最值得尝试的新AI功能和工具

这些里面哪个最让你兴奋？

Which of these sound most exciting to you?

一个是让Claude能用浏览器窗口帮你完成任务的新能力，一个是让ChatGPT能根据你的需要灵活切换临时模式的功能，还有一个新语音模型，不仅能准确转录你实际说的话，还能帮你整理内容、理解你的意图。

A new ability for Claude to use a browser window to be able to do tasks for you, the ability for ChatGPT to effectively switch in and out of temporary mode as suits you, A new voice model that not only can faithfully transcribe what you actually say, but even clean it up and get it your intent.

还有一个来自Google的新视频模型，可控性大大增强，能生成真正实用的视频。

A new video model from Google with massively more controllability for actually generating videos that you can use for real things.

或者一个完全不同的视频模型，生成视频的时间比观看视频的时间还短。

Or a totally different video model that takes you less time to generate the video than it does to watch it.

这些只是本周发布的一些新功能、模型和工具。

These are just some of the new features, models, and tools that released this week.

《AI每日简报》是一个每日更新的播客和视频节目，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.

在开始之前，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、Harbor、Blitzy 和 Hyperagent。

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

想要收听无广告版本，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.

想了解如何赞助本节目，请发送邮件至 sponsors@aideallybrief.ai。

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

最后，如果你听完今天的节目后心想：天哪，我想用所有这些功能，但感觉自己落后了。

And lastly, if you listen to today's episode and think to yourself, man, I wanna use all those features, but I feel behind.

我需要提升我的 AI 技能。

I need to catch up in my AI skills.

下一期超级智能训练项目，也就是在 AIDB 这里孵化的那些，将在九月份开始。

The next cohort of superintelligent training programs, which are incubated over here at AIDB, are coming up in September.

你可以在 AI Daily Brief 网站上找到相关链接，或者直接访问 training.besuper.ai。

You can find a link to them on the AI Daily Brief website, or you can just go to training.besuper.ai.

首先确认一下，Hugging Face 确实被出售了，而且买家就是 NVIDIA。

Kicking off confirmation that Hugging Face is being sold and that, yes, it is to NVIDIA.

这个消息是周三晚上晚些时候，在财报电话会议之后被曝出来的，消息人士说英伟达已经同意以129亿美元收购Hugging Face。

The information broke the story late on Wednesday night following the earnings call with their sources saying that NVIDIA has agreed to buy Hugging Face for $12,900,000,000.

关于这笔交易的传闻只是在上周末才开始出现，虽然当然有可能在外部知道之前就已经在进行了，但看起来这很可能是一次相当迅速的谈判。

Now rumors of the deal only started emerging over the weekend and while of course this could have all been going on before anyone externally knew, it seems like it might have been a pretty speedy negotiation.

很多报道都聚焦于高估值倍数，Hugging Face的年化收入只有1.5亿美元，这意味着英伟达实际上支付的价格大约是收入的80倍。

Much of the reporting focused on the strong valuation multiple, Hugging Face only has a 150,000,000 in annualized revenue meaning that NVIDIA is technically paying around 80 times revenue as a price.

作为参考，SpaceX以600亿美元收购Cursor，大约是收入的15倍。

As a point of reference, SpaceX's $60,000,000,000 acquisition of Cursor was around 15 times revenue.

不过，就像我前几天谈到这笔交易的可能性时说的，英伟达收购Hugging Face绝对不是冲着他们的收入去的。

Still, as I said the other day when I talked about the potential of this deal, NVIDIA is very much not buying Hugging Face for their revenue.

相反，这笔交易似乎证实了英伟达正在通过收购进入一个全栈开放商业模式。

Instead, the deal seems to confirm that NVIDIA is acquiring their way into a full stack open business model.

Hugging Face是开放模型的主要分发渠道。

Hugging Face is the premier distribution channel for open models.

英伟达上周宣布的与Poolside的60亿美元授权协议，让他们能够雇佣大约100名资深AI研究员来充实团队，而其他投资，比如与Perplexity的战略合作，也给了他们扎实的基础，以及一个应用层面的玩法。

NVIDIA's $6,000,000,000 licensing deal with Poolside announced last week allowed them to hire about a 100 veteran AI researchers to staff up their team, and other investments like their strategic partnership with Perplexity give them solid harnesses as well as an app player play.

而且，在大约六个月的时间里，NVIDIA 已经把他们的开源模型系列 Neutron 从一个技术演示变成了开源模型中的真正竞争者。

What's more, over the course of around six months, NVIDIA has taken their open model series Neutron from a tech demo to a real contender among open models.

Akash Gupta 认为，这既是进攻性的举措，也是防御性的举措。

Akash Gupta suggests that as much as it might be an offensive play, it's also a defensive play.

他写道：谷歌已经自己运行 TPU 十年了。

He writes: Google has run its own TPUs for a decade.

给 NVIDIA 开最大支票的那些实验室，正是最努力想摆脱它的。

The labs writing NVIDIA the largest checks are the ones working hardest to stop.

开源模型就是那个平衡力量。

Open models are the counterweight.

当 Meta、Mistral 或 DeepSeek 发布模型权重时，数百万开发者会下载、微调，然后绝大多数都在 NVIDIA 的 GPU 上通过 CUDA 运行。

When Meta or Mistral or DeepSeek releases weights, millions of developers download them, fine tune them, and serve them overwhelmingly on NVIDIA GPUs through CUDA.

定制芯片对这个生态系统毫无办法。

Custom silicon has no answer for that ecosystem.

我觉得这种防御性定位不仅关乎 NVIDIA 的芯片业务，也关乎它在那个我们一直在讨论的世界中的定位——公司会使用多种模型，而它们日常的许多核心流程不是运行在最前沿的模型上，而是运行在更开放、更轻量、更灵活、更可定制的模型上。

I think the defensive positioning is not just relative to NVIDIA's chip business, but also its positioning for the world that we keep talking about where companies simply use a diversity of models, and many of their common and core day to day processes are run not on the state of the art frontier models, but on something open, lighter, more nimble, more customizable.

不久前，NVIDIA 发现了这个缺口，从那以后就一直在朝这个方向猛冲。

A while ago NVIDIA spotted that opening and has been careening towards it ever since.

至于大家怎么看这件事，没人觉得 NVIDIA 这么做有什么难理解的。

Now in terms of how folks are receiving it, no one has a hard time understanding why NVIDIA made the move.

Fumbler 写道：NVIDIA 显然不只是想卖铲子，他们还想掌控建立在铲子之上的 AI 经济。

Fumbler writes: NVIDIA clearly don't just want to sell the picks and shovels, they want to own more of the AI economy built on top of them.

Gianluca 写道：NVIDIA 拥有 GPU，CUDA 拥有开发者，Hugging Face 拥有开源 AI 生态，黄仁勋正在慢慢把整个链条拼起来。

Gianluca writes: NVIDIA owns the GPUs, CUDA owns the developers, Hugging Face owns the open source AI ecosystem, Jensen is slowly putting the whole stack together.

NVIDIA 正在变得远不止是一家芯片公司。

NVIDIA is becoming much more than a chip company.

不过也有人在想，换了新东家之后，大家还会不会继续留在 Hugging Face。

Others though are wondering if people will stick around Hugging Face given its new ownership.

Todd Saunders 写道：Hugging Face 的价值一部分在于它被视为独立且大体中立的平台。

Todd Saunders writes: Hugging Face is valuable partly because it is perceived as independent and broadly neutral.

它的用户包括研究人员、初创公司、企业、云服务商、硬件公司，还有 NVIDIA 的竞争对手。

Their users include researchers, startups, enterprises, cloud providers, hardware companies, and competitors to NVIDIA.

现在，开发者们使用这个平台时，不会觉得它是在把他们往某个供应商那边推。

Today, developers can use the platform without believing that it is designed to push them towards one vendor.

