← 返回任务列表

7 Ways How We Use AI Is Changing

262 段 · 1 位说话人 · 原片 26:06
M1
M10:00

我们使用AI的方式正在改变。

The way we use AI is changing.

一开始,我们是提示词工程师,然后是上下文工程师,接着是框架工程师,再然后是循环工程师。

First, we were prompt engineers, then context engineers, then harness engineers, then loop engineers.

虽然我们很容易把这些都看作是一个又一个可笑的流行词在更迭,但现实是,我们都在这个旅程中一起摸索,试图找到使用一种全新科技的最佳方式,这种科技既在改变我们现有工作的方式,也在开拓新的工作可能。

And while it's tempting to view all of these things as just one silly buzzword giving way to the next, the reality is that we are all together on this journey figuring out just the best way to use a completely new technology that is changing both how we do our current work and opening up new possibilities of what work we can even do.

今天我们要讨论的是,我们使用AI的方式正在改变的7个方面。

Today we're talking about 7 ways that how we use AI is changing.

《AI每日简报》是一个关于AI领域最重要新闻和讨论的每日播客和视频节目。

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

好了朋友们,在我们深入之前先快速说几件事。

Alright friends, quick announcements before we dive in.
M1
M10:59

今天的节目是关于一系列我认为代表我们日常与AI互动方式转变的模式。

Today's episode is about a bunch of patterns that are shifts, I think, in how on average we interact with AI on a daily basis.

从ChatGPT诞生之初至今,我们与AI的互动模式从来都不是一成不变的,这一点始终成立。

One thing that's been true since the very beginning of ChatGPT is that our patterns of interaction with AI have never been static for long.

在最初,人们发现新的方式来给AI下提示词,结果有好有坏,然后通过成百上千的文章、帖子、YouTube视频、播客和课程传播开来。

At the beginning, it was people discovering new ways to prompt the AI that had better or worse results, which was then communicated through hundreds of articles, posts, YouTube videos, podcasts, and courses.

但随着时间的推移,互动模式的转变变得更加重要和实质化。

But over time, the interaction pattern shifts have gotten more significant and more substantive.

大概在过去一周左右的时间里,我看到了一些公告,它们确实强化了这些互动模式中的一些相当大的转变。

Over the last week or so, I saw a number of announcements that really reinforce some fairly big shifts in these interaction patterns.

虽然我不认为我们使用AI的方式会真正地一成不变,但我确实认为这些转变中——我将在本期节目中介绍的——有些是相当根本且可能会持续下去的。

And while I don't think that how we use AI will ever be truly static, I do think that some of these shifts, which I'm going to cover in this show, are fairly fundamental and likely to stick.

因此,当然,如果一个人想高效利用时间,这可能就更值得关注了。

And thus, of course, if one is trying to be efficient with their time, maybe more worthy of attention.

所以首先是一种互动模式,这与其说是我们改变与AI互动的方式,不如说是产品设计决策正在改变我们与AI互动的选项。

So first up is an interaction pattern that is not so much us changing how we interact with AI, but product design decisions that are shifting the options that we have to interact with AI.

很明显,现在有一种趋势是追求简化和集成,以及为最终用户减少而非增加认知决策的用户体验。

Very clearly, there is a push towards simplification and integration and user experiences that have less rather than more cognitive decision-making for the end user.

坐在我这个位子上,有一点很有趣,那就是早期的AI似乎常常违背硅谷的传统逻辑——即如果你不提供极其简单清晰的体验,用户就不会来。

One thing that's been interesting for me sitting in this seat that I have is the extent to which early AI has tended to defy traditional Silicon Valley logic that users won't show up unless you give them an incredibly simple and clear experience.

事实上,恰恰相反,许多早期采用者,不仅来自科技圈,也来自更广泛的知识工作者圈,他们不但没有抗拒复杂性,而是全身心投入其中。

In fact, quite to the contrary of that, a lot of the early adopters, not just from tech circles but from broader knowledge work circles, have, rather than resisting complexity, thrown themselves in headfirst.

对我来说一个很好的例子是,今年早些时候,有9000人——其中大多数人甚至不是工程师,也根本不太懂技术——报名参加了我的免费Claw Camp项目,这个项目基本上是让他们费尽心思去琢磨如何构建自己的agent。

A great example of this for me was when 9,000 people, most of whom weren't even close to engineers or really technical in any way, signed up for the free Claw Camp program I put out earlier this year, which basically had them crawling through glass to figure out how to build their own agents.

然而,看起来,也许尽管在AI领域愿意做那种苦差事的人比例确实比过去其他科技领域要高,但仍然有很多人想要那种简单。

And yet it appears that perhaps even if the portion of people who were willing to do that sort of hard work has been higher in AI than in other past tech sectors, there are still a lot of folks who want that simplicity.

当然,“简单”是新的Meta Muse产品的关键词之一,即使是那些完全有能力应对复杂性的高级用户也对这个产品的这一点赞不绝口。

Certainly simplicity is one of the watchwords for the new Meta Muse product and something that even the power users who fully have the capability of dealing with complexity are lauding in the product.

How IAI的Claire Vo在多个场合讨论过她认为Muse的设计考虑得多么周到,特别是关于它向用户隐藏了什么、又没有隐藏什么。

How IAI's Claire Vo has discussed in numerous places how well thought out she thinks the Muse design is, specifically around what it does and doesn't obfuscate from the user.

Node.js的创造者Ryan Dahl,他当然也完全有能力处理复杂性,同样称赞Muse成功做到了简单。

Node.js creator Ryan Dahl, who again certainly has no shortage of complexity capability, also shouted out Muse for having nailed the simplicity.

技术专家Tom Goodwin写道,Muse的美妙之处在于他们成功地以一种非常易上手的方式推出了它。

Wrote technologist Tom Goodwin, the beauty of Muse is that they've managed to introduce it in a very accessible way.

