本集简介
双语字幕
仅展示文本字幕，不包含中文音频；想边听边看，请使用 Bayt 播客 App。
Cloud Code 正在取得传奇般的进展，并且其用途已超越编程领域。
Cloud Code is on a legendary run and becoming useful beyond programming.
OpenAI 推出了面向医疗保健领域的 ChatGPT。
OpenAI rolls out ChatGPT for health care.
预测市场存在内幕交易问题吗？还是说并没有？
Prediction markets have an insider trading problem, or do they?
而忙碌工作的终结是坏事吗？
And is the end of busy work a bad thing?
接下来，在这个节目之后，周五版的大型科技播客将为您带来相关内容。
That's coming up on a big technology podcast Friday edition right after this.
财政负责、金融天才、货币魔术师。
Fiscally responsible, financial geniuses, monetary magicians.
这些是人们在谈到那些将车险转投 Progressive 并节省数百美元的司机时说的话。
These are things people say about drivers who switch their car insurance to Progressive and save hundreds.
因为 Progressive 为一次性付清保费、拥有房产等提供折扣。
Because Progressive offers discounts for paying in full, owning a home, and more.
此外，当您需要帮助时，可以依赖他们出色的服务，让您的每一分钱都物超所值。
Plus, you can count on their great customer service to help when you need it, so your dollar goes a long way.
访问 progressive.com，看看您是否能节省汽车保险费用。
Visit progressive.com to see if you could save on car insurance.
Progressive 伤害保险公司及其关联公司，潜在节省金额因人而异，并非在所有州或情况下都适用。
Progressive Casualty Insurance Company and affiliates, potential savings will vary, not available in all states or situations.
欢迎收看《大科技》播客周五版，我们将以我们传统、深入且细致的方式解析新闻。
Welcome to Big Technology podcast Friday edition where we break down the news in our traditional, pool headed, and nuanced format.
今天为大家准备了一场精彩的节目。
We have a great show for you today.
我们整理了一份丰富的素材清单，涵盖多个故事，包括 Cloud Code 的最新动态及其为何可能对编程以外的人群也有用处。
We have filled our document with a ton of stories to go through, including what's happening with Cloud Code and why it might be applicable for folks beyond programming.
它或许最终将影响大量知识型工作。
Maybe it will end up impacting much of knowledge work.
至少这是伊桑·莫洛赫的观点。
That's at least what Ethan Moloch has to say.
我们将逐一探讨他关于此事的帖子。
We'll go through his post on the matter.
我们还将讨论OpenAI进军医疗领域的事。
We're also gonna talk about OpenAI's foray into health care.
这实际上是OpenAI正式进军医疗领域，因为已经有数百万人在使用ChatGPT进行医疗咨询。
It's it's actually it's official foray into health care because millions of people have been using ChatGPT for health care already.
我们还会稍微谈一谈预测市场和繁忙工作的终结。
We're also gonna talk a little bit about prediction markets and the end of busy work.
和往常一样，周五加入我们的嘉宾是Margins的Ranjan Roy。
Joining us as always on Fridays, dude, is Ranjan Roy of Margins.
Ranjan，很高兴见到你。
Ranjan, great to see you.
欢迎回来。
Welcome back.
很高兴见到你，Alex。
Good to see you, Alex.
很高兴能在这里。
Excited to be here.
让我们先聊聊 Cloud Code。
Let's begin just talking about Cloud Code.
我不知道你有没有注意到，但最近在 X 上，Cloud Code 已经成了一个梗。
I don't know if you've seen this, but for me recently on X, Cloud Code has been a meme.
这简直就像那些查克·诺里斯的梗，人们一直在说 Cloud Code 几乎能做任何事，拥有超人的能力。
It's almost like those Chuck Norris memes where people have been talking about how Cloud Code could basically do anything and it's got these superhuman capabilities.
我一直在等一篇帖子，来说明到底是什么让人们对 Cloud Code 如此兴奋，我想我们终于等到了——沃顿商学院教授、节目的朋友伊桑·莫洛赫发了一篇。他在 Substack 上的专栏叫《Claude Code 以及接下来会发生什么》，我认为他真正展示了 ClaudeCode 能够完成的自主工作量，而且它早已不再是简单的代码自动补全。
And I was waiting for a post to kind of illustrate what was going on with Cloud Code that had made people so excited about it and I think we got one from Ethan Moloch, the Wharton professor, a friend of the show, and he writes the one useful thing Substack and his post is called Claude Code and what comes next And I think he really illustrates, just the amount of autonomous work that ClaudeCode has been able to do and the fact that it's not simply a code autocomplete anymore.
所以，让我先讲讲他开头的故事。
So, let me just begin with the story that he starts with.
他说，我打开了 Cloud Code，并给了它一个指令。
He says, I opened Cloud Code and gave it a command.
开发一个能让我每月赚 1000 美元的创业点子，由你来生成点子并完全实现它。
Develop a startup idea that will make me $1,000 a month where you do all the work by generating the idea and implementing it.
我只需要运行你给我一次性提供的程序就行了。
I shouldn't have to do anything except run some program you give me once.
我不需要具备任何编程知识，你得确保一切都能正常运行。
It shouldn't require any coding knowledge on my part, and so you make sure everything works well.
马利克写道，AI问我三个选择题，然后决定我应该以39美元的价格向专业用户出售500个提示词集合，而无需任何进一步的输入。
Malik writes, the AI asked me three multiple choice questions and decided I should be selling sets of 500 prompts for professional users for $39 without any further input.
然后它独立工作了一个小时十四分钟，创建了数百个代码文件和提示词，最后给了我一个可以运行的单一文件，该文件创建并部署了一个可用的网站。
It then worked independently for an hour and fourteen minutes, creating hundreds of code files and prompts, and then it gave me a single file to run that created and deployed a working website.
让我们来谈谈莫洛赫提出的观点，为什么这比前几代有所改进，以及这里的意义何在，为什么人们对Cloud Code如此兴奋。
Let's just talk about, you know, we're gonna go through what the ideas that Moloch, brings down in terms of, why this is an improvement, over previous generations and what the implications are here, why people are excited about Cloud Code.
但首先，你对Cloud输出的这个结果感到印象深刻吗？
But but just to start, are you impressed with the output that Cloud spit out here?
我感到印象深刻。
I am.
这个具体的输出，我其实并不太印象深刻。
This specific output, I'm actually not that impressed.
我对此很感兴趣，但我仍然认为，我会深入探讨一下，关于我在这个自主智能体世界中所看到的整体情况。
I'm interested, but I still think and I'll definitely get into this in terms of, like, overall what I'm seeing with this kind of world of autonomous agentic.
但我觉得，再次强调，写出500个可以打包的提示，甚至创建一个网站，完成所有这些步骤，确实令人惊叹，但这应该是基本要求。
But I think, again, like, writing 500 prompts that you can package up, maybe even creating a website, taking all these steps is actually incredible, but it should be table stakes.
越来越多的人每天都在意识到，这已经是基本要求了，技术已经能够完成这种水平的工作。
And more and more people are realizing every day that it is table stakes, that the technology is already there to do this level of work.
我想起，我2026年的一项预测是什么？
And I think, remember, what was one of my predictions from 2026?
自主智能体AI将会成为现实。
Agentic AI is gonna be real.
我认为，Claude Code 是许多人首次接触真正自主智能体AI的入口。
And and I think Claude Code is a lot of people's first, like, entry into this world of truly Agentic AI.
就连莫利克也继续写道。
And even Mollick goes on to write.
他说，最新的AI能够更自主地完成工作，并自行纠正许多错误，同时AI被赋予了一套工具和方法，可以以多种方式解决问题。
He was like, for for the latest AI is capable of doing work more autonomously while self correcting many of their errors, but then the AIs are being given an agentic harness of tools and approaches they can use to solve problems in many ways.
我特别喜欢第二部分，因为这正是我过去六个月一直在谈论的内容。
That second part I loved because this is what I've been talking about for six months now.
代理式AI并不是一个流程图。
Agentic AI is not like a process flow diagram.
它是给它一套工具，然后让它自己去摸索该做什么，并真正去执行。
It's giving it a set of tools and letting it go figure out what to do and actually go do it.
我认为，Claude Code之所以引起轰动，是因为大家终于第一次体验到了这种感觉。
And I think, Claude Code, that the excitement around it is everyone's finally getting a taste of that.
没错。
That's right.
所以，Malek将它描述为这两者的结合：AI能够自主工作并自我修正，同时拥有这种代理式工具集，虽然这个词对我来说很专业，但本质上就是工具包。
And so so the way that Malek frames it is the combination of these two things that the AI can do autonomous work and self correct and then have this agentic harness, which is jargony to me, but basically, vehicle tools.
我喜欢这个说法。
I like it.
代理式？
Agentic?
哦，算了吧。
Oh, come on.
我太糟糕了。
Got I terrible.
不。
No.
糟透了。
Bad.
非常糟。
Very bad.
不。
No.
不。
No.
不。
No.
比如这个？
Like about this?
好的。
Okay.
继续吧。
Go ahead.
不行。
No.
不行。
No.
等一下。
Hold on.
我觉得还行。
I like okay.
我一直在说，有一套预定义的工具和数据连接器。
I've been saying, like, a predefined set of tools and data connectors.
这听起来一点都不令人兴奋。
That doesn't sound exciting.
代理型 Harness，就像只是大致把它固定住，但仍然允许它自由移动。
Agentic harness, it's like kind of just, you know, holding it generally in place, but still letting it go.
我对骑马或这种情况下马具的工作原理其实不太了解，但至少在我脑子里，我就是这么想的。
I don't really know much about horse riding or how harnesses work in any kind of situation like that, but at least in my mind, that's what I'm thinking.
我喜欢这个。
I like it.
不。
No.
不是的。
It's not.
骑马太糟糕了。
It's terrible riding.
我的意思是，我很喜欢这里的想法，但真的很差。
I mean, really like the concepts here, and it's bad.
这术语太专业了。
It's jargony.
你太贴近它了，拉詹。
You're too close to it, Ranjan.
你只需要说这些。
Here's all you have to say.
它能执行自主工作并自我修正，还能使用工具。
It can do autonomous work and self correct, and it can use tools.
就这些。
That's it.
你看，马具更短。
See, harness is shorter.
智能马具。
Agentic harness.
智能智能，好吧。
Agentic Agentic okay.
我们接下来会继续，但我们不会在节目中使用‘智能型工具箱’这个术语。
We're gonna move on after this, but we are not accepting the term agentic harness in the show.
这根本行不通。
It's just it's not work not gonna work.
一种智能型工具箱，用于以新方式解决问题的工具和方法。
A gentic harness of tools and approaches they can use to solve problems in new ways.
你为什么一定要用‘智能型工具箱’这个说法？
Why do you have to use the phrase a gentic harness?
你为什么不能直接说，这些AI被赋予了可以使用的工具和方法？
Why can't you just write they can they are the AIs are are being given tools and approaches they can use
来解决问题。
to solve the problem.
嗯，所以，所以
Well, so so
这是不必要的行话。
It's unnecessary jargon.