这正是他们喜欢它的部分原因。

It's partially what they love about it.

英伟达的收购可能会破坏这种信任，但这也有可能大获成功。

An NVIDIA acquisition could threaten that trust, but this could be a home run.

普罗塔姆认为这也是一个更大的趋势。

Protham thinks that this is a bigger trend as well.

他写道，真正的问题不是英伟达为什么收购了Hugging Face。

They write, The real question isn't why NVIDIA bought Hugging Face.

而是当最大的玩家开始收购开源AI的栖息地时，开源AI会变成什么样？

It's what happens to open source AI when its biggest players start buying the places where it lives?

埃里克·哈特福德认为我们会看到一个机会。

Eric Hartford thinks that we will see an opportunity.

他写道：我为Hugging Face的退出感到高兴，但这将是开源AI的损失。

He writes: I'm happy for Hugging Face for their exit, but this is going to be a loss for open source AI.

Hugging Face 不再与硬件无关了。

HF is no longer hardware agnostic.

这为新标准的引领者崛起提供了机会。

This is an opportunity for a new standard bearer to arise.

不过，这种担忧在收购中总是会出现，但相比通常情况，这次收购的乐观情绪要多得多，这也说明了 NVIDIA 在建立 AI 社区信任方面做得有多好。

Still, that's the kind of concern that people always have around acquisitions, and relative to expecting that to be the case, I would suggest that there is a lot more optimism around this particular acquisition than you normally see, which shows what a good job NVIDIA has done building trust in the AI community.

Connor Dart 写道：我想不出比 NVIDIA 更适合收购 Hugging Face 的公司了，这很合理，因为它与 NVIDIA 的整体战略一致。

Connor Dart writes: I can't think of a better company to buy Hugging Face other than NVIDIA and it makes sense as it aligns with what they are doing as a whole.

我敢打赌，未来我们会看到 Hugging Face 上出现更多令人惊叹的功能，以及更多训练和使用模型的方法。

I bet we will be seeing loads of more amazing features and ways to train and use models on Hugging Face going forwards.

Ramiz Nam 写道：这真是绝配，也是向多元 AI 迈出的一大步。

Ramiz Nam writes: Such a great fit and such a great push towards Plural AI.

NVIDIA 卖的是算力。

NVIDIA sells compute.

Hugging Face 让获取各种将算力转化为价值的方式变得容易，降低了任何单一实验室、公司或国家垄断前沿 AI 的可能性。

Hugging Face makes it easy to access a plethora of ways to turn that into value, reduces the odds that any one lab, company or nation will monopolize frontier AI.

世界因为多元AI而变得更美好。

The world is better with Plural AI.

NVIDIA这周也发布了财报，业绩简直炸裂。

NVIDIA also reported earnings this week and kind of blew the doors off.

过去一年营收增长加速到106%，比第一季度末的增长率高出21个百分点。

Revenue growth accelerated to 106% over the past year, 21 points higher than their growth rate at the end of Q1.

本季度销售额达到962亿美元，距离单季度1000亿美元的里程碑只差一步，而目前只有五家公司达到过这个水平。

Quarterly sales for the quarter came in at $96,200,000,000 tantalizingly close to the $100,000,000,000 a quarter milestone that only five other companies have achieved.

不过，尽管预测超出预期，NVIDIA仍然指出第三季度增长将略有放缓，预计为89.5%。

Still, while forecasts were ahead of expectations, NVIDIA is still flagging that growth will slow slightly to 89.5% for Q3.

首席财务官Colette Crest表示，供应链问题将继续拖累增长，使得2027财年的增长率限制在70%。

CFO Colette Crest said that supply chain snarls would continue to weigh on growth, limiting NVIDIA to 70% for fiscal year twenty twenty seven.

NVIDIA还透露，一些大客户正在要求延长付款期限。

NVIDIA also disclosed that major customers are asking for more time to pay.

自由现金流同比下降了50%，为210亿美元，这一下降与应收账款增加50%相对应。

Free cash flow fell by 50% compared to a year ago coming in at $21,000,000,000 This fall was mirrored by a 50% rise in accounts receivable.

英伟达表示，原因是，引用一下，与某些投资级客户签订的大型多季度协议延长了付款期限。

NVIDIA said the cause was, quote, extended payment terms on large multi quarter agreements with certain investment grade customers.

当然，这对投资者来说是个大警示，应收账款周转天数从45天上升到了60天。

Certainly, this was the big red flag for investors, with days sales outstanding rising from forty five days to sixty days.

除了这些数字，英伟达还宣布正式重返中国市场。

Alongside the numbers, NVIDIA announced that they are officially back in the Chinese market.

他们报告说，根据新的出口管制规定，H200芯片的销量很小，但指出这个数字远低于特朗普政府允许的总量。

They reported a small volume of sales of H200 chips under new export control rules, but noted that the figure was well short of the total permitted by the Trump administration.

最终，虽然报告中有一些负面因素，但很难忽视那惊人的销售数字。

Ultimately, while there was room to find some negatives in the report, it's difficult to overlook the raw sales numbers.

英伟达去年夏天就已经在出货大量芯片，而自那以后他们的销量几乎翻了一番。

NVIDIA was already shipping an incredible number of chips by last summer and they've almost doubled sales since then.

为了强调这一点，亚马逊宣布将额外购买200万颗英伟达芯片，因为他们正在加速数据中心的建设。

Reinforcing that point, Amazon announced that they would be buying an additional 2,000,000 NVIDIA chips as they accelerate their data center build out.

英伟达没有透露他们每年总共销售多少芯片，但据信不到1000万颗，这意味着这是一笔巨大的额外订单，将在未来几年内保持强劲的需求。

NVIDIA doesn't disclose how many total chips they sell per year, but it's believed to be less than 10,000,000, meaning that this is a huge additional order that will keep demand solid for years into the future.

这份财报足以让投资者重新看好英伟达，盘后交易中股价上涨了4.8%。

The earnings were enough to drive investors back to NVIDIA, setting the stock up by 4.8% in after hours trading.

一个有趣的细节是，鉴于英伟达在AI生态系统中扮演的独特角色，公司暂停了与Neo Clouds的收入分成协议，因为他们正在重新考虑对行业支持的做法。

One interesting sub given the unique role that NVIDIA is increasingly playing in the AI ecosystem, the company has paused revenue sharing deal with Neo Clouds as they rethink their approach to industry backstops.

不到两个月前，英伟达宣布将推出新的融资方案，客户可以获得信贷支持，但需要分享一部分利润。

Less than two months ago, NVIDIA announced that they would start offering new financing deals where could receive credit support in exchange for a cut of the profits.

当时这被视为对英伟达之前模式的改进，之前他们保证数据中心需求，并希望这能转化为来自私人信贷公司或银行的更便宜融资。

This was seen at the time as an improvement on NVIDIA's previous model, where they would guarantee data center demand and hope that translated into cheaper financing from private credit firms or banks.

消息人士告诉《华尔街日报》，由于一些员工表达了担忧，这个项目已被搁置。

Sources told The Wall Street Journal that this program has been put on hold after some employees expressed concerns.

有人认为这种安排可能引起反垄断审查，而且对于英伟达对客户业务的控制程度也存在敏感问题。

There is a belief that the arrangement could attract scrutiny for antitrust and there were sensitivities around how much control NVIDIA would exert over customer businesses.

该报指出，暂停的具体原因无法得知，因此理论上这个项目未来可能会经过一些调整后重新推出。

The journal noted that the precise reason for the pause couldn't be learned and so theoretically the program could return in the future with some tweaks.

英伟达否认了这些报道，并表示：“我们在7月推出的新商业模式，旨在向快速增长的AI生态系统开放计算资源，目前仍然有效，并且由于需求旺盛，还在不断演变。”

NVIDIA for their part denied the reporting and stated: The new business model we introduced in July that opens up compute access to the fast growing AI ecosystem is still in place and continues to evolve due to high demand.

