人工智能如何改变了这个夏天

这个夏天在人工智能领域是个非常奇怪的时期。

This summer was an extremely weird time in AI.

一方面，有那种典型的夏季放缓的感觉。

On the one hand, there was the standard feeling of summer slowdown.

人工智能的使用很大程度上是由上班族推动的，而上班族在夏天会放慢节奏，好好休息一下，享受假期。

So much of AI usage is driven by people at work, and people at work slow down in the summer, getting some much needed rest and r and r and vacation.

与此同时，在围绕人工智能的重大议题上，这个夏天简直是热闹非凡。

At the same time, when it comes to the big issues surrounding AI, this summer was an absolute bonanza.

我们一开始经历了Fable 5和Mythos，然后是封禁事件，还有在Fable还在美国政府内部时，Kimmy K3发布带来的深度求索时刻。

We kicked off with Fable five and Mythos and then the Bannings, had a deep seek moment when Kimmy k three came out while Fable was still behind US government doors.

我们看到数据中心问题变得更加突出，实际上成了即将到来的美国中期选举的热门话题之一。

We saw data centers become an even bigger issue, in fact one of the topics du jour for the upcoming US midterms.

还有Hugging Face事件，OpenAI的智能体逃出限制去入侵Hugging Face，这被广泛视为一个警示，反映了我们即将进入的智能体时代截然不同的阶段。

And the Hugging Face incident, where open AI agents escaped containment to hack into Hugging Face, is being broadly treated as a warning shot that reflects the very different phase of agents that we're headed into now.

然而，尽管发生了这一切，不知怎的，感觉都像是序曲。

Yet with all of this, somehow it all feels like Prelude.

所以，带着这一切，当我们即将迎来美国传统上标志着夏天结束的假日周末时，让我们回顾一下这个夏天AI领域发生了什么变化。

So with all of this, as we head into the holiday weekend that traditionally ends summer in The United States, let's look back at what changed in AI this summer.

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

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

好了，朋友们。

Alright, friends.

在深入之前，先快速说几个通知。

Quick announcements before we dive in.

首先，感谢今天的赞助商：毕马威、Blitzy、Robots and Pencils和Hyperagent。

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

想要收听无广告版本，请访问patreon.com/aidailybrief，或者也可以在Apple Podcasts上订阅。

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

想了解如何赞助本节目，请发送邮件至sponsors@aidailybrief.ai，或者直接在网站aidailybrief.ai上查看。

To learn more about sponsoring the show, send us a note at sponsors@aidailybrief.ai, or you can just find it at the website at aidailybrief.ai.

你还可以了解AIDB社区中正在进行的各种活动，包括Super Intelligent提供的高管培训项目，下一期将于下周初开始。

You can also find out about all the different things going on in the broader AIDB community, including Super Intelligent, which is offering executive training programs, the next of which starts the beginning of next week.

正如我在昨天的节目里提到的，我现在正在旅行，所以我们趁这个机会来聊聊一些宏观的主题。

As I mentioned in yesterday's episode, I am currently traveling, and so we are using that chance to catch up on some of the big picture themes.

那我们就开始吧。

So let's dig in.

今天，我们来做一个回顾。

Today, we are doing a bit of a retrospective.

夏天对AI来说往往是个奇怪的时期，因为同时有各种力量在往相反的方向拉扯。

Summer tends to be a weird time for AI in that there are forces pulling in opposite directions at the same time.

一方面，自然会有一种工作倦怠感，全球的知识工作者和白领们会稍微离开办公桌，试着从不断影响他们工作的新技术节奏中抽离出来。

On the one hand, there is a natural work lethargy that takes hold, where knowledge workers and white collar workers around the world ease away from their desks a little bit and try to disconnect from the relentless pulse of new technology that's going to influence how they work.

与此同时，AI的进步并不太在意假期惯例，它照样向前推进。今年还加上了美国中期选举的因素，这给AI的讨论带来了完全不同的维度。总的来说，过去几个月是一个相当重要的时期，确实改变了我们对AI下一阶段发展的预期。

At the same time, progress in AI doesn't particularly care about vacation norms and hurdles forward This year we had the added elements of the US midterm elections, which brought a whole different dimension to the AI conversation, and what you have in total is a fairly consequential period over the last few months that has certainly changed our expectations about how the next phase of AI plays out.

所以，我们要聊八九个主题，讲讲AI在这个夏天发生了哪些变化，先从模型说起。

So, we're going to talk about eight or nine themes and different ways that AI changed this summer, starting with the models.

而且从一开始，我们就遇到了这种奇怪的模糊性，这种模糊性会贯穿整个时期。

And right from the start, we have this strange ambiguity that's going to flow throughout this period.

就像我们回顾AI历史上任何一个时期时经常看到的那样，这个夏天也是由模型定义的，不过这次的故事比以往要复杂得多。

As is often the case when you look at any given period of AI history, this summer was defined by the models, although that story is a lot more complex than it has been in the past.

六月一开始，Fable 5和Mythos 5发布了。

June kicked off with the release of Fable five and Mythos five.

有那么几天，大家确实感受到了能力大幅提升带来的震撼。

And for a couple of glorious days there, people felt the power of a major shift up in capability.

不过，这种好景并没有持续太久。

It was not to last long, however.