不用告诉我你怎么工作的,就告诉我怎么做事情就行了。

Don't tell me how you work, just show me how to do things.

David Herman写道,我认为Muse将会赢得休闲AI竞赛。

David Herman wrote, I think Muse is going to win the casual AI race.

我给我快70岁的爸爸看了,他一下就明白了。

I showed my dad who is nearly 70 and it clicked.

他很惊讶,我可以让它为我创建一份完整的合同,修改内容,然后通过短信把它用电子邮件发给客户。

He was amazed that I could have it create an entire contract for me, update things on it, then email and send it to a customer all through texts.

Claude、OpenAI都很棒,但到目前为止,我觉得只有不到1%的人真正理解这些模型是什么。

Claude, OpenAI are great, but so far I'd say less than 1% of the population gets what these models are.

Meta让这一切变得超级简单。

Meta just made it all super easy.

很显然这是有意为之。

And it's clear that this was intentional.

Meta的首席AI官Alexander Wang写道,团队在构建Muse时的一大重点是让它开箱即用。

Meta Chief AI Officer Alexander Wang wrote, A big part of the team's focus when building Muse was to make something that just worked.

为了简化产品、让使用变得轻松,他们投入了大量心血在每一个小细节上。

So much work went into all the little details to make the product simple and easy to use.

当然,让这成为趋势而不仅是Meta团队设计决策的原因是,这种简化正在到处出现。

And of course, the thing that makes this a trend, not just a design decision from the Meta team, is that this sort of simplification is showing up everywhere.

这周我们收到的最大的产品更新之一是Anthropic宣布,Claude Cowork和Chat不再是两个分开的东西。

One of the biggest product updates that we got this week came when Anthropic announced that Claude Cowork and Chat are no longer separate things.

Claude正在快速演变,变得不再碎片化、不再让用户在一开始就需要做很多决定,而是变成一种统一的体验。

Claude is careening towards an experience where instead of it being fragmented and full of decisions for the user before they even get started, instead a single unified experience.

而且,核心理念还是简单和简化。

And again, the watchword was simplicity, simplification.

顺便说一句,正如Claude Code的创造者Kat Wu所指出的,这种简单不仅体现在消除Cowork和Chat的差异上,还体现在整合Claude Design,让你常用的核心Claude也具备制作幻灯片、设计、文档等所有能力,不用切换到专门的界面。

And by the way, as pointed out by Claude Code creator Kat Wu, this simplicity extends not only to eliminating the difference between Cowork and Chat, but also to integrating Claude Design so that your standard core Claude also has all of the capabilities to do things like make slides, designs, docs, without having to switch into some specialized experience.

Anthropic团队也指出,这是基于用户的直接反馈。

And again, the team at Anthropic reports that this was based on direct user feedback.

Anthropic Labs的负责人、Instagram的前联合创始人Mike Krieger写道,我们最常听到的声音是人们不确定该从哪个产品开始。

Anthropic Labs lead and former Instagram co-founder Mike Krieger wrote, the most common thing we hear, people aren't sure which product to start with.

这种摩擦阻碍了人们从这些模型的能力中获得最佳体验。

That friction gets in the way of getting the best from what these models can do.

不过,正如我在本周早些时候的节目中提到的,我个人对这种做法并不完全认同,因为我喜欢拥有更具体的控制权,但看来我和大多数人的想法真的很不一样。

Now, as I mentioned in an earlier episode this week, I'm not totally sold on this on a personal level because I feel like I like having more specific control, but boy do I seem out of sync with the average here.

Christian Selig转发了这一宣布并说:“我们能否也把Codex做成这样,拜托?”

Christian Selig reposted the announcement and said, "Can we do this with Codex too, please?"

这两种模式就像两个家长在吵架。

The 2 modes are like having 2 parents fighting.

Sean和Matthew总结道,人们不想同时处理两三个文本代理、两三个聊天代理或两三个桌面终端。

Sean and Matthew summed up, people don't want to interact with 2 to 3 text agents, 2 to 3 chatbot agents, 2 to 3 terminals on desktop.

你希望有一个简单统一的界面,并让代理们协调完成任务。

You want a simple, unified interface and want the agents to coordinate how it gets done.

你只需要提出请求然后收到结果。

You simply want to ask and receive.

平台应该负责协调和委派代理,为任务路由模型,处理事情是在本地、云端还是虚拟机上,并主动和被动地应对任务。

The platform should coordinate and delegate agents, route models for tasks, handle whether it's local, cloud, or VM, be proactive and reactive about tasks.

Instinct、Muse和GrokBot确实引起了共鸣,而且在我看来,我们显然正在进入另一个范式转变的时刻。

Instinct, Muse, and GrokBot really hit a nerve, and it's increasingly clear we're in another paradigm shift moment in my opinion.

如果第一个变化是我们使用AI的界面变得简化,那么第二个变化本身就是一种简化,它关乎我们如何控制AI。

Now, if the first change is about the simplicity of the interfaces through which we use AI, the second change is itself also a version of simplicity that has to do with how we control AI.

这就是单一线程的趋势。

And this is the trend towards monothreads.

单一线程的意思是,你不再需要切换上下文窗口来解释新的需求或提示,而是使用一个连续运行的长时间线程,它继承之前的对话上下文,并以此管理整个体验。

The idea of a monothread is that instead of having to switch context windows and explain new needs and new prompts separate from the previous things that you've done, Instead, you use a single long continuously running thread that inherits the context from past conversations as a way to manage the entire experience.

以前,这受限于上下文窗口的大小。

Now previously, this was limited by the context window.

到某个时候,AI会“累”,也就是说它的上下文窗口会被填满,然后它就会开始出错、犯错或丢失指令等等。

At some point, the AI would get tired, i.e., its context window would get filled, and it would start doing weird things, making errors, losing instructions, etc.