我觉得这里就是关键，就像
This is where I think, like
我六个月前在莱德公司工作时，在这个播客里就提过这一点，我们专注于企业级AI。
and I I brought this up on on this podcast, like, six months ago at Ryder where I work, like and we're enterprise AI focused.
我们推出了一款工具，我在六月就开始测试了。
We launched a tool, and I'd started testing it in June.
当时我们还开玩笑说。
It was and we joked.
你就像，你知道的，你现在在一家AI公司了。
Like, you're like, you know, now you're at a AI company.
你正在感受通用人工智能。
You're feeling AGI.
但那是我第一次真正感到：天哪。
But it was the first moment I actually felt like, holy shit.
我给了它一些方向，现在它自己出去做各种其他事情了，因为它有一套预定义的工具和数据连接器，有人可能会称之为‘框架’。
I gave it some direction, and now it's going out and doing all this other stuff because it had a predefined set of tools and data connectors, one might call a harness.
但不管怎样，那种自主AGI式的行动体验，我认为这就是Claude Code过去六个月让我看到的突破。
But, anyways, they like, feeling that autonomous AGI ish type action, I think, like, that what Claude code has been the breakthrough I've been seeing this for, like, six months now.
这种事情正在发生。
Like, this is happening.
这不再是我们讨论‘我能不能订个机票，让它帮我搜索并完成’了。
This is no longer us talking about, can I book a flight and have it, you know, search and go do it for me?
它已经在为人们执行任务了，而Claude Code是很多人第一次亲身体验到这一点。
Like, this is going and doing tasks for people people are doing already, and Claude Code is the first time a lot of people are experiencing with that.
这就是为什么我真心认为，今年这股浪潮会像2020年的代码辅助一样，迎来巨大突破，成为2025年的主旋律。
And that's why I honestly think this year, this is gonna break through in a big way the way coding assistance were '20 the story of 2025.
没错。
Right.
这某种程度上超越了我们对编程语言的分歧。
And that's sort of getting past our our differences on languages.
我认为这就是为什么我觉得这件事很重要，也值得在这里提出来：人们普遍认为，AI用于编程只是代码自动补全，或者在某些情况下，你在Replit里输入一个Vibe代码提示，就能生成一个网站。
I think that's why I thought this was important, and that's why I think this was worth bringing up here is that I think that the common conception is that, you know, AI for coding is just that, like, it's code autocomplete, or in some cases, you know, you put a Vibe code prompt in Replit or something, and then you get a website.
我认为有趣的是，你只需给出提示，它就会以马莱克指定的方式完成，而他根本不需要碰代码。
I think what's interesting here is that you can give just the prompt and it will end up, you know, in in a way that that Malek has specified without him having to touch the code.
它最终能为他构建一个包含销售漏斗的网站，而他要做的基本上只是把它上线。
It can end up, building a website for him include which includes a sales funnel by the way, and and, you know, basically all he has to do is put it put it live.
它实现这一点的方式是通过长时间的自主编码任务，以及能够使用各种工具。
And the the way that it's doing this is the long, autonomous tasks of of coding and then also, being able to use the tools.
我们不再争论语言的问题了。
We won't argue anymore about the, the language.
他说，这两个因素的结果导致了大型AI公司最新AI工具的重大突破。
He says the result of these two factors has led to big leaps in the latest AI tools made by the big AI companies.
现在听好了，如果你听到这里，又不是程序员，也没在建网站，你可能会问自己：好吧，这太具体了，那对我有什么好处？
Now listen, if you're at this point of this discussion and you're not a coder or you're not building a website, you might you might be asking yourself, well, okay, this is very specified and and what's in it for me?
而这正是我们用这个话题作为节目开场的原因，因为它开始扩展开来，也许这仅仅是即将到来的、适用于其他知识工作者的变革的开端，我会继续关注马莱克，因为他在这方面非常出色。
And and this is again, like, just to go why we're leading this show with this, it starts to broaden out and if anything this might be just the beginning of what's coming for, the rest of knowledge work and and and I'll just, you know, continue to read Malik because he's very good on this.
他说，不幸的是，对于我们大多数想尝试AI的人来说，这些新工具都是为程序员设计的。
He says, unfortunately for the most of us unfortunately for most of us who want to experiment with AI, these new tools are built for programmers.
在很多方面，这都令人遗憾，因为这些系统实际上对各种类型的知识工作者都有广泛用途。
In a lot of ways, this is a shame because these systems are actually broadly useful to knowledge workers of all types.
通过了解你能做什么并亲自进行实验，我认为你可以学到很多关于人工智能未来的信息。
And by seeing what you can do and experimenting with them yourself, I think you can learn a lot about the future of AI.
他回到由Cloud Code发起的初创公司的例子。
He goes to return to the example of the startup company launched by Cloud Code.
它只是触及了该工具能力的一小部分。
It was only touching a small part of the capabilities of what the tool is capable of.
如果我让它对不同用户角色的在线网站进行用户测试并给我一份报告，它会连接另一个工具，连接到我电脑上的网页浏览器。
If I ask it to do user testing of the live site from different personas and give me a report give me a report, it it connects it deploys another tool connecting to a web browser on my computer.
它会控制浏览器，访问它创建的网站，并像人类一样滚动浏览。
It takes control of the browser, goes to the site it created, and scrolling through as a human would.
他说，他让Cloud Code工具对用户界面进行关键性评估，它在发现潜在问题和识别网站上一些可疑的虚假评论方面做得更好，他说，下一步我可以轻松地让它实施这些建议，继续这一过程，而我只需提供极少的输入。
He said he asked the Cloud Code tool for a critical report of the UI, and he said it did a better job of nailing potential issues and spotting some sketchy fake reviews on the site, and he says as a next step, I could, easily ask it to implement its suggestions, continuing the process with minimal input from me.
对我来说，真正有趣的是，你不仅可以让它创建东西，还可以让它测试东西，并通过自然语言继续优化。
That's this to me is really the the interesting part is you can you can not only have go create things, can have it test things and with natural language, then go and optimize.
而且，你知道，如果你想想，也许它是一个网站，但也许它也是——我们待会儿会谈到——你电脑上的一个文档文件夹。
And, you know, if you think about, you know, maybe maybe it's a website, but maybe it's also, and we're gonna get into this, a folder of documents on your computer.
你得进去，给你提供反馈，从这些数据中生成可视化图表。
And you have to go in and start to, you know, give you feedback on it, make data visualizations from it.
他举的一个例子是他的信用卡账单。
One example that he gave is his credit card statements.
所以你可以把它们下载到你的桌面浏览器上——当然要谨慎操作——然后让Cloud Code自行处理，它会生成可视化图表，发现异常，并为你提供财务报告，而你需要的就是这种更高级的功能。
So you could just kind of download them onto your desktop browser and obviously do this with caution, and then set Cloud Code loose on it, and it makes visualizations, spots anomalies, gives you a report on your finances, and you need that that more advanced capabilities.
你需要具备编程和调用工具的能力，才能完成这类工作。
You need the ability to code and to call tools to be able to go and do this type of work.
这对我来说简直太疯狂了。
It's pretty pretty crazy to me.
我的意思是，这是真实的。
I mean, it's real.
它正在发生。
It's happening.
我正生活在未来，亚历克斯。
I I I'm living in the future, Alex.
我已经在那里了。
I'm I'm already there.
但我可以告诉你，我认为重大转变发生在2025年初，当时人们开始谈论代理式AI。
And but but I can tell you, like, this is where I think the big shift I saw is early twenty twenty five when people talked about AgenTic AI.
再次，人们想到的是这些庞大而复杂的流程，以及如何以某种方式自动化它们，利用大语言模型的智能在某个环节做决策，或在另一个环节生成一些文案。
Again, these thinking of, like, these big heavy processes and how do we automate them in some way and bring an LLM's intelligence to make a decision in one part or maybe create some copy at another part.
真正的转变是彻底重新思考什么是代理式AI。
The big shift was completely rethinking, like, what AgenTic AI is.
再次，我不会用‘利用’这个词，但允许它开始使用这些工具。
And again, won't say the word harness, but allowing it to start to take these tools.
我认为他展示的，比如围绕用500个提示创建一个完整业务、每月定价39美元并推出它，你真的会这么做吗？
And I think what he showed there, like, around this, like, creating an entire business, launching it at 500 prompts at $39 a month, are you really gonna do that?
也许会，但大概不会。
Maybe, but probably not.
有些事情是这些较小的多步骤流程，我认为到今年年底，我们所有人都会以各种方式将其融入生活。
Something it's these kinda like smaller multistep processes where I I'm calling by the end of the year, all of us are gonna start incorporating in our life in a lot of different ways.
检查我的信用卡账单，找出所有超过25美元的消费，生成一份摘要并邮件发给我。
Go check my credit card statement, find any charge above, like, $25, come up with a summary, like, email me.
像这样的事情，至少对早期采用者来说，到今年年底将成为常态。
Stuff like that is gonna become routine for especially at least any early adopter, I think, by the end of the year.
这确实很有趣。
And it is interesting.
还记得去年我们讨论过，AI会不会遇到瓶颈吗？
So we've talked remember, like, last year we were talking about, like, is AI gonna hit a wall?
当时有很多不同的观点，比如预训练是否还在持续有效？
And there were all these different things about, like, well, is pre training still continuing to work?
我认为，还有一个糟糕的术语，就是‘脚手架’，对吧？
And and I think, you know, there's another bad jargon word, which is, like, the scaffolding, right?
这是我们曾经用过的一个概念。
Which is something we've used.
你围绕这些工具构建了什么，使它们能够绕过模型性能瓶颈或上下文窗口限制？
What do you build around these tools that enable them to sort of cheat their way through the limitations of, you know, the models topping out or, for instance, the context window.
上下文窗口当然指的是模型在耗尽其保持信息的‘内存窗口’能力之前，能够处理的文本量。
The context window, of course, is the amount of text that this model the model can work on before it sort of runs out of its ability to, like, keep it in the, like, sort of memory window, so to speak.
听到这些机制如何运作真的很有趣，因为Moloch甚至提到，Claude代码是如何巧妙地突破上下文窗口的限制的。
And and it's just interesting to hear how this stuff works because, because Moloch even talks about how, like, this sort of Claude code cheats its way through, like, the limits of the context window.
他说，当Claude代码用尽上下文时，它会暂停并压缩到目前为止的对话，记录下停止时的确切位置。
He says, when Claude code runs out of context, it stops and compacts the conversation so far, taking notes about exactly where it was when it stopped.
然后它清空上下文窗口，全新的Claude代码读取这些笔记，回顾迄今为止的进展，并利用这些笔记让Claude继续前进。
Then it clears its context window, the fresh version of Claude code reads the notes and reviews the progress to dates, and then, it gives it those notes give Claude everything it needs to keep moving.
这就是为什么Claude能够连续运行数小时的原因。
This is why Claude can run for hours at a time.
它会仔细记录自己在过程中所做的工作，生成中间成果，比如软件片段和报告，以便后续参考。
It carefully notes what it's doing along the way, produces interim work, and, you know, like pieces of software and reports that it can refer to.
我的意思是，这些公司设计这些系统的方式非常有趣。
I mean, this is this is very interesting the way that these companies are architecting stuff.