现在咱们先聊一下华尔街，信不信由你，Salesforce——我们稍后会在主节目里也会提到——凭借一份强劲的财报，正在引领SaaS行业的复苏。

Now staying on Wall Street for a minute, believe it or not, Salesforce who we will talk about in the main episode as well is leading a SaaS comeback with a strong earnings report.

六个月前，市场的一个主要论调是软件行业已经完了，公司很快就会用AI自己写代码来替代SaaS产品。

Six months ago, one major market narrative was that software was done for and companies would soon be vibe coding their own SaaS replacements.

这就是当时被广泛讨论的“SaaS末日”，如果你还记得这个词的话。

This is the much discussed SaaSpocalypse if you remember that term.

显然，那种程度的颠覆并没有发生，而且我们越来越多地看到，在传统软件产品中采用AI工具正在带来收入增长。

That level of disruption obviously did not play out and increasingly what we're actually seeing is adoption of AI tools within traditional software products is leading to a revenue boost.

周三晚上的财报就充分说明了这一点，Salesforce报告称，他们的Agent Force产品今年有望实现15亿美元的收入，高于第一季度预测的12亿美元。

That was certainly the story on Wednesday night when Salesforce reported that their Agent Force product is on track to deliver $1,500,000,000 in revenue this year, up from a forecast of $1,200,000,000 in the first quarter.

本季度总销售额为115亿美元，同比增长11%，所以AI收入在整体中占比仍然相对较小。

Overall sales were $11,500,000,000 for the quarter, growing at an 11% pace, so AI revenue is still a relatively modest part of the picture.

但这已经足以让市场将Agent Force视为一个成功且增长中的产品。

This was still enough though for the market to identify Agent Force as a successful and growing product.

周四股价暴涨22%，Salesforce现在距离完全收复SaaS末日期间的跌幅只差5%。

The stock was up a staggering 22% on Thursday, with Salesforce now just 5% away from filling the entire drawdown from the SaaSpocalypse period.

退一步看，我们一个一个地看到了一些之前跌得最惨的软件公司交出了不错的财报，然后股价又猛涨回来了。

Taking a step back, one by one we have seen some of the most beaten down software stocks deliver solid financials and surge back to life.

ServiceNow、Figma、Atlassian、CrowdStrike、Okta 这些公司在过去几个月里都实现了大翻身。

ServiceNow, Figma, Atlassian, CrowdStrike, Okta have all staged major comebacks over the past months.

Key Bank 的分析师 Jackson Atter 指出，这些公司甚至不需要发布惊人的财报就能带来巨大的回报。

Key bank analyst Jackson Atter noted that these companies haven't even needed to post blockbuster earnings to generate a huge return.

Atter 说：“我们都开始接受这样一个事实，就是事情会变得更持久，不会消失。”

Said Atter, We're all coming around to the idea that things are going to be more durable and are not going away.

这就是为什么 Salesforce 交出的成绩单虽然不错但不算惊艳，却引发了特别大的市场反应。

That's why you see Salesforce come out with fine but not stellar results, but you see a really outsized reaction.

贝尼奥夫在财报电话会议上当然也得意了一把，他说：“我真的想从最开始说起，聊聊所谓的‘SaaS 末日’。”

Benioff certainly took a victory lap on the earnings call saying: I really want to start here at the beginning and talk about the SaaS pocalypse.

怀疑者说席位会减少，但 Agent Force、Sales、Service 和 Slack 的席位都同比增长了。

Skeptics said that seats would decline, and Agent Force, Sales, Service, and Slack all seats grew year over year.

怀疑者说客户会流失，但客户流失率几乎降到了历史最低点。

Skeptics said customers would leave, but attrition was near its lowest level ever.

怀疑者说定价权会被削弱，但预订量环比增长了一倍多，而且所有细分市场、新业务和续约的合同期限条款都有所改善。

Skeptics said pricing power would erode, but bookings more than doubled quarter over quarter and contract length terms improved across all segments, new business, and renewals.

而且，通过模型上下文协议调用和CLA调用，平台上的智能体使用量激增了六倍。

And agentic use of the platform surged six fold via model context protocol calls and CLA calls.

应用并没有消亡。

Apps are not dying.

前AI领域的大卫·萨克斯迅速抓住这个消息发帖说：“AI资本支出是泡沫”和“SaaS已死”的说法今早被撕得粉碎。

Former AI's r David Sachs pounced on the news posting: The AI CapEx is a bubble and the SaaS is dead narratives getting shredded this morning.

Box的亚伦·利维写道：“本周的科技财报电话会议是一个重要的提醒，让我们看到软件和AI之间的关系有多么宝贵。”

Box's Aaron Levy writes: This week's tech earnings call were a critical reminder of how valuable the relationship is between software and AI.

软件和智能体将协同工作，最终将大幅扩大IT市场的整体规模。

Software and agents are going to work together and ultimately grow the size of the IT TAM dramatically as a result over time.

现在，把蓬勃发展的收入故事带到应用层。根据The Information获得的新数据，Cognition的年化收入已达到9亿美元（自年初以来增长超过3倍），这是编码智能体Devon的创造者，而且似乎整个行业都是类似的情况。

Now bringing the booming revenue story to the app layer, according to new numbers sourced by The Information, Cognition has hit $900,000,000 in annualized revenue (that's more than a 3x increase since the beginning of the year) for the creators of Coding Agent Devon and it appears to be a similar story across the board.

Higgs Field，一个视频和图像模型的接口封装器，今年也增长了两倍，年化收入达到7亿美元；Perplexity的运营率增长了两倍，达到7.5亿美元；Open Evidence，面向医生的聊天机器人，翻倍至3亿美元；甚至Manus在Meta收购的混乱收尾之后，也声称其运营率达到4亿美元，比去年翻了两番。

Higgs Field, an interface wrapper for video and image models, has also tripled this year to reach 700,000,000 in annualized revenue, Perplexity has tripled their run rate to 750,000,000, Open Evidence, the chatbot for doctors, has doubled to 300,000,000, and even Manus after the messy unwind of the Meta acquisition claims they have a run rate of 400,000,000 quadrupling from last year.

虽然总体数字还远落后于美国，但中国领先的AI公司也看到了类似的增长。

And if the total numbers still remain far behind The US, leading Chinese AI companies are seeing similar types of growth.

The Information再次报道，DeepSeek今年已经创造了7000万美元的收入，相比2025年全年数字增长了10倍。

The information again reports that DeepSeek has generated $70,000,000 in revenue this year, which is a 10x increase from their full year numbers for 2025.

现在，我们换个话题，快速了解一下政策方面的情况。

Now as we round the corner, a quick check-in on the policy side of the house.

一项关于AI监管的新行政命令在最近几周草案被悄悄传阅后，已经停滞不前。

A new executive order on AI regulation has stalled out after drafts were quietly circulated in recent weeks.

有报道称，特朗普政府最近提出了一项行政命令，将创建一个AI行业的自律监管机构。

Reports say that the Trump administration recently proposed an executive order that would create a self regulatory body for the AI industry.

你可能还记得，这个想法在DeepMind董事长Demis Hassabis在7月的一篇博客文章中正式提出后，开始获得支持。

You might remember this idea starting to gather support after DeepMind chairman Demises Sabes formally proposed it in a July blog post.

这个新机构将仿照FINRA，即金融服务业的自律监管组织。

The new body would be modeled on FINRA, the self regulatory organization for the financial services industry.

基本上，这种方法会要求AI实验室在安全问题上共享信息，对新模型进行安全测试，并制定管理该行业的法规。

Essentially, this approach would require AI Labs to share notes on security issues, conduct safety testing for new models, and create regulations that govern the industry.

