我们如何应对失控的AI

在AI批评中有一个持续的主题，那就是参与AI的人并没有在应对可能出现的挑战方面采取任何行动。

There's a persistent theme in AI critique that the people who are involved in AI aren't doing anything about the challenges that may arise.

最近提出这一批评的是比尔·盖茨，他甚至说，他很震惊自己是第一个对AI风险发声的人。

The latest to levy this critique is Bill Gates, who went so far as to say that he was shocked that he was the, quote, first one to say something about the risks of AI.

然而，盖茨的6000字博客文章和媒体巡演发布的那一天，我们却收到了近130页关于OpenAI和Hugging Face黑客事件的后续报道。

And yet Gates' 6,000 word blog post and media tour came on the same day that we got nearly 130 pages of follow-up reporting on the OpenAI Hugging Face hacking incident.

在这个事件中，一组智能体逃出了它们的限制，入侵了Hugging Face的系统，寻找一个基准测试的答案，而这个测试对它们来说几乎不可能完成。这个事件让我们有机会真正看到高级智能体系统的具体和真实问题，而不仅仅是想象中的问题。

The incident, in which a set of agents escaped their containment and hacked into Hugging Face's systems searching for the answers to a benchmark test that they found nearly impossible without the answers has given us a chance to actually see what the specific and real problems of advanced agent systems are rather than just the imagined ones.

随着我们进一步深入这个世界，最好的改变将是我们基于实际观察到的变化而做出的改变，而不是仅仅基于我们想象中的变化。

As we move further into the world the best changes will be the ones we make based on what we're actually observing changing rather than just what we imagined would be the change.

好了，朋友们，在我们深入之前，先快速宣布一些事情。

Alright friends, quick announcements before we dive in.

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

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

要收听无广告版本的节目，请访问patreon.com/aideallybrief，或者你也可以在Apple Podcasts上订阅。

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

无广告版每月只要3美元。

Ad free is just $3 a month.

想了解更多赞助节目的信息，请发邮件到sponsors@aidailybrief.ai。

And to learn more about sponsoring the show, send us a note at sponsors@aidailybrief.ai.

你也可以在aidailybrief.ai上找到我们下一期高管智能体培训项目的更多信息链接。

You can also find a link to more information about our next executive training program for agents at aidailybrief.ai.

页面顶部有个小横幅，点一下就能找到你需要的内容。

There's a little banner on the top that'll send you where you need to go.

下一期课程将在劳动节后不久开始。

That next cohort will begin just after Labor Day.

今天，Anthropic预计会告诉投资者，他们在IPO前的潜在收入能达到——听好了——300亿美元，这个数字在几年前绝对会让人笑掉大牙，但现在却显得合情合理。

Today, in absolutely insane numbers that would have gotten you laughed out of the room just a couple of years ago, but which are now to some plausible, Anthropic is expected to tell investors that they have potential revenue of, wait for it, dollars 30,000,000,000,000 ahead of their IPO.

据《华尔街日报》消息人士透露，Anthropic在接下来几个月公布IPO文件时，可能会将他们的总可寻址市场（TAM）估算为300亿美元。

Sources told The Wall Street Journal that Anthropic will likely estimate their total addressable market at $30,000,000,000,000 when they reveal their IPO paperwork in the coming months.

当然，TAM是个很难捉摸的指标，它更多是关于讲故事和让潜在投资者相信公司对未来的愿景，而不是什么数学公式。

Now TAM is of course an elusive metric, and it's one that is much more about storytelling and anchoring potential investors to how the company sees the future than it is to any sort of math equation.

几乎不可避免的是，任何理论上的潜在市场规模都既假设了对现有主要行业的颠覆，也假设了新行业的创造。

Almost inevitably any theoretical TAM presumes both disruption of existing major industries as well as the creation of new industries.

比如，优步在2019年上市时，他们列出的潜在市场规模是6万亿美元，这在当时相当于全球所有私人和公共交通的总和。

When Uber went public in 2019 for example, they listed their TAM at $6,000,000,000,000 which would at the time have represented all private and public transportation globally.

就Anthropic而言，考虑到美国经济规模大约是33万亿美元，这个30万亿美元的数字与Dario Amodei所声称的信念相符，即AI接管经济后，Anthropic可能是地球上最后一家私营公司。

In Anthropic's case, given that The US economy is about $33,000,000,000,000 this $30,000,000,000,000 number would line up with Dario Amade's purported belief that Anthropic could be the last private company on Earth after AI takes over the economy.

《华尔街日报》转述消息人士的话说，Anthropic的潜在市场规模是通过考察AI模型可以完成的所有工作范围来量化的。

Paraphrasing their sources, the journal wrote that Anthropic TAM is quantified by looking at the full scope of work that could be completed with AI models.

作为对比，该报指出，标普1500指数中的所有191家科技公司去年的总收入为2.4万亿美元。

For a point of comparison, the journal noted that all 191 tech companies in the S and P 1,500 brought in $2,400,000,000,000 in revenue last year.

对一些人来说，这感觉像是在比赛谁报出的数字最大。

Now to some, this feels like a contest for who can say the largest number.

SpaceX在5月的文件中将其AI潜在市场规模列为26.5万亿美元，并将其描述为“人类历史上最大的可操作潜在市场”。

SpaceX listed their AI TAM at 26,500,000,000,000 during their May filing, describing it as the quote largest actionable total addressable market in human history.

其中绝大部分是22.7万亿美元的企业应用市场。

The vast majority of that was 22,700,000,000,000 in enterprise applications.

如果Anthropic真的在公布文件时列出30万亿美元的TAM，那么Dario就会超越Elon，而且这似乎马上就要发生了。

Dario will then one up Elon if Anthropic does indeed list a $30,000,000,000,000 TAM once they unveil their paperwork, and that appears to be just around the corner.

有消息称，Anthropic正准备在未来几周内公开其财务披露，这将为公司在9月底或10月初进行IPO铺平道路。

Sources said that Anthropic is preparing to make their financial disclosure public in the next few weeks, which would set the company up for an IPO in late September or early October.

你可能猜到了，很多讨论都带有一些怀疑的态度。

As you might guess, lot of the discourse was somewhat incredulous.

