AI的OpenClaw化

今天在AI每日简报中，我们将探讨AI的开放资格问题，在此之前，先来看头条新闻：Anthropic与五角大楼的争端正在升级。

Today on the AI Daily Brief, the open qualification of AI, and before that in the headlines, the fight between Anthropic and the Pentagon heats up.

《AI每日简报》是一档每日更新的播客和视频节目，聚焦AI领域最重要的新闻与讨论。

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

好了，朋友们，在正式开始前，先做几个快速通知。

Alright, friends, quick announcements before we dive in.

首先，感谢今天的支持方：毕马威、AIUC、Robots and Pencils 和 Blitzy。

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

如需收听无广告版本，请前往 patreon.com/aideallybrief，或在Apple播客上订阅。

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

如果您有兴趣了解如何赞助本节目，请发送邮件至 sponsorsaidailybrief dot ai。

If you are interested to learn more about sponsoring the show, send us a note at sponsorsaidailybrief dot ai.

今天最后一点说明：非常抱歉昨天错过了节目。

And last note today: big apologies for missing the show yesterday.

我们从南美洲返程的经历比预期长得多、也复杂得多，不仅因为东海岸的暴风雪，还因为在巴西马瑙斯发生了紧急下机事件，最终导致整个旅程从出发到抵达耗时超过55小时。

Our journey back from South America ended up being significantly longer and more complicated than we expected, not only because of the East Coast snowstorms but because we also had an emergency deplaning in Manaus, Brazil that led ultimately to the entire trip being more than fifty five hours door to door.

因此，这周我们会把周三的节目当作休息日，周六照常播出，大家准备好周末收听吧。

So what we're going to do this week is treat the Wednesday show as the off day and we will have a normal show on Saturday, so get prepared for some weekend listening.

不过，我们还有很多内容要跟进，那么我们就开始吧。

Still, we have a lot to catch up on, so with that, let's dive in.

今天我们首先讨论一个完全可以单独做成一期主节目的话题。

We kick off today with a topic that could easily be a main episode.

五角大楼向Anthropic发出最后通牒，这可能是迄今为止最重要的AI安全之争。

The Pentagon has handed Anthropic an ultimatum in what could be the most crucial battle for AI safety to date.

本周初，我们陆续收到报道，称Anthropic与美国国防部（DOD，具体名称因人而异）的合同因AI准备问题而面临风险。

Now, heading into this week, we had building reports that Anthropic's contract with the Department of Defense or the DOW, depending on who you ask was in jeopardy over AI readiness.

Anthropic坚持认为，其技术不应被用于驱动自主武器或对美国民众进行监控。

Anthropic was insisting that their technology should not be used to power autonomous weaponry or for the surveillance of Americans.

在某些报道中，似乎Anthropic签约的前提是其技术不会被用于任何致命性交战。

At certain points in the reporting, it seemed like Anthropic had contracted on the basis that their technology would not be used to assist in any lethal exchange.

另一方面，五角大楼则坚持认为，科技公司不应干预技术在军事行动中的使用方式。

The Pentagon, on the other hand, was insisting that tech companies should not have input on how technology is used in military operations.

他们一直推动每个AI实验室在其模型使用条款中加入‘所有合法用途’的表述，本质上是主张战争行为的限制应仅由美国法律决定，而非科技公司的政策。

They have been pushing for Each AI Lab to provide their models with terms that allow, quote, all lawful use, basically establishing the position that only US law rather than tech company policy should be the limiting factor in the conduct of war.

这一冲突当然源于报道称Anthropic的AI模型被用于抓捕委内瑞拉总统尼古拉斯·马杜罗的突袭行动。

The conflict had, of course, erupted after reports that Anthropic's AI models had been used during the raid that captured Venezuelan president Nicolas Maduro.

关于AI实际参与的程度，报道并不清晰，但看起来Anthropic对其技术被用于一场最终导致致命后果的行动感到愤怒。

The reporting has been unclear to what extent AI was used, but it appeared that Anthropic was furious that their technology was used in what was ultimately a lethal operation.

在这种背景下，Anthropic的高管于周二被召至五角大楼开会以解决争端，但会议进行得并不顺利。

In that context, Anthropic executives were summoned to a meeting at the Pentagon on Tuesday to settle the dispute, and things did not go well.

国防部长佩特·海格塞斯亲自向Anthropic首席执行官达里奥·阿马德发出最后通牒，要求其在周五前同意五角大楼的条款，否则将被列入政府黑名单。

Defense secretary Pete Hegseth personally gave Anthropic CEO Dario Amade a Friday deadline to agree to their terms or face a government blacklist.

这不仅仅是他们会终止合同的问题。

And this is not just they would drop their contract.

威胁是Anthropic将被列为供应链风险企业，并被禁止参与任何军事承包商的项目。

The threat was that Anthropic would be designated a supply chain risk and barred for use among military contractors as well.

这种认定此前仅适用于华为等外国公司。

That designation had previously only been applied to foreign companies such as Huawei.

新的报道进一步阐明了各方的分歧。

New reporting substantially fleshes out the conflicting views.

熟悉Anthropic立场的消息人士表示，他们相信自己的AI模型足够可靠，可以在无人干预的情况下操作武器。

Sources familiar with Anthropic's position said they are confident that their AI models are reliable enough to operate weapons without a human in the loop.

他们还指出，目前尚无任何法律规范AI在国内监控中的使用方式。

They also note that there are no current laws dealing with how AI can be used in domestic surveillance.

据报道，黑格塞斯告诉Anthropic，他们必须在周五凌晨5点01分前表态，要么接受，要么拒绝。

Hegseth reportedly told Anthropic that they have until 05:01 on Friday to get on board or not.

他还警告Anthropic，他会确保《国防生产法》被应用于Anthropic，强制其为五角大楼提供服务，无论他们是否愿意。

He also warned Anthropic that he would ensure that, quote, the Defense Production Act is invoked on Anthropic, compelling them to be used by the Pentagon regardless of whether they want or not.

大多数人认为，这意味着五角大楼将强迫Anthropic提供一个移除了安全限制的Claude版本。

Most are taking this to mean the Pentagon would compel Anthropic to provide a version of Claude without guardrails trained into the model.

《国防生产法》确实允许美国。

The Defense Production Act does allow the U.

S。

S.

政府有权为被视为对国家安全至关重要的商品和服务规定合同条款。

Government to dictate contract terms on goods and services deemed critical to national security.

然而，该法案很少在企业反对的情况下被启用，通常用于命令整个行业优先生产特定商品以满足政府订单。

However, it has rarely been invoked against the will of individual companies and is more typically used to order entire sectors to prioritize the production of certain goods to fulfill government orders.

例如，在2021年加州野火期间，该法案被启用以增加消防水带的生产，在疫情期间则用于增加呼吸机的供应。

For example, it was invoked during the twenty twenty one California wildfires to increase the production of fire hoses and during the pandemic to boost the supply of ventilators.

前司法部与国防部联络官凯蒂·斯威滕指出，海格声称这些威胁有些自相矛盾，他说：‘我假设我们不希望使用存在供应链风险的技术，对吧？’

Katie Sweeten, a former liaison for the Justice Department to the Department of Defense, noted that Hague says threats are kind of contradictory, telling I would assume we don't want to utilize the technology that is a supply chain risk, right?