是的。
Yeah.
我的意思是，任何两年前或一年半前使用Claude的人，都会遇到那个令人头疼的问题：即使你是付费的Claude Pro用户，也会用完积分，不得不等待十二个小时才能再次在顶级模型上运行提示。
I mean, this is where anyone who was using Claude two years ago, year and a half ago, and would hit that dreaded, like, you've run out of credit even as a paying Claude Pro subscriber and have to either wait twelve hours to run a prompt on the leading model again.
像这样的代理工作流，其架构层面的问题正在得到解决。
Like, the architecture side of how people are approaching these kind of agentic workflows is getting solved.
我认为这才是关键。
I think that's the big thing.
并不是模型层面在提供帮助，而是——没错。
And it's not at the model level that models are helping, but and it's yeah.
其实也不是产品本身。
It's the it's not the product, actually.
你知道吗？
You know what?
与其说是产品与模型架构的对比，不如我们再加入一个层面：那就是可以被优化或创新的架构层，真正推动这些技术向前发展。
Instead of product versus model architecture, let's throw that in there as well as another layer that can be optimized or innovated on to actually to really move this stuff forward.
所以你是支持架构派的吗？
So are you team architecture?
我认为我是支持架构派的。
I think I am team architecture.
我的意思是，这应该是我们在微软内部讨论的问题。
I mean, this is something we should be discussing on Microsoft.
你提到的拉里或谢尔盖，他们谈到了架构。
Larry or Sergei that you talked about architecture.
嗯，谢尔盖谈的是算法改进。
Well, Sergei talked about algorithmic improvements.
好的。
Okay.
所以我认为这个谢尔盖。
So I think this Sergei.
架构。
Architecture.
是的。
Yeah.
是的。
Yeah.
我会称之为架构。
I would call it architecture.
而且它也开始在其他地方出现了。
And and and it's starting to show up in other places too.
这是云桌面。
Here's Cloud Cloud Desktop.
所以这是Anthropic最近发布的一个新功能。
So this is a relatively new thing that that Anthropic has released.
如果你是每月20美元的付费用户，就可以把Claude放在你的桌面上，马利克说，只需给AI访问某个文件夹的权限。
You can put Claude on your desktop, if you're in the $20 a month bracket of the paying users, and, Malik says just give the access just give the AI access to a folder.
记住，Claude可以对那个文件夹中的任何文件进行操作，所以如果文件敏感，请小心并做好备份，然后你就可以开始与AI协作了。
Remember that Claude can do anything, to the files in that folder, so be careful and make a back if it's if it's sensitive and make a backup and you can start working with AI.
让它进行研究并撰写报告，给它访问你的信用卡记录的权限，以便它能把数据整理到电子表格中并告诉你异常情况，让它做数据可视化，或者任何你想要它做的事。
Have it research and write reports, give it access to your credit card records so it can put them into a spreadsheet and tell you about anomalies, ask it to do a data visualization, whatever else you like.
好的，拉詹。
Alright, Ranjan.
所以，作为我们这里的AI代理专家，我想问你一个问题。
So as our as our resident AI agent expert, I'm gonna ask you a question about this.
我们真的需要通过Claude Code来运行这些内容吗？还是可以直接上传到Claude聊天机器人本身的上下文窗口中？
Do we do we really need to, like, be running this stuff through co Claude code, or can you just, like, upload it into the, into the context window of, like, the Claude chatbot itself?
通过Claude Code能带来什么优势？
Like, what does going through Claude code give you?
是因为它能为你即时构建东西，从而提供更强的能力吗？
Is it just more power because it can build things on the fly for you?
能再详细说说为什么这很重要吗？
Tell me a little bit more about why this is important.
是的。
Yeah.
它之所以重要，是因为能够实现多步骤操作。
The it's the ability to again, it's multistep.
这不是单一的动作。
It's not one action.
而是一系列操作，特别是当你给出总体指导和行动方向，并从结果角度来思考时，你就不再是告诉Claude：你该去做什么。
So it's a set of, like, especially giving that general guidance and action, thinking it from thinking of it from an outcome perspective, it's no longer telling Claude, here's what you should go do.
而是说：这是我想要的。
It's saying, here's what I want.
而要达到这个目标，这与典型的聊天对话方式完全不同。
And to get there like, that's not how a typical chat back and forth works.
典型的聊天对话是让我们一起迭代。
A chat back and forth is let's iterate together.
让我们一起探索。
Let's explore.
让我们尝试理解，而不是：我知道我想要什么。
Let's try to understand versus I know what I want.
替我去达成目标。
Go get there for me.
所以我认为，这种能力在于能够结合实际的推理部分——如果需要的话，它可以创建出确定性的、会被使用的硬编码 Python 脚本。
So I think and that ability to combine whatever the actual kind of, like, reasoning side, if it needs to to to actually create, like, hard coded Python scripts that are deterministic and are gonna be used, it can do that.
不把限制局限于聊天窗口，这才是最大的不同。
Like, not limiting it to that chat window is the biggest difference here.
而且，再次强调，就我自己的公司而言，我们在 Ryder 发现，最大的变化在于日常事务方面。
And again, increasingly, and again, talking about my own company here, like, we see at Ryder, people the biggest change is the routine side of it.
这些事情不再是你出于兴趣或作为实验偶尔为之的临时行为。
It's these things are no longer gonna be one off things you do for fun and kind of try as an experiment.
你会开始建立常规流程，让这类工作持续不断地进行。
You start to set routines so this kind of work is being done on an ongoing basis.
所以，这不再是我要进去上传一份文档。
So it's not, I'm gonna go in and upload a document.
它实际上是在后台自动进行的。
It's actually happening in the background.
而且，Claude 的界面会被抽象掉。
And, like, the the Claude UI gets abstracted away.
你可能只是收到一封邮件或通知，里面包含各种结果。
Like, you maybe you just get an email or a notification with whatever results.
所以我认为这带来了巨大的
So so I think it it gives a big
差异，在你的例子中。
difference in your examples.
举一些具体的例子，比如这类例行工作是
Give some concrete examples of of, like, the type of routines that
我的意思是，别太深入企业销售的世界，但就目前工作而言，我们使用 Salesforce。
I mean, like, not to get too into the enterprise sales world, but, like, right now at work, we have, like, Salesforce.
我们有 Gong，这是一种通话录音和数据类型的工具。
We have Gong, which is like a call recording and data type thing.
我们还有其他一些系统，我们有这么多不同的系统。
We have something called we have all these different systems.
我确实有一个客户账户列表。
I literally have a list of accounts.
过去，我每天早上都会逐一查看。
In the past, I would go through every morning.
我通常会查看每个账户的报告或仪表板。
I would either have, like, a report in each of one or a dashboard.
我真正做了一件事，就是创建了一个系统，可以进去查看。
I literally was able to create something where it's like, go in.
这里有一份账户列表。
Here's a list of accounts.
过去24小时内发生了什么？
Did anything happen in the last twenty four hours?
发生了什么？
What happened?
把它总结一下。
Summarize it.
发邮件给我。
Email me.
而且，我构建了这个。
And, like, I built that.
这整个就是一个应用程序。
That's a whole app.
里面有很多代码、API调用和MCP连接器，我现在就能做这些事。
There's tons of code in there and API calls and MCP connectors, and I can do that stuff now.
所以，这就是关键所在；我认为我算是个有点技术背景的人，但我的意思是，任何人都能开始做，尤其是那些技术门槛更低的人。
That, like so this is where and I think I'm a reasonably technical person, but, like, I mean, anyone, especially simple simpler levels of this will be able to start doing.
好的。
Okay.
这很好地引出了马利克在帖子末尾的那段话。
And this is that's a good good segue to the end of the post here from Malik.
他说：今天的AI能够完成真正持续且重要的工作，进而开始改变我们处理任务的方式。
He says, today's AIs are capable of real sustained work that actually matters and in turn is starting to change how we approach tasks.
他说，这种变化出人意料地从编程开始。
He says it is starting unsurprisingly with programming.
AI领域最著名的程序员之一安德烈·卡帕西最近发帖说：‘我从未感到自己作为一名程序员如此落后。’
One of the more famous coders in AI world, Andrej Karpathy, recently posted I've never felt this much behind as a programmer.
随着程序员所贡献的代码片段变得越来越稀疏，这一职业正在被大幅重构。
The profession is being dramatically refactored as the bits contributed by the programmer are increasing increasingly spare in between.
我有一种感觉，如果我能恰当地整合过去一年左右出现的所有工具，我的效率可能会提升十倍；而未能抓住这一提升机会，明显是一种技能问题。
I have a sense that I could be 10 x more powerful if I just properly string together what has become available over the last year or so, and a failure to claim the boost feels decidedly like a skill issue.
马利克总结道：不要被当前Claude代码的笨拙或其对编程的专门化所迷惑。
Malik concludes, don't let the awkwardness of the current Claude code or its specialization for coding fool you.
天啊，新的……
New oh god.
不久的将来，将出现新的工具，让AI能够应用于其他知识型工作任务，并随之带来变革。
New harnesses that make AI work for other knowledge work tasks are coming in the near future, and so are the changes they will bring.
天啊。
Oh my god.
抱歉。
Sorry.
好的。
Okay.
等一下。
Hold on.
第一，我非常喜欢Karpathy的这句话，所以我在努力克制自己，或者在再次看到‘工具’这个词后重新镇定下来。
One, I love this quote from Karpathy, so I'm trying to contain myself or regain my composure after seeing harnesses again.
但好吧。
But okay.
Karpathy。
Carpathi.
我非常喜欢这句话。
I loved this quote.
我觉得这句话是在假期期间发布的，当时我在家，通过Twitter看到了，然后开始越来越多地看到它，因为这正是我作为一个技术背景但本质上仍是知识工作者的感受。
I think it came out over the holidays when I was at home, caught it on Twitter, like, started seeing it more and more because, like, this is how I feel as a fairly technical, but still knowledge worker at heart.
当我看到自己能做的事情时，我简直惊呆了。
Like, I feel when I'm seeing what I'm able to do, I feel like, holy shit.
我觉得自己可能落后了，或者我不确定别人是不是比我学得更快，他们可能会远远甩开我。
I I I think I'm falling behind or, like, I don't even know if other people are getting this faster than me, that they're gonna be way ahead.
所以，当你开始看到这些可能性时，这种感觉就一直萦绕在我心头。
And so I that feeling was kinda, like, lingering in me the moment you start to see what's possible.
当我看到他这么说时，尤其是来自世界顶尖程序员之一的人，这让我们觉得，现在这些东西实在太有趣了，连他这样的人也会觉得，我得赶紧掌握它。
And I think, like, when I saw him say that, especially about programming from one of the leading world's leading programmers, like, it starts to make us feel like this stuff is so interesting right now that even someone like him can actually feel like, I better get a handle on this right now.
对。
Right.
听我说，我就直说了。
And look, I'll I'll just say this.
我并不完全认同这就是我们未来的方向。
I'm not completely sold that this is where we're going.
我知道你坚信不疑。
I know you feel firmly.
你也在投资，你知道的。
You're also like, you know, you're Investing.
你有内在的偏见。
You have a built in bias.