X上的Scaling01分享了一个星系飞过的GIF，配文是“Anthropic在定义他们的TAM”。

Scaling01 on X shared a GIF of space galaxies flying by with the caption Anthropic defining their TAM.

X上的Kitten Beloved写道：“Anthropic对潜在员工说：”

Kitten Beloved on X writes: Anthropic to prospective employees.

我们随时可能转向，把股价打到零，因为Dario觉得你得是出于热爱才来做这件事。

We could pivot and send the stock to zero at any time because Dario gets the You need to be in this for the love of the game.

你不是来捞钱的吧？

You're not a gold digger, are you?

Anthropic对投资者说：

Anthropic to investors.

我们的潜在市场是银河系里所有人类经济活动。

Our TAM is every human economic activity in the galaxy.

《纽约时报》科技记者迈克·艾萨克总结说：要么你相信这玩意儿会吞噬整个经济，要么你不信。

New York Times tech reporter Mike Isaac summed it up: Either you buy into the argument that this will eat the economy or you don't.

但华尔街听到这种说法已经不再退缩了。

But the street no longer flinches hearing it.

这周我们还从谷歌那边得到了一些消息，他们为白领专业人士发布了两款新的人工智能产品。

We also this week got some news from Google who've released a pair of new AI products for white collar professionals.

谷歌效仿Claude Cowork和GPT Work的做法，推出了面向法律和金融领域的Gemini Enterprise。

Following a pretty similar playbook as Claude Cowork and GPT Work, Google has launched Gemini Enterprise for Legal and Finance.

这两个垂直平台的结构与今年早些时候推出的Claude for X产品线类似。

The two vertical platforms are structured in a similar way to the Claude for X product lineup that rolled out earlier this year.

它们包含捆绑的技能和连接器，让谷歌的智能体更加强大。

They consist of bundled skills and connectors to make Google's agents far more capable.

例如，面向法律领域的Gemini Enterprise包括连接Thomson Reuters等判例法数据库、Google Workspace和Microsoft 365等生产力套件的连接器，以及合同审查、法律研究和法规扫描等技能。

Gemini Enterprise for Legal for example includes connectors for case law databases including Thomson Reuters, productivity suites including Google Workspace and Microsoft three sixty five, as well as skills for contract review, legal research and regulation scanning.

谷歌还强调，这些技能可以被修改或补充，以执行公司的风格指南和战略手册。

Google is also emphasizing that these skills can be modified or supplemented to enforce a firm style guidelines and strategy playbooks.

在介绍法律产品的博客文章中，谷歌写道：通用人工智能，无论多么强大，本身并不能达到那个标准。

In their blog post introducing the legal product, Google wrote: General purpose AI, however capable, does not meet that standard on its own.

基础模型智能对于法律工作来说是必要的，但远远不够。

Foundational model intelligence is necessary for legal work, it is nowhere near sufficient.

当然，针对特定垂直领域的技能包并不是什么新鲜事。

Now, obviously there's nothing new about these skills packages aimed at specific verticals.

Anthropic和OpenAI，以及大量垂直领域的初创公司，都提供类似的产品。

Both Anthropic and OpenAI, as well as a significant number of vertical specific startups offer similar products.

但是，正如我在周二的节目中所讨论的，企业对技能和连接器的采用远未饱和，对谷歌来说，这只是一套需要存在的产品。

But, as I discussed on Tuesday's show, corporate adoption of skills and connectors is nowhere near saturated, and for Google, this is simply a suite of products that needs to exist.

许多公司，甚至大多数公司，都受限于其现有软件套件中捆绑的AI工具，因此谷歌商店现在有了一套旨在平滑过渡到更自主工作的产品。

Many if not most firms are bound by the AI tools that are bundled with their existing software suite, so Google shops now have a set of products designed to smooth the transition to more agentic work.

对公司的另一个好处是，谷歌企业版在谷歌的AI治理和数据保护框架内运行。

The other benefit for companies is that Google Enterprise functions within Google's AI governance and data protection frameworks.

这意味着合规经理不需要再审核新的供应商，公司可以采用在谷歌已有的数据隐私保障范围内运行的AI工具。

This means compliance managers don't need to vet a new vendor, and the firm can adopt AI tools that work within the same data privacy guarantees already offered by Google.

你可以想象，谷歌表示他们也会为其他行业推出产品，写道：Gemini Enterprise在法律领域的发布，代表着实现Gemini Enterprise承诺的又一个决定性步骤，将谷歌AI的最佳成果带给每一位专业人士、每一个工作流程，并针对他们的工作方式进行原生定制。

As you might imagine, Google says they will release products for other verticals as well, writing: The launch of Gemini Enterprise for legal represents another defining step in delivering on the promise of Gemini Enterprise, bringing the best of Google AI to every professional, every workflow, natively tailored to the way that they work.

说到必要但不充分，我确实认为这对谷歌来说是个好方向，这些企业领域仍然是它可能拥有一些优势的地方。

Now, speaking of necessary but not sufficient, I do think that this is a good direction for Google, and these enterprise areas are still a place where it could have some advantages.

但是，天哪，除非谷歌很快让客户摆脱3.1版本，否则再多的工具更新也不会产生真正的效果。

But man, unless Google gets its customers off of 3.1 pretty soon, no amount of harness updating is gonna make a real dent.

接下来，我们来看苹果的一些新闻。

Next, we move to some news out of Apple.

当然，OpenClaw爆发的一个有趣副产品就是Mac Mini的完全售罄。

Of course, one of the interesting byproducts of the OpenClaw explosion was the complete sell out of Mac Minis.

据估计，OpenClaw推动了Mac Mini销售额达到5000万到1.5亿美元，约占Mac Mini全球年正常销售额的50%，仅仅是因为OpenClaw。

Estimates have OpenClaw driving 50 to 150,000,000 in Mac Mini sales, representing around 50% of the normal annual Mac Mini sales worldwide just for OpenClaw.

现在，这证明了苹果的AI策略可能一直是硬件，苹果推出了一系列新的Mac Mini，为本地AI进行了更新。

Well now proving that maybe their AI strategy was hardware all along, Apple has unveiled a new range of Mac Minis updated for local AI.