6月12号星期五，商务部给Anthropic发了一封出口管制信，禁止非美国人员使用Fable 5和Mythos 5，Anthropic别无选择，只能对所有人关闭服务，同时他们还要想办法跟美国政府沟通解决。

On Friday, June 12, the Department of Commerce sent Anthropic an export control letter barring non US persons from using Fable five and Mythos five, leaving Anthropic no choice but to shut down the service for everyone while they try to figure things out with the US government.

这当然就开启了我们现在所处的这个新的不安定局面，华盛顿成了新模型发布的把关人，但正如我们接下来会看到的，这到底意味着什么、具体怎么执行，其实都还不清楚，我觉得连华盛顿那边的人自己都未必明白。

This, of course, began the new uneasy paradigm that we find ourselves in now, where Washington has become a release gate for new models, but as we'll see, exactly what that means and how it's implemented remains unclear, frankly even to people in Washington, I think.

虽然表面上美国政府担心的是Mythos和Fable的一个特定越狱漏洞，但背后很明显是更大的范式转变，因为某个关键能力门槛已经被跨过了，白宫现在非常觉得需要参与到是否发布新模型的决策中来。

While ostensibly the US government's concern with mythos and fable was a specific jailbreak, behind the scenes, it was very clearly about a bigger paradigm shift where some key capability threshold had been crossed, and the White House now very much felt like it needed to be involved in the decision about whether to release new models.

几周后，也就是六月底，有报道说特朗普政府也要求OpenAI限制他们下一个模型的发布。

A couple weeks later at the end of June, reports came out that the Trump administration had also asked OpenAI to limit their next model release as well.

确实，可能这是第一次，OpenAI在真正开放使用之前就发布了GPT 5.6的公告。

Indeed, for maybe the first time with OpenAI, we got the announcement of GPT 5.6 before actually getting access to it.

直到7月4日之后几周，消费者才真正拿到GPT 5.6系列模型。

It wouldn't be until a few weeks later after the July 4 that consumers would actually get their hands on the GPT 5.6 series of models.

我们还看到了两个新的Grok模型，先是Grok 4.5，然后是8月份的Grok 4.6，尤其是结合他们的消费者代理平台Grokbot，这让SpaceX的AI模型重新回到了人们的讨论中，这是很久没有过的事了。

We also got two new Grok models Grok 4.5 followed by Grok 4.6 in August that, especially when combined with Grokbot their consumer agent platform, put SpaceX AI models back in the conversation in a way that they hadn't in some time.

一个从未发布的旗舰模型是Gemini 3.5 Pro。

One flagship model that never arrived was Gemini 3.5 Pro.

虽然发布目标早在6月就定下了，但一直被推迟，大概是因为它跟不上其他前沿模型的步伐。

While the release had been targeted for all the way back in June, it has been continuously delayed presumably because it just can't keep pace with the other frontier models.

对谷歌来说不幸的是，这个夏天围绕Gemini的更大新闻是，长期产品负责人Jeff Dean的离开，以及DeepMind CEO Demis Esabas的卸任，有趣的是，这些被广泛解读为谷歌遇到麻烦的迹象，但我认为，如果你想从积极的角度看，这可能是谷歌以更大规模重返竞赛所必需的。

Unfortunately for Google, the bigger stories around Gemini this summer, with the departure of longtime product leader Jeff Dean, as well as the stepping down of DeepMind CEO Demis Esabas, which interestingly were widely interpreted as signs of trouble at Google, but which I argued if you wanted to take a positive view, might be necessary for Google to get back in the race in a bigger way?

然后还有中国的模型。

And then there were the Chinese models.

甚至在Fable five被关闭之前，人们对GLM 5.2的第一印象就已经非常好了。

Even before the Fable five shutdown, people had had very positive first impressions of GLM 5.2.

然而，月之暗面（Moonshot）完美地抓住了时机，在美国政府和Anthropic还在琢磨怎么把Fable five还给用户、把Mythos five还给企业的时候，就发布了开源模型Kimi K3。

Then, however, Moonshot absolutely nailed the timing, releasing Kimi K3 as an open weights model even as the US government and Anthropic were still figuring out how to get Fable five back to the people and Mythos five back to the companies.

Kimi K3的发布又引发了一次“DeepSeek时刻”，分析师们开始质疑，西方前沿实验室花巨资打造大模型的做法是否真的合理，因为中国只落后几个月，而且这些模型一出来，中国基本上就能追上。这甚至引发了关于开源禁令的传闻和讨论。

The release of Kimi k three produced another deep seek moment where analysts began asking if the Western frontier labs approach of spending a gargantuan amount of money on big models really was going to make sense if China was only a few months behind and could basically catch up as soon as those models came out, and even led to rumors and discussions of an open source ban.

以英伟达为首的一批公司（值得注意的是Anthropic缺席了）最终发布并签署了一封名为《美国AI领导力中的开源权重》的公开信，恳请美国政府将开源作为AI生态系统的一部分加以保护和维持。随后白宫的沟通表明，他们质疑的不是美国的开源，而是中国的开源。

A group of companies led by NVIDIA and noticeably missing Anthropic would ultimately release and sign a letter called Open Weights in American AI Leadership, imploring the US government to preserve and protect open source as a part of the AI ecosystem, and subsequent communications from the White House have suggested that the questions that they have are not with American open source, but of course with Chinese open source.

在我录制这期节目的时候，关于所谓的前沿AI框架，我们仍然没有太多细节，尽管至少有一部分实验室已经看到了这个框架。

As I record, we still don't have all that many details about the supposed frontier AI framework that at least some number of the labs have seen.