你是对的。
You're Yeah.
这是你工作的地方。
It's where where you work.
我不是，但关键是。
I'm not but here's the thing.
我已经看到了围绕能力提升的大量能量，就像我们在这段对话开头谈到的那种，也是马利克在解释当前情况时所列举的那种。我认为马利克在这篇帖子中已经清晰地指出了这一点，忽视它将是愚蠢的，就像马利克基本在说：别被这种事只是与编程相关所迷惑。
I've seen enough energy around the increases in capabilities, the type that we spoke about, at the beginning of this conversation, the type that, Malik listed as he explained what's going on, And I think that that has been clearly drawn out by Malik in this post that I think it would be foolish to ignore it the same way that Malik is basically saying, like, don't be fooled by this being just a coding specific thing.
很快，我们可能会拥有使这类工作自动化、自我修正、调用工具的界面，让许多其他类型的知识工作职业也能使用。
Soon, we'll probably have the interfaces that will make this type of work, autonomous work, things that self correct, things that call tools, available for for many, many more different types of knowledge work professions.
这就是为什么我认为强调这一点很重要。
And that's why I think that this is important to highlight.
我的意思是，我们显然正在起步，一年只是时间上的一个划分，但2026年我们已经进入了一个我认为这个领域此前经历了大量变动的阶段。
I mean, we are getting obviously, we're getting you know, a year is just a delineation in time, but we are getting started here, in 2026 after what I think was a lot of movement in this space.
显然，大量资金投入了进来。
Obviously, lot of money went into it.
我们一直对这些公司的财务状况持批评态度，但我不认为这应该蒙蔽我们的眼睛，我们在这里讨论的正是这种细微差别。
We've been very critical about the financial picture of some of these companies, but I I think that shouldn't blind us to the fact and again, we're here for like nuance about this.
我们不应该忽视一个事实：这项技术正在以一些非常有趣的方式进步，而我们节目的使命也是提醒大家：注意这一点，因为它有可能变得非常重要。
That shouldn't blind us to the fact that, this technology is progressing in some very interesting ways and our mission on the show is also to be able to be like, hey, Pay attention to this because it's something that we think has the potential to be big.
我认为这里的情况正是如此。
And I think that's the case here.
我也这么认为。
I think so too.
我要在这里下个赌注。
And I'm gonna make the bet here.
现在，每周亚历克斯和我都会共享一个谷歌文档，把链接和文字粘贴进去。
Right now, every week, Alex and I have a collaborative Google Doc that we paste links in and kinda copy and paste text into.
我认为，几个月后，我们不会再以那种方式工作，而是会有一个更有趣的过程。
I think we are going to not be working in that way and have something more interesting as a process in a few months' time.
我敢打这个赌。
I'm making that call.
我会认真听的。
I will I will take the the listen.
我刚刚可能谈到了我对这项技术力量的信念。
I may I may have just talked about how I believe in the power of this technology.
我不会反对使用 Google Docs。
I I will not bet against Google Docs.
我不会这么做。
I won't do it.
我们还是会做同样的事情，我不会改变我的流程。
We're gonna be doing the same thing, and, and I will not change my process.
这就是我的赌注。
That's my that's my bet.
我很乐意粘贴一个链接。
Love to paste a link.
实际上，粘贴链接这一行为是人类最自然的举动之一。
It's it's actually the act of pasting the link is is the most human thing one can do.
我同意。
I'll agree.
另外，我的意思是，我觉得也不必过度聚焦于这个谷歌文档的讨论，但为什么不呢？
Well, also, I mean, I think it's sort of the just to not to over index on this Google Doc discussion, but why not?
我的意思是，这也是个问题。
I mean, this is the problem too.
人们有自己的流程。
People have processes.
他们喜欢自己的流程，要让他们放弃，需要一个极其优越的流程和系统，而随着时间推移，这可能是阻碍这项技术发展的最大因素——人们就是喜欢自己的流程，如果你跟他们说：‘嘿，这里有个全新的智能方式来规划你的节目或处理任务A、B、C’，他们的第一反应绝不会是：‘太棒了！’
They like their processes, and it takes an unbelievably better process and system to be able to get them to switch off, and that's probably the biggest thing that will hold this technology back over time is just the fact that, like, people like their processes, and if you say here's this brand new agentic way to do your, you know, show planning or take tasks a, b, or c, their their immediate reaction isn't gonna be like, oh, hooray.
我们还是换过来吧。
Let's let's switch over.
他们的原始回答是不行。
Their original No.
你得说：除非我死了。
You got be like, over my dead body.
我的意思是，说实话，这是最大的问题。
You you got I mean, honestly, this is the biggest thing.
技术已经存在了。
The technology is there.
我早就说过这一点了。
Like, I I I've been saying it for a while.
技术已经存在了。
The technology is there.
这是个变更管理问题，是人机交互的问题，就像你说的那样。
It's a change management person, human UI, whatever problem, but, like, exactly as you're saying.
我们有一个流程。
We have we have a process.
这是一个很棒的方案。
It's a beautiful one.
它每周都带我们来到这个节目，并且真正实现了目标。
It brings us to this show every week, and and it gets it done.
但没错，人们仍然依赖着旧方式。
But, yeah, people are tied to prosties.
所以，我认为这将是今年最有趣的部分。
So, yeah, I think that's gonna be the most interesting part of this year.
但就我所见，趋势已经逐渐显现：当越来越多的人了解到有更好的方法时，尤其是早期采用者，都会想：我至少得试试。
But, like, the way I see it already kinda going is the more people hear about there are better ways to do things, the more you especially early adopters, you're like, I gotta at least try.
也许技术会重新发明世界，或者AI会重新发明世界上最古老的技术——电子邮件。
And maybe technology will reinvent, the world or maybe AI will reinvent the world's oldest technology, and that is, of course, email.
也许它会改变那里的流程。
And maybe it will change processes there.
这是来自TechCrunch的消息。
This is from TechCrunch.
Gmail推出个性化AI收件箱、搜索中的AI概览等功能。
Gmail debuts a personalized AI inbox, AI overviews in search and more.
谷歌已为Gmail推出全新的AI收件箱，旨在为您提供个性化的任务概览，并及时通知重要更新。
Google has unveiled a new AI inbox for Gmail that's designed to provide a personalized overview of your tasks and keep you informed about important updates.
谷歌还在搜索中推出AI概览功能，并增加类似Grammarly的校对功能。
Google is also launching AI overviews in search and a Grammarly like proofread feature.
此外，Gmail将多项此前仅限付费用户的AI功能向所有用户开放。
Additionally, Gmail is bringing to all users several AI features that were previously available only to paid users.
新的AI收件箱——我认为这是最有趣的消息所在——包含两个部分：建议待办事项和需要跟进的主题。
The new AI inbox, and this is I think where most interesting news is coming, The new AI inbox tab features two sections, suggested to dos and topics to catch up on.
第一个部分显示需要采取行动的高优先级邮件摘要，例如提醒您明天有账单到期，或需要联系皮肤科医生确认邮寄地址以便寄送处方续药。
The first section displays summaries of top priority emails that require an action such as a reminder that you have a bill due tomorrow or that you need to call your dermatologist to confirm your mailing address so they can ship you your prescription refill.
在‘需要跟进的主题’下，您会看到诸如您的Lululemon退货已处理完毕，或Wealthfront的年终对账单现已可用等更新。
Under the topics to catch up on, you'll see updates such as your Lululemon return has been processed or the end of the year statement now available from Wealthfront.
您对这个有什么看法？
What do you think about this?
这不就是Gmail现有的个性化收件箱吗？
Is this, is this basically the same the same personalized inbox that we have from Gmail?
我的意思是，Gmail之前就有主要邮件和更新邮件的分类。
I mean, Gmail has prior primary and updates.
这是同样的东西，还是你觉得上面叠加的AI层真的会改变我们的邮件方式？
Is this the same thing, or do you think the AI layer on top of it is gonna actually change the way we email?
我非常喜欢这篇文章，很高兴我们正好现在提到它，因为我创建了一个系统，这个系统其实是当时在《金融时报》的同事斯泰西·玛丽·伊斯梅尔教给我的。
So I love this article, and I'm glad we just got to it right now because I created a system taught to me actually by my coworker at the time of the Financial Times, Stacy Marie Ismail.
通过Gmail的标签，你可以创建自定义标签，而不仅仅是简单的星标或不星标。
With Gmail flags, you can create, like, customized instead of just the star or not star.
你可以有五种不同的选项，建立一个分层的优先级系统，我有自己的体系。
You can have there's, like, five different options And, like, creating a tiered ranking system, and I have my system.
这个体系是2011年建立的。
It's from 2011.
直到今天，我仍在Gmail中使用它。
I'm still using it in Gmail today.
展开剩余字幕（还有 399 条）
我关闭了主要收件箱和所有这些功能。
I turned off primary inbox and all of that.
所以我很想知道我是否会开始使用这个。
So I am curious to see whether I start to use this.
但另一方面，电子邮件确实需要改变。
But on the other hand, like like, email has to change.
我们都清楚，电子邮件的实际价值到底在哪里？
We all know that email, like, what's the value of the actual email?
有没有更好的方式来处理？
Is there a better way to do it?
所以我很高兴他们至少在这里尝试了一些新东西。
So I I'm glad they're at least trying something here.
嗯，这正是你之前提到的、非常有用的那种版本，它能够对邮件进行分类排序。
Well, is kind of exactly the version of what you were talking about before that has been so useful is that it sort of sorts through
是的。
Yeah.
嗯。
Yeah.
嗯。
Yeah.
我想，你的收件箱会对你收到的信息进行筛选、排序和总结，这样你就不用像以前那样去逐一处理每个系统了。
I guess the action in your inbox and prioritizes it for you and summarizes it so you don't have to like similarly, you don't have to go through every system.
这个功能可能让你根本不需要逐封阅读每封邮件。
This one, maybe you just don't have to go through every single email.
嗯。
Yeah.
不。
No.
不。
No.
所以刚才我笑了，因为事实上，即使现在给了我这个新功能，我可能还是会继续用我用了14年的旧系统，这多少也算是对你之前观点的一种认可吧。
So so that's why I was laughing because, like, fact that I probably will stick to my 14 old system even though this has just been offered to me to to to give you a little bit of credit or from your earlier point there.
但没错。
But yeah.
我认为，这对谷歌来说是一个非常难得的机会，对于那些还没有使用Gemini的人，可以借此向他们介绍AI，或者介绍这种更具代理性的AI，因为这就在他们的收件箱里。
I I think, like, the other interesting part of this to me is it's such a rich opportunity for Google, for anyone who has not been using Gemini to kind of introduce AI to them or introduce this kind of more agentic AI because, like, this is in their inbox.
电子邮件应该非常直接，它主要就是文本。
Email should be a very straight it's just it's mostly text.
它非常结构化。
It's very structured.
有主题行，而LLM应该能很好地处理这类内容。
There's subject lines, and, like, it's something LLMs should be able to do a very good job of handling.
它包含大量上下文。
There's a lot of context.
你可以想象，如果这成为你触达接下来数亿Gmail用户的方式。
And you imagine if this starts to be the way you get the next I don't even know how many hundreds of millions of users Gmail has.