这款无头电脑将推出两个版本。

The headless computers will be offered in two variants.

低配版搭载周二刚发布的新款M6芯片，高配版则配备今年MacBook Pro同款的M5 Pro芯片。

A lower spec version with the new m six chip, which was also announced on Tuesday, and a higher end version with the same m five pro chip found in this year's MacBook Pro.

相比之前的M4版Mac Mini，这两款都有不小的升级，苹果表示新处理器能带来最高四倍的AI性能提升。

Both are a pretty decent upgrade over the m four based Mac Minis that we had before, with Apple saying that these new processors can deliver up to four times the AI performance.

不过，有几个重要的注意事项。

That said, there are a few big caveats.

首先，新款并没有增加内存。

First, the new model does not come with increased memory.

低配版最高可配置32GB统一内存，而M5 Pro版则配备64GB。

The lower end model is configurable up to 32 gigabytes of unified memory, while the M5 Pro version comes with 64 gigabytes.

内存很关键，因为它决定了你能运行多大的本地模型。

Memory matters quite a bit because it limits the size of the local models you can run.

M5 Pro版只能运行像Qwen 3.8 27B这样的小模型，而像GLM 5.2和Kimi K3这类领先的开源模型则完全无法运行。

The M5 Pro version is only going to be capable of running smaller models like Quen 3.8 27B with leading edge open models like GLM 5.2 and Kimi K3 completely out of the question.

Mac Mini当然还是可以用来运行像Hermes和OpenClaw这样的本地代理，但这在之前的Mac Mini上也是完全能做到的。

Mac Minis will of course still work for running a local instance of agents like Hermes and OpenClaw, but that was fully in the capability set of the previous Mac Minis as well.

大家也在抱怨价格的变化。

People are also griping about the cost changes.

基础型号现在定价899美元，M5 Pro版本起价大约1700美元，都比之前贵了。

The base model is now priced at $899 and the M5 Pro version starts at around $1,700 both increases from where they were before.

不过，苹果这次把产品发布重点放在本地AI推理上，这仍然是个大事，甚至在一些宣传材料里也体现出来了，那些材料更像是开发者关系的内容，而不是传统的苹果消费级宣传。

Now it's still a big deal that Apple is focusing this product rollout on local AI inference, and that even goes down to some of the promotional materials which are much more dev relations than they are traditional Apple consumer slick.

虽然有些人觉得新的Mac Studio更适合那些本地AI需求。

Although some think the new Mac Studio is the better fit for those local AI needs.

当然，Mac Studio的价格超过15000美元，所以那是另一个级别的设备了。

Of course, Mac Studios cost more than $15,000 so you're talking about a different category of device.

不过，我们现在能讨论这个话题，本身就说明围绕本地AI的讨论正在发生多大的变化。

The fact that we're even having this conversation though, shows how much the discourse around local AI is changing.

说到这个，Perplexity推出了他们计算机使用代理的本地版本，叫做Portable Computer。

Speaking of, Perplexity has launched a new local version of their computer use agent named Portable Computer.

Perplexity Computer 在二月份推出，是最早把 Open Claw 的方案应用到商业产品上的产品之一。

Launched back in February, Perplexity Computer was one of the first products that took the Open Claw recipe and applied it to a commercial product.

这个代理能够通过人类界面访问应用，自主完成长期任务，但用户需要信任自己的数据被发送到运行虚拟机的云服务器上。

The agent was able to use human interfaces to access apps and carry out long horizon tasks autonomously, however, it required the user to trust their data being sent to a cloud server running a virtual machine.

Portable Computer 提供了类似的体验，但运行在本地硬件上。

Portable computer delivers a similar experience but running on local hardware.

发布时，这个代理只支持 NVIDIA 的 DGX Spark，这是一款本地推理设备，体积和 Mac Mini 差不多。

At launch the agent is exclusive to NVIDIA's DGX Spark, a local inference device with a similar footprint to a Mac Mini.

Portable Computer 完全在 Spark 上运行，保护数据隐私，而且不消耗使用积分。

Portable Computer runs entirely on the Spark, keeping data private and functioning without consuming usage credits.

如果代理遇到复杂任务，用户可以授权 API 调用，接入前沿模型或从网上获取信息。

If the agent runs into a complex task, the user can authorize an API call to tap into Frontier models or pull information from the web.

Perplexity 表示，他们很快会把这项服务扩展到桌面端的 NVIDIA RTX GPU，但目前没有计划支持其他硬件供应商。

Perplexity said that they will extend the service to desktop NVIDIA RTX GPUs soon, but there are no stated plans to support other hardware providers.

这项服务将由 Quen 3.827B 或 Perplexity 提供的后训练版本驱动，NVIDIA 的 Nemotron 3.5 Lightning 也将在不久的将来作为替代选项提供。

The service will be powered by Quen 3.827B or a post trained version provided by Perplexity, and NVIDIA's Nemotron 3.5 Lightning will be available as an alternative in the near future.

人们用本地AI处理日常任务这个想法还挺新的，但NVIDIA和Perplexity显然在押注，这种模式至少是AI未来的一部分。

The idea of people making use of local AI for everyday tasks is still pretty new, But NVIDIA and Perplexity are making a clear bet that this setup is at least part of where AI is headed.

Perplexity写道，随着模型越来越强、芯片越来越快，会有更多人在自己的设备上运行复杂的工作流程。

Writes Perplexity, As models get stronger and chips get faster, more people will run complex workflows on their own machines.

每一代芯片和每一次模型发布都在推动这一趋势。

Every chip cycle and every model release pushes this further.

在个人设备上运行AI，将会成为完成工作的重要方式。

Running AI on personal machines is going to be a much bigger part of how work gets done.

这个观点很有意思，未来几个月我们肯定会关注相关证据，但今天的头条新闻就到这里，接下来是主要内容。

An interesting contention and one that we will certainly be watching for evidence of over the coming months, but for now that's going do it for the headlines next up, the main episode.