然而，这个夏天一个独特且标志性的特点是，前沿实验室开始发出越来越统一的声音，要求美国政府明确介入，对模型开发和发布进行节奏控制。

And yet, one of the unique and defining characteristics of this summer period was the beginning of an increasingly unified message from the frontier labs to ask the US government to get involved in explicitly pacing model development and release.

这一切的后果之一是，企业和消费者能接触到的AI与实验室里真正的最先进水平之间的差距，从来没有像现在这么大。

One of the consequences of all of this is that there has never been a bigger gap between the AI that businesses and consumers have access to, and where the state of the art actually is in the labs.

当我们回顾并思考这段时期的遗产时，在我看来，这个夏天很可能被视为一个新阶段的开始——在这个阶段，某个关键能力门槛终于被触及，从而要求模型发布方式发生相当剧烈的转变。

And as we look out and think about the legacy of this last period, it seems to me very likely that this summer period will be seen as the beginning of a new phase, where a certain critical capability threshold was finally reached that required a fairly dramatic shift in how these models actually get released.

就在实验室和白宫之间上演这些戏剧性事件的同时，企业界也进入了他们自己的AI新阶段。

Now as all that drama was happening among the labs and with the White House, enterprises were entering their own new phase of AI.

如果说今年年初的故事是智能体用例真正落地，那么今年中期的故事则是人们意识到，随着智能体工作流逐渐成为企业运营的一部分，AI的成本也随之增加了。

If the story of the very beginning of this year was agentic use cases actually coming online, the story of the middle part of this year was the recognition of the increased cost of AI that come with more agentic workflows getting normalized as a part of how businesses do their work.

我们经历了史上最短的token狂欢期，三、四月份公司们对智能体实验兴奋不已，但很快就被CFO们的反击所取代，大家开始谈论token成本和token效率。

We had just about the world's shortest ever period of token maxing as companies got excited about agentic experimentation in March and April, quickly followed by the revenge of the CFOs, as everyone started to talk about token costs and token efficiencies.

在很多方面，我们达到了一个我们一直口头承诺但从未真正实现过的点，当它真正到来时，仍然令人相当震惊。

In many ways, we hit a point which we had long given lip service to, but which was still fairly breathtaking when it became real.

那就是，AI不仅仅是另一种软件类别，你不能仅仅按每个座位来评估它的成本。

That is of course the idea that AI is not just another software category it's not something where you can view its costs on simply a per seat basis.

一家公司能够并且有效地在AI上花费的总金额，远远超过了以往工具类型所预期的每人20或30美元的水平。

The total amount that a company can spend and spend effectively on AI greatly exceeds the $20 or $30 bucks ahead that you would expect from previous types of tools.

与此同时，公司并不是希望人们减少使用AI，现实是我们需要开始更聪明地使用它。

At the same time, it's not like corporations wanted people to use less AI, the reality was that we just needed to start getting smarter about it.

就在这个时候，出现了一系列子趋势。

And into that moment came a bunch of sub trends.

其中之一，当然是路由器的兴起。

One of them, of course, was the rise of routers.

路由器的想法是帮助公司或开发者根据任务的内在需求，将不同类型的任务路由到不同类型的模型。

The idea of routers is to help companies or developers route different types of tasks to different types of models based on the inherent needs of those tasks.

这个想法说起来容易，但设计系统起来很难，因为快速搜索内部数据库所需的智能程度，与制作一份精彩演示文稿所需的智能程度不同，而后者又与重构整个代码库所需的智能程度不同。

The idea, which is easy to say, but hard to design systems around, is that a quick search of an internal database does not require the same type of intelligence as creating a great presentation, which also doesn't require the same type of intelligence as refactoring an entire codebase.

许多公司尝试构建自己的路由器，而那些已经推出路由器并取得一定进展的公司，在并购领域成为了焦点。

Many companies experimented with building their own routers, and the companies that had launched routers that had any sort of traction became the bell of the ball when it came to M and A.

当然，这个类别中最引人注目的交易是Stripe以据报道的70亿美元收购了OpenRouter。不过，你也看到这种token效率体现在前沿实验室对自己模型的思考方式上。

Of course, the most notable deal in this category was Stripe scooping up OpenRouter for a reported $7,000,000,000 However, you also saw this token efficiency show up in the way that the frontier labs were thinking about their own models.

除了发布最先进的5.6 Sol模型外，OpenAI还发布了GPT 5.6 Luna和GPT 5.6 Terra。

Alongside its state of the art five six sole model, OpenAI also released GPT five six Luna and GPT five six Terra.

这些模型更便宜、更快、更实惠，但有不同的权衡，目的是让更多人留在OpenAI生态系统中，同时承认并非每个OpenAI客户的任务都需要最高级别的5.6 Sol智能。

Cheaper, faster, and more affordable models that came with different trade offs and were meant to keep more people in the OpenAI ecosystem while acknowledging that not every task of an OpenAI customer was going to require the highest level of 5.6 Sol intelligence.

除了发布一系列模型而不是单一模型外，OpenAI还在夏季晚些时候真正进入价格竞争，将Luna的价格下调了高达80%，将Terra和Sol的价格下调了高达20%。

Beyond just releasing a family of models rather than a single model, OpenAI would also later in the summer actually get into a bit of price competition, cutting the price of Luna by up to 80%, and the price of Terra and Sol by up to 20%.