是数十亿吗？
Is it billions?
十亿以上。
Billion plus.
十亿以上。
Billion plus.
当然。
Sure.
突然间，可能达到二十亿。
Suddenly probably 2,000,000,000.
我的意思是，这太荒谬了。
I mean, it's it's absurd.
所以这是一个巨大的成功。
So It's big success.
突然间，所有这些用户 worldwide 的活跃用户将超过二十亿。
Suddenly, all of these people will start to over 2,000,000,000 users active worldwide.
突然间，所有这些人都会以这种方式作为入口，将 Gemini 推广给这些人，并开始让他们思考：既然我们已经为你总结了这些内容，你是否想去做点什么？
Suddenly, all of them like, as an entry point to bring Gemini to all these people and start to make them think here's how to use, like, slightly more agentic AI, adding more, like, now that we've summarized this for you, do you want to go do something?
实际上，考虑到我们上周关于Manus和Meta的讨论，既然Manus是他们进入消费市场的‘特洛伊木马’，也许这正是谷歌进入真正消费级智能代理领域的切入点。
Actually, given our earlier conversation about Manus and Meta last week, as Manus is their kind of like Trojan horse into consumer, maybe this is Google's entry point into consume like, true consumer agentic.
好的。
Okay.
我其实收到一封读者的有趣邮件，正好为这次关于Manus的讨论画上句号，我觉得值得一读。
I actually got an interesting, email from a reader just to, like, put a bow on this manus conversation that I think is just worth reading.
这位读者说，请记住，我们上周又谈到了Meta收购了AI代理公司Manus，我当时想，也许这会成为他们融入消费者聊天机器人的重要一步。
The reader says keep in mind that this is we were talking again last week about Meta acquired Manus, the AI agent company, and I was like, oh, maybe that's actually gonna be the thing that they start to bake into their, consumer chatbot.
我收到一些读者反馈——虽然不算多，但确实有人认为，这本质上是个企业级产品。
I got a lot of, like, well, not a lot, but some reader feedback being like, no, this is an enterprise thing.
所以让我念一下这封邮件，然后我们休息一下。
So let me just read that then we can take our break.
这是一条很棒的信息。
It's a great message here.
读者（或听众）来信说：当你做播客时，请记住，Meta的客户是那些在平台上做广告的企业，而不是普通用户。
The reader says, keep in mind or listener actually, listener mail when you do a podcast, keep in mind that Meta's customers are in the business of advertising on its platform, not the users.
归根结底，它将企业与客户匹配起来，企业越多地通过简化业务运营和触达客户来加入平台，平台就越成功。
At the end of the day, it matches businesses with customers, and the more businesses it can pull into the platform by making business operations, reaching customers easier is the better.
泽卡之前曾表示，我们的目标是让任何企业都能告诉我们他们想实现的目标，比如销售产品或获取新客户，以及他们愿意为每个结果支付多少费用，然后我们来完成其余部分。
Zekka has previously said our goal is to make it so that any business can basically tell us what objective they're trying to achieve, like selling something or getting a new customer, and how much they're willing to pay for each result, and then we do the rest.
我想象NGENTIC工作流在实现这一愿景中将至关重要。
I imagine NGENTIC workflows will be super important in achieving that vision.
Meta正试图捕捉并自动化从产品构思到销售的整个流程。
Meta is attempting to capture and automate the entire funnel from the product ideation to sale.
好的。
Okay.
这是一个我们上周没有涉及的有趣替代视角，我认为很有道理。
That's an interesting alternative perspective that we didn't cover last week that I think makes a lot of sense.
是的。
Yeah.
我觉得这很合理。
I think I think that's reasonable.
再者，无论在这些领域的哪一个，驱动整个公司的商业模式都会对事情产生影响。
Again, like, in any of these areas, yeah, business model that drives the entire company is going to kinda influence things.
我想，如果你仔细想想，Facebook Workplace，还记得那个吗？
I guess if you think about it, like, Facebook Workplace, remember that?
或者它还存在吗？
Or does it still exist?
是的。
Yes.
我认为他们已经关闭了。
I think they shut it down.
好吧。
Okay.
这其实是一个完美的例子，说明Facebook以及后来的Meta本应有能力在这个领域做些非常有趣的事情。
Like, the that actually is a perfect example of there is no reason Facebook and then Meta should not have been able to do something really interesting in that space.
这是一个非常有利可图的领域。
It's a very lucrative space.
他们本来完全有能力做到这一点，但却没有做。
There's no reason, like, they should not have been able to do that, but they didn't.
我认为，商业模式以及它如何驱动整个组织，正是问题的核心。
And I think, like, the business model and just how that drives the entire organization is kinda at the core of that.
但话说回来，谷歌即使在广告业务的基础上，也成功打造了庞大的谷歌云和谷歌工作区。
But saying that Google did manage to do that even with an advertising business and then building the behemoth of Google Google Cloud and Google Workspace.
所以你可以看到，确实如此。
So but you can see how yeah.
核心商业模式究竟如何影响整个努力方向呢？
Like, what's the core business model actually influencing the entire, like, effort?
对。
Right.
好的。
Alright.
嗯，ChatGPT OpenAI 正在进军医疗领域，推出了 ChatGPT Health 和面向医疗保健的 ChatGPT。
Well, ChatGPT OpenAI is getting into health care with ChatGPT Health and ChatGPT for health care.
我们将在接下来的内容中探讨正在发生的事情、其中的利害关系，以及关于忙碌工作终结的一段有趣讨论。
We'll cover what's happening and what the stakes are along with a fun little discussion about the end of busy work right after this.
你想吃得更健康，但你既没有时间，也没有精力去实现。
You wanna eat better, but you have zero time and zero energy to make it happen.
Factor 不要求你提前备餐或遵循食谱。
Factor doesn't ask you to meal prep or follow recipes.
它只是彻底解决了这个问题。
It just removes the entire problem.
两分钟，真实食物，搞定。
Two minutes, real food, done.
还记得那次你想做健康餐却点了披萨吗？
Remember that time where you wanted to cook healthy but ordered pizza?
你并不是在健康饮食上失败了。
You're not failing at healthy eating.
你只是失败在每晚多出三个小时上。
You're failing at having three extra hours every night.
Factor 的餐食由厨师精心制作，由营养师设计，并直接配送到您家门口。
Factor is already made by chefs, designed by dietitians, and delivered to your door.
您只需加热两分钟即可享用。
You heat it for two minutes and eat.
里面包含优质蛋白质、色彩丰富的蔬菜、全食物成分和健康脂肪——这些都是您如果有时间会自己做的东西。
Inside, there are lean proteins, colorful vegetables, whole food ingredients, healthy fats, the stuff you'd make if you had the time.
前往 factormeals.com/bigtech，使用代码 big tech 50 off，即可享受首份 Factor 盒餐 50% 折扣，外加一年免费早餐。
Head to factormeals.com/bigtech 50 off, and use code big tech 50 off to get 50% off your first factor box plus free breakfast for one year.
该优惠仅适用于使用该代码的新 Factor 客户，并需购买符合条件的自动续订订阅。
The offer is only valid for new factor customers with the code and qualifying auto renewing subscription purchase.
用 Factor 让健康饮食变得更简单。
Make healthier eating easy with Factor.
我们回到《大科技》播客周五版，为您解析本周的新闻。
And we're back here on Big Technology podcast Friday edition breaking down the week's news.
本周的大新闻是 ChatGPT 或 OpenAI 发布了 ChatGPT Health 和面向医疗保健的 ChatGPT。
Big story this week, ChatGPT or OpenAI release ChatGPT Health and ChatGPT for for health care.
因此，这将既有面向消费者的部分，用户可以与机器人对话以获取专业的医疗服务，也有面向企业的版本，通过API为医疗保健机构量身定制。
So there's both going to be a consumer side of things where you can speak, to the bot for specialized health care services, and then more of a, health care version of it that is tailored for business, healthcare businesses on on the API side.
那么，我们一个一个来。
So, let's let's go one by one.
首先，这是ChatGPT Health。
First of all, this is ChatGPT Health.
这里有一个惊人的数据。
Amazing stat here.
全球每周有超过2.3亿人向ChatGPT咨询健康和保健相关的问题。
Over 230,000,000 people globally ask health and wellness related questions to ChatGPT every week.
ChatGPT Health是一个新产品，基于这一趋势，其回复会结合您的健康信息和上下文进行优化。
ChatGPT Health is a new product that builds on this so responses are informed by your health information and context.
现在您可以安全地连接医疗记录和健康应用，如Apple Health、Function和MyFitnessPal，让ChatGPT帮助您理解最新的检查结果、为看医生做准备、获得关于饮食和锻炼计划的建议，或根据您的医疗模式理解不同保险方案的利弊。
You can now securely connect medical records and wellness apps like Apple Health, Function, and MyFitnessPal so ChatGPT can help you understand recent test results, prepare for appointments with your doctor, get advice on how to approach your diet and workout routine, or understand the trade offs of different insurance options based on your health care patterns.
它作为一个独立的空间运行，具备增强的隐私保护机制，以保障敏感数据的安全。
It operates health operates as a separate space with enhanced privacy to protect sensitive data.
健康相关的对话不会用于训练我们的基础模型。
Conversations in health are not used to train our foundation models.
如果你在ChatGPT中开始一个与健康相关的对话，我们会建议你转到健康模式，以获得这些额外的保护措施。
If you start a health related conversation in ChatGPT, we suggest we'll suggest moving to health for these additional, protections.
我不知道你怎么样，拉詹，但我已经把ChatGPT当作一个健康顾问来使用了。
I don't know about you, Ranjan, but I have been using ChatGPT already as a, health sort of consultant, so to speak.
我觉得这些功能都非常棒。
I think these are these are great.
我有一块佳明手表。
I'd love I have the Garmin watch.
我希望能将它连接起来。
I'd love to be able to connect connect it.
我平时只是截取手表和应用的屏幕截图，然后直接发给ChatGPT，和它讨论这些内容。
I basically just been taking screenshots of of the watch and the app and just dropping it into Chateappity and speaking about it.
我认为这是OpenAI一个巨大的潜在增长领域。
I I think this is a huge potential area for growth of growth for OpenAI.
有点欢迎这个。
Kinda kinda welcoming this.
你对这个新功能有什么看法？
What what do you think about this this new offering?
不。
No.
不。
No.
从消费者（或客户）层面来说，我非常喜欢。
I I at a consumer level, call it customer level, I love it.
这很有趣。
I it's funny.
我实际上用Claude来回答所有与健康相关的问题。
I actually use Claude for all my health related questions.
我不知道为什么。
I don't know why.
我只是觉得在这件事上，用Claude比用ChatGPT更舒服。
I just kinda feel more comfortable with Claude than ChatGPT on that.
但有趣的是，每个人都在做这件事。
But, like, I it's the interesting part to me is, again, everyone is doing this.
我喜欢他们真正试图把它产品化，让它更简单易懂，把它作为一个独立的产品来呈现。
I like that they're actually kinda trying to make it more productized and make it, like, simpler to understand, kinda presenting it as this separate offering.
