Arena AI 在 8 个月内实现 1 亿美元年化营收

Arena AI的年收入刚刚突破一亿美元。

Arena AI has just hit a $100,000,000 run rate.

这距离他们推出付费评估才过去八个月。

This is only eight months after they launched paid evaluations.

我想仔细分析一下Arena到底做什么，他们为什么特别，为什么独一无二，又为什么增长这么快。

I wanna break down what Arena does, why they're special, why they're unique, and why they're growing so fast.

Palantir正在利用NVIDIA的Nemotron开源模型为美国政府开发AI系统。

Palantir is tapping NVIDIA's Nemotron open models for US government AI.

这件事挺有意思，背后还藏着不少故事和风波。

This is interesting, and a lot of drama is behind this story as well.

伊丽莎白·沃伦和斯科纳隆正在推动一项新法案，拟禁止AI公司出售健康数据。

Elizabeth Warren and Skanolone are reviving a bill to ban AI firms from selling health data.

我们一会儿会详细聊聊这个。

We'll get into the details on that.

Flexion Robotics正在训练数百个类人机器人自主完成办公室跑腿任务，而中国的CXMT则拿下了一笔30亿美元的内存供应合同，客户是腾讯。

Flexion Robotics is training hundreds of humanoids to run office errands autonomously, and China's CXMT is landing a $3,000,000,000 memory supply deal with Tencent.

我们知道，内存是AI基础设施建设中的关键一环，如今内存成本上涨，几乎推高了所有电子产品的价格。

We know memory is one of the critical pieces of the AI infrastructure build out, and there's a lot of issues going on with memory increasing the costs of basically all electronics today.

如果你用过Claude这类AI模型，却因为没法生成图片、音频或视频而感到沮丧，我建议你试试AI Box的MCP连接器。

If you've ever been using an AI model like Claude and have been frustrated that it doesn't create images or audio or video, I'd love for you to try the MCP connector for AI Box.

这是我自己的创业项目。

That's my own startup.

你只需要给Claude发一个链接，它就能把你平时用的其他AI工具全都接入进来。

We essentially allow you with one link, give it to Claude, and it can bring any of the AI models that you use for everything else into Claude.

这样一来，你就能在Claude里直接使用ChatGPT的图像生成功能。

So you can use ChatGPT's image generation inside of Claude.

你还可以在Claude里调用ElevenLabs的音频生成，或者Google的Video Generation v3来生成视频。

You can use Eleventy Labs audio generation inside of Claude or Google v o three's video generation inside of Claude.

我们目前支持接入80多种不同的AI模型，把它们的能力都整合起来。

So all of the capabilities of the different AI models, there's 80 different ones that we allow you to connect.

你可以把这些功能全部整合到Claude、ChatGPT、Gemini、Cursor，或者你日常用的任何AI工作平台里。

You can bring them all inside of Claude or ChatGPT or Gemini or Cursor or any of your other places where you really do all of your work, all of your workspaces with AI.

如果你想了解详情，可以访问 aibox.ai/mcp。

So if you wanna check that out, it's aibox.ai/mcp.

这个 MCP 连接器用起来特别简单。

It's a super easy MCP connector.

你只要在连接器里把这条链接发给 Claude，登录你的 AI Box 账号，就能立刻启用所有这些功能。

So you give Claude this link inside of the connectors, and you log in to your AI Box account, and now you have all of these capabilities.

我专门做了一个网站，详细说明了怎么操作，入门费用是每月 8.99 美元。

I have a whole website where I explain how this works, how you can do it, and it is $8.99 a month to get started with it.

价格非常亲民，我希望这能帮你释放更多创意，让你在 Claude 或其他 AI 工作平台里轻松调用这 80 个不同的模型和功能。

So it's super cheap, and I hope that really unlocks a lot of creativity for you to be able to get 80 different AI models and capabilities inside of Claude or whatever else you are building with.

好的。

Okay.

我们来聊聊 Arena 这家公司最近的情况。

Let's talk about what's going on with Arena.

这家公司现在的估值已经达到了一亿美元。

This is a company that is now at a $100,000,000.