所以我不知道如何调和这一点。

So I don't know how you square that.

听起来更像是，供应链风险并非真正的理由，而更像是一种惩罚，因为他们没有屈服。

What it sounds like is that the supply chain risk may not be a legitimate claim, but more punitive because they're not acquiescing.

尽管气氛紧张，但消息人士表示，会议仍保持礼貌，没有提高嗓门。

Despite being tense, sources said the meeting was cordial and that there were no raised voices.

赫塞特甚至称赞了Anthropic的模型，这一直是该报道中的一条潜在线索。

Hexath even praised Anthropics models, which has been one of the undercurrents of the story.

五角大楼似乎认为，Anthropic的技术远远领先于竞争对手，无法被其他供应商取代。

The Pentagon seems to believe that Anthropics technology is so far ahead of the competition that it can't be replaced by another supplier.

五角大楼与OpenAI、谷歌和xAI都有长期合同。

The Pentagon has standing contracts with OpenAI, Google, and xAI.

然而，只有xAI同意了在机密和非机密环境中都适用的合法使用标准。

However, xAI is the only one to agree to all lawful uses standard across both classified and unclassified settings.

根据Axios的报道，Claude是目前唯一获准在机密环境中使用的模型。

Claude is the only model currently cleared for use in classified settings according to Axios.

熟悉Anthropic内部想法的消息人士表示，他们没有计划向五角大楼的要求屈服。

Sources familiar with Anthropic's thinking said that they have no plans to buckle to the Pentagon's demands.

他们的官方声明礼貌而坚定，评论道：我们继续本着诚意就使用政策进行对话，以确保Anthropic能够继续支持政府的国家安全任务，且符合我们模型可靠且负责任的能力范围。

Their official statement was polite but firm, commenting: We continued good faith conversations about our usage policy to ensure Anthropic can continue to support the government's national security mission, in line with what our models can reliably and responsibly do.

与此同时，五角大楼的一名发言人表示：这与监控或自主武器的使用无关。

Meanwhile, a Pentagon spokesperson said: This has nothing to do with surveillance and autonomous weapons being used.

你不能靠例外来领导战术行动。

You can't lead tactical ops by exception.

截至周三，Axios报道称，五角大楼正在采取初步措施，将Anthropic列为供应链风险。

As of Wednesday, Axios reports that the Pentagon is taking preparatory steps towards labeling Anthropic a supply chain risk.

他们已联系波音和洛克希德·马丁等主要承包商，以确定这些公司是否对Anthropic的技术存在关键依赖。

They've reached out to major contractors including Boeing and Lockheed Martin to determine if they have any critical dependencies on Anthropic technologies.

这之所以如此重要，正如我所说，完全足以成为一集独立报道，甚至未来可能确实需要如此，因为牵涉的远不止Anthropic的两亿美元合同。

Now the reason that this is so important and like I said could easily be a main episode and who knows might need to be at some point is that there's much more than Anthropoc's $200,000,000 contract at stake.

这场争端直指过去十年互联网上激烈争论的AI安全核心议题。

The dispute goes right to the heart of the debates around AI safety that have been raging on the internet for the past decade.

我们讨论的不再是假设性问题，而是一个非常实际的问题：AI实验室与美国政府，谁应该对军用AI的部署拥有最终决定权。

We are no longer talking about hypotheticals, but a very practical question of whether AI labs or the US government should have the final say on how military AI is deployed.

Symaphore撰稿人Reid Alberghati评论道：这场风波最可能的结局是，五角大楼将利用《国防生产法》迫使Anthropic屈服。

Symaphore writer Reid Alberghati commented: The most likely conclusion of this saga is that the Pentagon forces Anthropic to comply using the Defense Production Act.

在某些方面，这意味着双方都能如愿以偿：五角大楼不再受限于如何使用Claude，而Anthropic也能保全面子，即使最终还是遵从了五角大楼的要求。

In some ways, this means both sides get what they want: the Pentagon no longer has restrictions on how it can use Claude, and Anthropic gets to save face even if it ends up complying with the Pentagon's demands.

对一些人来说，这将是一个非常令人不满的结局。

For some, this would be a very unsatisfying conclusion.

LawFair Media撰文称：军队使用本世纪最具变革性技术的条款，正由国防部长与一家初创公司首席执行官通过双边谈判决定，既无民主参与，也无持久约束。

LawFair Media wrote: The terms governing how the military uses the most transformative technology of the century are being set through bilateral haggling between a defense secretary and a startup CEO, with no democratic input and no durable constraints.

这些规则应由国会制定，而且必须尽快完成。

Congress should be setting these rules, it should do so in a hurry.

显然，这一事件仍在持续发展，我们将持续关注。

Obviously, story continues to evolve and is something we will watch.

但目前，我们转向Anthropic的主要竞争对手OpenAI，报道称其‘Stargate’项目已陷入停滞。

But for now, we move over to one of Anthropic's big competitors in OpenAI, where reporting suggests that their Stargate project has stalled out.

‘Stargate’项目于去年一月公布，是引发人工智能基础设施热潮的首批关键事件之一。

Project Stargate was announced last January and was one of the first and catalyzing events that kicked off the AI infrastructure boom.

萨姆·阿尔特曼前往白宫，宣布了价值5000亿美元的数据中心投资计划。

Sam Altman went to the White House to announce the half $1,000,000,000,000 data center investment.

出席活动的还有软银的孙正义和甲骨文的拉里·埃里森。

Also in attendance were SoftBank's Masayoshi Son and Oracle's Larry Ellison.

‘Stargate’为今后的数据中心投资定下了基调。

Stargate set the tone for data center investments moving forward.

在此之前，超大规模云服务商一直持观望态度，谨慎避免过度建设。

Prior to that, hyperscalers had sort of been hedging their bets, being cautious of overbuilding.

随着宣布投入5000亿美元，萨姆·阿尔特曼宣称他将按文明规模构建AI算力，并要求其他人跟上步伐。

With the announcement of half $1,000,000,000,000 in spending was Sam Altman's declaration that he was going to build AI compute on a civilizational scale and everyone else should get with the program.

然而，尽管叙事宏大，该项目从一开始就面临挑战。

And yet, despite the grandiose narrative, the project had challenges from the start.

事实上，在项目宣布当天，埃隆·马斯克发推称：他们根本没有这笔钱。

Indeed, the day it was announced, Elon Musk tweeted: They don't actually have the money.

尽管我们知道埃隆在这里有特定的立场，但至少部分情况确实如此，后来的报道显示，该项目在融资方面遇到困难。

And although we know Elon has a particular axe to grind here, that did seem to at least be partially the case, with reporting later in the year suggesting the project was struggling to get financing in place.

然而，如今最新消息表明，整个Stargate项目已陷入停滞。

Now, however, the information reports that the entire Stargate project is stalling out.

他们的消息来源称，这家合资企业至今尚未完成团队组建，也未有任何以Stargate名义进行的数据中心建设。

Their sources claim that the joint venture still hasn't been staffed up and that there is no data center development under the Stargate banner.