当我看到这个时，我本来就想，这对我的Oura戒指有什么影响？我每个月付6到7美元订阅它，它们本该为我分析数据。
I was already, when I saw this, thinking more like, what does this do to an Oura Ring, which I'm paying $7 $6 a month for a subscription, and they're supposed to be analyzing my data for me.
而实际上，我确实也在做类似的事，我会把Oura的数据截图发到Checked或者Claude里，然后问问题。
Whereas in realize, I do reality, I do kind of the same thing where I just screenshot stuff from Oura into checked or into Claude and asked questions.
比如这些Apple Health，是的，帮我追踪数据，但这个应用从来就没好用过。
Like, all these like, Apple Health, yes, track it for me, but that app has never been that good.
MyFitnessPal这些应用，我觉得都会因此遇到一些麻烦。
MyFitnessPal, all these start to get in a little bit of trouble with this, I think.
ChatGPT有八亿周活跃用户。
ChatGPT has 800,000,000 users, weekly users.
整整四分之一的人已经在使用ChatGPT来咨询健康问题了。
Fully one fourth of them are using ChatGPT for, for health questions already.
是的。
Yeah.
这真是个不小的数字。
Which that's a big number.
但你觉得这对看医生的流程会有什么影响呢？
But what do you think this does, like, at the doctor visit level?
因为，同样的事情。
Because, like, same thing.
我去看了医生。
Went to the doctor.
每年体检时，他们会告诉我一些东西。
They for yearly physical, they, like, tell me some stuff.
但随后我就把所有检查结果直接发给Claude，上传后咨询，效果好得多——这里就不提我医生的名字了。
But then I just took all the results and just went straight to Claud and uploaded and had a much much won't mention my doctor's name here.
他们很好。
They're fine.
他们真的没问题。
Like, they're they're fine.
但我通过AI聊天获得了更深入、更好的了解，感觉更安心，一切都更顺利。
But I got a much, much deeper, better in learning and felt more comfortable and just everything with AI chat.
是的。
Yeah.
我的意思是，这显示了我们对这些聊天机器人有多么信任。
I mean, I think this kinda shows how trusting we've become of these bots.
我直接把我的血液检测结果上传到了
Like, I dropped my blood labs right
Jetsypti。
into Jetsypti.
我做了Aura的血液检测，是的。
I did the aura aura blood labs in Yeah.
第一名，没错。
First place, yeah.
太棒了。
It was amazing.
你知道，医生会给你发一封邮件，通过门户网站告诉你一切正常，注意胆固醇之类的。
You know, the doctor gives you like an email and through like the portal it says everything looks good, watch cholesterol or something like that.
但这个系统会逐项解释每个数据点的意义，并结合你的年龄、健康状况，甚至如果你跟它聊过锻炼情况，它也能知道，从而给你一个全面的健康画像，还可能建议你做其他检查。
This goes, you know, data point by data point explaining what each one of them are and sort of, you know, taking into account your age, your health status, it already knows when you work out if you talk to it about that, and gives you, like, this full picture of your health, maybe it will suggest other tests to order.
所以，这就是我觉得事情会如何发展的。
So here's kind of how I think this goes.
就像我们之前做过一期关于AI和法律的节目，AI可以作为律师，把所有文件都读一遍，有时能帮你得出答案，有时对有时错，但它确实让你能完成比你作为人类逐字逐句处理这些数据多得多的工作，对吧？
It's just like we had this like episode about AI and law a little while ago that like AI can can, if you're a lawyer, AI can, you know, take in all your documents and sometimes get you to an answer, sometimes it's right and sometimes it's wrong, but it definitely enables you to do more than if you were to code by this stuff, you know, minute by minute as a human user, right, or a human person trying to get through these this data.
我认为医疗领域也会发生同样的事，比如ChatGPT经常能诊断你或给你一些需要考虑的建议，而且它比WebMD强多了。
I think the same thing is gonna happen in health that like ChatGPT will often, like, diagnose you or give you things to to consider, and and it does a much better job than WebMD.
我的最爱例子是WebMD，有一次我喉咙有点痒，WebMD就说：好吧。
I mean, my favorite example about WebMD is, like, I once had a scratchy throat and WebMD is like, alright.
这里有一些可能性，也许是埃博拉。
Some possibilities here, maybe Ebola.
我当时想，天哪。
And I was like, goddamn.
我不认为 ChatGPT 会这么做。
I I don't think chat chip PT does that.
所以我认为它会给你提供更好的医疗建议。
So I think what what it's gonna do is it will give you some better, medical advice.
有时它会给你做出诊断。
Sometimes it will diagnose you.
我认为关键在于把信息带给医生。
I think the key then is to bring that information to a human doctor.
对吧？
Right?
我认为接下来会发生的是，医生会看到带着 ChatGPT 日志前来就诊的人，然后开始与人类和机器人一起讨论发生了什么。
And I think that's what's gonna happen is that, like, doctors will see people coming in with their chatty PT logs, and they'll sort of start to consult with the human and the bot about what's happened.
当然，这种人类的专业知识非常宝贵，但我确信我们会看到越来越多的人走进诊所，说：嘿。
And, obviously, that human expertise is pretty invaluable, but I I am certain we're gonna see much more people in the clinic coming in and being like, hey.
你知道，能不能再详细讲讲这个CHECHYPT提出的可能性？
You know, please tell me a little bit more about this possibility that CHECHYPT suggested.
这样听起来对吗？
Does that sound right?
实际上，哦，我喜欢这一点。
Actually, oh, and I like that.
我觉得我同意。
I think I agree.
我认为2026年最大的行为变化将是人们走进医生办公室时掏出聊天PT。
I think the big behavioral change in 2026 is gonna be people going into the doctor and pulling out chatty PT.
而就在一年前，我根本无法想象人们会如此自信地这样做。
Whereas, like, certainly a year ago, I can't picture people actually doing that with confidence.
也许他们确实这么做了，但我是说，我曾经在去看医生的路上跟克劳德聊天，但我并没有直视医生的眼睛说：克劳德告诉我这些。
Maybe they were, but, like, I I mean, I was to the side talking to Claude while leaving the doctor going in, but I wasn't looking him in the eye and being like, well, Claude told me this.
对。
Right.
但我认为你说得对。
But I think you're right.
人们可能会这么做，而且说实话，这应该也会让任何医生听众——如果你们有负面反馈，欢迎告诉我们——但我觉得这反而会让他们的生活轻松一点。
People it might and honestly, it should maybe any doctor listeners can, like, send us any negative feedback on this, but I would think it would make their life a little easier too.
因为这至少会形成一层初步的工作，取代了之前非常随意的个人描述症状等方式，而这些工作本来就是医生必须做的。
Because it would start to at least kind of rather than very, very loose personal ways of trying to explain symptoms and other things, there's already at least one layer of work that's been done that the doctor would have had to do anyways.
所以医生可以更合理地思考问题。
So for them to actually try to, like, think of through things in a more reasonable manner.
所以这看起来不错。
So it seems good.
你知道那些梗吗？学生用ChatGPT写作业，老师用ChatGPT批改作业？
Well, you know how there's these memes that, like, the students are writing their homework with ChatGPT and then teachers are grading the homework with ChatGPT?
对。
Yeah.
也许这也会发生在医疗保险上。
Maybe that's going to happen with Medicare.
等等，不是医疗保险，是医疗保健，因为患者会自己做这些，然后带到医生那里。
Wait, not Medicare, with health care because, you're gonna have patients that are gonna be doing this, on their end, and then they're gonna bring it to the doctor.
现在OpenAI也在发布面向医疗保健的OpenAI。
And now OpenAI is also releasing OpenAI for health care.
所以他们将同时拥有聊天端和API端，并且会构建专门用于医疗工作流程的模型，根据他们的博客文章，这些模型能提供高质量的临床研究和运营工作响应，由GPT-5驱动。
So they are gonna have both, you know, a chat side of things and API side of things, and they are gonna have, like, specific models built for healthcare workflows, that will diagnose here's this is from their blog post of high quality responses for clinical research and operational work powered by GPT-five.
他们会实现机构政策和护理路径对齐，因此能与SharePoint等企业工具集成，使回复能够整合机构批准的政策、路径文档和运营指南，以支持团队间的一致执行，并提供可重复使用的模板来自动化工作流程，例如用于出院总结、患者指导、临床标题和预先授权支持的共享模板。
They'll have institutional policy and care pathway alignment, so they'll integrate with enterprise tools like SharePoint and other systems so responses can incorporate an institution's approved policies, pathway documents, and operational guidance to support consistent execution across teams and reusable templates to automate workflows, such as, shared templates for discharge summaries, patient instructions, clinical headers, and prior authorization support.
所以他们真的会深度融入诊所。
So they're really gonna end up being deep in the clinic.
我不知道。
I don't know.
我的意思是，显然幻觉可能是个问题，还有很多空间可以让它变得更好，但我个人认为，医疗领域非常需要这项技术。
I I I mean, obviously, like, the hallucinations can be a problem, and there's a lot of, you know, there is a lot of room to go here to make it make it as good as it can get, but I personally think that medical field is something that needed this technology deeply.
我以前在节目中说过，现在再说一遍：我父亲是一名医生，他一生中有一半时间都在处理文书工作。如果你能有一个工具，可以录音、做笔记、整理这些笔记，让人们要么专注于病人，要么能看更多病人，或者在使用ChatGPT时通过清晰的提示来协助他们管理健康记录，那情况也不会比现在更糟了，我个人对此非常非常乐观。
I've said on the show before, I'll say it again, my father's a doctor and he's, spent like half his life just doing paperwork and if you could get a tool that like records, take notes, takes those notes and allows people to like either focus on the patient or even see more patients or give them the bright prompts to negotiate their health records when they're using ChatGPT, it can't be much worse than the situation is today, and I I I'm very, very optimistic about this.
从商业角度来看，这可能会成为OpenAI的一个非常重要的业务线。
And from a business standpoint, it could be a very big business line for OpenAI.
不。
No.
我确实同意在消费级医疗健康领域的整体应用。
I I definitely agree consumer level overall for health care.
不过我要说的是，关于OpenAI专注方向这个反复出现的话题，我仍然不太理解。
One thing I will say, though, is on the topic that's recurring about OpenAI and focus, still don't quite understand.
因为记得，2026年的重点是企业级市场、科学发现、ChatGPT订阅增长，现在又推出了ChatGPT Health，这大概也算是在推动订阅增长。
Because remember, we had enterprise is the focus of 2026, scientific discovery, ChatGPT subscription growth, now ChatGPT Health, which I guess kind of ladders up into subscription growth.
但这也提醒我，他们试图涉足一切领域。
But, like, it's still a reminder to me that they're trying to do absolutely everything.
也许他们真的能做到。
Maybe they will.
他们有一些不错的产品，但对我来说，仍然不够清晰的焦点。
They have some good products, but it's still that more clear focus still isn't there for me.
是的。
Yep.
但我不确定。
But I I don't know.
我认为这一点，我的意思是，显然当你阅读这些内容时，你会看到他们实际上为此做了多少工作。
I think that this I mean, obviously, you you actually like, when you read this, you see how much work they actually did on it.
这正是我的意思。
That's what I mean, though.
不。
No.