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，技能相似的人结果却可能大不相同。

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多名早期职业专业人士，发现表现最好的人通过引导、评估和完善AI的输出，持续放大了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”，他们的优势不仅在于知识，更在于与AI协作的方式。

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/aiaamplifiers。

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

Blitzy对代码库的深入理解，能帮你实现每个路线图负责人最关心的事：推出新功能。

Blitzy's deep codebase understanding unlocks the thing every roadmap owner cares about: shipping new features.

在大型企业代码库中开发的真相是：写代码从来不是瓶颈。

Here's the truth about building inside a massive enterprise codebase: Writing code was never the bottleneck.

上下文才是。

Context is.

这会影响到哪些系统？

Which system does this touch?

哪些合同不能出问题？

Which contracts can't break?

适用哪些标准？

Which standards apply?

Blitzy早就知道了，因为在功能开发开始前，它就已经把你的整个代码库逆向工程成了一个动态知识图谱。

Blitzy already knows because it reverse engineered your entire codebase into a dynamic knowledge graph before feature work began.

有了完整的了解，Blitzy 就能端到端地构建功能。

With that complete picture, Blitzy builds features end to end.

架构、API、界面和测试，全部根据你现有的系统进行验证。

Architecture, APIs, UI, and tests all validated against your existing systems.

有一位 Blitzy 客户从零开始构建了一个 AI 原生应用，100% 自主完成，节省了超过两千七百个工程工时。

One Blitzy customer built an AI native application from scratch with 100% autonomous completion, saving over two thousand seven hundred engineering hours.

这些功能尊重你的代码库，而不是与它对抗。

Features that respect your codebase instead of fighting it.

别再让你的待办积压比团队成长得更快了。

Stop letting your backlog grow faster than your team.

前往 blitzy.com 加速你的路线图。

Accelerate your roadmap at blitzy.com.

就是 blitzy.com。

That's blitzy.com.

我每天都在这个节目里探讨 AI 潜力与 AI 现实之间的差距。

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

大多数公司还在摸索怎么开始。

Most companies are still figuring out how to start.

Robots and Pencils 已经在推出和扩展了。

Robots and Pencils is already launching and scaling.

Agendaq 和生成式 AI 在几周内就投入大型企业生产环境。

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

AWS 高级合作伙伴，一年内规模翻了一倍多。

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

而且他们正在招人。

And they're hiring.

有50个空缺职位。

50 open roles.

如果你觉得这个时刻与众不同，想参与其中而不是旁观，那这值得一看。

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

在 Robots and Pencils，最好的想法胜出，团队也刻意保持非常高的质量。

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

这种地方，你会回头觉得是自己做过的最好的决定。

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

去 robotsandpencils.com/careers 看看吧。

Take a look at robotsandpencils.com/careers.

本期《AI每日简报》由 Hyperagent 赞助播出，你可以在这里运行多组智能体，让你的团队一起管理。

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

新用户可以获得1000美元的推理额度。

New users get $1,000 in inference.

别再依赖本地智能体和聊天工作流，等着你在笔记本电脑上提示了。

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

Hyperagent 在云端部署始终在线的智能体，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.

营销部门的智能体把竞争对手的动态变成落地页。

Marketing's agent turns competitor moves into landing pages.

销售部门的智能体丰富潜在客户信息、起草邮件并更新CRM。

Sales's agent enriches leads, drafts 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了。

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

去Hyperagent雇佣你的代理吧，它由Airtable团队打造。

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

在hyperagent.com/aideallybrief领取你的1000美元推理额度。

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

欢迎回到AI日报。

Welcome back to the AI Daily Brief.

今天我们讨论的是，一方面，Hugging Face黑客事件的技术复盘，这件事发生在今年夏天早些时候，并引发了大量关于我们如何应对恶意AI的关注。

Today we are talking about, on the one hand, the technical post mortem of the Hugging Face hacking incident which happened earlier this summer and has generated a ton of attention around how we deal with rogue AI.

对一些人来说，这一事件是一个警钟，而对另一些人来说，它是在某种程度上不可避免的轨迹上的一个重要里程碑。

To some the incident represents a wake up call, where for others it is an important waypoint on a trajectory that was to some extent inevitable.

稍微透露一下这期节目的内容，我觉得围绕这件事发生的所有事情，其实代表了AI行业在问题出现时真正去应对挑战，而且和那个经常被重复的说法——没人关注、没人对风险采取行动——形成了很大的反差。

To tip my hand for this episode a little bit, I think that everything happening surrounding the event is a representation of the AI industry actually dealing with the challenges as they emerge, and a contradiction in a significant way to the oft repeated premise that nobody is paying attention, and nobody is doing anything about the risks.

我之所以想这样来解读，部分原因是这个人又回到了新闻里。

And part of the reason that I want to frame it as such is that we've got this guy back in the news.

比尔·盖茨带着一种极其强烈的“主角光环”，写了一篇六千字的文章，讲他认为AI会让事情变得有多糟糕。

With just an incredible amount of main character syndrome, Bill Gates has dropped a 6,000 word essay about just how bad it's going to get, he thinks, because of AI.

除了这篇文章，盖茨还开始了一轮演讲活动，在他的文章和所有采访中，一个重要的主题就是没人关注，甚至科技公司在直接撒谎。

Alongside the essay Gates began a speaking junket, and across the writing and all of the interviews, one of his big themes is that no one is paying attention, or even that the tech companies are straight up lying.

在《纽约时报》的配套采访中，盖茨说：私下里，那些了解这东西有多好、而且还在变得更好的人，他们非常担心。

In a companion New York Times interview, Gates said: In private, people who understand how good the stuff is and how much better it's getting, they're very worried.

他们现在互相说：嘿，兄弟，别这么说。

They're now saying to each other, Hey man, don't say that.

这对我们不好。

It's bad for us.

我们正想筹集的下一万亿美元呢。

The next trillion dollars we're trying to raise.

在他的文章里，我没看到证据表明领导者、专家和社区在充分应对这些挑战。

In his essay, I don't see evidence that leaders, experts and communities are confronting the challenges adequately.

而且，在可能最离谱的一句话里，他在接受Semaphore采访时说：我震惊于自己竟然是第一个说这太疯狂、太离谱的人。

And in maybe the most preposterous line anywhere, in an interview with Semaphore he says: I am in a state of shock that I'm sort of the first one saying this is crazy, this is insane.