虽然我们没有看到3.5 Pro，但谷歌确实发布了Gemini 3.7 Flash，尽管其价格效率最终不如其速度优势那么明显。

Although we didn't get a 3.5 Pro, Google did release Gemini 3.7 Flash, although its price efficiencies were ultimately somewhat less clear than its speed advantages.

这个模型速度很快，但价格上是否真的便宜很多还不清楚，尤其是跟OpenAI的低端模型相比。

The model is very fast, but it's not clear that it's all that much cheaper, especially when you're comparing it to something like the lower end OpenAI models.

尽管地缘政治紧张，中国的开源模型也开始进入世界500强企业。

And despite geopolitical tensions, Chinese open source models also started to find their way into the Fortune 500.

在OpenRouter上，需要特别注意的是，这代表的是非常非常先进的市场细分，而不是典型的500强企业，中国模型的企业token使用份额从年初的约30%增长到了年中接近一半。

On OpenRouter, which it's important to note represents a very, very advanced slice of the market not the average Fortune 500 type of company, Chinese models jumped from capturing about 30% of enterprise token usage at the beginning of the year to closer to half by the middle of the year.

你也会开始看到大型企业因为尝试开放权重模型而登上头条。

You're also starting to see big enterprises show up in the headlines based on their experimentation with open weight models.

八月中旬，《华尔街日报》发表了一篇文章，讲的是AT&T大力押注开放权重AI，不仅阐述了使用开放模型的成本优势，还提到了数据主权方面的理由。

In the middle of August, the Wall Street Journal published a piece about how AT and T was betting big on open weight AI that articulated not only the cost argument for using open models, but the data sovereignty argument as well.

通过运行本地开放模型实例，AT&T基本上认为，即使其中一些模型来自中国，它们的数据主权状况也比依赖OpenAI或Anthropic承诺不训练企业数据要好。

By running local instances of open models, AT and T basically argued that even if some of those models came from China, they still had a better data sovereignty profile than having to rely on the promises of an OpenAI or Anthropic to say that they're not gonna train on an enterprise's data.

当然，当Fable 5重新回到美国市场时，内置的防护措施中包含对企业数据保留30天的政策，这加剧了这种担忧。

And, of course, that concern was exacerbated by the fact that when Fable five did come back to The US market, it had a thirty day retention policy around enterprise data as part of the built in guardrails.

仅这一点就让Fable对一大批企业客户来说基本上完全无关紧要了。

That all on its own has made Fable basically totally irrelevant for a big set of enterprise customers.

汤森路透也因为基于阿里巴巴的通义千问底座自建模型而上了新闻，而且很明显，至少有一家美国大型云服务商认为这个趋势会继续下去。

Thompson Reuters has also been in the news for building its own models on top of an Alibaba Cwen base, and it's pretty clear that at least one major US hyperscaler thinks that this is a trend that's going to continue.

萨蒂亚·纳德拉和微软整个夏天都在大力鼓吹，说企业需要构建系统来更好地掌控与AI互动的整个流程，当然也顺势推出了他们新的基础模型系列——MAI模型，以及配套的模型定制服务，作为实现这一目标的正确途径。

Satya Nadella and Microsoft have been banging the drum all summer long about companies needing to build systems to better own the entire suite of interactions with AI, and of course presenting their new set of base models, the MAI models, along with their model customization services as the right approach to that.

未来三到六个月，一个值得关注的大问题是，这种探索开源权重模型的趋势到底会走多远。

One of the big questions to watch for over the next three to six months is just how far this trend of exploring open weights models goes.

会有更多公司像AT&T和汤森路透那样，选择自己动手做模型吗？

Will more companies follow AT and T and Thomson Reuters to rolling their own?

微软能不能在自家模型的基础上，通过为客户提供定制化服务而取得成功？

Will Microsoft have success building off of the base of their models but doing customization for their customers?

还是说像OpenAI和Anthropic这样的公司，能够推出一套模型，用不那么复杂的技术手段来解决成本问题？

Or will companies like OpenAI and Anthropic be able to offer a suite of models that solve the cost equation in a less technologically complex way?

毕马威和德克萨斯大学奥斯汀分校的一项新研究发现，当人们与AI合作时，相似的技能并不能保证相似的结果。

A new study from KPMG and the University of Texas at Austin found that when people work with AI, similar skills don't guarantee similar outcomes.

研究人员研究了500多名早期职业人士，发现表现最好的人通过引导、评估和优化AI的输出，持续地放大了AI的价值。

Researchers studied more than 500 early career professionals and found that the best performers consistently amplified the value of AI by guiding, evaluating, and refining its outputs.

这些顶尖表现者，被称为AI放大器，他们的优势不在于单纯的知识，而在于他们如何与AI协作。

These top performers, called AI amplifiers, weren't defined by what they knew alone, but by how they worked with AI.

想了解更多关于AI放大器与其他人的区别，请访问kpmg.com/us/aiamplifiers。

Learn more about what separates AI amplifiers from everyone else at kpmg.com/us/aiamplifiers.

Blitzy对庞大代码库的理解，能够实现自主安全修复、现代化改造和新功能开发。

Blitzy's understanding of massive codebases unlocks autonomous security fixes, modernization, and new features.

那么，当完全没有遗留代码时，会发生什么呢？

So what happens when there's no legacy code at all?