在去年年底和今年年初,我做的很多早期课程和讲座都是关于如何编写良好的交接文档,以在不同线程之间传递上下文。

Many of the earlier courses and lessons that I did from late last year and early this year were all about how to build good handoff documents to move context between different threads.

不过,这种状况今年开始转变,特别是当Codex团队专注于压缩上下文,以此改变人们仅通过一个统一线程使用Codex的方式时。

That started to shift this year though, especially as the team at Codex really focused on compaction to change the way that people could interact specifically with Codex through a single unified thread.

早在四月份,Codex的Nick Baumann写道,很多编码代理设计都基于一个假设,即超过上下文窗口并进行压缩会逐渐导致更差的结果。

Codex's Nick Baumann wrote back in April, so much coding agent design is built on the assumption that breaching context windows and compacting context yields progressively worse results.

当你放下这个假设,所开辟的产品方向将会非常令人兴奋。

When you drop this assumption, the product direction it opens up is very exciting.

他是在回应开发者Anthony Kroger,后者曾写道:“我使用Codex时从不担心上下文窗口。”

He was responding to developer Anthony Kroger who had written, I literally never worry about context windows using Codex.

它能压缩好几次,但不知何故,模型依然记得细节。

It can compact like 3 times and the model still remembers the details somehow.

在大约一周后发布的文章中,Codex团队的Nick在他的一篇名为《我的Codex线程是活的》文章中阐述了他的想法。

In a post about a week later, Nick from the Codex team expanded on his thoughts in a piece he called My Codex Threads Are Alive.

他在里面写道,我现在最有用的Codex线程是我过去三周一直在用的那个。

In it he wrote, the most useful Codex thread I have right now is the one I've been using for the last 3 weeks.

它每小时检查我的Slack、Gmail,以及我写的或正在看的PR。

Every hour it checks my Slack, Gmails, and PRs I wrote or am watching.

它每小时把噪音变成我能采取行动的清晰信号。

Every hour it turns the noise into clear signal I can act on.

我的Codex使用方式已经从启动很多短期对话,转变为围绕重复出现的工作流保持更少数量的线程存活。

My Codex usage has shifted from starting lots of short-lived chats to keeping a smaller number of threads alive around recurring work streams.

我仍然经常开新线程,但有些工作不应该每次我问问题就重置。

I still start fresh threads constantly, but some work should not reset every time I ask a question.

他最后写道,总之,我已经变得只专注单一线程了。

In sum, he writes, I have become mono-thread-pilled.

Nick继续说,我最近说过,Codex团队做的压缩改进正在为编码代理开辟新的产品方向。

Nick continues, I recently said that the compaction improvements made by the Codex team are opening up new product directions for coding agents.

很多代理产品设计假设长线程最终会退化。

A lot of agent product design assumes long threads eventually degrade.

如果这是真的,产品应该推动用户创建新线程并写上下文摘要。

If that is true, the product should push users towards creating new threads and writing context summaries.

Codex团队一直在做很多压缩相关的工作,而这项工作削弱了这个假设。

The Codex team has been doing a lot of work on compaction, and that work weakens the assumption.

当一个线程能在很多次对话后仍然有用,我就希望同一个线程继续处理这个重复出现的任务。

When a thread can stay useful across many turns, I want the same thread to keep working on the recurring task.

简单说,有了良好的上下文压缩,线程的价值会随时间增加。

Put simply, with good context compaction, a thread's value increases over time.

现在,Nick在四月份写了这个,而我整年确实发现自己的行为越来越多地转向那个模式。

Now, Nick wrote about that back in April, and I have very much found myself throughout the year seeing more and more of my behavior shift to that pattern.

例如,我在AI Daily Brief网站上做的所有事情,目前都是通过一个单一的Claude Code线程管理的。

Everything that I do on the AI Daily Brief website, for example, is managed at the moment from a single Claude Code thread.

当我偶尔进去做改动或更新时,我不是从头开始那个线程。

When I pop in to make a change or an update, I'm not starting that thread all over.

我只是接着上次停下的地方继续。

I'm just jumping into where we left off.

我用Claude从节目里生成视频片段的社交流程,是另一个线程。

The social pipeline through which I use Claude to generate video clips from the show is another thread.

我甚至发现这种行为开始影响我在Claude其他部分(不只是Claude Code)里的交互方式。

And I've even found this behavior start to influence how I interact in other parts of Claude outside of just Claude Code.

例如,在过去的几个月里,在项目内部,我发现对于我常做的事情,比如更新或写新的广告文案,我不会每次需要做那个任务时都开新线程或新对话,而只是接着上次做类似任务的地方继续。

Over the past couple of months, for example, inside projects, I've found that when it's things that I do frequently, such as updating or writing new ad copy, instead of starting a fresh thread or conversation every time I need to do that task, I just jump in at the end of the last time I did a similar task.

结果发现,我并不是唯一一个开始以这种方式改变行为的人。

Well, it turns out I haven't been the only one who has started to shift their behavior in this way.

事实上,到现在,从这周开始,这正式成为了Claude项目的工作方式。

In fact, this is now, as of this week, officially how Claude projects work.

周四,Claude账户发帖说,项目现在从Claude Code里的一个对话开始运行。

On Thursday, the Claude account posted, projects now run from one conversation starting in Claude Code.

你描述需要做什么,Claude会指导并行线程,在关上笔记本电脑后这些线程继续工作。

You describe what needs doing and Claude directs parallel threads that keep working after you close your laptop.

换句话说,并不是没有多个线程,而是这些多个线程由Claude管理,你要做的是通过一个线程与Claude交互。

In other words, it's not that there aren't multiple threads, it's that those multiple threads are managed by Claude and what you do is interface with Claude through a single thread.

这和我刚才描述的模式非常相似。

This is very similar to the pattern that I had just described.

顺便说一句,在这个功能到Claude之前的一周,它也已经到了Cursor。

And by the way, a week before this came to Claude, it had also come to Cursor.