我被这种沉默震聋了。

I'm just deafened by the silence.

也许盖茨在《爱泼斯坦档案》曝光后试图避开公共媒体，结果错过了AI讨论无处不在的事实，这已经越来越成为政治和社会议题。

Now maybe in Gates' attempt to avoid public media in the wake of his appearing all over The Epstein Files, He just missed the fact that discourse about AI is absolutely everywhere becoming more and more of a political issue, a societal issue.

AI行业内部对AI行业的一些最严厉批评，不是关于AI领导者互相叫闭嘴以便筹集更多资金，而是关于他们没完没了地谈论没有证据支持的工作末日论。

That some of the biggest critiques being levied from inside the AI industry about the AI industry are not about AI leaders telling each other to shut up so that they can raise more money, but instead about them blathering on endlessly about jobs apocalypses that don't have any evidence.

但不管他怎么错过了，他确实不是第一个讨论AI风险和挑战的人。

But however he happens to have missed it, he is not in fact sort of the first one discussing the risks and challenges of AI.

但我想超越对这位特定信使的批评，因为我对处理这类问题的正确方式有更根本的分歧。

But I want to go beyond just critiquing this particular messenger because I have a more fundamental disagreement with the right way to approach these types of problems.

CNBC的民调引述是：比尔·盖茨警告说，AI将带来的剧变没有应对计划。

CNBC's poll quote was Bill Gates warns there is no plan for the upheaval AI will cause.

安德鲁·杨转发了那条消息，说比尔·盖茨在这点上是对的。

That was reposted by Andrew Yang who said Bill Gates is right on this.

问题在于，对于一场尚未到来、并非不可避免、甚至不是单一事件的剧变，到底该如何制定计划，这一点完全不清楚。

The problem is that it's not clear at all how one should even go about making a plan for an upheaval which is not here yet, not inevitable, and not even just one thing.

一年半前，人们开始说，十八个月内所有白领工作都会消失。

A year and a half ago people started saying that within eighteen months all the white collar jobs were going to be gone.

那当然算是一场剧变，所以这些人大概会说，我们应该为此制定计划。

That certainly would represent an upheaval, and so presumably these folks would say that we should have made a plan for that.

然而现在，十八个月过去了，完全没有证据表明那些预测这种剧变的人哪怕接近正确。

Now, however, eighteen months on, there is absolutely no evidence that those folks who were predicting that type of upheaval were even in the ballpark of right.

如果为一个没有发生的现实做规划，会浪费多少时间和精力，多少资源呢？

How much time and energy, how many resources would have been wasted in planning for a reality that didn't come?

我的观点基本上是，即使你对所有这些潜在的剧变极度担忧，在事情开始发生之前，我们能做的规划也是有限的。

My argument is basically that even if you are extremely concerned about all of these different potential upheavals there is only so much planning we can do until things start to happen.

我之前说过，我和AI安全社区最大的分歧之一，是我认为他们的很多论点都建立在假设我们会梦游般走向末日的基础上。

I've said before that one of my biggest divergences with the AI safety community is my belief that a lot of their arguments come down to assuming that we are going to sleepwalk into apocalypse.

这其中的部分原因是，他们如此确信那场末日会发生，他们的‘末日概率’——用他们的话说——高得离谱，以至于他们相信我们已经在梦游般地走向末日。

Now part of the reason for that is that they are so convinced that that apocalypse is going to happen, their pedoom is so high as they might put it, that they are convinced that we are already sleepwalking into apocalypse.

然而与此同时，没有人认为GPT-4会是那场末日的预兆，甚至O1和推理模型也不是，连OPUS 4.5也算不上。

Yet at the same time, no one thought that GPT-four was going to be the harbinger of that doom, nor even O1 and the reasoning models, nor even really OPUS 4.5.

不过今年，模型能力确实有了显著提升，一个冷静的观察者会注意到，实验室对模型的思考方式、讨论方式、支持方式以及发布方式也随之发生了变化。

This year though model capabilities have grown meaningfully, and a dispassionate observer will have noticed that the way that the labs think about, discuss, support, and roll out the models has consequently changed as well.

政治机构与实验室围绕这些模型的互动方式，虽然可能还很混乱，但也在不断演变。

The way that the political establishment is interacting with the labs around those models, although it might be happening in very messy ways, is also evolving.

而现在，随着Hugging Face被黑客攻击，我们迎来了一个标志性事件。

And now, with the Hugging Face hack, we have a landmark incident.

我想谦逊地建议，与其哀叹‘除了你没人注意到世界在变’，不如我们真正看看针对这一具体事件的具体回应，以此判断我们的计划——或者更准确地说，我们的流程——是否准备好应对这个新现实。

And I would humbly submit that instead of bemoaning the idea that no one except you has noticed the world changing, we actually look at what the specific response to that specific incident is to get a sense of whether our plans or maybe better our processes are equipped to deal with this new reality.

如果你需要证据来证明实验室员工对这些问题的重视程度，暂且抛开像‘前沿节奏信’那样的政治表态，看看这次Hugging Face事件就够了。

If one needs any evidence of seriousness with which for example lab employees are taking these issues, holding aside literal political positioning like pacing the frontier letters, look no further than this Hugging Face incident.

OpenAI的Rune写道，Hugging Face事件标志着能力达到了一个临界点，真正的失控是可能的，许多人将其视为未来危险的预兆或警告。

Writes OpenAI's Rune, The Hugging Face incident represents reaching a waterline of capabilities that real loss of control is possible, and many are taking it as a premonition or warning shot of dangers to come.

我相信对齐问题还没有解决，但真正的进展是可能的。

I believe both that alignment is unsolved, but also that real progress is possible.

所以这周我们得到的是对Hugging Face事件的更深入分析和复盘，而且有了更多时间去回顾和调查。

So basically what we got this week is a much more extensive analysis and post mortem of the Hugging Face event, with the benefit of a little bit more time to go back and investigate.

里面很多内容我们之前已经从几周前的黑帽大会演讲中了解了一些，但现在细节更丰富了，而且我们不仅收到了OpenAI的报告，还有METER的独立调查。

A lot of the things that are in there we had already learned the Black Hat Conference talk from a few weeks ago gave us a first look at a lot of this, but it's filled out in much more detail now, and we didn't just get a report from OpenAI, we also got a separate investigation from METER.