财政部长斯科特·贝森特，不知为何在AI政策中扮演关键角色，公开支持了这个想法。

Treasury Secretary Scott Bessent, who for whatever reason has a key role in AI policy, publicly supported the idea.

据报道，高级官员在过去一个月里起草并传阅了一份行政命令，旨在创建这个新机构。

Reportedly senior officials drafted and circulated an executive order that would create this new body over the past month.

虽然目前还不完全清楚这项努力为何停滞不前，但一些消息来源指出，前白宫AI事务负责人大卫·萨克斯的反对可能是原因之一。

And while it's not totally clear exactly why the effort has stalled out, some sources flagged opposition from former White House AI czar David Sachs as a potential cause.

在上周末的《All In》播客节目中，萨克斯称AI版的FINRA是“糟糕的主意”，并说这是“开源模型禁令的特洛伊木马”。

During last weekend's episode of the All In podcast, Sachs called FINRA for AI a, quote, horrible idea and a Trojan horse for an open source model ban.

他认为，一旦这个机构成立，前沿实验室会推动制定适用于所有模型的统一标准，

He believes that once the body is set up, the Frontier Labs will push for standards that apply equally to all models, I.

即

E.

开源模型开发者将无法满足的标准。

Standards that open model developers will find impossible to comply with.

萨克斯还指出，FINRA在促进竞争和创新方面名声并不好。

Sachs also pointed out that FINRA doesn't have a good reputation for promoting competition and innovation.

很多想进入金融行业的初创公司认为，FINRA主要关心的是用监管来保护现有企业。

Many start ups looking to enter the financial industry view FINRA as largely concerned with using regulation to protect incumbents.

不管怎样，目前还没有针对AI的FINRA，但这并不意味着行业自律的努力没有在进行。

Whatever the case for now, no FINRA for AI, but that doesn't mean that industry self regulatory efforts are not happening.

值得注意的是，OpenAI、Anthropic以及超过100家公司呼吁采取紧急的新方法来应对网络防御。

Notably, OpenAI and Anthropic and more than 100 other companies have called for an urgent new approach to cyber defense.

这一呼吁以公开信的形式发出，签署方涵盖了科技、金融和工业领域的公司。

The call came in the form of an open letter signed by companies across the tech, finance, and industrial sectors.

除了谷歌、微软和AWS，名单上还包括Visa、通用汽车和普华永道等公司。

Alongside Google, Microsoft, and AWS, the list also includes companies like Visa, General Motors, and PwC.

信中警告说：我们加强网络防御的时间窗口有限。

The letter warned: We have a limited window to strengthen cyber defenses.

未来几个月，随着全球模型能力越来越强，AI驱动的网络攻击将变得更加普遍和复杂，我们社区所依赖的公司和公共服务——从医院到水处理厂，再到支撑互联网的基础设施——都面临风险。

In the coming months, AI enabled cyberattacks will become far more widespread and sophisticated as models around the world become increasingly capable, the companies and public services our communities depend on from hospitals to water treatment plants to the infrastructure that powers the internet are at risk.

如今AI的进步已经为防御者提供了新的方法，来修复多年来积累的弱点。

Today's AI advances are already giving defenders new ways to fix weaknesses that have accumulated for years.

如果我们果断行动，就能利用防御者的窗口期，让我们的数字世界变得更加安全。

If we act decisively, we can use the defenders' window to make our digital world much more secure.

这封信提出了一个集体应对的方案。

The letter proposes a collective response.

信中指出，现状已经行不通了，因为所有计算机系统里都积压着长期存在的漏洞、未修补的弱点和技术债务。

It asserted that the status quo isn't going to cut it, with long standing bugs, unpatched vulnerabilities, and technical debt across all computerized systems.

信中还呼吁，要让网络防御者能够使用更先进的AI工具，并加强跨行业的协调合作。

The letter called for cyber defenders to be provided with access to better AI tools as well as better coordination across industries.

信里强调，需要全球性的应对措施，建立新的伙伴关系，提高安全标准，并为新兴的网络威胁找到新的解决方案。

A global response is necessary, it argues, requiring new partnerships to raise security standards and find new solutions for emerging cyber threats.

对于这封公开信，山姆·奥特曼认为安全比竞争更重要，他写道：这是AI网络防御的关键时刻。

Remarking on the open letter, Sam Altman positions security as more important than competition, posting: This is a critically important moment for cyber defense with AI.

现在没有太多时间可以浪费了。

There is not much time to act.

如果你们想和我们或我们的任何竞争对手或合作伙伴合作，我们都很欢迎，但请认真对待这一刻。

We are happy if you want to work with us or any of our competitors or partners, but please take this moment seriously.

只有紧急而集中的集体响应才能奏效。

Only an urgent and intense collective response will work.

为了强调这一点，伊桑·莫洛克教授写道：你的组织在开放权重模型和工具可用之前，没有投入足够的精力来加强网络安全。

Putting an exclamation point on it, Professor Ethan Moloch writes: Your organization is not spending enough of its effort bolstering cybersecurity during the window before open weights mythos class models and harnesses become available.

Hugging Face的事件表明，即使没有故意的恶意行为者，你也可能面临AI黑客攻击的风险。

The Hugging Face incident shows us you don't even need intentional bad actors to be exposed to AI hacking.

就像我说的，这周结束时的头条新闻相当重磅，但现在让我们进入主要部分，聊聊你现在就可以使用和尝试的一堆新功能和工具。

Like I said, quite a meaty headlines to end the week, but for now, let's move into the main episode and talk about the slew of new features and tools you could be using and trying right now.

KPMG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多名早期职业专业人士，发现表现最好的人通过指导、评估和完善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，他们的定义不仅仅在于他们知道什么，而在于他们如何与AI合作。

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.

如果你听这个节目，你很可能有自己的观点。

If you listen to this show, you likely have a thesis.

可能是企业采用，可能是算力，也可能是某个特定的实验室。

Maybe it's enterprise adoption, maybe it's compute, maybe it's a specific lab.

Harbor Capital的AI实验室生态系统ETF让你通过五只主动管理型ETF来表达你的观点，每只ETF都专注于围绕一个主要实验室的生态系统：Anthropic、OpenAI、DeepMind、Meta或SpaceX AI。

Harbor Capital's AI Lab Ecosystem ETFs let you express it via five actively managed ETFs, each seeking exposure to the ecosystem around one major lab: Anthropic, OpenAI, DeepMind, Meta, or SpaceX AI.

你对AI竞赛的看法，以ETF的形式呈现。

Your view of the AI race in ETF form.

Harbor Capital Advisors的AI实验室生态系统ETF系列为投资者提供了一种投资他们认为最有可能成功的AI生态系统的方式。

Harbor Capital Advisors' AI Lab Ecosystem ETF Suite gives investors a way to invest in the AI ecosystem they believe is best positioned for success.

在你投资的任何地方搜索Harbor AI实验室生态系统ETF，或关注HarborCapital的X账号了解更多信息。

Search Harbor AI Lab Ecosystems ETFs wherever you invest or follow HarborCapital on x to learn more.

访问harborcapital.com获取包含投资目标、风险、费用、支出及其他重要信息的招股说明书。

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

投资前请仔细阅读并考虑这些信息。

Read and consider it carefully before investing.

风险包括本金损失以及与人工智能相关的风险。

Risks include principal loss and artificial intelligence related risks.

Harbor ETF由Foresight Fund Services, LLC分销。

Harbor ETFs are distributed by Foresight Fund Services, LLC.

Harbor与AI Daily Brief没有关联，这些基金与任何AI实验室也没有关联、赞助或背书关系。

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

这是一则付费广告，并非个性化投资建议。投资涉及风险，包括可能损失本金。

This is a paid advertisement and not personalized investment Investing involves risk, including possible loss of principal.