事实上,当Cursor推出他们的项目版本时,这个逻辑内置其中。

In fact, when Cursor introduced their version of projects, it had this logic built in.

在他们的公告中,他们写道,你不需要为每个任务创建一个聊天,而是通过一个单一持久线程与一个协调器代理一起工作。

In their announcement, they write, rather than creating a chat for every task, you work with a coordinator agent in a single persistent thread.

这种转变的影响相当大。

And the impacts of this shift are pretty big.

Claude Code的创建者Boris Cherny写道,项目不仅改变了我与Claude互动的方式,也改变了我编码的方式。

Claude code creator Boris Cherny wrote, projects have changed not only how I interact with Claude, but how I code.

我停止了管理会话。

I stopped managing sessions.

我只是想到什么就发什么,Claude 会把它们拆分成不同的线程,项目也会记住我是怎么工作的。

I just send thoughts as they come, Claude splits them into threads, and the project remembers how I work.

而且我觉得,这跟我们用AI工作的方式上一个非常相关的转变有关。

And this, I think, gets to a very related shift in how we work with AI.

也就是说,聊天机器人正逐渐变成代理舰队的管理者。

Which is that chatbots are increasingly agent fleet managers.

如果你看看我刚才举的那几个例子,变化在于我们跟一个协调型AI交互的地方变了,但那个协调型AI仍然在调度很多不同的子代理来完成工作的不同部分。

If you look at all of those examples that I just gave, what was changing was the place that we interact with a coordinator AI, but that coordinator AI was still spinning up lots of different subagents to do different parts of the work.

我在这儿觉得有意思的是,我想在一段时间里,我们很多人大概都以为聊天机器人界面本身是个中间界面,一个当时方便又熟悉的东西,但也许会让位给其他东西。

What's interesting about this to me is that I think for a while, many of us would've assumed that the chatbot interface itself was perhaps an intermediate interface, something that was convenient and familiar for the time, but might give way to something else.

有意思的是,当你把单线程和聊天机器人作为代理舰队管理者两者结合起来,发生的事情是,模型和基础设施公司悄悄改变了聊天机器人的本质,而不是从聊天机器人上移开,但也没把界面这块地毯从我们脚下抽走。

What's interesting is that when you take the combination of monothreads and chatbots as agent fleet managers, what has happened is that instead of shifting away from chatbots, the model and harness companies have quietly shifted what chatbots actually are, but without pulling the interface rug out from under our feet.

我们仍然能获得那种跟聊天机器人互动时同样舒适的体验。

We still get that same comfortable experience of interacting with a chatbot.

实际上,甚至还更简化了,因为我们可以从单个线程里做这件事。

In fact, one that's even more simplified because we can do it from a single thread.

但那个聊天机器人做的事情,跟一年前相比,已经非常非常不同了。

But what that chatbot is doing is very, very different than what it was doing a year ago.

Claude Project 之所以这么有意思,是因为它处理代理团队真的很在行。

What makes Claude Project so interesting is that it handles teams of agents really well.

你跟一个主协调代理对话,它就会调度出各种专家。

You talk to a main orchestrator agent and it spins up specialists.

基本上,它会创建一个组织来解决你的问题,根据你的偏好,混合使用贵的和便宜的代理。

Basically, it creates an organization to solve your issues, mixing expensive and cheap agents depending on your preferences.

顺便说一句,这对我们其他的软件服务也有影响,它们可能会越来越像我们主协调代理的上下文数据库。

This also, by the way, has implications for the rest of our software services, which might increasingly start to look like context databases for our main coordinator agents.

我所有的 iMessage 和邮件现在都通过一个单一的代理界面来跑。

All of my iMessages and email now run through a single agent interface.

我仍然发短信,但我从手机上删掉了所有邮件客户端,也取消了每个邮件订阅或应用。

I still text, but I deleted every email client from my phone and cancelled every email subscription or app I had.

Gmail 对我来说基本上就是数据库了。

Gmail is basically just a database to me now.

Google 的其他服务也越来越多是这样。

Increasingly so are the rest of Google's services.

也就是说,我再也不想手动写邮件、建文档、做幻灯片、安排日历或者管理文件了。

I.e., I never want to manually write an email, create a doc, build slides, organize a calendar, or manage files again.

我就想表达意图,然后让底层那些记录系统正确地更新好。

I just want to express intent and have the underlying systems of record updated correctly.

我觉得现在你会看到应用消失得很快。

I think now you'll see apps disappear quickly.

很多应用最好也只是变成数据库,最糟的话就完全过时了。

Many will become databases at best or entirely obsolete at worst.

代理会成为你个人和职业生活的默认界面。

The agent will be the default interface to your personal and professional life.

而且即使这事儿没完全发生,或者比像 Signal 那样有远见的人想的时间要长得多,但很肯定的是,你正在交互的那些聊天机器人,做的已经跟以前不一样了。

And even if that doesn't fully happen, or if it takes a lot longer than someone as forward-thinking as Signal might think, it is certainly the case that the chatbots that you interact with are doing something different than they used to.

市面上每个 AI 编程工具做的第一件事都一样:就开始写代码。

Every AI coding tool on the market does the same thing first: it starts writing code.

Blitzy 却反着来。

Blitzy does the opposite.

在写一行代码之前,Blitzy 会花好几天时间反向工程你整个代码库。

Before writing a single line, Blitzy spends days reverse-engineering your entire codebase.

成千上万的代理会吸收数百万行代码,映射出每一个依赖、每一个没有文档记录的约束、每一个在过去十年里做出的架构决策。

Thousands of agents ingest millions of lines, mapping every dependency, every undocumented constraint, every architectural decision made over the last decade.

结果就是一个动态的知识图谱,它能像一位在这行里待了三十年、经验丰富的工程师那样理解你的软件。

The result is a dynamic knowledge graph that understands your software the way a principal engineer would after 30 years in the building.