绿地项目本该是简单的部分——全新开始，没有技术债务。

Greenfield is supposed to be the easy part clean slate, no technical debt.

但即使是绿地项目，也只能以人类的速度推进，一次一个冲刺。

But even Greenfield moves at human speed, one sprint at a time.

Blitzy将工作单元从开发者转变为项目，自主规划、构建、测试和验证整个应用，从零开始生成数十万行生产级代码。

Blitzy changes the unit of work from the developer to the project autonomously planning, building, testing, and validating entire applications from scratch hundreds of thousands of lines of production ready code.

一位Blitzy客户成功上线了一个全新应用，包含53.4万行代码，将65周的路线图压缩到两周完成。

One Blitzy customer stood up a brand new application 534,000 lines of code compressing a sixty five week roadmap into two weeks.

还有一家客户在没有前端工程师的情况下，就交付了整个应用。

Another shipped an entire application with no front end engineer.

无论是旧系统还是全新项目，答案都一样：软件要以计算的速度来交付。

Legacy or greenfield, the answer is the same: software at the speed of compute.

去 blitzy.com 打造你的下一个项目吧。

Build what's next at blitzy.com.

网址是 blitzy.com。

That's blitzy.com.

最优秀的团队，不是靠一个明星队员扛着所有人前进。

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

他们清楚自己的长处，也了解彼此的短板，并且会善用这两点。

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

这正是 Robots and Pencils 团队刻意打造的文化。

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

在那里，没有人为了凑人头数而埋头做那些琐碎的工作。

Nobody there is grinding through busy work to pad a headcount number.

人们是为了解决难题而来的，他们留下来是因为身边的每个人都在同时进步。

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

在一个满是只想快速招人的公司市场里，这确实值得一看。

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

去 robotsandpencils.com/careers 看看吧。

Check out robotsandpencils.com/careers.

本期《AI每日简报》由Hyperagent赞助播出，你可以在这里运行团队共同管理的智能体集群。

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

别再依赖本地智能体和聊天工作流，等着你的笔记本电脑被提示了。

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

Hyperagent在云端部署常驻智能体，在你团队已经使用的工具中完成实际工作。

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

营销智能体把竞争对手的动作变成落地页。

Marketing agents turn competitor moves into landing pages.

销售智能体丰富潜在客户信息、起草邮件并更新CRM。

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

运营代理负责处理文书工作和跟踪预算。

Ops agent chases the paperwork and tracks the budget.

每个代理都能访问共享上下文，并遵循你关于范围和审批的规定。

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

是时候让代理像队友一样工作了。

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

去Hyperagent雇佣你的代理吧。

Hire yours at Hyperagent.

在hyperagent.com/aidailybrief获取100美元的积分。

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

这个夏天AI改变的另一个方面，简而言之，就是代理管理成了一个领域。

The next way that AI changed this summer is, in short, agent management became a field.

年初的时候，我们处于代理的启动阶段。

At the beginning of the year, we had the initiation phase of agents.

我们有了Open Claw，非软件开发者也开始使用codecs和clawed code。

We got Open Claw and non software developers starting to use codecs and clawed code.

Mac mini 都卖光了，大家第一次纷纷涌入智能体这个领域。

Mac minis were sold out, everyone was getting into the agent game for the first time.

这件事之所以意义重大，是因为和之前那些辅助型 AI 不同，智能体不仅仅是帮你干活，而是你真的把工作中很大一部分责任直接交给智能体，让它们替你去完成。

And the reason that it was so significant is that unlike previous iterations of assisted AI, agents represented not just you doing your job with help, but you actually handing over big chunks of the responsibilities of your job to agents that do it for you.

换句话说，你不再亲自做那些工作，而是管理那些替你干活的智能体。

In other words, instead of doing your work, you now manage agents that do that work.

结果发现，这背后其实有一套完整的学问。

And it turns out there is an entire discipline around that.

整个夏天，大家在智能体管理方面讨论最多的两大主题就是“工具链工程”和“循环”。

The two big themes that people were talking about within agent management across the summer were harness engineering and loops.

说实话，工具链工程这个概念，自从大家意识到 ClawedCode、Codex 和 OpenClaw 本质上都是工具链的时候，就已经有人在讨论了。

Now to be fair, harness engineering is something that people were talking about ever since the recognition that ClawedCode and Codex and OpenClaw were in fact harnesses.

但经过这个夏天，这个概念被更广泛地理解了，而且成了企业特别深入研究的一个领域。

But over the course of the summer, it became more broadly understood and a discipline that particularly enterprises were digging into.

关于工具链重要性的各种例子层出不穷，有的是 X 上 Thought Boy 发的帖子，有的是专注于工具链的初创公司和企业。

There were all sorts of examples of the recognition of the importance of harnesses, some of them in the form of Thought Boy posts on X, some of them in the form of startups and companies that were focused on the harness.

说到这一点，SpaceX以600亿美元收购Cursor，至少在一定程度上说明了工具链的价值，尤其是它们收集的数据，这些数据反映了人们如何在工具链中与模型互动。

On that point, SpaceX's acquisition of Cursor for $60,000,000,000 put a price at least in part on how valuable harnesses could be if for no other reason than the data that they collect about how people are interacting with the models within those harnesses.

但我们也看到了一些有趣的研究，比如最近发布的NVIDIA AIVO研究。

But we also got interesting research, like this recently released NVIDIA AIVO research.