Blitzy对庞大代码库的深刻理解，能够自动完成安全修复、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，如果完全没有遗留代码，会发生什么？

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

绿地项目本该是容易的部分。

Greenfield is supposed to be the easy part.

从零开始，没有技术债务。

Clean slate, no technical debt.

但即便是全新项目，进展速度也还是人的速度，一次一个冲刺。

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

Blitzy 把工作单元从开发者变成了整个项目，自主规划、构建、测试并验证整个应用，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.

Blitzy 的一个客户搭建了一个全新的应用，有53.4万行代码，把65周的路线图压缩到了两周。

One Blitzy customer stood up a brand new application 534,000 lines of code, compressing a 65 roadmap into two weeks.

另一个客户在没有前端工程师的情况下，交付了整个应用。

Another shipped an entire application with no front end engineer.

无论是遗留代码还是全新项目，答案都一样：软件以计算的速度运行。

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

在 blitzy.com 上构建你的下一个项目。

Build what's next at blitzy.com.

网址是 blitzy.com。

That's blitzy.com.

本期《AI 每日简报》由 Hyperagent 赞助，你可以在 Hyperagent 上运行你的智能体团队，并让团队一起管理。

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.

别再依赖本地代理和聊天工作流，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在云端部署始终在线的代理，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.

营销部门的代理能把竞争对手的动作变成落地页。

Marketing's agent turns competitor moves into landing pages.

销售部门的代理能丰富线索、起草邮件并更新CRM。

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

运营代理负责跟进文书工作并跟踪预算。

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则。

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加入你的团队了。

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

去Hyperagent雇佣你的代理吧，这是Airtable团队打造的。

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每日简报。

Welcome back to the AI Daily Brief.

就像几乎不可能跟上AI新闻的狂潮一样，任何一个人也很难在某一周内跟踪所有的新功能、新模型和新工具。

Just like it is nearly impossible to keep up with the onslaught of AI news, so too is it extremely difficult for any one person to keep track of all of the new features and models and tools that come out in any given week.

大多数时候，我们的注意力只放在那些可能更根本地改变我们与AI互动方式的大型模型发布上，尽管在2026年期间，我们也越来越擅长处理更新。

Mostly our attention just goes to when there are big model announcements that might more fundamentally shift how we interact with AI, although over the course of 2026 we've also gotten a lot more adept at harness updates as well.

不过，有时候还是值得去盘点一下模型和工具公司发布的所有功能更新、生活质量提升和新能力，看看其中有没有哪些能对我们工作方式有帮助。

Still, sometimes it's worth going through and just taking an inventory of all of the feature updates and quality of life upgrades and new capabilities that the model and harness companies have released to see if any of them can be relevant for the way that we work.

基于这个想法，今天我们要过一遍，嗯，大概十几个我最近一周左右看到的更新，这些更新可能对你近期非常有用和实用。

And with that in mind, today we are going through, I don't know, a dozen updates that I have seen over the last seven days or so that are things that you might find extremely useful and practical in the immediate term.

首先，我们有一个关于Claude的更新。

First up, we have an update for Claude.

Claude Cowork 现在内置了浏览器，让基于网页的任务更方便。

Claude Cowork now has a built in browser to make web based tasks more convenient.

如果你的任务涉及网站上的操作，比如填写表单、使用网页应用，会话旁边会打开一个浏览器窗口供 Claude 使用。

If your task includes something that happens on a website, like filling out a form, interacting with a web app, a browser window will open alongside your session for Claude to use.

这个功能也可以用于代理式浏览，类似于 Perplexity 的 Comment 或 The Browser Company 的 Dia 等 AI 浏览器早期吸引人的一些功能。

The feature can also be used for agentic browsing, similar to some of the early features that caught people's attention from AI browsers like Perplexity's Comment or Dia from The Browser Company.

Claude 指出，无需安装任何新东西，它直接内置在桌面应用中，并且与你的个人浏览器和登录状态保持隔离。

Claude points out that there is nothing new to install, that it's built directly into the desktop app and stays separate from your own browsers and log ins.

换句话说，这是它自己的浏览器。

In other words, this is its own browser.

它不会接管你的浏览器。

It's not taking over yours.

不过，对于那些想留在自己浏览器里的用户，也可以通过 Claude 和 Chrome 使用这个功能。

Although for those who wanna stay in their own browser, that feature exists via Claude and Chrome.

对很多人来说，这个功能真是盼了很久了。

For many, this one was a big old finally.

Every的Dan Chipper说，他们终于做到了。

Dan Chipper from Every says they did it folks.

MercadoLibre的Santi写道，早就该这样了。

Santi from MercadoLibre writes it was about time.

非常好的功能。

Very good feature.

Vibub Sicinte指出了为什么这看起来像是一个小功能更新，但当你把它和ChatGPT的工作以及其他新东西如GrockBot结合起来时，这一切加起来就是一个相当有意义的转折点。

Vibub Sicinte gets at why although this seems like a small feature update, when you combine it with ChatGPT work and other new things like GrockBot, it all adds up to a fairly meaningful inflection.

他写道，AI在同一个月里从写文字变成了替你浏览网页，或者正如Enums所说，感觉AI助手和AI操作员之间的界限变得模糊了很多。

AI, he writes, went from writing text to browsing the web on your behalf in the same month or as Enums put it, feels like the line between AI assistant and AI operator just got a lot thinner.

Sally Stockholm写道：这感觉比只是Claude有了浏览器要大得多的更新。

Sally Stockholm writes: This feels like a much bigger update than just Claude got a browser.

有趣的是，Claude现在真的能帮你完成网页上的工作，而不只是告诉你该怎么做。

The interesting part is that Claude can now actually do the web work instead of just telling you how to do it.

需要研究点什么、浏览网站、填写表格或者完成任务吗？

Need to research something, navigate a website, fill out forms, or complete a task?

浏览器会在Cowork里打开，Claude会和你一起处理整个过程。

The browser opens inside Cowork and Claude handles the process alongside you.

我们正在慢慢告别那种只给答案的AI。

We're slowly moving away from AI that just gives answers.

下一阶段是AI直接打开标签页，把事情搞定。

The next phase is AI that opens the tab and gets the job done.

不过也有人指出，Anthropic这次来得有点晚。

Now as some pointed out, Anthropic is at least a little late to the party here.

TokenGremlin说OpenAI早就有了类似功能，ChatGPT工作区已经用了好几周了。

TokenGremlin says OpenAI has already had basically this and ChatGPT work for weeks.

Julia的CEO Rahul提到了他们七月份发布的浏览器代理，基本上就是说欢迎加入。

Julia's CEO Rahul points to the browser agent that they released back in July and basically said welcome to the party.

不过对很多人来说，这个功能明显的灵感来源是GrockBot。

Still for many, the obvious inspiration for this was GrockBot.

GrockBot发布的时候，能在自己的浏览器里工作这一点，是大家觉得特别强大、而且很可能被其他实验室模仿的功能之一。

When GrockBot was released, the ability for it to work in its own browser was one of the things that people latched onto as really powerful and frankly most likely to be copied by all the other labs.

马丁·卡萨多说，我觉得ChatGPT的工作方式更像高级的RPA。

A 16 Martin Casado writes, I find ChatGPT work more like fancy RPA.

它运行得不错，但更偏向流程导向。

It works well, but it is more workflow oriented.

而GrockBot真的就像一个虚拟的Go工作者。

Work Rockbot really is a virtual go worker.

给它一个自己的浏览器，通过Slack聊天，它能很好地适应你的风格和偏好。

Give it its own browser, chat over Slack, it does a great job adapting to your style and preferences.