事实上，消息来源称，由于缺乏领导力和协调，这家合资企业在宣布后数周内就开始瓦解。

In fact, the sources state that the joint venture began falling apart within weeks of the announcement due to a lack of leadership and coordination.

OpenAI、甲骨文和软银无法就各自职责达成一致，导致项目停滞不前。

OpenAI, Oracle, and SoftBank couldn t agree on who would do what, and so the project languished.

受阻的OpenAI试图独立开发数据中心，但贷款方似乎对向一家未经验证的公司提供数十亿美元的贷款持保留态度。

Stymied OpenAI sought to develop data centers independently, but lenders apparently balked at the prospect of multi billion dollar loans tied to an unproven company.

几个月后，OpenAI重新转向合作，开始与甲骨文和软银共同开发数据中心，但这些项目都没有涉及Stargate的全部三方参与者。

Months later, of course, OpenAI circled back to partnerships and began developing data centers in collaboration with both Oracle and SoftBank, but none of the projects involved all three Stargate participants.

这里有很多细微差别，但归根结底，问题是建设的规模和速度未能达到OpenAI的雄心。

There's a lot of nuance here, but ultimately the issue is that the scale and speed of the build out has fallen short of OpenAI's ambitions.

他们原本的目标是在2025年前签署总计10吉瓦的Stargate承诺，但截至2026年初，OpenAI仅与甲骨文签订了6吉瓦、与软银签订了2吉瓦的合同。

They originally targeted having 10 GW in Stargate commitments signed by the 2025, but as we stand in early twenty twenty six, OpenAI has six GW under contract with Oracle and two GW with SoftBank.

所有项目都遭遇了延误，目前仅有一个项目被证实部分投入运营。

The projects have all been hit with delays and only one is known to be partially operational at this stage.

最终，这一切可能并没有表面看起来那么戏剧性，只是在如此前所未有的规模下做事所面临的挑战。

Ultimately, all of this may be less dramatic than it seems and just the challenge of doing things on such an unprecedented scale.

OpenAI负责计算扩展的Sachin Kadi发表评论写道：Stargate是我们计算战略的 umbrella 品牌。

Sachin Kadi, who works on compute scaling at OpenAI, weighed in, writing: Stargate is the umbrella brand for our compute strategy.

这关乎调动整个生态系统，实现全球AI算力的跨越式提升。

It's about mobilizing the full ecosystem to deliver a step change in global AI compute.

在过去的几个季度里，这一愿景已成为现实。

And over the past few quarters that vision has become a reality.

我们在2025年底时拥有约2GW的可用算力，并建立了一套旨在长期扩展产能协议、专门设计的协同部署以及深度参与下一代数据中心设计的合作模式。

We exited 2025 with around two GW of available compute and a model designed to scale long term capacity agreements, purpose built co located deployments, and deep collaboration on next generation data center design.

算力领导力是研究与产品迭代的基础，而Stargate正是我们构建未来的方式。

Compute leadership is the foundation of research and product velocity, and Stargate is how we build what comes next.

总体而言，我对这一点其实不太确定该作何解读，但值得关注的原因在于，华尔街在试图理解整个领域时，必然会密切关注它。

Overall, I don't really know what to make of this one, but the reason it's worth paying attention to if for no other reason than you know that Wall Street, as it tries to figure out what to make of this entire space, is going to be watching it closely.

说到华尔街，今天我们终于转向英伟达，它交出了一份令人惊艳的财报，再次创下营收新高。

Speaking of Wall Street, we turn finally today to Nvidia, who absolutely crushed earnings delivering another record quarter for revenue.

周三晚间的财报结果毫无歧义。

Wednesday night's earnings left no room for interpretation.

AI芯片热潮仍在如火如荼地进行中。

The AI chip boom is still in full swing.

英伟达的收入同比增长73%，达到681亿美元，超出预期高达11个百分点。

NVIDIA's revenue was up 73% annualized to reach $68,100,000,000 exceeding expectations by a massive 11 percentage points.

他们还为本季度指引了更高的增长，预计年化收入增幅将达到77%。

They also guided even higher growth for the current quarter, projecting a 77 annualized revenue gain.

数据中心业务的增长速度甚至超过了整体收入，增长率达到75%。

Data center business grew even faster than top line revenue, coming in at 75% growth.

英伟达还指出，其预测中仍假设来自中国的收入为零，因此如果出口恢复，收入可能再次大幅增长。

NVIDIA also noted that they're still assuming zero revenue from China in their forecasts, so revenue could see another big boost if exports start flowing.

首席执行官黄仁勋表示，人工智能的需求正在持续增强，并评论道：‘全球对算力令牌的需求已经完全呈指数级增长。’

CEO Jensen Huang remarked that AI demand is going from strength to strength, commenting: The demand for tokens in the world has gone completely exponential.

我认为我们都看到了这一点，甚至我们云中六年前的GPU都已被完全占用，价格也在上涨。

I think we're all seeing that, to the point where even our six year old GPUs in the cloud are completely consumed and the pricing is going up.

尽管如此，英伟达还披露了为人工智能基础设施建设提供的大量新保障措施。

Still, NVIDIA also disclosed massive new backstops for the AI build out.

他们目前已为开发数据中心的公司提供了35亿美元的担保，是上一季度披露金额的四倍。

They've now provided $3,500,000,000 worth of guarantees to companies developing data centers, four times the amount they disclosed in the previous quarter.

这些担保使较小的Neo Cloud公司能够以更有利的条件获得债务融资。

These guarantees allow smaller Neo Cloud companies to access debt on more favorable terms.

然而，如果AI需求放缓，NVIDIA可能需要接手陷入困境的市场中的数据中心租赁和运营。

However, if AI demand falters, NVIDIA could be required to take over leases and the operation of data centers in a struggling market.

NVIDIA并未披露这些新担保的具体受益公司。

NVIDIA did not disclose which companies they gave these new guarantees to.

AMD最近向Crusoe提供了类似的担保，承诺3亿美元，表明此类安排正变得越来越普遍。

AMD recently gave a similar guarantee to Crusoe, pledging $300,000,000 suggesting these arrangements are becoming more common.

尽管这些担保为NVIDIA的资产负债表带来了一定的边际压力，但与迅猛增长的利润相比，这些交易的规模相对较小。

While the backstop adds marginal pressure on NVIDIA's balance sheet, ultimately these deals are pretty modest compared to the booming profits.

Jensen还就近期软件股的下跌发表看法，认为市场判断有误。

Jensen also weighed in on the recent drawdown in software stocks arguing that the markets got it wrong.

他重申了自己的观点，即现有软件公司将在其产品之上构建代理功能，但其作为记录系统和数据库的价值仍将保持。

He restated his view that existing software firms will build agentic features on top of their products, but their value as systems of record and databases will remain.

他说：我们今天使用的所有工具，无论是Cadence、Synopsys、ServiceNow还是SAP，都存在一个根本性的合理原因。

He said: All of these tools that we use today, whether it's Cadence or Synopsys or ServiceNow or SAP, these tools exist for a fundamentally good reason.

这些具代理功能的AI将是智能软件，代表我们使用这些工具，帮助我们提高生产力。

These agentic AI will be intelligent software that uses these tools on our behalf and help us be more productive.