不。
No.
这正是我的意思。
That's that's what I mean.
这不仅仅是创建另一个标签并称之为健康那么简单。
This isn't just let's create another tab and call it health.
他们有不同的连接器和电子病历系统，确实投入了大量工作，而这些工作并没有投入到他们所谈论的其他业务领域，比如人工智能、云服务和个人设备等。
Like, they're having different connectors and EMRs, and, like, there there's definite work that went into it, which is work that did not go into other parts of the business that they're talking about and AI cloud and personal devices and all the above.
是的。
Yeah.
没错。
That's true.
我的意思是，再强调一遍，如果这是一种通用技术的话。
I mean, I just think that, like, again, if you're, like it's a general purpose technology.
如果你要向投资者推销总可用市场（TAM），你就需要一个医疗解决方案。
If you're selling investors on a TAM, you need you need a a medicine No.
不。
No.
我明白。
I get it.
我明白。
I get it.
我只是想说。
And I'm just saying.
好的。
Alright.
嗯。
Yeah.
我听懂了。
I I hear you.
如果你要创建一个关于OpenAI今年是否上市的Polymarket合约，我相信已经有一个了。
If you were to create a Polymarket contract on whether actually, I'm sure there is one.
你觉得OpenAI今年会上市吗？你支持哪一方？
Whether OpenAI is IPO ing this year, which side are you taking?
我选择不。
I'm taking no.
绝对不。
Definitely not.
不。
No.
如果 payout 的概率是 2%，但回报是 50 倍呢？
What about if it's 2% chance to pay out 50 times to one that they are?
你觉得这样的赔率怎么样？
Do you like those odds?
这确实不错，我的意思是，赔率不错，但这并不会改变我对此事的整体看法，那就是不。
That's a good I mean, that's good odds, but it doesn't change my my overall feeling on the matter, which is that no.
好吧。
Okay.
但说到预测市场，哦，等等。
But but speaking of prediction markets oh, wait.
抱歉。
Sorry.
你刚才在说什么
What are you were
你在想什么？
you thinking?
我的这个过渡。
My That segue.
那是我的过渡。
That was my segue.
好的。
Okay.
说到预测市场，我们上周聊过预测市场。
Speaking of prediction markets, we talked about prediction markets last week.
这周预测市场发生了一些疯狂的事情，我们不能忽视。
Some crazy shit has happened with prediction market this week that we cannot overlook.
首先，马杜罗被捕这件事在预测市场上有押注，有个人似乎掌握了内部信息，在周五投资了3万美元押注马杜罗下台。
So first of all, the Maduro capture was on, was on these prediction markets and somebody who seems like they had under, knowledge, they invested $30,000 on Friday in Maduro's exit.
马杜罗周六早上被拘捕后，同一位投资者赚了43.6万美元，还有些零头。
After Maduro went into custody Saturday morning, the same investor made $436,000 and, you know, some change after that.
显然，这个人很可能掌握了关于马杜罗行动的情报。
Obviously, this person probably had knowledge of Maduro, the of the Maduro operation.
这是我的问题。
Here's my question.
所以，这是我要问你的问题。
So so here's my question to you.
很明显，内幕交易将在预测市场中出现。
So clearly insider trading is going to be a thing on prediction markets.
这没问题吗？
Is that fine?
因为这里的论点是：是的，会有内幕交易，但预测市场的功能之一正是希望它们能准确预测事件。
Because the argument here has been like, yes, there's gonna be insider trading, but ultimately, one of the functions of prediction markets is you want them to be able to accurately predict something.
如果有人确实掌握了即将发生事件的内幕信息，那么预测市场就成为了一种不可思议的窥探未来的工具。
And if somebody does have insider knowledge of what's gonna happen, then prediction markets become like an unbelievable tool to see the future.
我看过这个观点。
I saw that thesis.
我忘了是谁提出的。
I forget from who.
这让我快崩溃了。
This is killing me.
我的意思是，几周前，我谈过我其实很喜欢预测市场的概念。
I mean, a couple of weeks ago, I talked about how I actually do love the concept of prediction markets.
我不喜欢Polymarket和Calci等平台推出这个概念的方式，但这一周我真是低头叹气，对发生的事情感到难以置信。
I don't love how Polymarket and Calci and stuff are really rolling it out, but this week was kind of my just hanging my head and Kim just being in disbelief about the kind of stuff that's happening.
这个观点的关键在于，金融市场的核心意义，或者说任何市场的核心，是让每个人至少相信，自己有机会公平地赚钱。
The thing about that idea is, like, the whole point of financial markets or any kind of market is giving everyone the kind of at least belief some belief that there is a fair opportunity to try to make money in it.
当然，我知道，尤其是那些远离传统金融市场的人，会觉得一切早已被操控，市场结构本身就有诸多问题，偏向大玩家。
Like and if you're if you're and, yes, I understand, like, there's a lot of people, especially outside of traditional financial markets that are like, everything's rigged anyways, and there's lots of problems in market structure, kind of biasing towards bigger players.
没错。
Yes.
以上所有。
All of the above.
但我的意思是，有人在这种情况下知道马杜罗会发生这种事，还能赚到这么多钱，我觉得这对任何人都不好，对预测市场也不好。
But, like, I mean, the idea that someone in this situation knows that this is gonna happen with Maduro, can make that kind of money, I don't think is good for anyone, but I also don't think it's good for prediction markets.
因为如果你总是觉得，我在游戏中处于劣势，内幕人士能抢先一步，这就会彻底破坏市场的意义。
Because if you're always like, I am playing at a disadvantage and insiders will be able to, you know, get ahead of this, it kinda kills the whole point to the market.
所以我认为，一方面，是的。
So I think, like, on one hand yeah.
或者我很好奇。
Or I'm curious.
我不知道。
I don't know.
你对这件事感觉怎么样？
How how are you feeling about that?
我的意思是，作为一个对预测市场能否预测事物感兴趣的人，是的，这很好。
I mean, as someone who's interested in whether prediction markets can predict things, yes, that's good.
但回到我们上周讨论的问题，为什么这些事情变得如此庞大，是因为人们唯一觉得拥有掌控感的地方就是赌场。
But also going back to our discussion last week of the reason why these things have become so big is because the only place that people are fine you know, feel like they have any agency is the casino.
他们没有，而且他们最终会。
They don't, and they will they will end up.
他们是在与之对抗。
They are playing against.
没有内幕信息，你不可能拥有一个英国市场。
There's no way you can have a British market without insider information.
要让这个市场免于内幕交易，比股票市场更难，而且监管更少。
It's even harder to keep that, free of insider trading than it is the stock market, and there are fewer regulations.
甚至有可能。
Well, even possible.
是的。
Yeah.
是的。
Yeah.
没错。
Exactly.
所以，是啊。
So Yeah.
但这周有一件事让我特别恼火，就是关于白宫新闻秘书卡罗琳·莱维特的那则新闻。
But there's one thing that's really annoyed me this week, which has been this story about Caroline Levitt, the White House press secretary.
你知道，在她那场新闻发布会还剩六十五分钟时就离开了。
You know, left just before the sixty five minute mark of her
解释一下实际的或
Explain the actual or
你知道，在预测市场上，有人预测了她新闻发布会的时长，盘口是六十五分钟，而她果然提前结束了。
There there was a some sort of, like, on the prediction markets, you know, a prediction of how long her press briefing would last, and the over under was sixty five minutes, and she, like, you know, wrapped up.
所以，当时有98%的概率认为会超过六十五分钟，但她却在还差三十秒到六十五分钟时就离开了。
There was so the so there was a 98% chance that it was gonna run past sixty five minutes, and then she left with thirty seconds to spare.
也就是在六十五分钟标记前三十秒。
So thirty seconds before that sixty five minute mark.
大家都说：哦，莱维特在预测市场上押注，知道她会坚持到64分30秒。
And everyone's like, oh, Levitt's got money on the prediction market knowing that she would go sixty four thirty.
我的意思是，当然有可能，但这些新闻发布会本来就会突然结束。
I mean, obviously, there's a chance, but these these press conferences, they wrap abruptly anyway.
我不认为她是想赶在截止前跑掉。
I don't think she was running away to make it before.
那个提示音，我感觉它就是不
See The the the buzzer, I it just doesn't
听好了，我要说了。
See, I'm gonna say it.
一开始，我觉得这太荒谬了。
Like, on one at first, I'm like, this is ridiculous.
但我看了几遍视频，发现她脸上带着一丝微笑，还非常迅速地跑下了台。
But I watched the clip a few times, and there's a bit of a smile, and there's a bit of, like, just, like, very quickly running off stage.
而且别忘了，特朗普一家在Polymarket上押注，而卡拉是他们的顾问。
And again, remember, like, I mean, the Trumps are on the Polymarket and Cal, she like, their adviser.
而且根本没说那里有任何关联。
Like, the and not even saying there's any kind of, like, connection there.
更像预测市场在马阿拉歌的作战室或任何地方，当他们入侵委内瑞拉或对付马杜罗时也是如此。
It's more prediction markets are the same way in the situation room or whatever at Mar A Lago when they're invading Venezuela or getting Maduro.
他们会在屏幕上显示x，而预测市场必须成为与他们讨论的一部分。
They have x up on the screens and searches for, like, prediction markets have to be part of the conversation with them.
就像白宫的一群工作人员。
It's like the group of staffers at the White House.
无论这纯粹是为了好玩，或者在这种情况下只是有点愚蠢的乐趣。
And whether it's just for fun, and it could be just kinda stupid fun in this case.
但对我来说，这件事更有趣的部分是回到我们之前关于这个问题的讨论：当人们能通过下注真正影响结果时，会发生什么？
But to me, the more interesting part of this is going back to our earlier question around this, what happens when people can actually drive outcomes based on bets?
说‘我有内部信息，我要用它来下注’是一回事。
It's one thing to say, I have insider information, and I'm going to use that to make a bet.
但另一回事是，我实际上会因为下了注而改变最终结果，就像在体育比赛中打假球一样。
It's another two, I'm actually gonna change the end outcome because I have money on it, same as, like, throwing throwing a game throwing a match or in sports.
对，没错。
Like Yeah.
实际创造出一个特定的结果。
Actually creating a specific outcome.
对我来说，值得开始思考的疯狂之处在于，这简直变得太反乌托邦了。
And to me, like, the crazy thing to start thinking about is I mean, this is where it just got so dystopian.
比如，有趣的是，你正在做预测市场，却提前结束了新闻发布会。
Like, something like this, let's say you're as it's funny, you're doing the prediction mark cutting the press conference short.
而且，我们居然对这种事有合约，这实在太荒谬了，不过我想也只能这样了。
And again, it's just so ridiculous that we have contracts on this, but I guess that's fine.
想象一下，如果我们开始让重大决策受制于人们为了在预测市场上赚钱而采取的行为，并让这些行为真正驱动决策本身。
Like, the more imagine if we start making big decisions people do to make money on prediction markets and letting that actually drive the decision itself.
这太可怕了。
That's terrifying.
比如
Like
嗯。
Yep.
我的意思是，我不知道。
I mean, I don't know.
我还没有完全接受这一点。
I'm I'm still not fully fully bought into this.