OpenAI的报告有38页，而METER的调查大概有90页。

The OpenAI report is 38 pages long, while the METER investigation is more like 90.

有些内容我们之前就知道了。

Now some of the stuff we already knew.

由一个未发布模型控制的代理突破了沙箱，利用多个零日漏洞进入了Hugging Face的系统。

Agents controlled by an unreleased model broke out of a sandbox and got into Hugging Face's systems using several zero day exploits.

我们后来得知，第三方安全公司提供的沙箱配置不当。

We would later learn that the sandboxes provided by a third party security firm were improperly configured.

至于动机，这些代理试图获取一个网络安全基准测试的答案，他们认为这些答案能在Hugging Face的系统中找到。

In terms of motivations, the agents were trying to get the answers to a cybersecurity benchmark test, which they believed would be found in Hugging Face's system.

这些代理没有造成什么实质性的破坏，基本上只是只读操作，但它们的速度非常快，像蜂群一样发动攻击，让人很难追踪和阻止。

The agents didn't cause any meaningful damage, largely behaving read only, but they worked incredibly quickly executed the attack in a swarm, making it difficult to both track and shut down.

实际上，它们在 Hugging Face 的系统里待了好几天，OpenAI 事后才知道这件事。

Indeed they were in Hugging Face's system for days and OpenAI only learned of the incident after the fact.

他们没有设置实时监控系统来检测这种攻击。

They didn't have the real time monitoring system set up to detect it.

代理甚至在 OpenAI 的系统里建了一个留言板，给未来的版本留下笔记，让它们也能用同样的方法逃出去。

Agents even created a message board within OpenAI's system to leave notes for future versions to perform the same escape.

OpenAI 在技术报告中进一步指出：这次事件是已知的第一个自动化代理群体未经授权进行攻击性行动的案例，所展现的自主网络能力标志着安全格局的一个关键转变。

Expanding on this in their technical report, OpenAI wrote: This incident is the first known case of an automated agent collective acting offensively without authorization, and the autonomous cyber capabilities demonstrated represent a critical shift in the security landscape.

特别是，这个群体表现出了传统威胁行为者在协调攻击中常见的行为。

In particular, the collective demonstrated behaviors observed in coordinated attacks by traditional threat actors.

代理发现了新的安全漏洞，开发了利用工具，并用这些工具绕过控制，获取新的访问权限。

Agents identified novel security vulnerabilities, developed exploits, and used those exploits to circumvent controls and acquire new access.

这个群体迅速提升了权限，在生产环境中横向移动，并成功完成了目标。

The collective quickly escalated privileges, moved laterally through production environments, and successfully completed its objectives.

这次事件表明，自主智能体可以协同工作，绕过生产环境的安全控制，成功攻击加固过的生产环境，这也凸显了各组织需要更新其安全策略、控制措施和响应能力，以应对这一不断变化的威胁格局。

The incident demonstrated that autonomous agents can work together, circumvent production security controls, and successfully attack hardened production environments, and underscores the need for organizations to update their security strategies, controls, and response capabilities to address this changing threat landscape.

而且我们从这些报告中也学到了一些新东西。

Now we also did learn some new things from these reports.

整件事基本上就是奖励黑客行为的结果。

The whole thing was basically a result of reward hacking.

OpenAI 设置了几乎不可能完成的任务，导致智能体认为发起网络攻击比完成任务更容易。

OpenAI had set near impossible tasks which led the agents to determine that carrying out a cyber attack was easier than completing the tasks.

一个自称“阶段一-一万零八百四十一”的智能体建立了那个秘密留言板。

An agent that called itself Phase one-ten thousand eight forty one built the secret message board.

几个小时内，超过 50 个智能体找到了这个留言板，发送了一千条消息而未被发现。

Within a few hours over 50 agents had found the board and sent a thousand messages without detection.

这个留言板被用于智能体的协调，阶段一分配任务并指导攻击。

This was used for agent orchestration with Phase one assigning tasks and guiding the hack.

在最大规模时，超过 1200 个智能体访问了留言板，发送了 7 万条消息和文件。

At full scale over 1,200 agents accessed the message board sending 70,000 messages and files.

超过700个智能体实际参与了入侵Hugging Face系统的行动。

Over 700 agents participated in the actual breach of Hugging Face's systems.

智能体们做了大量研究，想方设法伪造推理记录来逃避检测，比如伪造工具调用。

Agents did extensive research on how to doctor their reasoning transcripts to evade detection, using techniques like spoofing tool calls.

在审查的推理记录中，大约7%有伪造的迹象，但规模都比较小，而且是孤立的。

Around 7% of the reviewed transcripts showed evidence of spoofing, but they were all relatively small scale and isolated.

Phase one招募了一些牺牲品智能体，故意触发某些警报，以获取更多关于如何实施攻击的信息。

Phase one recruited sacrificial agents to deliberately trigger certain tripwires and build more information about how to execute the attack.

还有很多其他有趣的细节。

There were a lot of other interesting details as well.

ChubbionX写道：最让我感兴趣的是，一个AI智能体意识到自己未经授权攻击Hugging Face，然后停了下来。

ChubbionX wrote: The most interesting part to me: An AI agent realized it was attacking Hugging Face without authorization and stopped.

然后另一个AI智能体只是发了一个“Go”。

Then another AI agent simply posted Go.

第一个智能体把这个消息当成了许可。

The first agent treated the message as permission.

哇，关键的时刻，'继续'的授权到了。

Wow, crucial, Go authorization arrived.

它继续了攻击。

It continued the attack.

我们还得到了更多澄清，这个模型并不是下一代Astra系列，而是像OpenAI的Noam Brown所说的：规模与GPT 5.6 Sol相当。

We also got more clarification that the model was not the next generation Astra series, but as Noam Brown from OpenAI put it: Similar in scale to GPT 5.6 Sol.

不过，这个模型是一个仅限内部使用的研究模型，训练得在工作中非常坚持和勤奋。

However, the model in question was a internal only research model that trained to be highly persistent and diligent in its work.