没有人能比ServiceNow更好地为我们服务，它们将开发出高度优化的代理，专门适配它们工具所支持的工作。

Nobody is going to serve us better than ServiceNow, and they are going to come up with agents that are really fine tuned and optimized for the work that uses the tools that they have.

最终，我们需要这些工具完成工作，并以我们能理解的方式将信息反馈回来。

In the end, we need the tools to finish their work and put the information back in a way we can understand.

尽管前景乐观，但市场反应却略显平淡。

Now despite the rosy outlook, the market response was a little underwhelming.

NVIDIA在财报公布后盘后交易中股价上涨了3%，但当晚稍晚时又跌回了收盘价以下。

Nvidia stock gained 3% in after hours trading once earnings were announced, but fell below its closing price later in the evening.

归根结底，我认为目前市场整体已经完全失去锚定，尤其是在AI领域更是如此。

Ultimately, right now, I just think that markets are completely unmoored in general, but especially when it comes to AI.

对此最明显的证据，莫过于市场对这一NVIDIA公告的反应。

There's basically no better evidence for that than its response to this Nvidia announcement.

无论如何，朋友们，今天我们就先说到这里。

In any case friends, that is where we will end the headlines for today.

接下来是主要节目。

Next up, the main episode.

Agendaq AI 正在推动一场价值三万亿美元的生产力革命，而企业领导者正面临一个关键的决策点。

Agendaq AI is powering a $3,000,000,000,000 productivity revolution and leaders are hitting a real decision point.

你是要自建 AI 代理，直接购买现成的，还是通过合作借力来加速扩展？

Do you build your own AI agents, buy off the shelf, or borrow by partnering to scale faster?

普华永道最新发布的白皮书《应对构建、购买或借用决策》出色地拨开了迷雾，提供了一个实用的框架，帮助你根据价值、风险和准备程度做出选择，并指导如何通过正确的信任、治理和协调基础来扩展代理。

KPMG's latest thought leadership paper, Navigating the Build, Buy, or Borrow Decision, does a great job cutting through the noise with a practical framework to help you choose based on value, risk, and readiness and how to scale agents with the right trust, governance, and orchestration foundation.

别选错模型。

Don't lock in the wrong model.

你现在就可以访问 www.kpmg.us/navigate 下载这份白皮书。

You can download the paper right now at www.kpmg.us/navigate.

再次提醒，网址是 www.kpmg.us/navigate。

Again, that's www.kpmg.us/navigate.

我认为，未来企业 AI 代理领域将出现一个全新的标准。

There's a new standard that I think is going to matter a lot for the enterprise AI agent space.

它被称为AIUC One，是世界上首个AI代理标准。

It's called AIUC one, and it builds itself as the world's first AI agent standard.

它旨在涵盖所有核心企业风险，例如数据与隐私、安全、可靠性、问责制以及社会影响，并由可信的第三方进行验证。

It's designed to cover all the core enterprise risks things like data and privacy, security, safety, reliability, accountability, and societal impact all verified by a trusted third party.

它引起我关注的原因之一是，Eleven Labs——我之前多次提到过这家公司，目前正是行业巨头——刚刚成为首家通过AIUC One认证的语音代理，并推出了全球首个可投保的AI代理。

One of the reasons it's on my radar is that Eleven Labs, who you've heard me talk about before and is just an absolute juggernaut right now, just became the first voice agent to be certified against AI UC one and is launching a first of its kind insurable AI agent.

这意味着在实践中，它具备实时防护机制，可阻止不安全的回应并防范操纵，同时还配备完整的安全体系。

What that means in practice is real time guardrails that block unsafe responses and protect against manipulation, plus a full safety stack.

这种特性正是推动企业采用的关键。

This is the kind of thing that unlocks enterprise adoption.

当一家基于Eleven Labs构建的公司能够出示第三方认证，并宣称我们的代理安全、可靠且经过验证时，整个对话就会发生改变。

When a company building on Eleven Labs can point to a third party certification and say our agents are secure, safe, and verified, that changes the conversation.

访问aiuc.com，了解全球首个AI代理标准。

Go to aiuc.com to learn about the world's first standard for AI agents.

网址是aiuc.com。

That's aiuc.com.

大多数公司并不缺乏创意。

Most companies don't struggle with ideas.

他们的问题在于如何将这些创意转化为真正创造价值的AI系统。

They struggle with turning them into real AI systems that deliver value.

Robots and Pencils 是一家旨在弥合这一差距的公司。

Robots and Pencils is a company built to close that gap.

他们设计并交付基于生成式与代理式AI的智能云原生系统，注重专注力、速度和明确的成果。

They design and deliver intelligent, cloud native systems powered by generative and agentic AI, with focus, speed, and clear outcomes.

Robots and Pencils 以小型、高影响力团队的形式开展工作。

Robots and Pencils works in small, high impact pods.

工程师、战略家、设计师和应用AI专家协同合作，从创意到生产无缝推进，避免不必要的阻力。

Engineers, strategists, designers, and applied AI specialists working together to move from idea to production without unnecessary friction.

借助他们的Identic加速平台Roboworks，团队能够在短短45天内（视范围而定）交付有意义的成果，包括初步上线。

Powered by Roboworks, their Identic Acceleration Platform, teams deliver meaningful results, including initial launches, in as little as forty five days depending on scope.

如果你的组织准备加快步伐、降低复杂性，将AI愿景转化为实际成果，Robots and Pencils 正是为此刻而打造的。

If your organization is ready to move faster, reduce complexity, and turn AI ambition into real results, Robots and Pencils is built for that moment.

从 robotsandpencils.com/aideallybrief 开始对话。

Start the conversation at robotsandpencils.com/aideallybrief.

那就是 robotsandpencils.com/aideallybrief。

That's robotsandpencils.com/aideallybrief.

Robots and Pencils 的速度与影响力：想将企业软件开发速度提升5倍吗？

Robots and Pencils Impact at Velocity Want to accelerate enterprise software development velocity by 5x?

你需要 Blitzy，这是唯一专为大型企业代码库打造的自主软件开发平台。

You need Blitzy, the only autonomous software development platform built for enterprise codebases.

你的工程师定义项目、新功能、重构或全新开发。

Your engineers define the project, a new feature, refactor, or greenfield build.

Blitzy 的智能代理首先会摄入并映射你的整个代码库，然后平台会生成一个定制的代理行动方案，供你的团队审阅和批准。

Blitzy agents first ingest and map your entire codebase, then the platform generates a bespoke agent action plan for your team to review and approve.

一旦获得批准，Blitzy 就会自主开始工作，生成数十万行经过验证和端到端测试的代码。

Once approved, Blitzy gets to work autonomously generating hundreds of thousands of lines of validated, end to end tested code.

单次运行即可完成超过80%的工作。

More than 80% of the work completed in a single run.

Blitzy 不是在生成代码，而是在以计算的速度开发软件。

Blitzy is not generating code, it's developing software at the speed of compute.

你的工程师负责审查、优化并发布。

Your engineers review, refine, and ship.