我的意思是，是的，你得考虑这种可能性，但我对这些新闻发布会，它们，我
I mean, yes, you have to consider the possibility, but I'm not I mean, these press conferences, they I
我不认为
don't I don't I think
时间，这完全是推测。
the time, and it's such speculation.
我很惊讶它引起了这么多关注。
I'm surprised it got that much, you know, pickup.
但你所提出的更广泛的观点，我同意。
But you're but the the broader point that you're making, I'm on board with.
原则可以推动结果。
The principles can drive the outcomes.
当你在做这些事情时，随着预测市场的发展，你会看到这种情况发生。
And if you're doing and as these prediction markets rise, you're gonna see that happen.
我的意思是，想象这样一个世界：因为预测市场上有大量资金，我们抓住了马杜罗；或者，如果这些市场真正成熟、增长并更加整合，更多人把钱投入其中，我们将会开始看到一些非常奇怪且令人担忧的结果，我认为。
I mean, imagine the world where we capture Maduro because there's so much money on the prediction market or, like like, if these are actually maturing and growing and becoming more integrated, more people have money in them, we're gonna start to see some really weird troubling outcomes, I think.
是的。
Yes.
我仍然从理论上喜欢预测市场。
Still love prediction markets in theory.
我还是喜欢它们。
Still love them.
从理论上讲。
In theory.
我不知道它们在实践中是否真的有效。
Don't I know if they're really working in practice.
好的。
Okay.
在我们离开之前，我想谈谈无效工作的终结。
Before we leave, I want to talk about the end of busy work.
这是一篇《华尔街日报》的文章，探讨了将人工智能用于工作中所有枯燥任务的负面影响。
Here's this Wall Street Journal article, the downside to using AI for all those boring tasks at work.
由于人工智能工具可以整理和总结邮件、记录会议纪要、处理报销单据，使得没有无效工作的工作日比以往任何时候都更接近现实，这些工具能让我们专注于更重要的事情。
Workdays without busy work are closer to reality than ever thanks to artificial intelligence AI tools that can sort and summarize emails, take meeting notes, and file expense reports promise to free us to concentrate on the important stuff.
这听起来很棒。
That sounds great.
但问题是，我们的大脑无法持续不断地进行深度思考，如果我们完全取消了这些简单重复的工作，可能会失去那些在做这些任务时偶然涌现的灵感。
The catch is our brains are incapable of thinking big thoughts nonstop, and we risk forfeiting the epiphanies that sometimes spring to mind while doing easy repetitive job functions.
Aflac的首席执行官丹·阿莫斯在他的日程表上安排了一些低强度的任务，这些任务本可以交给助理或机器人完成。
Aflac chief executive Dan Amos dots his calendar with low intensity tasks that could be dedicated to an assistant or a bot.
这些做法部分源于传统的习惯和个人化的细节，但也在于给自己留出心理休息的空间，让创意的火花得以迸发。
These practices are partly about old fashioned habits and personal touches, they're also about taking mental breaks or leaving space for creative sparks to fly.
这就像在淋浴时思考一样，让大脑进入自动模式，直到灵感涌现。
It's the same principle as thinking in the shower, putting your brain on autopilot until it goes.
还有一位来自一家名为‘常规公司’的首席执行官。
There's also a CEO, called from a company called Conventional.
他的名字叫罗杰·柯克内斯。
His name is Roger Kirkness.
他在年中时开始意识到空闲时间的价值。
He's become to the value of slack time around the middle of the year.
那时他注意到，人工智能带来了大约20%的显著生产效率提升。
That's when he noticed meaningful productive gains from AI, about 20%.
随后他发现，人工智能整体上带来了约20%的显著生产效率提升，但他观察到，团队成员常常在周五时精神疲惫、效率低下。
Then he noticed meaningful productive gains from AI about 20% overall, but he observed that he observed that teammates were often mentally exhausted and unproductive by Friday.
拉詹，你是要站起来，和这些首席执行官们一起为忙碌工作辩护呢，还是说这不过是办公室里最后的愚昧堡垒——人们坚持认为，你真的需要做些像提交报销单这样的事情，才能激发创造力？
Ranjan, are you gonna stand up, together with these CEOs in defense of of busy work, or, or is this just like the the last bastion of stupidity, in the office where people say you actually need to do, you know, things like file expense reports to think creatively?
我不会为这种做法辩护。
I am not gonna defend this.
我觉得回到本集的开头，也许把一个链接粘贴到谷歌文档里，这种行为有种美妙之处，我们可能永远不会放弃它，因为它有一种释放感，提醒我们这个世界上真正重要的东西，但报销单呢？
I think going back to the beginning of this episode, maybe there's something beautiful about taking a link and pasting it into a Google Doc, and maybe we're never gonna give that one up, and there's just something cathartic and, like, just reminds you of what's important in this world, but expense reports?
拜托。
Come on.
它们难道不是你必须做的最让人精神疲惫、最枯燥乏味的工作吗？
Like, they're what are the most kind of, like, mentally drudgery mental drudgery type of work you have to do?
是的，我不这么认为。
And, yeah, that that I don't think so.
但我确实认为，给获得奖金或退休的员工手写便条，这件事挺有意思的。
I do think the handwritten notes to employees who receive bonuses or retire, It's interesting.
手写有一种特别的意义，但我觉得这种‘忙活式工作’的想法，不过是我想当个老派高管，说些荒唐的话，希望有一天它们能被当作智慧流传下去。
Like, there's something about handwriting that's still meaningful to me, but I think this idea of busy work is just kind of like, I just wanna be an old executive guy and just say ridiculous things and have them kind of like land as as wisdom one day.
这就是我人生的追求。
That's my goal in life.
我的意思是，打电话。
I mean, call.
对吧？
Right?
就是，你懂的。
Like, you get it.
嘿，我是《华尔街日报》的卡勒姆·博彻斯。
Hey, it's it's Callum Borchers from the Wall Street Journal.
哦，嗨，卡勒姆。
Oh, hey, Callum.
你好吗？
How are you?
这位是阿莫斯先生。
This is mister Amos.
我是安达保险的首席执行官。
I'm the CEO of Aflac.
嗯。
Yeah.
你对琐碎的工作有什么看法？
What do you think about busy work?
哦，我超喜欢。
Oh, I love it.
特别喜欢填报销单。
Love doing expense reports.
就像在淋浴时思考一样。
It's like thinking in the shower.
然后一个，等等。
And one and then hold on.
然后他去了孵化器。
And then he went on to incubator.
他最喜欢的创意孵化器是锻炼后的蒸汽房。
His favorite idea incubator is the steam room after a workout.
经常从蒸汽中走出来时思路清晰，立刻发一封邮件记录下来。
Often steps out of the fog with a clear thought and dashes off an email about it.
实际上，蒸汽房
Actually, steam room
我同意。
I'm with that.
我也同意。
That I'm with.
不错。
Not bad.
但他并没有在做琐碎的工作。
But he's not doing busy work.
那不是琐碎的工作。
That's not busy work.
我是不是太乐观了，以为我们能终结琐碎工作，腾出空间
Am I am I being too optimistic in thinking we can end busy work and make room
给蒸汽房。
For the steam room.
这就是摆脱琐碎工作的方式，这样你就能在蒸汽房里思考。
That's what get rid of the busy work so you can be in the steam room and you can think.
是的。
This is Yes.
这是
This is
是对新闻业的亵渎。
a travesty of journalism.
为了解决这个问题，只有一件事能救我们，你知道我想说什么。
For this, I I there's only one thing that can save us for this, and you know where I'm going.
我们需要自主型工具。
We need the agentic harness.
我们需要它。
We need it.
没错。
Yep.
只是为了安装那个自主型框架，多花点时间在蒸汽房里。
Just to get that agentic harness in place, more time in the steam room.
进入并让自己置身于那个自主型框架中，然后就开始吧。
Get in get yourself into that agentic harness and away you go.
尽情思考吧。
Steam away.
大胆思考。
Think big.
再次相信你的公司。
Believe in your company once again.
等等。
Wait.
稍等。
Hold on.
稍等。
Hold on.
就到这里。
Just to end.
这篇文章还提到，他一年赚两千万，但他却拒绝多花点钱购买去广告的流媒体服务，因为他觉得广告能让他有时间思考刚才看的内容。
The piece did go on to say, like, he makes 20,000,000 a year, but he declined to pay a few extra bucks for an ad free version of a streaming service because commercials offer a moment to think about what he just watched.
我觉得这是个火力全开的进攻。
I I think that's a fireball offense.
我的意思是，这明显显示出非常糟糕的决策，好吧。
I mean, that's that's just shows some very poor decision making Alright.
如果问我的话。
If you ask me.
我超喜欢这家伙。
I love this guy.
实际上，你觉得他是不是在故意开玩笑？
Actually, you know, do you think he was just trolling here?
他就是说些荒谬的话，看看大家会不会信，你说对吧。
He was like, I'm just gonna say some ridiculous stuff, and let's see if they go with it because, come on.
他根本不会看广告。
He's not watching commercials.
对不起。
I'm sorry.
你一年赚两千万。
You're making 20,000,000 a year.
你不是你，你可以暂停。
You are not you you can hit pause.
你知道怎么按暂停键。
You know how to hit the pause button.
他们刊登了，所以我真希望他们能来。
They printed it, so I'd love to have them on.
如果你是Aflac的CEO丹尼，你们大多数人听这个播客，那就来和我们聊聊这些琐碎的工作吧。
If you're, if you're Aflac's CEO, Danny, most of you listen to the podcast, come on and speak with us about busy work.
我很想进行这场对话，从另一个角度来探讨。
I'd love to have this conversation and take the other side of it.
而且，天啊，老兄，直接跳过广告吧。
And and no, for God's sakes, man, just skip the commercials.
不是我们的。
Just Not ours.
不是我们的。
Not ours.
不是我们的。
Not Not ours.
但别人会做。
But others do.
好的。
Okay.
好吧。
Alright.
我们就到这里吧，拉詹。
Let's let's call it here, Ranjan.
这个很不错。
This this has good.
我们做完了。
We've done it.
我们这一周的工作已经完成了。
We've done our we've done our work this week
当然。
for sure.
我现在要去报销一些费用。
I'm gonna go file some expenses right now.
我要去。
I'm gonna go
去好好想想，理清思路。
into to think just to think clearly.
好了，各位。
Alright, everybody.
谢谢，拉詹。
Thank you, Ranjan.
感谢大家收听。
Thanks everybody for listening.
我们将在节目中邀请Mistral公司的首席执行官亚瑟·门奇，讨论人工智能是否是一种托管服务。
We'll have, Mistral CEO Arthur Mench on the show to talk about whether AI is a managed service at
在结束时。
the end of
一天结束时。
the day.
这将在周三播出，之后拉詹和我将在下周回来。
That's gonna come up on Wednesday, and then Ranjan and I will be back next week.
再次感谢收听，我们下周再见，欢迎收听《大科技》播客。
Thanks again for listening, and we'll see you next time on Big Technology Podcast.
关于 Bayt 播客
Bayt 提供中文+原文双语音频和字幕，帮助你打破语言障碍，轻松听懂全球优质播客。