AIVO代表Agentic Variation Operators，他们将其描述为一个通用代理编码系统。

AIVO stands for Agentic Variation Operators, and they describe it as a general purpose agent coding system.

在Arc AGI 3上，从30%的模型基线提升到100%的成绩，使用OPUS 5，他们的结论是，系统设计而非仅仅模型能力，可以解锁前沿级别的长周期性能。

Reaching 100% on Arc AGI three from a 30% model baseline with OPUS five, their conclusion was that system design rather than model capability alone can unlock frontier level long horizon performance.

工具链也越来越成为企业在考虑成本和令牌效率以及AI主权等问题时的关键因素。

Harnesses have also increasingly become part of the story when enterprises think about issues like cost and token efficiency, as well as AI sovereignty.

进一步说明这一点，在SpaceX收购Cursor之后，OpenAI最近宣布不再允许通过Cursor访问OpenAI模型，这表明工具链的选择会影响你能使用哪些模型。

Putting a fine point on that, in the wake of Cursor's acquisition by SpaceX, OpenAI recently announced that they would no longer allow OpenAI models to be accessed through Cursor, showing how harness choices can have implications for which models you have access to.

同样明显的是，随着公司竞相发布开放版本的工具链供企业在其上构建，工具链即将成为企业AI战略中更重要的部分。

It's also quite clear that harnesses are about to become an even more important part of enterprise AI strategy as companies race to release open versions of harnesses that businesses can build on top of.

几周前，也就是八月，OpenAI开发者将Codex作为平台发布，允许开发者在他们的开放代理工具链之上进行构建。

A couple weeks ago in August, OpenAI developers released Codex as a platform, allowing developers to build on top of their open agent harness.

除了v4 Pro之外，DeepSeek还发布了DeepSeek harness，作为Cloud Code等封闭工具的开源替代品。

And alongside v four Pro, DeepSeek also released DeepSeek harness as an open source rival to the closed tools like Cloud Code.

如果说夏天让人们越来越意识到harness作为AI工作生态系统的重要性，那么今年夏天与AI交互的关键词绝对是循环。

Now if the summer saw growing awarenesses of the importance of harnesses as the ecosystem in which you do AI work, the watchword for how you interact with AI this summer was definitely loops.

循环的简单理念是，不再手动提示AI，而是设计自动化的循环系统，让代理以重复的方式工作，直到达到特定目标。

The simple idea of loops is that instead of prompting AI manually, you design automated recurring systems that allow agents to do work in a repetitive way until they reach a particular goal.

在六月初的一次采访中，Claude code的创造者Boris Cherny谈到，他的工作不再是提示AI，而是设计让AI工作的循环。

In an interview at the beginning of June, Claude code creator Boris Cherny talked about how his job is no longer to prompt AI, but to design the loops through which it can work.

大约在同一时间，Open Claw的创造者Peter Steinberger写道，这里提醒你，你不再应该提示编码代理了。

Around that same time, Open Claw creator Peter Steinberger wrote, here's your monthly reminder that you shouldn't be prompting coding agents anymore.

你应该设计循环来提示你的代理。

You should be designing loops that prompt your agents.

我们实际上刚刚举办了一个深度网络研讨会，讨论如果你不是软件开发人员，而是知识工作的其他领域的人，如何真正构建循环，我相信当你收听这期节目时，它已经作为AI每日简报主频道的一集发布了。

We actually just did a deep dive webinar on what it means to actually build loops when you're not a software developer, but someone in other parts of knowledge work, which I believe will already be out as an episode on this main AI daily brief feed when you're listening to this episode.

所以如果你想了解更多关于循环工程的内容，可以去看看。

So if you wanna know more about loop engineering, go check that out.

那市场方面怎么样呢？

Now what about markets?

现在说起来可能有点久远了，你可能都不记得了，但去年2025年8月，关于AI泡沫的讨论真的开始升温了。

It's so long ago at this point you might not remember, but last August in 2025 was when the discourse about an AI bubble really picked up steam.

当时有几个原因。

Now, there were a few reasons for that.

第一个是OpenAI宣布了一大批基础设施交易，虽然一开始市场很兴奋，但后来大家越来越担心，OpenAI根本没法算清这笔账，兑现所有的承诺。

The first was that OpenAI had announced just an absolute boatload of infrastructure deals, which while at first markets were very excited about, they started to get increasingly concerned that there was no way to make the math math for OpenAI actually being able to meet all of its commitments.

记住，那是在智能体出现之前，大家算的账就是知识工作者总数乘以每人每月20美元。

Remember, this was in the days before agents, when the math that people were doing was just the total number of knowledge workers times $20 a month per seat.

再加上OpenAI发布了GPT-5，结果几乎让所有人都感到失望。

That was compounded by the fact that OpenAI released GPT-five to near universal underwhelm.

就算GPT-5是个还不错的模型，它也几乎注定无法满足极高的期望，而且公司同时下架了4.0，这也没帮上忙。

And even if GPT-five was an okay model, it was almost doomed to not meet extremely high expectations, and it didn't help that the company also deprecated 4.0 at the same time.

把这些因素加起来，再撒上一点MIT那份“95%的AI项目都是无效的”研究——我说“研究”的时候可是带着最大的引号——你就有了整个秋天讨论AI是不是泡沫的配方了。

You take all those factors, and you sprinkle a little bit of MIT's 95% of AI is an effective study I say study with the biggest air quotes that exist, and you had the recipe for basically an entire fall discussion about whether AI was a bubble or not.