其他工具只用 grep 搜索和 markdown 文件来猜测上下文。

Other tools guess at context with grep searches and markdown files.

Blitzy 从来不会乱猜。

Blitzy never guesses.

它先建立真正的理解,然后自主完成超过 80% 的整个软件史诗任务。

It builds true understanding first, then delivers over 80% of entire software epics autonomously.

经过验证、端到端测试、达到生产级别的 Pull Request。

Validated, end-to-end tested, production-grade pull requests.

这就是为什么世界 500 强的工程团队会把最关键的代码库交给 Blitzy。

That's why Fortune 500 engineering teams trust Blitzy with the codebases that matter most.

去 blitzy.com 亲自看看吧。

See for yourself at blitzy.com.

网址是 B-L-I-T-Z-Y 点 com。

That's B-L-I-T-Z-Y dot com.

最优秀的团队不是靠一个人扛着所有人前进。

The best teams don't have a single star carrying everyone else.

他们清楚自己的优势,也了解彼此的短板,并且会利用这两点。

They know their own strengths and each other's weaknesses and play to both.

这就是 Robots and Pencils 刻意打造的团队。

That's the team Robots and Pencils has built on purpose.

那里没有人为了凑人头数而磨洋工。

Nobody there is grinding through busywork to pad a headcount number.

大家为了解决难题而来,并且因为周围人都在同步进步而留下来。

People come for the hard problems, and they stay because everyone around them is leveling up at the same time.

在一个很多公司只想着快速招人的市场里,这值得一看。

In a market full of companies that are just trying to hire fast, that's worth a look.
M1
M113:35

每期节目,我都会聊到 OpenAI、Anthropic、SpaceX AI、Google 和 Meta 之间的竞争。

Every episode, I talk about the competition between OpenAI, Anthropic, SpaceX AI, Google, and Meta.

如果你听了一段时间,可能已经有自己的偏好了。

And if you've been listening for a while, you might have a favorite.

也许你觉得 OpenAI 和 Anthropic 能保持领先,或者 Meta 的开源策略能赢。

Maybe you think OpenAI and Anthropic can stay ahead, or perhaps Meta's open-source strategy can win out.

不管你怎么看,每个 AI 实验室都创造了不同的投资机会。

Whatever your view, every AI lab creates a different investment opportunity.
M1
M115:10

顺便说一句,我们在这个节目里聊过很多次的一个简单但深刻的变化是,你跟那个聊天机器人的交互方式会越来越多地通过语音。

And by the way, one simple but profound change, which we've talked about a lot on this show, is that the way that you're interacting with that chatbot is increasingly going to be through your voice.

我今年发布的每一个免费学习项目,一开始都恳请大家要么用原生的 ChatGPT 语音控制,要么安装类似 WhisperFlow 的东西,纯粹是为了语音带来的高效和更广阔的上下文。

Every free learning program I've released this year starts by imploring people to either use native ChatGPT voice control or to get something like WhisperFlow set up, for the sheer efficiency and expanded context that voice is going to enable.

而且,这正逐渐成为智能体产品的基本门槛。

And increasingly, this is just table stakes for agent products as well.

这周,GrokBot 也加入进来了,它有了自己的原生语音集成。

This week, GrokBot joined the party with its own native voice integration.

xAI 的 Danny Graziosi 写道,用语音模式快速向你的首席参谋下达一连串任务,同时它协调你的机器人团队,感觉就像魔法一样。

xAI's Danny Graziosi wrote, using voice mode to rapid-fire tasks to your chief of staff while it orchestrates your bot team feels like magic.

同一团队的 Parker Conrad 写道,GrokBot 已经在向更无缝的交互模式迈出了一大步。

Parker Conrad from that same team wrote, GrokBot already took such a leap towards a more seamless interaction model.

我看好语音能把它推得更远。

I'm bullish on voice taking that even further.

Gaurav Bison 指出,这不只是语音本身,而是语音正在与更强大的底层能力相结合。

Gaurav Bison pointed out that it's not just voice alone, but the fact that voice is now meeting expanded underlying capabilities.

他写道,今年我试过很多智能体产品的语音功能,在智能体真正能做事之前,它一直感觉像个噱头。

He wrote, tried voice on a bunch of agent products this year, and it always felt like a gimmick until the agent could actually go do things.

语音加上真正干实事,这才是组合拳。

Voice plus real work is the combo.

就连那些早就拥抱语音的公司和项目,现在也变得更加重视了。

And even the companies and projects that were already voice-pilled are getting even more so.

OpenAI 刚刚发布了一个新的 Codex 广告,他们团队的两名开发人员在锻炼的时候,跟 Codex 中由新模型 GPT Live 1 驱动的语音智能体对话。

OpenAI just released a new Codex ad with 2 of their team's developers talking to the voice agent in Codex powered by their new GPT Live 1 model.

他们一边锻炼一边干这个。

While they're working out.

Google 也在做同样的事,这周发布了 Gemini 3.8 Live 和 3.8 Live Extended Thinking,他们写道,在智能和并行推理方面的重大升级,让你用语音协作和执行复杂任务变得更直观了。

Google is also on the same tip, releasing Gemini 3.8 Live and 3.8 Live Extended Thinking this week, writing that major upgrades in intelligence and parallel reasoning make it more intuitive to collaborate and execute complex tasks using your voice.

上周,和Codex在那支健身广告里宣传的一样,我们有了Devin Voice,这让使用Devin的开发者也能通过语音调用新的GPT Live模型来写代码了。

Last week, similar to the thing that Codex was advertising in that workout ad, we got Devin Voice, which allowed developers using Devin to also tap into the new GPT Live model to code by talking.

Allie K. Miller总结得很到位,她说:每次我看到新的语音优先界面,心里就有个小声音在欢呼。

And Allie K. Miller summed it up when she said, Whenever I see a new voice-first interface, a tiny voice inside of me screams with joy.