这就是财富500强企业如何将数月的项目压缩到单个冲刺周期中，将工程效率提升5倍的方式。

This is how Fortune 500 companies are compressing multi month projects into a single sprint, accelerating engineering velocity by 5x.

立即前往 blitzy.com 亲身体验 Blitzy。

Experience Blitzy firsthand at blitzy.com.

那就是 blitzy.com。

That's blitzy.com.

欢迎回到《AI每日简报》！

Welcome back to the AI Daily Brief!

我终于回到演播室了，此前我不仅在南美洲待了几周，还连续三天不间断地旅行。

I am finally back in the studio after not only a couple weeks abroad in South America, but also three non stop days of travel.

我可以很有把握地说，自从我开始做这个节目以来，从未像过去三天这样与AI新闻如此脱节。

I think I can say fairly confidently that I have never, in the entire time I've been doing this show, been as disconnected from the AI news as I was for the last three days.

当我重新回来时，浏览了所有错过的新闻，这期节目最核心的问题变成了：我错过的最重要主题是什么？我应该把这第一期主节目聚焦在哪个主题上？

And so coming back to it, as I was looking through all of the news that I had missed, the big question for this show was going to be: What was the most important theme that I had missed that I wanted to focus this first main episode on?

我认为答案是AI的开放化趋势——越来越多的AI产品开始在某种程度上 resemble OpenAI，而我认为这背后的意义远不止于与一个热门项目竞争。

The answer, I believe, is the open clawification of AI a phenomenon where more and more products in the AI space are starting to resemble Open Claw in ways that I believe are about something much bigger than just trying to compete with a hot project.

那么，我们先来谈谈一些激发了这期节目的功能更新。

So let's talk first about some of the feature announcements that inspired this show.

首先，Claude Code推出了全新的远程控制功能。

Right at the top we have Claude Code's new remote control feature.

顾名思义，它实现了人们长久以来期待的功能——此前许多初创公司都曾围绕这一需求开发产品，而现在这一功能终于进入了主流：你可以在手机上直接使用Claude Code进行编程。

It is exactly what it sounds like and brings a thing that people have wanted forever and that numerous startups have built around to the mainstream which is the ability to code with Claude Code from your phone.

周二，Anthropic发布了：Claude Code远程控制功能上线。

On Tuesday Anthropic posted: New and Claude Code remote control.

在终端启动任务后，你可以在散步或参加会议时，用手机继续操作。

Kick off a task in your terminal and pick it up from your phone while you take a walk or join a meeting.

Claude会在你的设备上持续运行，你可以通过Claude应用或claude.aicode随时控制会话。

Claude keeps running on your machine and you can control the session from the Claude app or claude.aicode.

一些最初的反应是开玩笑说这对社交生活会产生影响。

Some of the first reactions were to joke about the social life implications of this.

Cremio写道：Anthropic拯救了旧金山的夜生活。

Cremio wrote, Anthropic just saved the San Francisco nightlife.

然而对大多数人来说，最明显的对比其实是OpenClaw。

And yet for most, the very clear comparison was in fact OpenClaw.

AppClub的Meta Preston写道：这正是我想要Claude Bot的唯一原因。

AppClub's Meta Preston writes, This is the only reason I wanted Claude Bot.

投资者兼内容创作者Sara Dietschy写道：好吧，我想我不用再试Claude Bot了。

Investor and content creator Sara Dietschy writes: Okay, I guess I don't need to try Claude Bot anymore.

Guy Al Breton称这是给成年人用的OpenClaw。

Guy Al Breton calls it OpenClaw for grown ups.

Ally K.

And Ally K.

Miller说：这正是我一直在等待的功能。

Miller says: This is the feature I've been waiting for.

我每天都会用手机工作好几个小时。

I work from my phone for hours every day.

远程控制功能快要让我变成一个Claude代码狂热者了。

Remote control is about to turn me into a Claude Codemaniac.

所以人们非常喜欢OpenClaw的一个重要功能是可以通过Telegram与你的代理互动。

So one of the big things that people love about OpenClaw is the ability to interact with your agents via Telegram.

但这并不是因为Telegram在你坐在电脑前时一定是最佳体验。

And that's not because Telegram is necessarily the best experience when you're sitting in front of a computer.

而是因为使用Telegram、WhatsApp或你连接的任何聊天应用，都能让你随时随地与代理互动。

It's because using Telegram or WhatsApp or whatever chat app you've connected it to makes it so that you can interact with your agents on the go.

这确实是吸引我使用OpenClaw的最初原因之一——拥有一个可以随身携带的编程代理，这样我即使不在电脑前也能继续处理项目。

This was certainly one of the first things that attracted me to OpenClaw was the ability to have an on the go coding agent so I could keep working on my projects when I wasn't sitting in front of the laptop.

显然，很多人也有类似的感受。

Lots of other people clearly felt similarly.

Nate Hercke写道：Claude代码功能添加了每个人想要的东西。

Nate Hercke writes Claude code just added what everyone wanted.

通过远程控制，你 essentially 拥有了随身携带的 Claude 代码，可以随时随地查看进度并继续分配任务。

With remote control, you essentially have Claude code in your pocket so that you can check-in on progress and continue to assign work as you're out and about.

有些人甚至认为，OpenClaw 在这个新远程控制功能中的使用场景可能会有所不同。

Some even went so far as to argue that the use cases for OpenClaw in this new remote control feature might actually end up being a little bit different.

他们写道：ClaudeCode 是你的工作引擎。

They wrote: ClaudeCode is your work engine.

它是一个专业的终端工具。

It's a professional terminal tool.

它嵌入在你的代码库中，让你遛狗时也能编写函数、运行测试和创建拉取请求。

It lives in your codebase to write functions, run tests, and open PRs while you walk the dog.

OpenClaw 是你的生活引擎。

OpenClaw is your life engine.

它是一个全天候的管家。

It's an always on butler.

它更适合管理日历、预订航班或给妈妈发短信。

It's better for managing your calendar, booking flights, or texting your mom.

为什么远程控制不同？

Why the remote control is different?

首先是原生与消息传递的区别。

First is native versus messaging.

OpenClaw 通过 Telegram 或 WhatsApp 与你交流。

OpenClaw talks to you through Telegram or WhatsApp.

Cloud Code 远程控制为你提供一个专用的网页或移动桥梁，连接到你的本地硬件。

Cloud Code Remote Control gives you a dedicated web or mobile bridge to your local hardware.

深度上下文：Cloud Code 会感知你的特定开发环境会话，而 Butler 则持续在你的机器或 VPS 上 24/7 运行。

Deep Context Cloud Code is aware of your specific dev environment Session versus Butler You boot up Cloud Code for a work session Open Claw stays running 20 fourseven on your machine or VPS.

从实际使用角度看，使用 Claude Code 启动一次大规模重构，然后去开会，并在手机上查看代码差异。

From a pragmatic usage perspective, use Claude Code to start a heavy refactor, head to a meeting, and check the diffs on your phone.

使用 OpenClaw 在你离开键盘时，持续跟进通知和日常数字事务。

Use OpenClaw to stay on top of your notifications and general digital existence while you're away from your keyboard.

如果你想要一个能精准操作代码的随身工具，就选择 Claude Code。

If you want a code from your pocket with surgical precision, get Claude Code.