关于ChatGPT版本的这项功能，官方ChatGPT账号写道，你可以让它为新公寓设置水电服务，预约DMV或护照办理，通过保险查询报销费用，根据你的时间找到并预约网络内的医生，查看你的汽车注册何时到期并准备续期文件，比较你的租房保险政策和你正在给房东发的邮件中提到的问题，找到并保存符合你条件的公寓列表，只需上传照片就能补货，等等等等。

Now in terms of ChatGPT works version of this feature, the official ChatGPT account, they write, you can ask it to set up utilities for a new apartment, book a DMV or passport appointment, check reimbursement costs through your insurance, find and book an in network doctor around your availability, check when your car registration expires and prepare the renewal paperwork, compare your rental insurance policy with an issue you're emailing your landlord about, find and save apartment listings that match your criteria, restock something just by uploading a photo, and so on and so forth.

还有大约十几个其他想法。

There are about a dozen other ideas.

鉴于我们一直在讨论AI在生活和工作之间的平衡，我——

Given our ongoing conversations about the balance between AI for life and AI for work, I.

嗯。

E.

消费级AI和企业级AI的对比，我觉得有意思的是，他们把这些更偏向生活化的任务作为ChatGPT Work的新功能来推广。

Consumer AI versus B2B AI, it's interesting to me that they're pitching all of these much more life oriented tasks as a new capability for ChatGPT Work.

OpenAI的开发者体验负责人Romaine Hewitt写道：ChatGPT Work现在在云端有自己的电脑了。

OpenAI's head of developer experience Romaine Hewitt writes: ChatGPT Work now has its own computer in the cloud.

当它能上网并为你采取行动时，可能性几乎变得无限。

When it can use the web and take action for you, the possibilities start to feel almost limitless.

另外，ChatGPT还有一个小的生活质量升级，公司最近意识到，他们需要一种更好的方式让人们能在临时性和永久性之间切换。

Now one other small quality of life upgrade from ChatGPT, the company realized recently that they need a better way for people to switch back and forth between temporality and permanence.

他们在X上发帖说：我们听到了你们的反馈，有时候临时聊天会变成值得保存的内容。

On x, they posted, we heard your feedback that sometimes a temporary chat becomes something worth saving.

现在你可以选择用你现有的记忆、自定义指令和插件来个性化临时聊天，如果你想保留它，也可以保存到历史记录中。

Now you can choose to personalize a temporary chat with your existing memories, custom instructions, and plugins, and save it to your history if you want to keep it.

临时聊天不会创建新的记忆，除非你选择保存，否则它们不会出现在侧边栏中。

Temporary chats don't create new memories and stay out of your sidebar unless you choose to save them.

换句话说，对于那些对隐私有些担忧、可能更倾向于默认使用临时聊天功能而不是默认保存所有内容的人来说，这个功能现在让临时聊天也能轻松引入相关的ChatGPT上下文，并且在合适的时候升级为非临时聊天。

In other words, for people who have a little bit more concern around privacy and who would want to be perhaps default using the temporary chat feature as opposed to saving everything that they do by default, this feature now makes it easier to pull in relevant ChatGPT context even into temporary chats, as well as to upgrade them to non temporary when it makes sense to.

还有一个重要的体验升级，说实话，这个升级在新闻价值上绝对会远超它的分量，我是在开始录制前才注意到的。

Another big quality of life upgrade, one which will frankly punch way above its weight class in terms of its newsworthiness, I noticed just before I started recording.

OpenAI的Jason写道，现在插件里可以连接多个Gmail账户了。

Jason from OpenAI writes, multiple Gmail accounts available in plugins now.

也就是说，不再只能连接一个Gmail，你可以连接多个。

In other words, instead of connecting only one Gmail, you can connect multiple.

这长期以来一直是很多重度用户的心愿单上的项目。

This has long been on many power users' wish lists.

考虑到像我们这样的人，日常工作中通常会跟五六个不同的邮箱打交道，How I AI的Claire Vaux最近就特别兴奋地说，这其实是她最喜欢的新Grock Bot功能，因为它不限制你只能连接一个Gmail或一个Slack。

Given that the normal course of business for people like this, folks, is to have, I don't know, a half dozen different emails that you interact with in some meaningful way, How I a i's Claire Vaux was recently raving about how this was, in fact, her favorite feature of the new Grock Bot that it didn't constrain connectors to a single Gmail or a single Slack.

哎呀，这个功能我不管是谁先做的，只要大家都做就行。

And, boy, howdy, is this one where I don't care who did it first as long as they all do it.

说到这个，很多人其实很乐意使用大公司提供的任何工具，不管是OpenAI、Anthropic、SpaceX AI、谷歌还是其他什么公司。

Speaking of that, many of you will always be quite happy to use whatever tools the big companies give us access to, whether it's OpenAI or Anthropic or SpaceX AI or Google or whomever.

但对于那些想要更多控制权，或者可能被OpenClaw这样的工具所赋予的灵活性吸引的人来说，关注一下Hermes agent是值得的。

But for those who want more control and who maybe got hooked on the flexibility that something like OpenClaw enabled, it's worth keeping an eye on the Hermes agent.

Hermes团队显然在密切关注所有工具和模型公司的一举一动，因为只要那些公司一发布什么有用的东西，很快就会出现在Hermes上。

The team over at Hermes is clearly watching everything that the harness and model companies are doing because as soon as anything useful gets announced from those companies, inevitably it finds its way over to Hermes extremely quickly.

比如，这周四他们宣布，Hermes代理现在可以无缝地替你浏览网页。

So for example, on Thursday this week, they announced that the Hermes agent can now seamlessly browse as you.

他们写道，开启真实配置文件浏览后，你的代理会使用你现有Chrome配置文件的副本，以你的登录状态来操作。

They write, turn on real profile browsing and your agent acts with your logins from a managed copy of your existing Chrome profile.

这大约是在十天前，也就是Grock Bot发布之后，他们为Hermes桌面版推出了Bot模式。

This follows from about ten days ago right after the Grock Bot announcement when the company introduced Bot Mode for Hermes desktop.

他们写道，在Bot模式下，你的代理配置文件会变成一系列命名的机器人。

With Bot Mode they write, your agent profiles become a series of named bots.

每个机器人都有自己的角色、模型、记忆、技能和头像。

Each bot has its own role, model, memory, skills, and profile picture.

机器人可以使用任何模型，甚至还能互相交流。

Bots can use any model and even communicate with each other.

一次构建一个专业机器人，就能永久使用。

Build a specialist bot once to use it forever.

为了说明为什么这可能对某些人有吸引力，制作Hermes的新闻研究账号写道：Hermes Agent也有自己的远程电脑，只是你不一定非要用它，而且每月只需几美元，而不是200美元。

Articulating why this might be of interest to some folks, the news research account who makes Hermes writes: Hermes Agent has its own remote computer too, you just aren't locked in to using it and it costs a few dollars per month instead of 200.

还有其他一些重大区别吗？

Some other big differences?

可以连接任何提供商、订阅或模型，完全免费在你自己的机器上使用，或者使用News Portal上的几个免费模型之一，在机器人模式和经典会话模式之间切换，自我改进的代理能学会更便宜、更有效地执行任务，开源所以你可以随意修改代理本身以满足你的需求。

Connect any provider, sub, or model, use it completely free on your own machine or with one of several free models on News Portal, switch between bot mode and classic sessions mode, self improving agents that learn to execute tasks cheaper and more effectively, open source so you can modify anything you want about the agent itself to suit your needs.

再说一次，这对每个人来说都不是必需的，但我认为从消费者选择的角度来看，Hermes确实在紧跟所有主要框架的步伐，随着这些新功能升级和能力的上线。

Again, this is not going to be necessary for everyone, but I think it is pretty rad from a consumer choice perspective that Hermes is very much keeping pace with all of the major harnesses as these new feature upgrades and capabilities come online.

顺便说一句，Hermes不只是抄袭，他们自己也在做一些非常酷的实验。

And by the way Hermes isn't just copying things, they're also doing some really cool experimentation on their own as well.