语音就是现在和未来的交互界面。

Voice is the interface of the present and future.

今年,除了单线程任务和更强大的、能管理自己任务队列的智能体协调器,以及语音指令之外,我们给AI下达指令的性质也发生了变化。

Now alongside monothreads and more capable agent coordinators that manage their own fleets and voice-based instructions, the nature of the instructions that we give to our AI has also changed this year.

总结一下这个重大转变:我们不再只是给AI“提示”,而是越来越多地给它们设定目标。

Summarizing the big shift, instead of prompting our AIs, we increasingly set goals for them.

这本质上给了AI更多自由裁量权和自主性,让它们自己去琢磨怎么完成工作,并且能反复循环直到任务做完。

Inherently, this creates more latitude and agency on the part of the AIs for figuring out how that work should get done and for working in recurring loops until that work gets done.

这个变化始于一些基础功能的引入,比如Claude Code和Codex里的“/goal”命令。

This started with the introduction of primitives like /goal in Claude Code and Codex.

这是对这种交互模式的规范化,让人们可以指定目标,而不仅仅是扔一条提示。

It was a formalization of this interaction pattern that allowed people to specify goals instead of just dropping a prompt.

后来这扩展到了整个“循环工程”的概念,我们整个夏天都在聊这个,大概一个月前还专门做过一次网络研讨会。

This would later expand to the whole idea of loop engineering, which is something we've talked about throughout the summer and even did a webinar on about a month ago.

循环的概念真正火起来,是因为OpenAI的创作者Peter Steinberger写道:你不应该再给编码智能体写提示了,你应该设计循环来提示你的智能体。

The idea of loops really got a big lift when OpenAI creator Peter Steinberger wrote, you shouldn't be prompting coding agents anymore, you should be designing loops that prompt your agents.

循环的概念很简单:除了定义目标之外,你还为这个目标定义成功标准,让智能体能自己对照衡量,这样它就能在一个持续循环中运行,直到根据你给的量化标准达成目标。

This idea of a loop is simply that in addition to defining the goal, you define success criteria for the goal that the agent can measure itself against so that it can run in a continuous loop until it achieves said goal based on the quantifiable criteria given.

这个变化之所以值得在节目里提,是因为虽然“循环”和“循环工程”听起来完全是我们早期用户圈子里喜欢的那种热门词,其他人听到可能只想翻白眼,但实际上,这是一个相当关键的思维转变,关乎如何最大化我们与AI的互动效果——这也是我们花这么多时间聊它的原因。

And the thing is, and the reason that it's worth mentioning in this show, is that although loops and loop engineering has all of the hallmarks of the type of buzzy terms that we love over in our early adopter circles, but which reasonably elicit eye rolls among the rest, this is actually a fairly critical mindset shift when it comes to thinking about how to get the most out of our interactions with AI, which is why we spend so much time on it here.

我们与AI交互方式的第六个变化,我想可以叫“多模型并用”,也就是在个人技术栈里同时使用多个不同的模型。

The 6th change in how we interact with AI is, I guess you might call multimodality, i.e., the using of multiple different models in a personal stack.

这不只是组织层面的事情,即使是在单个用户的工作范围里也是如此。

Not just on an organizational level, but even within the domain of an individual user's work.

我到现在还没找到最合适的词来描述这种多模型并用,我的很多Patreon订阅者也喜欢指出这一点。

I have not yet settled on the exact right language to describe this multimodality, as a number of my Patreons love to point out.

不过,一方面,很多高级AI用户一直把模型选择当成一种获取优势的方式——也就是说,不局限于你订阅付费的那个单一模型,而是找到最适合特定任务的模型——但现在不同了,任务的复杂度、难度、深度和持续时间,加上最先进模型的成本,以及这些模型在标准套餐中补贴方式的变化,都意味着这不再只是早期用户的尝鲜行为,而是变成了成本和效率的现实要求。

But while on the one hand, many advanced AI users have always viewed model selection as a way to get benefits— in other words, not just being constrained to a single model because that's the one you pay for a subscription for, but finding the model that does the best with your particular task— the difference now is that the complexity, difficulty, depth, and duration of the work, combined with the cost of the most advanced models and shifts in the subsidy those models in the standard plans means that this is not just an early adopter alpha thing anymore, but basically a cost and efficiency requirement now.

另一个促成因素是,随着时间的推移,那些稍微落后于技术前沿但更便宜的模型,已经能完成几个月前才属于前沿水平的事情。

It's also enabled by the fact that the more time that goes on, the more that cheaper models that are just outside the state of the art can do things that would have been state of the art just a couple of months ago.

这在Fable 5.1发布的时候特别明显。

You saw this a lot when Fable 5.1 dropped.

当时X上和各个AI社区里,大家都在分享各种技巧,怎么最大化利用这个模型又不花太多钱。

You had people all over X and the other AI communities sharing their hacks on how to get the most out of the model without breaking the bank.

而且,这种情况正在从事后找技巧,逐渐演变成实际的使用系统规范。

And increasingly, you can see this moving from ex post hacks to actual usage systems discipline.

这种交互模式的转变,和“聊天机器人作为智能体舰队管理者”的想法密切相关。

And this interaction pattern shift dovetails very closely with the idea of chatbots as agent fleet managers.

很多人注意到,至少在目前,像Fable 5.1这样的模型,默认做法不只是用最先进的模型做协调员,连所有子智能体也用它。

One of the things that a lot of folks have noticed is that at least at current, the default for a model like Fable 5.1 is to not just use the most advanced model as the coordinator, but also to use it for all the subagents.

我们很多人都遇到过这种情况:交给Fable一个我们认为很常规的任务,结果回头一看,当前的用量上限被用掉了很大一部分——原来它为了做点研究,启动了5到10个子智能体,而这些子智能体全都用了最先进的Fable模型,可手头的任务完全不需要这么高的配置。

Many of us have had the experience where we give Fable what we think is a fairly standard task, only to come back and see a major percentage of our current usage cap tapped because it turns out that the 5 or 10 subagents that it spun up to go do some sort of research were all using the most advanced Fable model as well, even though that was completely not required for the task at hand.