他们最初是在加州大学伯克利分校创建的，本质上是一个AI模型排行榜。

They were originally created in UC Berkeley, and it's basically an AI leaderboard.

对吧？

Right?

你可以去他们平台测试不同AI模型之间的表现。

Like, this is the company where you can go and test different AI models against each other.

他们现在年收入已经达到了一亿美元，而距离他们去年九月推出付费评估服务才刚过去八个月。

They've hit this $100,000,000 annual revenue, and this is eight months after they launched their paid evaluations just back in September.

这比他们今年一月的收入翻了三倍，那时候他们才三千万，说实话，一月那会儿我就觉得已经很惊人了。

This is triple what they were doing in January, which was 30,000,000, which honestly, even in January, I thought it was really impressive.

这家公司现在直接跟Scale AI和Merkur竞争后训练市场的资金，因为他们提供基于上亿次人类模型对比的结构化分析服务。

So the company is now directly competing with Scale AI and Merkur for post training dollars because they're selling labs structured analytics built on a on 10,000,000 plus human model comparisons.

简单说，他们总共完成了两轮融资，累计融资额达两亿五千万美元。

So, basically, what's going on is they've just raised $250,000,000 total across two different rounds.

他们在一月完成了1.5亿美元的A轮融资，公司估值达到17亿美元。

They So did a $150,000,000 series a in January at a $1,700,000,000 valuation.

这笔融资来自十六Z、克莱默·珀金斯和法利希斯。

They did that from a sixteen z, Climber Perkins, and Fallicious.

他们赚钱的方式是，公开排行榜依然免费。

So the way that they actually are making money is that the public leaderboard is still free.

如果你还没试过，可以去网站上，它会把两个不同的AI回复并排展示出来。

So essentially, if you haven't tried this before, you go to the site and it puts two different AI model responses side by side.

很多人用这个，因为实际上等于白嫖AI服务。

A lot of people use this because you you basically get free AI usage out of it.

不用花钱买ChatGPT，你来这儿，它会给你两个回复让你选。

Instead of having to pay for ChatGPT, you can go pay this and it will give you, you know, two responses side by side.

你挑出更喜欢的那个，其实就是在帮系统判断哪个回答更受欢迎。

You pick which one you like better, and you're helping, you know, tell it which ones are are the most popular.

然后这些排行榜就出来了，你就能看到，天啊，ChatGPT的新模型居然赢过了Anthropic的模型之类的。

Those leaderboards are then, you know, created so you can see, you know, oh my gosh, ChatGPT's new model is, you know, beating the open the, you know, the the model from Anthropic, whatever.

这就是为什么这么多排行榜公司能运转起来，为什么大家愿意用它们。

So that's kind of where a lot of these leaderboard companies, how they kind of run and why people use them.

但他们真正的盈利方式是推出了一款AI评估产品包，把数据提供给不同的AI实验室。

But how they're actually making money is that they have an AI evaluations product package, which is gonna show, data basically to different AI labs.

比如OpenAI、Anthropic和谷歌，当他们的模型在平台上接受用户投票时，平台会告诉他们哪些方面输给竞争对手了。

So like OpenAI, Anthropic, and Google, when they have their models in there getting voted on, they'll show them what areas their models are losing on to competitors, right?

他们会说：你的模型不错，但一遇到医疗问题，你的表现就差了；遇到金融问题，或者这类问题时，也一样。

So it's like, Hey, look, your model's good, but anytime a healthcare question gets asked, yours is doing poorly, or anytime a finance question, or anytime a question related to this.

他们精准指出模型的短板，帮助这些公司优化强化学习的方向。

So they're telling them exactly where their models are lacking, and they're helping them to guide reinforcement learning.

这对OpenAI和谷歌来说，价值巨大。

So it's a, you know, it's a massive value for OpenAI and Google.

他们愿意为此支付大笔费用。

They're gonna pay a ton of money for that.

对用户来说，这也很划算。

It's a great value for people.

既能免费使用AI，又能有趣地看到哪家公司的模型在排行榜上表现最出色。

They get free AI usage, and also it's interesting for all of us to see what companies are doing the best in the leaderboards.