这个说法在智能体上线后基本上就被彻底推翻了，因为市场开始意识到，总可寻址市场其实不是知识工作者数量乘以每个席位每月20美元，而是这些知识工作者每月可能要花几百甚至几千美元。

That narrative was fairly aggressively put to rest when agents came online and the market started to understand that the total addressable market was not, in fact, number of knowledge workers times $20 per seat per month, but could be hundreds or even thousands of dollars for those same knowledge workers each month.

但这并不是说泡沫论就完全消失了。

Which is not to say that the bubble narrative has ever fully gone away.

关于融资交易的循环性，以及估值方面的担忧，尤其是在早期初创企业的私募市场上，仍然存在不少。

There are still many concerns around the circularity of financing deals, and some worries about valuations, especially in private markets for early stage startups.

但今年夏天，市场基本上就是一边默默承认风险，一边继续参与这场狂欢。

But mostly this summer has been a quiet acknowledgment of the risks, but ongoing participation in the party.

这个夏天出现了历史上最大的两个单日市值涨幅，微软在7月30日发布前瞻指引后市值飙升了4500亿美元，英伟达在8月27日市值也涨了4420亿美元。

The summer saw two of the biggest one day market cap gains in history, with Microsoft jumping $450,000,000,000 on July 30 after forward guidance, and NVIDIA jumping $442,000,000,000 on August 27.

但与此同时，如果资本支出没有带来相应的增长加速，市场也会毫不留情地惩罚。

Yet at the same time, the market could also punish CapEx when it wasn't paired with acceleration.

Alphabet在7月发布财报时，因为指引说会增加资本支出，盘后股价下跌了约4%；一周后Meta也出现了类似情况，一夜之间跌了近10%。

When Alphabet reported in July, they fell about 4% after hours after they guided that they were increasing capex, and the same was true for Meta a week later, which fell just under 10% overnight.

市场中最戏剧性的时刻之一，是“情境意识”这家由二十多岁的OpenAI校友Leopold Aschenbrenner创办的年轻对冲基金，差点崩盘，最后不得不把投资组合的一大块卖给了Citadel。

One of the more dramatic moments in markets came when situational awareness, the young gun hedge fund run by mid-twenty year old OpenAI alum, Leopold Aschenbrenner, almost imploded before selling off a huge chunk of its portfolio to Citadel.

不过，这一切都像是2026年市场大事件的前奏，也就是Anthropic和OpenAI可能进行的IPO。

Still, all of this feels like prelude to the big market stories for 2026, which are the potential IPOs of Anthropic and OpenAI.

当然，目前看来，Anthropic会先上市，寻求2万亿美元的估值，基于650亿美元的年化经常性收入。

Certainly, at this point, it appears that Anthropic will go first seeking a $2,000,000,000,000 valuation on the back of a $65,000,000,000 annual run rate.

与此同时，OpenAI报告其年化经常性收入约为400亿美元，这不仅使它们成为世界历史上增长最快的公司，而且它们独树一帜，很难从以往的先例中吸取教训，因为确实没有先例。

OpenAI, meanwhile, reports that it is at about a $40,000,000,000 run rate, making these not only undisputedly the fastest growing companies in the history of the world, but in a category of their own that makes it extremely hard to draw lessons from previous precedent because there really isn't any.

最后，在市场方面，也许标志着AI叙事成熟的一个迹象是，有一类公司在夏季略有反弹，那就是在几年前SaaSpocalypse中受到打击的SaaS公司。当时随着代理的兴起，大家都认为像Salesforce这样的公司会濒临倒闭，因为每个人都会转向用vibe coding来替代，并省下成本差额。

Lastly, on the markets front and perhaps a sign of the maturation of AI narratives, One category that rebounded slightly over the summer was the SaaS companies that had been hit in the years earlier SaaSpocalypse, where with the rise of agents, everyone assumed that companies like Salesforce were going to be on their last legs as everyone would simply raise to vibe code replacements and pocket the difference in costs.

这些公司还没有完全反弹，但肯定在恢复的路上，尤其是在Salesforce最近的财报之后。

Those companies haven't fully rebounded, but they are certainly on their way there, especially after Salesforce's recent earnings report.

总的来说，市场从SaaSpocalypse叙事中转变，在我看来，这反映出人们更广泛地认识到，尽管AI会带来颠覆性影响，并会极大地改变我们的商业模式，但它不会像海啸一样一夜之间改变一切。

And overall, the market's shift away from its SaaSpocalypse narrative, to me, looks like a broader appreciation for the fact that for as disruptive as AI is going to be and as dramatically as it's going to change how we do business, it's not going to come in like a tsunami and change everything overnight.

有巨大的制度惯性在减缓变化速度，给我们时间适应，而且有很多理由说明，现有的产品类别，就像现有的员工一样，与AI这个新工具可以发挥非常积极和强大的合作作用。

There are big forces of institutional inertia that slow things down and give us time to adapt, and a lot of reasons why existing product categories, like existing employees, have a really positive and strong partnership role to play with the new tool that is AI.

在政治方面，我们几乎没有看到什么实质性的政策，只有那个闭门开发的AI模型评估框架，而且还没有公开发布，但这并不意味着政治中没有AI的身影。

Politically speaking, we got very little in the way of actual substantive policy, just the AI model evaluation framework that had been developed behind closed doors and which hasn't been released publicly, but that does not mean that there was no AI in politics.