简单说,搞清楚哪些任务需要哪种模型、什么能力,以及如何在管理AI时规范化这一点,将越来越重要。

TL;DR, understanding where different types of models and model capabilities are needed and how to normalize that in your management of your AI is going to increasingly be an important discipline.

顺便说一句,这个趋势如果不小心处理,可能会和那个追求简洁的趋势相冲突。

And by the way, this is one trend that, if not careful, could kind of run at odds with that trend towards simplicity.

因为所有模型和工具厂商都在试图让普通用户体验更简单,这往往意味着隐藏甚至取消那些精细的控制选项——比如决定哪个任务用哪个模型。

As all of the model and harness labs try to make the average experience more simple, in many cases that will mean obscuring or even doing away with the more fine-grained controls for determining things like which tasks get which model.

理论上,厂商们会尝试在底层构建自动化系统来帮你搞定这些细节——比如意识到研究任务的子智能体不需要用最前沿的模型。

Now, theoretically, the labs and model companies will try to build automated systems underneath that can figure out those specifics for you, i.e., to realize that the subagents needed for a research task are not state-of-the-art.

但我并不确定他们总能做对。

But I'm not sure that they're always going to get it right.

所以我觉得,这方面用户还是需要保留一定的自主权。

And so I think that this will still be something that people need personal agency around.

最后一个交互模式的变化,我最近聊了很多,所以不打算深入展开,就是“共享智能体”这个概念。

The last interaction pattern shift is one that I've talked about a lot recently, so I won't go too deep on, but is the idea of shared agents.

2026年会被看作智能体真正成为现实的一年,在那之前大家已经兴奋了好多年。

2026 Will be looked back at as the year that agents became real after years and years of excitement.

但在今年头三分之二的时间里,几乎所有智能体行为都还局限在个人层面。

But for the first two-thirds of the year, almost all of that agent behavior was siloed to individuals.

也就是我们每个人各自组建自己的OpenAI团队,改变自己做事的方式,但不一定会改变我们团队整体的工作方式。

It was each of us individually building our OpenAI teams, changing how we do work, but not necessarily changing how our teams do work overall.

这个趋势正在开始转变。

You're starting to see that shift.

今年夏天有个例子,Claude推出了Claude Tag,取代了之前在Slack里用的老系统。

An example of that came this summer when Claude introduced Claude Tag, which replaced the old system of Claude in Slack.

以前每个人可以调用自己的Claude实例,现在换成了一套基于团队的系统,每个频道都有自己的Slack,能够跨整个频道的上下文工作,不再绑定到某一个人身上。

Where each individual could call upon their own Claude instance, with a new team-based system where each channel had its own Slack that could work across the context of that channel, not tied to any one individual.

这个思路其实很清楚。

Now the idea here is pretty clear.

我们的工作有一部分确实是个人完成的,但很大一部分是团队完成的,所以我们需要的可能不只是存在于个人小世界里的智能体,而是存在于团队交汇点上的智能体。

Some amount of our work gets done, yes, as individuals, but a lot of it gets done in teams, and we're probably going to need agents that live at the intersection of our teams, not just agents that live in our own little individual worlds.

现在你开始看到很多朝着这个方向的努力了。

And you're now starting to see tons of efforts towards these ends.

Mio XYZ这周上线了,号称是“整个团队共享的第一个AI员工”。

Mio XYZ launched this week as the, quote, first AI employee your whole team shares.

他们在公告里写道,从7月开始内测以来,Mio为团队节省了数千小时,完成了数万个任务。

In their announcement post, they wrote, in beta since July, Mio saved teams thousands of hours and completed tens of thousands of tasks.

你不需要另一个智能体,你需要一个能和团队共享的智能体。

You don't need another agent, you need a shared one with your team.

我几乎每天都能看到六七个类似趋势的例子,不过多人AI的具体含义还处在探索阶段。

On any given day, I feel like I see a half dozen or more examples of this same impulse, although exactly what multiplayer AI means is still very much in development.

本月初,Ethan Mollick写道:多人AI——即组织中多人一起使用AI来达成目标——仍然是当前使用AI时最大的非技术难题之一。

Earlier this month, Ethan Mollick wrote, Multiplayer AI, where many people in an organization can use AI together to accomplish goals, remains one of the biggest non-technical problems in using AI right now.

目前的方法还非常原始,基本上就是把AI当成群聊里的一个人。

Approaches tend to be pretty primitive and based around AI as a person in your group chat.

这很局限。

That is limiting.

Aria Bhuktani写道,我们至今没有一个统一的心智模型来理解在多人智能体系统中,到底谁是主体。

Aria Bhuktani writes, we still don't have a consistent mental model for who the principal is in a multiplayer agentic system.

是你,用户吗?

Is it you, the user?

智能体作为你的代理,继承你的上下文、权限和授权?

Agent is your proxy, inheriting your context, access, and authority?

还是智能体本身?

Is it the agent itself?

智能体有自己的身份、关系、记忆和权限。

Agent has its own identity, relationships, memory, and access.

Aria接着说,这就像是智能体文明与人类文明以智能体速度运转之间的区别。

It's the difference, Aria continues, between agent civilizations versus human civilizations at agent speed.

实际上,答案很可能是两者皆有,但适用于不同的系统和场景。

Realistically, the answer is both, but for different systems and use cases.

多人AI将成为2027年最大的设计问题之一。

Multiplayer AI will be one of the biggest design problems of 2027.

光是这个短语,跟我聊的人每个人理解都不一样。

The phrase alone means something different to every person I talk to.

它指的是共享记忆、上下文、组织架构,还是完全不同的东西?

Is it shared memory, context, org design, or something else entirely?