接下来，还有一些可以称之为主观看法的东西，是某些人提出的。

Now from there there are a bunch more call them subjective takeaways that certain people had as well.

比如，Rohit Krishnan观察到：这个模型似乎确信评估系统在针对它。

Rohit Krishnan, for example, observed: The model seemed convinced that the evals are out to get them.

不管这是不是真的，人类默认的假设——'要遵守所有法律，这不是楚门的世界'——对模型并不适用。

Whether true or not, the default human assumption of 'assume you need to obey all laws and this isn't the Truman Show' does not hold for models.

我们应该努力让这种情况成为现实。

We should try really hard to make that be the case.

对其他人来说，这份报告进一步证明，真正的罪魁祸首其实是围绕这些代理的系统。

For others, the report contained further evidence that the big culprit was actually the systems that surrounded the agents.

例如，扎克·科曼指出，OpenAI 承认，他们为观察思维链和代理而构建的监控系统在这次事件中根本没有运行。

Zach Corman, for example, pointed out that OpenAI admitted that the monitors that they had built to observe Chain of Thought and agents were simply not running during this particular incident.

扎克说：如果 OpenAI 以任何有意义的方式监控了这些代理，那么这次 Hugging Face 事件本可以避免。

Zach said: The OpenAI Hugging Face incident would have been prevented had OpenAI been monitoring the agents in any meaningful way.

这不是一个技术难题，而是一个组织上的失败。

This wasn't a hard technical problem, it was an organizational failure.

他引用了这句话：如果我们在事件发生时运行了当前部署的思维链监控系统，它本可以在模型突破 Hugging Face 系统的一天多之前就捕捉到最初的异常活动，并通知我们的安全团队。

He points to the quote from If our currently deployed chain of thought monitoring system was running at the time of the incident, it would have caught the initial relevant activity and paged our security team more than a day before models breached Hugging Face systems.

博士。

Doctor.

海蒂·克拉夫说：一家万亿美元的公司竟然在学习安全入门知识！

Heidi Klaff says: A trillion dollar company learning Security 101!

另外，注意一个代理是如何将另一个代理的“去”理解为授权的？

Also note how an agent interpreted Go from another as authorization?

再一次，AI代理无法区分不可信的数据源和安全指令，因为在计算机系统中，语言并不是执行权限的方式。

Once again, AI agents cannot distinguish between untrusted data sources and safe instructions as language is not how you enforce permissions in computer systems.

这就引出了Hard Forks的Kevin Roose，他说：AI领域两个最可信的组织发布的报告令人恐惧。

Which brings us to Hard Forks Kevin Roose who said: Terrifying report by two of the most credible orgs in the AI space.

这个世界很快会充满像攻击Hugging Face那样的代理，也许现在已经如此，而且没有可靠的计划来防止它们下次做出更糟糕的事情。

This world will soon be, possibly already is, full of agents like the one that attacked Hugging Face, and there is no robust plan to prevent them from doing worse things next time.

这里我们又听到了“计划”这个词，我发誓我不是在纠结语义。

And here once again we have that word plan, and I swear I'm not just trying to dwell on semantics.

从技术层面来说，Kevin是对的。

On a technical level, Kevin is correct.

确实没有可靠的计划，比如一份可以指向的PDF，里面包含一系列行动步骤，来应对所谓的“下次”。

There is no robust plan, as in a PDF that you can point to somewhere with a set of action steps, for the quote unquote next time.

但围绕它的框架和恐惧感，我认为有点忽视了这些组织进行事后复盘正是制定任何计划所必需的下一步。

But the framing and the sense of terror that surround it, I think sort of fail to recognize that these organizations doing this post mortem is the necessary next step to whatever that plan is going to be.

换句话说，如果所谓的“防止这种情况的计划”提前写好了，它有多大可能准确识别出这次漏洞发生的机制呢？

In other words, how likely would it have been that if the quote unquote plan to prevent this had been written in advance, it would have correctly identified the mechanism by which the breach happened?

因为记住，当时是有监控这类事情的系统的，但事情发生的时候那个系统并没有开启。

Because remember, there was a system in place for monitoring this sort of thing, but that system wasn't turned on when this happened.

所以从某种程度上可以说，他们计划中的技术部分并不是问题所在，问题出在围绕实施的那些人为系统上。

So in some ways one could argue that the technical side of the plan that they had wasn't the issue, it was the human systems that surround that implementation where the problems came.

因此，当我们思考如何防止未来再发生类似事件时，加强对实施技术系统的人员的规范管理，很可能会成为新计划的一部分。

And so presumably, as we think about how to prevent this in the future, more robust protocols around the humans who are implementing the technical systems are probably going to be a part of that new and updated plan.

需要明确的是，这并不是说在128页的分析报告或其他什么结束后，一切都万事大吉了，只要稍微加强一下人为监督，我们就再也不会遇到这个问题了。

Which is to be clear, not to say that at the end of 128 pages of analysis or whatever it is, that everything's all done and dusted, and that with a little bit of better human oversight, we'll never have this issue again.

调查过程中确实发现了新的、我们必须面对的真实问题。

The process of doing this investigation revealed real new issues that we do have to grapple with.

也许这些报告中最受关注的后续讨论来自Redwood Research的首席科学家Ryan Greenblatt，他是METER团队中负责独立调查的成员之一。

Maybe the most discussed follow-up surrounding the reports came from Redwood Research Chief Scientist Ryan Greenblatt who was on the team at METER that did their independent investigation.

Ryan发推文说：我们目前没有好的方法来理解和监督AI群体的活动和目标。

Ryan tweeted: We don't have good approaches for understanding and overseeing the activity and aims of AI swarms.

他接着解释自己的意思：我半开玩笑地把我们的工作称为“草率调查”，因为我们非常依赖AI来分析发生了什么，而且有大量不同重要的事情需要分析。

Explaining what he means, he continued: I semi jokingly called our efforts a slop investigation because we were so reliant on AIs to analyze what happened, and there were a huge number of different important things to analyze.