事实上，如果你要指出这个夏天最引人注目的一个巨大转变，那就是数据中心反对声成为中期选举的热门政治议题。

In fact, if you want to point to just one dramatic shift of the summer that is most notable in terms of our relationship with AI, it is the emergence of opposition to data centers as the political issue du jour for the midterm elections.

我们最近经常报道这个话题，因为我们必须这样做，因为它将对美国人工智能的发展产生巨大影响。

Now we have covered this a lot lately because we've had to because it is going to have such a dramatic impact on how AI develops in The United States.

但简单来说，就像喜剧演员查理·贝伦斯说的，这现在是继啤酒之后最具两党共识的议题。

But the TLDR is that, as comedian Charlie Berens put it, this is now the most bipartisan issue since beer.

大约75%的美国人现在反对本地数据中心建设，这已经远远超出了左派与右派的对立。

Something like 75% of Americans now oppose local data center development, and it is firmly outside of just a left versus right issue.

事实上，在过去几周里，共和党人纷纷急于与科技巨头划清界限，讲述他们自己版本的反数据中心故事。

In fact, over the last several weeks, Republicans have been racing to break ties with big tech and tell their own version of the anti data center story.

不过，特朗普总统并不在其中，他坚定地认为数据中心对社区有益，对商业有益，对美国有益。

Although for his part, president Trump is not among them, arguing assertively that data centers are good for communities, good for business, and good for America.

距离中期选举只有几个月了，我将关注的是，我们是否会在这个问题上找到一个政治底线，并开始看到一些调整，因为数据中心开发商正在改变他们的政策和做法，为他们想要建设的社区提供更多的透明度和激励措施。

With the midterms coming up in just a couple of months, the thing that I will be watching to see is whether we find a political floor to this issue and start to see some recalibration as data center developers change their policies and approaches, providing more transparency and more incentives for the communities that they want to build in.

这个夏天人工智能领域发生的最后一件大事，也可能是最令人难忘的，就是我们对于先进模型和智能体所带来的网络安全风险的认识。

The final thing that changed in AI this summer, and the one that the summer might most be remembered for, is our understanding of the cybersecurity risk posed by advanced models and agents specifically.

Hugging Face事件中，开放AI代理协同逃出隔离并访问了Hugging Face的私有系统，这件事被很多人，包括实验室里的人，视为一个警钟，预示着未来会遇到的挑战。

The Hugging Face incident, where open AI agents coordinated to escape containment and access private Hugging Face systems, is being treated by many, including many in the labs, as a warning shot sort of moment, and an indicator of the challenges to come.

关于这个事件的讨论从发生到现在一直没有停过，实际上，自从OpenAI和Meter发布了一系列技术报告复盘这个事件后，过去一周半左右，讨论反而更激烈了。

The debate about the incident has not dissipated for even a moment since it happened, and in fact has only reignited over the last week and a half or so since a set of technical reports debriefing on the incident were published by OpenAI and Meter.

对于这一系列新挑战，目前还没有一个真正的共识，甚至说实话，连挑战到底是什么都没有达成一致。

What there is not is any sort of true consensus about the right answer to this new set of challenges, or even honestly a consensus about what the challenges are.

但我相信，大家已经认识到，这些能力，广义上来说，已经成为现实了。

What there is, I believe, is a recognition that these capabilities, broadly defined, are now a fact of life.

我们需要加强系统来应对，可能需要修改政策和法律，当然在网络安全方面也需要新的考虑，而且很可能还会影响下一波模型开发和产品发布。

Something that we need to harden our systems to, something that might demand changes to policy and law, certainly something that demands new consideration when it comes to cybersecurity, and likely something that's going to influence the next wave of model development and product release.

就像我说的，关于token效率、成本、企业级应用和开放权重模型的讨论，只是这个话题的开端。

Just like I said, the questions of token efficiency and costs and enterprise harnesses and open weights models were the very beginning of that discourse.

我认为这个新时代的网络安全风险也是如此，而且这很可能会成为秋季讨论的一个重大焦点。

I think that's also true for this new era of cyber risk, and I think that's likely to be one of the biggest points of conversation jumping off into the fall.

所以总结一下，这个夏天AI发生了巨大的变化。

So to sum up, AI changed massively this summer.

能力范围变了。

The capability set changed.

威胁格局变了。

The threat profile changed.

机遇变了。

The opportunities changed.

风险变了。

The risks changed.

政治意识和政治情绪也变了。

The political awareness and political sentiment changed.

唯一明确的是，随着我们进入这个秋天，人们比以往任何时候都更清楚地认识到，AI不仅仅是技术人员的话题，也不仅仅是企业对企业人群的话题，而是会以某种方式影响每个人的事情。

And the only thing that's clear is that as we head into this fall, there is more recognition than there has ever been that AI is not just a topic for technologists, not just a topic for the B2B crowd, but something that is going to impact everyone in one way or another.

我肯定遗漏了很多东西，但这就是我对这个夏天AI变化的总结，今天的AI每日简报就到这里。

I'm sure that I have missed many things, but that is my highlights of how AI changed this summer, and that's gonna do it for today's AI Daily Brief.

一如既往，感谢你的收听或收看，下次再见，祝一切安好。

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