能代理的业务人员。

In other words, while the engineers might help the financial pacing reports move from two days to ten minutes, it is in many cases going to be the business folks, newly influenced by these agentic techniques and maybe even building and working with some agents themselves, who figure out the best way to spend the other one day, twenty three hours, and fifty minutes.

而在大多数情况下，这并不会意味着做更多同样的事。

And in many if not most cases, that won't be doing more of the same work.

而是会去做全新的、平行的、过去一直想做却从未实现过的工作。

It will likely be doing new work, orthogonal work, work that was always dreamed but never possible before.

我相信，真正改变企业的，恰恰是这些新发现——它们不是来自效率提升本身，而是来自把效率提升的成果重新投入创造新价值。

And I believe it will be, in fact, those new things that are uncovered the output not of the productivity itself, but the reinvestment of the gains of the productivity that really changes the business.

不过，这种探索方式确实能帮助更多人和组织在这个AI时代蓬勃发展。

Still, this is the sort of experimentation that is going to help more people and more organizations thrive in this era of AI.

这种合作方式将不再让潜藏的认知努力成为决定谁能在AI时代胜出的唯一因素。

This is the type of collaboration that is going to not make latent cognitive relationships with effort be the only factor determining who thrives with AI.

布鲁克斯表面上认同这种内在动机并非固定不变的观点。

Brooks gives lip service to the idea that those intrinsic levels of motivation are not necessarily fixed.

但你能明显感觉到——如果我说得不对，抱歉大卫——他其实并不真这么认为。

But you can almost telland sorry to David if I'm misinterpretingbut it feels to me like you can almost tell that he doesn't really believe it.

他只是最后加了点乐观的包装，但很明显，他觉得只有那些能坚持长跑的人才能挺过这场转变。

He's giving a nice spin on things to end with some optimism, but it's clear he basically thinks that the marathoners are the only ones who survive this transition.

如果他真这么想，那我不同意。

If that is his belief, I disagree.

我认为，在教育和职场中，我们很久很久都没对人提出过什么高要求了。

I think that in most cases, in both education and in work, we haven't really asked people for much for a very long time.

我们从来没逼过他们。

We haven't stretched them.

我们从来没挑战过他们。

We haven't challenged them.

我们也没激励他们去迎接挑战。

We haven't incentivized them to be challenged.

我们给他们一堆零散的任务，理由往往让人摸不着头脑，还告诉他们，只要在规定时间内完成这些任务就算成功。

We give them discrete buckets of tasks, often to be done for nearly inscrutable reasons, and tell them success is doing those tasks in the time that they have allotted.

这或许对公司的日常运转够用，但绝对没法激发人的真正潜力。

That might be fine for corporate functioning, but it certainly doesn't maximize people's true potential.

我觉得，如果我们真正把AI用好了——也就是真正去支持它，就会发现，人的潜力远比大多数人想象的要大得多。

And I think if we do AI wellby which I mean actually supporting itwe will find far more potential in people to be maximized than most people realize is there.

总之，这些想法值得你这个周末慢慢消化，不过今天就先说到这里，我们的AI每日简报到此结束。

Anyway, something to chew on for the rest of this weekend, but for now, that is gonna do it for today's AI Daily Brief.

一如既往，感谢你的收听或观看，我们下次再见，保重。

Appreciate you listening or watching as always, and until next time, peace.