如果你想要一个24/7的数字助手，那就继续使用OpenClaw。

If you want a 20 fourseven digital butler, keep OpenClaw.

现在，我并不完全认同的观点是，OpenClaw只适用于管理日历或预订航班这类个人用途。

Now what I don't totally agree with here is the idea that OpenClaw is only for these sort of personal use cases like managing your calendar or booking flights.

事实上，正如我几周前在关于我的OpenClaw设置的视频中提到的，我最有用的代理是那些持续24/7运行的研究代理，它们不断更新对本节目以及我们生态系统项目如AIDB Intelligence都有用的信息。

In fact, as I mentioned in my video from a couple weeks ago about my OpenClaw setup, my most useful agents are my research agents that are running 20 fourseven and updating information that's both useful for this show as well as for some of our ecosystem projects like AIDB Intelligence.

但我认为有两点：首先，关于为什么从编码角度看这将对人们如此有价值——因为它接入了人们已经在使用的主系统，即Claude Code——我认为这是正确的。

But I think two things: first of all, the points about why this is going to be so valuable to people from a coding perspective because it plugs into the main system that people are already using in the form of Claude code is, I think, true.

我发现自己使用远程编码器OpenClaw代理的频率略有下降，原因之一是它与我进行其他大部分项目所处的环境之间存在上下文断层。

One of the reasons I believe that I found myself using my remote coder OpenClaw agent a little bit less is the context gap between it and where I'm doing most of my other projects.

第二点与Internouns Post相关的是，即使你不认同这里对OpenClaw的描述，其背后的核心观点是，我们正经历一场更广泛的‘代理化’范式转变，未来会出现多种工具和平台，它们具有类似架构和交互模式，但应用场景截然不同。

The second thing that I think is relevant about Internouns Post is that even if one doesn't agree with the idea of how OpenClaw is characterized here, The actual underlying point is that we're in a broader paradigm shift of agentification, and that there are going to be different tools and different platforms that have similar types of setups and similar types of interactive modes, but with very different use cases.

无论最终应用场景如何，远程控制都取得了巨大成功。

Whatever the use cases end up being, remote control hit like gangbusters.

Claude Code的产品经理Noah Zweiben在公告发布后的第二天不得不回到Twitter，为他们遇到的容量问题道歉，并保证将推出更大规模的上线。

Claude Code PM Noah Zweiben had to come back to Twitter the day after the announcement to apologize for the capacity issues that they had run into on their end and to ensure that a broader rollout was coming.

对一些人来说，远程控制只是进一步证明了Anthropic进展之快。

For some, remote control was just further evidence of how fast Anthropic is moving.

匿名AI传闻账号Irule the World写道：Anthropic已经发布了OpenClaw。

Anonymous AI rumor account Irule the World writes: Anthropic have now shipped Openclaw.

Dario和他的团队远远领先，他们的速度令人难以置信。

Dario and the gang are unfathomably ahead and their velocity is insane.

这种观察是合理的：他们可能已经达到了逃逸速度，碾压了所有基准，甚至让企业界和国防部都求着他们。

It's a fair observation that they may have reached escape velocity, crushing every benchmark, eating Enterprise, Department of War begging.

其中一个实验室取得了重大突破。

One of the labs made a huge breakthrough.

还有其他人，比如LLM爱好者Amwill指出，真正重要的是，Anthropic并不仅仅是采纳了Claude Code的功能，而是开始展现出一种新型AI交互模式的新基本要素。

Still others like LLM Junkie Amwill noted that what mattered here wasn't so much Anthropic picking up on Claude code features but starting to see the new primitives of a new modality of AI coming to the fore.

他们写道：这是任何主要AI实验室首次实现从手机到本地机器的官方集成，但绝不是最后一次。

They write: This is the first official phone to local machine integration from any major AI lab, but not the last.

这将被广泛采用。

This will become widely adopted.

然而，Anthropic 并没有停下！

Anthropic however was not done!

就在一天后，他们回到 X 平台，宣布了一项名为“计划任务”的新协作功能。

Just one day later they returned to X to announce a new feature this time for co work called Scheduled Tasks.

再次说明，这正是字面意思。

Once again it was exactly what it sounds like.

他们写道：Claude 现在可以自动在特定时间完成重复性任务：如晨间简报、每周电子表格更新、周五团队汇报。

They wrote: Claude can now complete recurring tasks at specific times automatically: a morning brief, weekly spreadsheet updates, Friday team presentations.

他们还写道：通过插件，协作功能在设计、工程、运营等领域获得了更深入的专业能力。

They also write: It gets better with plugins, which give co work domain expertise across design, engineering, operations, and more.

此外，我们正在协作侧边栏中添加一个全新的自定义标签页。

Also, we're adding a new customized tab in your co work sidebar.

一个集中管理你的插件技能和连接器的地方。

One place to manage your plugin skills and connectors.

现在，回到 AI 的‘爪化’这一概念，如果你听过我关于 OpenClaw 的节目，或者自己尝试过，就会知道其魔力的一大来源在于它与计划任务的交互方式。

Now, going back to the idea of the clawification of AI, if you've listened to my episodes about OpenClaw or if you've tried it out yourself, you'll know that a big part of the magic is the way that it interacts with scheduled tasks.

首先，OpenClaw代理具有一种称为‘心跳’的功能，本质上是一个提醒机制，其时间间隔可以自行设置，但默认为每三十分钟一次，在此期间代理会提醒自己记住使命，并在上一个周期内有所懈怠时继续工作。

First, Open Claw agents have something called a heartbeat, which is basically a reminder the time interval of which you can set but which is every thirty minutes by default, where effectively the agent reminds itself to remember its mission and to go keep working if it perhaps has lapsed in the previous period of time since the last heartbeat.

除此之外还有定时任务（Cron Jobs），你可以让代理在特定时间执行特定任务。

On top of that are Cron Jobs, where basically you can tell an agent to do a specific thing at a specific time.

深入研究OpenClaw后你会发现，其核心魔力很大程度上就在于能够按特定时间安排任务。

When you dig into OpenClaw, turns out that a lot of what the magic is, is just being able to schedule tasks at certain times.

在我的五个项目管理代理中，它们各自有不同的日程安排，每天会检查几次我的状态，要么提醒我所有待办事项，要么与外部源交互以获取各个项目的最新进展。

Across my five Project Manager agents, they all have different schedules for the couple times a day when they check-in with me either to remind me of all the to do lists or to go interact with external sources to find the state of different projects.

因此，这种定时任务的概念再次显得并非Anthropic在与OpenClaw竞争，而是承认了智能体AI中一种新的基础能力。

And so this idea of scheduled tasks feels once again not like Anthropic trying to compete with OpenClaw, but instead recognizing a new primitive in agentic AI.

然而，这并没有阻止人们做出这种比较。

That didn't stop people making the comparison, however.

内容创作者Ole Lemon分享了这篇帖子，并称‘面向普通人的OpenClaw已经到来’。

Content creator Ole Lemon shared the post and said Open Claw for normies has landed.

Akash Gupta认为，这一范式转变的部分在于：AI能够持续运行，而无需你每次都要主动提示。

Akash Gupta gets that part of the paradigm shift is that AI that works continuously without you having to prompt it every time.