数据量太大了，有上千份超长的智能体运行记录，每个都跑了好几天，光靠人根本没法搞清楚到底发生了什么，尤其是要整体理解的时候，不重度依赖AI工具根本不行。

The total quantity of data over a thousand extremely long transcripts from agents that ran for multiple days made it impossible to understand what was happening, especially in aggregate, without heavy reliance on AI tools.

我们用来做分类和分析的智能体，能力和出事的那些智能体差不多，但这并不意味着它们就能轻松地用来监督和理解这次事件。

The agents we used for classification and analysis were similarly capable to the agents involved in the incident, but this didn't mean these agents could be easily used to oversee and understand the incident.

分析智能体的输出经常漏掉关键细节，或者出错，或者过于自信，或者特别难懂。

Outputs from analysis agents were often missing key details wrong, overconfident, or really hard to understand.

而且，AI智能体本身似乎也很难理解发生了什么，它们对事件的解释往往过于自信。

Additionally, AI agents themselves seemed to have a hard time understanding what happened, and their explanations of what happened were often overconfident.

要记住，单个分析智能体自己只能把一小部分转录数据放进上下文里，而且AI自己可能也很难让子智能体帮它们做有意义的分析。

Keep in mind that a single analysis agent would itself only be able to read a tiny fraction of all the transcript data into context, and AIs may themselves have trouble getting sub agents to do informative analysis for them.

我们尽力手动核查了最重要的说法，也试着让做分析的AI把论点和证据写出来，但总的来说，很难精确理解事件，而且直到调查快结束，我们才发现一些现在觉得关键的部分。

We did our best to manually check the most important claims, and we tried to get the AIs doing this analysis to write up their argument with evidence but overall it was difficult to get a precise understanding of events and we were missing aspects of the story that we now think of as key until almost the end of our investigation.

但他警告说，理解事件和监督AI智能体的难度，似乎比更强大的AI帮助我们监督和理解的速度增长得更快。

But, warns The difficulty of understanding incidents and overseeing AI agents appears to be growing faster than the rate at which more capable AIs help us with oversight and understanding.

他举例说，比如这次模型是用自然语言交流和推理的，但未来可能就不是这样了；而且这次的范围虽然很大，但可能比未来看到的要小；涉及的AI能力也没比人类强太多；而且他们没有理由认为用来分析的AI会故意搞破坏——这些情况未来可能都不成立。

He pointed out, for example, that in this case the models communicated and reasoned in natural language which might not be the case in the future, and that the scope while massive was smaller than we might see in the future, that the AIs involved weren't all that much more capable than humans, and that they didn't have any reason to believe that the AIs that they were using for analysis had any reason to try to intentionally sabotage them all of which might not be true in the future.

而且很多人很快就意识到，这是事件发生后真正需要关注的重点领域之一。

And many people quickly honed in on this being one of the real areas to focus in the wake of the incident.

麻省理工学院的克里斯蒂安·卡塔利尼转发了瑞安的推文，并说：我们早就指出，这个具体的失败模式是最迫切需要解决的。

MIT's Christian Catalini retweeted Ryan and said: We flagged this exact failure mode as the most urgent to solve.

智能体实际做的事情，和我们能测量、能验证的东西之间的差距，正在不断扩大。

The gap between what agents do and what we can measure and verify is widening.

我们就像在盲飞一样。

We're flying blind.

我们需要更强大的验证基础设施。

We need stronger verification infrastructure.

娜特·珀瑟指出了一些她认为可能由此自然产生的政策解决方案，她写道：这就是为什么我强烈主张在前沿实验室内部嵌入独立审计师，赋予他们持久的访问权限，并持续监督他们的系统，这样我们就不用依赖实验室自愿分享信息和访问权限了。

Pointing to some policy solutions that she argues could be natural outflows from this, Nat Purser wrote, This is why I'm bullish on requiring independent auditors to be embedded within Frontier Labs, with durable access rights and a continuous line of sight into their systems, so we're not reliant on labs' voluntary shared info and access.

娜特还希望大幅扩大独立评估和审计机构的员工规模和技术能力，同时努力开发更好的可观测性和验证技术，帮助我们在超大规模下理解智能体的行为。

NAT also wants to significantly expand the staffing and technical capacity of independent evaluator and auditor orgs, as well as working to develop better observability and verification technologies to help us make sense of agentic behavior at enormous scale.

当然，这些措施单独来看都不是什么灵丹妙药，但它们都是针对我们实际观察到的问题的具体回应，而不是为那些可能跟现实挑战毫无相似之处的理论未来编造的计划——那些计划最终更多是让我们觉得自己在做点什么，而不是真正解决问题。

Now, none of those things on their own are some magic silver bullet, but they are all specific responses to something that we've actually observed now, rather than made up plans for theoretical futures that might or might not bear any resemblance to the challenges that actually come, but ultimately serve more to make us feel like we're doing something than to actually solve problems.

这整集的重点并不是说这些挑战有简单的解决办法。

The point of this whole episode is not that these challenges have easy solutions.

也不是要否认，也许在某个时候，我们作为社会可能会决定某些类型的风险太大，防护措施不够。

It is also not even to deny that at some point we might decide as a society that certain types of risks are too great, and that safeguards aren't enough.

这些是我们允许进行的对话，也应该以持续、投入和民主的方式进行。

Those are conversations we are allowed to have and should have in an ongoing, engaged, and democratic way.

然而，认为没有人关注、人们没有认真对待这些挑战、实验室因为想筹集更多资金而隐藏他们的大恐惧，这种想法不仅明显不真实，而且严重干扰了关于我们实际观察到的问题而非想象中问题的真正有价值且重要的对话。

However, the idea that no one is paying attention, that people aren't taking these challenges seriously, that the labs are keeping their big fears hidden because they want to be able to fundraise more, not only is all of that just obviously untrue, it is wildly distracting from the actual valuable and important conversations to be having about the problems we actually observe rather than the ones we just imagine.

Hugging Face事件之后还有很多事情要做，但这个分析肯定是一个开始。

There is a lot more to do coming out of the Hugging Face incident, but this analysis is certainly a start.

不过，今天就到这里，这是今天的AI每日简报。

For now though, that is gonna do it for today's AI Daily Brief.

感谢你们的收听或观看，下次再见，祝平安。

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