他们有Hermes HUD模式，基本上是一个无缝覆盖层，你可以在做其他事情（比如玩游戏）时跟踪工作生活中的事情。

They have a Hermes HUD mode which is basically a seamless overlay where you can keep track of things going on in your work life as you're doing other stuff like playing a game.

官方新闻账号写道：我现在整天玩游戏，在那个小文本框里完成工作。

The official news account writes: I just play games all day and get my work done in that little text box now.

如果说Hermes在光谱的一端，那么Salesforce和Claude的新合作肯定在另一端。

Now if Hermes is on one end of the spectrum, the new collaboration between Salesforce and Claude has to be on the other.

Salesforce和Anthropic正在合作，他们把这个合作叫做，听好了，Claude Force。

Salesforce and Anthropic are teaming up for a partnership that they are calling, wait for it, Claude Force.

本质上，这是一套新的插件和技能，让Salesforce用户可以通过Claude代理访问他们的CRM。

At its core, this is a new set of plugins and skills to allow Salesforce users to access their CRM using Claude agents.

他们可以用Salesforce的数据作为上下文，而且得益于Salesforce的AI框架，他们称之为AI Force，可以在尊重现有治理护栏的同时执行代理操作。

They can use Salesforce data as context and thanks to Salesforce's AI harness they call AI Force, agentic actions can be taken while respecting existing governance guardrails.

在他们的新闻稿中，Salesforce写道：几十年来，企业软件要求用户手动浏览静态界面来完成工作。

In their press release Salesforce writes: For decades enterprise software required users to manually navigate static UI to get work done.

现在，代理可以直接访问数据、工作流程和规则，软件变成了一个驱动所有界面的系统。

Now that agents can access the data, workflows and rules directly, software is a system that powers every interface.

这一举措凸显了我们与CRM等系统交互方式的转变，让代理能更自然地做到这一点。

The move underscores that there is an shift in how we interact with things like CRM, making it easier for agents to more natively do so.

Salesforce首席执行官马克·贝尼奥夫说：在这里，界面就是AI，让你能动态构建自定义应用，并回答任何企业问题。

Salesforce CEO Mark Benioff said: Here, the UI is the AI, allowing you to build custom apps dynamically and answer any enterprise question.

贝尼奥夫继续说：仅靠概率智能无法运营一家公司，确定性系统也无法推理。通过将Claude非凡的推理能力与企业所依赖的可信数据、工作流程和治理相结合，我们正在提供一个能思考、推理和行动的动态界面。

Probabilistic intelligence alone doesn't run a company, and deterministic systems don't reason Benioff continues: By fusing Claude's extraordinary reasoning with the trusted data, workflows, and governance every enterprise runs on, we're delivering a dynamic interface that thinks, reasons, and acts.

这就是未来所有企业的运作方式。

This is how every business will run.

而且没错，他们还为此拍了一支新的马修·麦康纳广告。

And yes, it even got a new Matthew McConaughey ad.

对有些人来说，尤其是当达里奥和贝尼奥夫一起出现在CNBC上时，这成了开玩笑的好时机。

Now for some, especially as Dario appeared next to Benioff on CNBC, this was a moment for cracking jokes.

企业家Varun Ram Ganesh写道：Anthropic在融资时说，我们的市场是美国经济，我们是终极智能，我们要干掉整个应用层，我们要造成50%的失业率，让整个人类劳动力变得多余。

Entrepreneur Varun Ram Ganesh writes: Anthropic while raising VC money, our TAMs The US economy, we are the ultimate intelligence, we are killing the entire app layer, we are going to create 50% unemployment, and make the entire human workforce redundant.

而现实中的Anthropic：很高兴宣布与Salesforce合作推出Claude Force。

Anthropic in reality: Excited to announce Claude Force in partnership with Salesforce.

但这种玩笑大大低估了企业界对Salesforce这类工具的依赖程度。

But this sort of joking massively undersells just how much of the enterprise world actually runs on things like Salesforce.

Thomas Fanning写道：Claude Force这个动作足够引人注目，让我重新思考之前的判断。

Thomas Fanning writes: Claude Force is notable enough to make me rethink my priors.

竞争信号正在发生变化。

The competitive signal is shifting.

除了头条新闻之外，我觉得最值得注意的是，Anthropic的首席商务官说，CRM集成只是众多合作中的第一个。

Beyond the headline, what I find most notable is that the Anthropic CCO saying that CRM integration is the first of many.

实际上，软件供应商本来就应该一直是前沿实验室最大的客户。

Realistically, software vendors always should have been the Frontier Labs' largest customers.

但到目前为止，看起来Anthropic和OpenAI似乎决定直接面向客户。

Up to now, however, it seemed like Anthropic and OpenAI had decided to go direct instead.

他们的策略正在改变，在一个动态的市场中，当事实发生变化时，观点也必须随之改变。

Their strategy is changing and in a dynamic market, when the facts change, view has to change.

Anthropic显然意识到，在企业中，代理护栏、治理和基础设施还不存在，而且未来一到两年内也不会存在，因此很难直接展示软件。

Anthropic is clearly recognizing that it's hard to display software outright when the agentic guardrails, governance and infra doesn't exist in the enterprise and won't for another one to two years.

现在的问题是：客户是否仍然希望将数据从记录系统中移出，以原生方式构建并避免支付软件平台API费用，还是他们会接受在现有软件基础设施上启用AI的实用性？

The question now is: Do customers still want to move data out of systems of record to build natively and avoid paying software platform API fees, or will they accept the utility of enabling AI on an existing software infrastructure?

叙事正在转变，直觉上，这些原生集成应该能为客户带来他们想要的AI价值，而无需付出太多努力。

The narrative is shifting and intuitively these native integrations should deliver customers the value they want from AI without as much of a lift.

鉴于这一叙事发展的规模，目前很难对企业软件持悲观态度。

Hard to be bearish on enterprise software at present given the magnitude of this narrative development.

数据、深度嵌入的工作流程、专业知识和市场地位，仍然是软件的核心优势，而Anthropic发出的这个市场信号，也承认了这些优势的持久性。

Data, operationally embedded workflows, expertise, incumbency are still the motes for software and Anthropic is sending a market signal acknowledging the persistence thereof.

而且你看，虽然这只是几个零星的例子，但我跟一些真正使用Salesforce这类系统的人聊过这个Claude Force的合作，他们都非常兴奋。至少有一位原本打算自己开发替代Salesforce的方案，或者至少想探索一下这条路的人，基本上就停下来了，想先看看这个方案是不是就能满足他们的需求，值不值得花这个钱。

And look, obviously it's just a handful of anecdotal examples, but the folks that I spoke to about this Claude Force deal that actually interact with systems like Salesforce were extremely excited about it, and at least one who had been planning on potentially building some alternatives to Salesforce, or at least exploring down that path, basically threw the brakes on that so they could figure out if this just did what they wanted in a way that it justified the cost.

所以，虽然X上的那些网友可能觉得好笑，但在真正的企业界，我想你们有些人现在心里想的是，这是我到目前为止说的所有事情里最重要的一笔交易。

So while the x posters may get their yucks, over in the real enterprise world, some of you I imagine are sitting here thinking that this is the biggest deal of anything I've said so far.

我们已经聊了很多功能了，但还有几个模型值得一提，这周它们主要集中在音频和视频方面。

Now already we've gone through a ton of features, but there are a couple models that are worth noting, and this week they were clustered around audio and video.

Gemini这周又上了新闻，因为他们发布了新版的语音转文字模型Gemini 3.5 Transcribe。

Gemini was back in the news this week with the release of a new version of their speech to text model Gemini 3.5 Transcribe.

这个模型支持超过85种语言，谷歌说他们在易用性上取得了很大的进步。

The model supports more than 85 languages and Google says they've made some huge strides in usability.