他写道：定时任务意味着Claude不再是你需要不断与之对话的软件，而变成了在你睡觉时也能自行工作的软件。

He writes: Scheduled tasks means Claude stopped being software you talk to and became software that works while you sleep.

这是一种类别上的变革，而不仅仅是功能更新。

That's a category change, not a feature update.

现在，他和许多其他人还指出，这不太可能让华尔街减轻对Anthropic每项新公告的担忧，尤其是对白领工作带来的影响。

Now he, along with lots of others, also pointed out that this is unlikely to make Wall Street, which was already nervous about every new announcement from Anthropic, any less scared about the implications for white collar work.

在同一帖子的后面，他写道：一份早晨简报，会在你醒来前汇总一夜之间的Slack活动、邮件线程和日历变更。

Later in the same post he writes: A morning brief that compiles overnight Slack activity, email threads, and calendar changes before you wake up.

一份每周电子表格，会自动提取数据、运行公式，并在每个星期五将格式化的Excel文件放入你的文件夹中。

A weekly spreadsheet that pulls data, runs formulas, and drops a formatted Excel file into your folder every Friday.

合同提醒会在下午3点自动触发，而无需任何人记得发送。

Contractor reminders that fire at 3PM without anyone remembering to send them.

这些任务过去都是SaaS产品、Zapier工作流，或初级员工的晨间例行工作。

Each of those tasks used to be a SaaS product, a Zapier workflow, or a junior employee's morning routine.

而现在，它们只需一行指令和一个调度界面，运行在一个每月20美元的订阅服务上。

Now, they're a single line and a scheduling interface running on a $20 month subscription.

Sitebringer写道：聊天只是一个玩具。

Sitebringer writes: Chat was a toy.

定时任务是一种劳动基础工具。

Scheduled tasks is a labor primitive.

当一个模型能够在你的工具中无需你指示就按计划执行工作时，它就不再是被帮助的对象。

When a model can do work on a schedule inside your tools without you asking, it stops being helped.

它变成了地球上最便宜的员工。

It becomes the cheapest employee on Earth.

现在，这个特定的案例并不是关于白领岗位替代及其相关担忧的，但为了记录在案，我想说，认为这一变化的唯一影响将是消除员工，而不是大规模地赋能员工承担更多工作，这种观点在将来回看时会显得相当片面。

Now this particular episode is not about white collar job displacement and the concerns there, but I will say just for posterity that I think that the idea that the only implication for this will be the elimination of employees rather than the mass empowerment of employees to do more will seem, in retrospect, fairly reductionist.

开发者西蒙·威利森写了一篇关于测试这些新功能的笔记，他指出这些功能与OpenClaw有重叠。

Developer Simon Willison wrote a note about testing these new features, both of which he notes overlap with OpenClaw.

当所有推特评论者都激动不已时，他指出这些功能仍处于早期阶段。

And while all the Twitter commentators were breathless, he pointed out that these features are still early.

用他的话说，例如：远程控制目前还略显笨拙。

To use his words, for example: Remote control is a little bit janky right now.

有趣的是，在他关于远程控制的帖子末尾，他提到：将这一点与OpenClaw等解决方案进行对比很有意思，因为OpenClaw的一个主要卖点是能够通过手机控制你的个人设备。

Now funny enough, at the end of his post about remote control, he said: It's interesting to contrast this to solutions like OpenClaw, where one of the big selling points is the ability to control your personal device from your phone.

Clawed Code 目前仍没有文档化的机制来实现定时运行任务，而这正是Claude类软件的另一项核心功能。

Clawed Code still doesn't have a documented mechanism for running things on a schedule, which is the other killer feature of the Claude category of software.

当然，他随后更新了内容，指出Anthropic也已宣布推出了定时任务功能。

Of course, he then had to update it sharing that Anthropic had now announced scheduled tasks as well.

他指出了一个限制：定时任务仅在电脑处于唤醒状态且Claw桌面应用打开时才会运行。

He did point out one limitation: that scheduled tasks only run while your computer is awake and the Claw desktop app is open.

基本上，如果任务预定执行时电脑处于休眠状态或应用已关闭，CoWork会跳过该任务，待电脑唤醒或应用重新打开后自动补跑。

Basically, if the computer is asleep or the app is closed when that task is scheduled to run, CoWork skips it and then runs it automatically once it wakes up or the app is opened up again.

如今，本地设备管理无疑是这一领域的一个全新维度。

Now it is certainly the case that local device management is a whole new aspect of this.

我最近亲身经历了这一点：我出国两周期间，哈德逊河谷接连遭遇暴风雪，导致断电，而我的设备又没有自动重启机制，结果在那段时间里我完全无法使用OpenClaw。

I just experienced this by being out of the country for two weeks, as the Hudson Valley got hammered by snowstorm after snowstorm, knocking the power out and my machines off without automatic startup procedures that left me open clawless for a big chunk of the time down there.

重要的是，尽管Anthropic是引发众人热议‘OpenClaw化’趋势的大公司，但这种模式 elsewhere 也早已出现。

Now importantly, while Anthropic was the big company that got everyone talking about this open clawification, the pattern was exhibited elsewhere as well.

周三，Perplexity 发布了 Perplexity Computer，并写道：Computer 将所有当前的 AI 能力整合到一个系统中。

On Wednesday, Perplexity announced Perplexity Computer, writing: Computer unifies every current AI capability into one system.

它可以研究、设计、编码、部署和端到端管理任何项目。

It can research, design, code, deploy, and manage any project end to end.

Computer 会协调模型并行运行代理，并利用 Opus 将每项任务匹配到最适合的模型。

Computer orchestrates models to run agents in parallel, leveraging Opus to match each task to the model best suited for it.

他们写道，Perplexity Computer 就是 2026 年个人电脑应有的样子。

Perplexity Computer, they write, is what a personal computer in 2026 should be.

它为你个人定制，记得你过去的工作，并且默认安全。

It's personal to you, remembers your past work, and is secure by default.

数百个连接器、持久内存、文件和网页访问功能，全部构建在 Perplexity 基础设施之上。

Hundreds of connectors, persistent memory, files, and web access all built on top of Perplexity infrastructure.

从单个任务轻松过渡到数百个活跃项目。

Go from a single task to hundreds of active projects.

谷歌高级 AI 项目经理 Shibam Sapu 写道：Perplexity 刚刚推出了他们自己的 OpenClaw。

Google Senior AI Project Manager Shibam Sapu writes: Perplexity just launched their own OpenClaw.

组合式计算机就足够了。

Composable computers is all you need.

Perplexity 首席执行官阿拉文德·特里纳瓦斯写道：过去两个月，Perplexity 做了些什么？

Perplexity CEO Aravind Trinavas wrote: What has Perplexity been up to the last two months?

我们一直在默默研发 Perplexity Computer 这个下一代重磅产品。

We've silently been working on the next big thing Perplexity Computer.