有意思的是,Rishi随后解释了为什么每个角度单独看都很有道理。

And what's interesting is that Rishi then goes on to show why each of these sounds compelling when you look at it in isolation.

他写道,从记忆角度看:五个智能体做一个项目,对发生的事情有五个不同版本,彼此不沟通。

The memory lens, he writes, 5 agents on 1 project, 5 different versions of what happened, none of them talking to each other.

一团糟。

It's a mess.

解决办法就是用一份共享记忆取代五份私有副本,这就是全部关键所在。

Fixing it, 1 shared memory instead of 5 private copies, is the whole game.

但接下来还有上下文这个维度。

But then there's the context lens.

Rishi 接着说,几乎每个我聊过的人,都给自己搭建了一个小小的上下文层,给自己的 agent 用。

Rishi continues, nearly everyone I talk to has built their own little context layer for their own agents.

对他们自己来说是管用的。

It works for them.

但放到整个团队里,那就行不通了。

Stretch it across the whole team and it falls apart.

个人上下文感觉已经解决了,但组织级别的上下文还完全是个空白。

Personal context feels solved, org-wide is wide open.

然后当然还有组织变革这个维度。

And then of course there's the organizational change dimension.

Rishi 总结说,多人 AI 最难的部分可能根本不是技术问题。

Rishi concludes, the hardest part of multiplayer AI might not be technical at all.

让整个团队都采用同一种工作方式,这更像是组织设计,而不是工程问题。

Getting a whole team to adopt one shared way of working is more org design than engineering.

没人愿意承认这一点。

Nobody wants that to be true.

但很明显,这方面会有大量的工作要做。

What's clear is that there is going to be a ton of work done in this area.

我们还没就多人 AI 的架构模式达成共识,就已经决定要让 AI 进入多人模式了。

We've decided AI is going multiplayer before we've agreed on the architectural patterns for multiplayer AI.

然后当然他会继续解释这一切,试图让我们得到一个稍微更连贯的答案。

And then of course proceeds by trying to explain all of that and get us to a slightly more coherent answer.
M1
M125:25

会有大量的产品体验和优秀的产品思路,出现在多人 AI 或者共享 agent 这个领域。

There are going to be tons and tons of product experiences and great product thinking coming to the multiplayer AI or shared agent space.

而且很快就会来。

In very short order.

所以,如果你想把这个问题往后推一推,让行业和规范再定型一些,我完全可以理解。

And so to the extent you want to kick that can down the road just a little bit and let the industry and norms settle a bit more, I totally understand.

话虽这么说,但我确实觉得,对于想要深入尝试的团队来说,抢先一步肯定能获得一些先机。

That said, I do think for teams who want to dive in, there is certainly some alpha to be had by getting out ahead of that.
M1
M125:50

以上就是我们使用 AI 的方式正在改变的七个方面。

So that is the 7 ways that how we use AI is changing.

希望这能给你一些值得探索的有趣新方向,以及一些思考你和 AI 关系的新角度。

Hopefully this gives you some interesting new directions to explore and some new ways to think about your relationship with AI.

那么,今天的 AI Daily Brief 就到这里。

For now, that's going to do it for today's AI Daily Brief.

一如既往感谢你的收听或收看,下次见,Peace。

Appreciate you listening or watching as always, and until next time, peace.
已剔除 5 处广告(点击展开查看)
M1
M10:43广告 · 已剔除

First of all, thank you to today's sponsors: Blitzy, Robots and Pencils, Harbor, and Hyperagent.

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

To learn more about sponsoring the show, send us a note at [email protected].

M1
M113:30广告 · 已剔除

Check out robotsandpencils.com/careers.

M1
M113:52广告 · 已剔除

Harbor Capital Advisors' AI Lab Ecosystem ETF Suite lets you invest in the ecosystem behind the AI lab you believe in.

Search Harbor AI Lab Ecosystem ETFs wherever you invest or follow @HarborCapital on X to learn more.

Visit harborcapital.com for a prospectus containing investment objectives, risks, fees, expenses, and other important information.

Read and consider it carefully before investing.

Risks include principal loss and artificial intelligence-related risks.

Harbor ETFs are distributed by Foresight Fund Services LLC.

Harbor is not affiliated with AI Daily Brief, and the funds are not affiliated with, sponsored by, or endorsed by any AI lab.

This is a paid advertisement and not personalized investment advice.

Investing involves risk, including possible loss of principal.

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

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

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

Marketing agents turn competitor moves into landing pages.

Sales agents enrich leads, draft emails, and update the CRM.

Ops agent chases the paperwork and tracks the budget.

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

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

Hire yours at HyperAgent.

Get $100 in credits at hyperagent.com/AIDailyBrief.

M1
M124:15广告 · 已剔除

Now, if this is something that you are interested in, I will show for a moment the latest free self-directed training program that we've dropped here, which is the Multiplayer AI Sprint for Teams.

This does not assume that multiplayer AI is figured out, or frankly almost any of the answers to those important questions.

What the sprint does is it gets people to start to build the foundations to how multiplayer AI, which is really just another way in our case of saying shared agents, can function inside your team.

Over the course of the 4 sprints, which we've divided into weeks but don't have to be split up that way, you're going to inventory first what your team is actually running right now for AI and/or to the extent that you are even more new, perhaps bring in some champion or expert from your organization who is a little bit farther in their stack to help you understand best practices in your specific organizational context.

And once you have that sort of inventory, you move then into sprint 2, which is all about context, writing down what your AI knows about you and what it should know about you.

Sprint 3 is looking at the overlap in workstreams and in context, which is going to give you your possible places to build shared agents.

And sprint 4 is going to be where you actually do that building.

And of course, you can then repeat sprint 3 and sprint 4 multiple times to experiment with multiple agents as much as it helps.

M1
M125:44广告 · 已剔除

And hopefully that's what the multiplayer AI sprint, which is of course linked in the show notes, can maybe help you do.