这个模型可以逐字转录，如果你需要直接转写的话；或者你也可以用智能翻译模式，它会去掉语气词，精简啰嗦的表述，处理自我纠正，最后给你一个干净利落的用户意图版本。

The model can function in verbatim mode if you need a direct transcription, or you can use it in smart translate mode which strips away filler words, condenses rambling thoughts, and handles self corrections to deliver a clean version of user intention.

你甚至可以自定义词汇表，让模型更好地处理专业术语和俚语。

You can even create custom vocabulary to enhance the model's ability to deal with jargon and slang.

谷歌表示，这个模型在嘈杂环境中表现很好，还能区分最多三个不同的说话人。

Google says the model performs well in noisy environments and can distinguish between up to three different speakers.

如果你喜欢用语音和AI互动，这个模型可以在谷歌的AI套件里使用，同时也提供API，供第三方应用调用。

The model can be used across Google's AI suite if voice mode is your preferred way of interacting with AI and is also available as an API to power third party apps.

这其实和WhisperFlow这类产品很像，但它不只是做一个终端应用，更是为其他人提供的基础设施，让他们可以在上面继续开发。

This is then playing very much in the space of something like WhisperFlow, although built not just to be an end application in itself, but also a key piece of infrastructure for other people who can build on top of it.

AI开发者Jan Kronberg写道：这种发布很多人可能会错过。

AI developer Jan Kronberg wrote: This is the kind of release a lot of people are going to miss.

以前，语音输入基本上要求你像写文章一样说话，要有标点、结构清晰，还要尽量减少中途修改。

Until now, voice input required you to basically speak the way you write punctuation, well structured sentences, fewer corrections along the way.

Gemini 3.5 Transcribe把这个完全颠倒过来了。

Gemini 3.5 Transcribe flips that around.

你按自己真实的思考方式说话，系统会生成你本来想写的文字，边听边帮你整理和清理。

You speak the way you actually think and the system produces the text you probably wanted to write, cleaning it up and organizing it as you speak.

顺便说一句，要把这种清理做好真的很难。

Now, as an aside, it is really hard to do that cleanup well.

比如说，我在WhisperFlow上经常在轻度清理和无清理两个版本之间来回切换。

For example, I am constantly switching back and forth on WhisperFlow from the version where it does light cleanup to the version where it does no cleanup.

无清理版本看起来确实像没有标点符号的漫谈，有时会变得不清楚，但即使在最轻的清理设置下，WhisperFlow也会把我用作连接词而不是填充词的词语当作填充词去掉。

No cleanup really does look like rambles without good punctuation and sometimes gets unclear, but even on the lightest cleanup setting, WhisperFlow strips away words that I am using as connectors rather than filler assuming that they are filler.

无论如何，就因为这个原因，我很期待看看谷歌在这里能提供什么不同的东西。

In any case, for that reason alone, I'm excited to see if Google offers something different here.

Space xAI也发布了Grok Voice Think Fast 2.0，并指出它现在在人工智能分析语音到语音指数中排名第一。

Space xAI also dropped Grok VoiceThink Fast 2.0, noting that it's now number one on the artificial analysis speech to speech index.

他们写道，这个指数衡量的是语音代理能否对听到的语音进行推理，解决真实的客户问题，并正确使用代理工具完成任务。

They write, this index measures whether voice agents can reason over the speech it hears, resolve real customer issues, and correctly complete tasks using agent tools.

他们继续说，在Starlink，我们每天使用Grok Voice处理超过15000个入站客户支持和销售电话。

At Starlink, they continue, we're using Grok Voice to resolve over 15,000 inbound customer support and sales calls a day.

Grok诊断硬件问题，发送替换件，并通过语音电话和聊天每周完成超过3000个订单。

Grok diagnoses hardware issues, ships replacements, and fulfills over 3,000 orders a week via voice calls and chat.

很多人展示了演示，探索速度和延迟的升级究竟如何改变了在代理环境中使用语音的可能性。

A bunch of people showed demos, exploring just how much the speed and latency upgrades really change what you can do with voice in the agentic context.

不过，我们还从谷歌那里得到了一个新的视频模型。

We also, however, got a new video model from Google as well.

这个模型叫 Gemini Omni 1.1 Flash，他们写道，通过这次更新，你可以扩展场景、指定镜头的起始和结束帧、添加视频输入参考、把喜欢的片段升级到4K，还能在360p下快速测试想法。

The model is called Gemini Omni 1.1 Flash, and they write that with this update, you can extend your scenes, specify starting and ending frames of a shot, add video input references, upscale your favorite takes to four ks, and test ideas quickly in 360p.

换句话说，这次更新百分之百不只是关于所谓的“更好的视频”，而是大大提升了视频生成的实用性。

In other words, this is 100% an update that is not just about quote unquote better video, it's about massively more usability of video generation in general.

现在，我们又得到了一些新的基准测试结果。

Now once again we got some new benchmarks.

Arena 表示，Omni 1.1 Flash 在文本转视频的竞技场中排名第一，在图像转视频的竞技场中排名第二，不过对于新视频模型来说，效果好不好还得看实际表现。

Arena says that Omni 1.1 Flash hit number one in the text to video arena and number two in the image to video arena, although when it comes to a new video model, the proof is in the pudding.

看起来人们的第一印象是，这确实是对之前 Omni 模型的一次升级。

It seems like people's first impression is that this is indeed an upgrade over the previous Omni model.

Choblin 写道：最大的升级是语音，听起来明显更自然，而且生成的视频也比 Omni Flash 更逼真。

Choblin writes: The biggest upgrade is the voice and it sounds noticeably better, more natural, and it also generates more realistic videos than Omni Flash.

不过也有人发现，当把它和 Seed Dance 2.5 比较时，它并不算明显的改进，甚至可能根本没有改进。

Some found however that when they compared it to Seed Dance 2.5, it wasn't as clearly an improvement if it was an improvement at all.

在视频方面，Fall还推出了H3 Max，他们宣传说这个模型对提示的理解更好，而且速度也快得多。

On the video front, Fall also introduced H3 Max, which they pitched as having better prompt understanding and being just way, way faster.

他们声称平均延迟是3.49秒，相比之下，C Dance 2.5是68秒，Gemini OmniFlash是26.7秒。

They claimed an average latency of three point four nine seconds compared to C Dance two point five sixty eight seconds and Gemini OmniFlash's twenty six point seven seconds.

Andreessen Horowitz的Justine Moore写道：这是Foul推出的一个非常厉害的视频模型。

Justine Moore from Andreessen Horowitz writes: This is an insane model from Foul.

我无法形容在短短15秒内生成一个长视频片段的感觉有多震撼。

I can't describe how wild it feels to generate a long clip in just fifteen seconds.

这将会彻底改变这个领域。

This will be transformative for the space.

Adam Holter写道：细节上有点模糊，但这个模型在1.3秒内就生成了，成本只要20美分。

Adam Holter writes: Very fuzzy on the detail here but this generated fast in one point three seconds for 20¢.

这太离谱了。

That's ridiculous.

H3 Max确实是个大事件。

H three max is a big deal.

Alex，那个做AI广告的家伙写道：H3 Max的初步印象。

Alex the AI ads guy writes, h three max first impressions.

生成时间只要十五秒。

Fifteen second generation time.

音频效果太棒了。

Audio is amazing.

声音听起来很好，音效也很棒。

Voices sound good and sound effects too.

对提示词的遵循度简直不可思议。

Incredibly prompt adherent.

你让它做什么，它就做什么。

It does exactly what you tell it to.

还需要更多测试，但H3 Max在平衡性和价格上很难被超越。

Needs more testing, but it's hard to beat the balance and price of h three max.

Ethan Moloch写道：AI视频的一个新里程碑被突破了。

Writes Ethan Moloch, a line in AI video was crossed.