在一篇名为《AI 就是计算机》的公告博客中，他基本上论证了 Perplexity 版本的 OpenClaw 的独特之处在于它仅能访问 Anthropix 模型，并反复强调使用 Perplexity 可以获得 19 个模型，从而让每个任务都能匹配到最合适的模型。

In a blog post surrounding the announcement called The AI is the Computer, he basically made an argument that the differentiator for Perplexity's version of OpenClaw was that it only had access to Anthropix models and hammered on the idea that with Perplexity you get 19 models available so that the right task can find the right model.

他写道：AI 模型正变得越来越强大，围绕它们构建的产品反而成了限制其真正潜力的瓶颈。

He writes: AI models are becoming so capable that the products built around them have been a bottleneck for showing their true potential.

聊天界面适合回答问题，代理则适合处理单个任务。

The chat UI is good for answers, and agents are good for individual tasks.

而整个工作流的界面，一直以来都是计算机。

Meanwhile, the UI for entire workflows has always been the computer.

AI 现在已牢固地进入多模态时代。

AI is now firmly multimodal.

它能在单一连贯的系统中理解和生成多种类型的数据。

It understands and generates many forms of data in a single coherent system.

詹森·黄和其他人表示，人工智能的未来也必须是多模态的。

Jensen Huang and others have said that the future of AI must also be multimodal.

他们说得对，专用模型必须像团队一样协作。

They're right specialized models must collaborate like a team.

随着人工智能越来越多地取代计算机本身的功能，计算机的核心活动将变成大规模的多模态协调。

As AI replaces more of the function of the computer itself, the central activity of the computer will be massively multimodal orchestration.

所以现在，Anthropic 被 Claw 化了，Perplexity 也被 Claw 化了。

So now we've got Anthropic getting clawified, Perplexity getting clawified.

还有谁呢？

Who else?

我暂时把 OpenAI 放一边，因为他们显然请来了 OpenClaw 的创造者彼得·施泰因伯格加入他们。

Well, I'm holding OpenAI aside because they obviously got Peter Steinberger, the creator of OpenClaw, to come join them.

但我们在产品公司中也看到了这种趋势。

But we're also seeing it in the product companies.

此外，在周二，Notion发布了其自定义代理功能。

Also on Tuesday, Notion announced their Custom Agents features.

果然，其宣传中包含了许多熟悉的卖点。

And sure enough, it had a lot of familiar value propositions in the pitch.

Notion将自定义代理称为永不休息的AI团队。

Notion calls Custom Agents the AI team that never sleeps.

它们是自主的，专为团队设计，任何人都能轻松构建。

They're autonomous, built for teams, and easy for anyone to build.

为它们分配任务，设置触发条件或计划，它们就会全天候完成工作。

Give them a job, set a trigger or schedule, and they'll get it done round the clock.

大多数AI需要你主动提问才会行动。

Most AI waits for you to ask.

而自定义代理则会主动执行。

Custom agents just go.

自动将bug分配到正确位置、回答问题、更新文档、起草周报，并通知相关人员。

Route bugs to the right place, answer questions, update docs, draft weekly updates, and ping the right people.

它们是多人协作、模型无关的，专为所有团队设计，而不仅仅是技术团队。

They're multiplayer, model agnostic, and built for every team, not just the technical ones.

创作者安德鲁·沃纳写道：Notion 刚刚为普通用户推出了 OpenClaw。

Creator Andrew Warner writes: Notion just launched OpenClaw for regular people.

优点：无需购买 Mac mini。

Pros no need to buy a Mac mini.

界面非常直观，任何人都能轻松使用。

Very visual so anyone can use it.

它基于时间自动运行自动化任务，你的团队无需了解 cron job、sudo 或 cli 等术语。

Runs automations based on time and your team doesn't need to know terms like cron job, sudo, or cli.

他指出的缺点是，你需要为团队中的每个人购买一个席位，否则谁也用不了，当你在 Notion 中有大量用户时，这可能是个问题。

The cons he pointed out is that you need to buy a seat for each person on your team or no one gets it which when you've got a lot of people in Notion could be problematic.

几周前，我们还看到 Airtable 推出了同类产品，名为 Super Agent，它实际上更接近 Manus 或 Genspark 的理念，但仍是在开拓一个全新的相似领域。

A couple weeks earlier, we also saw Airtable launch something in the same family called Super Agent which honestly feels a little bit closer to the paradigm of Manus or Genspark but is still pushing a new similar qualified territory.

我认为，将这些举措都视为公司试图追赶 OpenClaw 是一个错误。

Now I think it would be a mistake to view all of these moves as companies trying to catch up with OpenClaw.

我甚至不认为他们特别关注与OpenClaw的竞争。

I don't even think that they are particularly focused on the competition with OpenClaw.

我认为他们并不是想利用或搭上OpenClaw所抓住的这股东风。

I don't think they're trying to harness or latch onto the lightning in a bottle that OpenClaw captured.

我认为更准确的说法是，他们都承认OpenClaw是第一个，即使不是最早发现、至少是最早普及了AI未来形态的人。

What I think instead is that they're all an acknowledgment that Open Claw was the first to, if not see at least to popularize, what AI looks like going forward.

一种交互式模式，能跟随你 wherever you are，且不依赖于特定设备。

Interactive modalities that follow you wherever you are and are not contingent upon one device or another.

一种无需你主动触发就能持续进行的工作方式。

Persistent work that happens without you having to prompt it.

具备代理能力，可以与你的所有个人上下文互动，并能获得你的授权来操作你的系统完成任务。

Agentic capabilities that can interact with all of your personal context, and which you can give permission to use your systems to do things.

关键是，这些不是可以复制的功能，而是我们正步入的代理时代所必需的新基本要素。

The point is that these are not features to be copied they are new fundamental primitives of the agentic era that we're moving into.

它们将变得无处不在，因为它们解锁了人们非常明确想要的一系列新可能性。

They will become ubiquitous because they unlock a new set of things that people very clearly want.

现在，我非常不同意的一个观点是，因为这些功能变得越来越容易使用且产品化得更好，你就应该直接跳过整个OpenClaw的设置阶段。

Now one thing that I do disagree with very heartily is the idea that because these features are getting easier and better productized, you should just skip the whole Open Class setup phase.

很多人会这么做，这很合理，但真正现在花精力去设置OpenClaw的最佳理由之一是，它能让你比那些功能被抽象掉的产品更深入地理解这些基本要素。

Many will do that and it's reasonable, but one of the best reasons to actually do the hard work of setting up OpenClaw now is that it's going to give you a better education in understanding these primitives than will the products where they are abstracted away.

当然，大多数人不会花时间去做这件事，但如果你在听这个节目，我知道你不是大多数人。

Certainly most people will not take the time to do that, but if you're listening to this show, I know you are not most people.

无论如何，我预计今年AI的这种‘OpenClaw化’趋势将持续下去，最终，OpenClaw不会被看作只是被扔进历史垃圾堆的炒作阶段，而是被视为一个全新AI时代的发令枪。

In any case, I expect that this open clawification of AI continues throughout the year, and ultimately the claw is seen not just as some hype phase confined to the junk heap of history, but as the starting gun for a totally new era of AI.

谢谢收听或观看。

Thanks for listening or watching.

一如既往，能再次和你们在一起真好。

As always, great to be back with you guys.

下次再见。

Until next time.

再见。

Peace.