所以默克这家竞争对手的年营收今年达到了十亿美元，而Handshake AI的训练团队规模在三个月内从五亿增长到了大约十亿美元。

So Mercker's annualized revenue, which is one of their competitors, hit $1,000,000,000 this year, and Handshake AI's training army grew from 550,000,000 to about $1,000,000,000 in three months.

我觉得这正好展示了这些后训练市场的规模，以及它们实际能赚到的钱。

I think this just kinda shows the scale of a lot of these post training markets, how much money they're actually able to make.

咱们来聊聊Palantir最近的情况。

Let's talk about what's going on with Palantir.

他们基本上已经选定英伟达的Nemotron开源模型，用于美国政府的所有AI项目。

So they have just basically selected NVIDIA's Nemotron open models to use at the US government for all of their AI.

当然，现在 Anthropic 和 OpenAI 那边闹得挺凶，美国政府也得介入，比如把他们的模型撤下来，或者告诉他们，嘿……

Of course, there's a ton of drama with Anthropic and even OpenAI right now, and the US government having to, you know, pull their models or tell them, like, hey.

你们不能发布模型，除非我们给出一个准确的评估，诸如此类。

You can't release your model till we give it, like, an accurate assessment, whatever.

对吧？

Right?

所以有一件事挺有意思，我觉得特别跟国防部有关，还有他们跟 Anthropic 之间的那些矛盾，比如把 Anthropic 列为供应链风险，因为他们不让对方做那些想做的事之类的。

So one thing that's interesting, I think, particular with the Department of War and a lot of the beef that they had with Anthropic and them, you know, classifying Anthropic as a supply chain risk because they wouldn't let them do all the things that they wanted to and stuff.

因此，为美国政府做大量软件开发的Palantir，选择了使用一个开源模型。

Because of that, Palantir, who does a lot of the development for the US government, a lot of software development has selected to use an open source model.

NVIDIA有自己的NeMotron模型。

So NVIDIA has NVIDIA has their NeMotron model.

他们正在用这个开源模型运行大量任务。

They're using this open source model, and they're using it to run a lot of stuff.

现在，美国政府内部的许多AI工作都依赖它。

Now a lot of AI, you know, work inside of the US government.

开源模型特别适合这种场景，政府之所以这么青睐，是因为它不需要把数据传回任何中心服务器。

And the reason why an open source model is great for this particular use case or why the US government would be so happy to use it is because there is no centralized, you know, servers that the the model's gotta go back to.

没人能跳出来说：

There is no one that can say, hey.

你不能用它来做X、Y、Z这些事。

You can't use it for x y z reasons.

这是个开源模型。

It's an open source model.

你可以随意使用它。

You can do whatever you want with it.

所以他们在做的一切事情上，都不用去申请审批。

So in basically everything that they're doing, they will have they won't have to go and get it signed off.

我记得有个著名的对话，是国防部门负责人和Anthropic的CEO达里奥之间的，他们说：我们需要它能完成X、Y、Z这些任务。

I remember this kind of this famous conversation between the head of war and the CEO of Anthropic, Dario, and they were saying like, hey, we need it to be able to do x y z tasks.

达里奥回应说：哦，如果你们真要这么做，从技术上讲，这违反了我们的服务条款。

And Dario was like, oh, well, if you're gonna do that, like, technically that's against our terms of service.

然后对方问：那要是真遇到紧急情况，必须做关键决策呢？

And then they're like, okay, well, what if, like, there was a moment when we really needed to make some sort of, you know, critical decision on this?

达里奥说：哦，那你随时给我打电话，如果真有特殊情况，需要绕过规则，打个电话就行，我来帮你解锁。

And he and Dario was like, oh, well, could call me and, like, I could if there's, like, an exception to, like, a rule, just call me and, like, I'll unblock something for you.

对方说：明白了。

And they're like, k.

但我们不想在进行高度机密的作战行动时，还得打电话给你求解锁。

Well, we don't wanna call you and unblock something if we're the middle of, like, a confidential, you know, battle thing, blah blah blah.

很多人对这件事都有自己的看法，但我们听到的主要是这两种观点。

So, I mean, a lot of people have a lot of opinions on this, but these are the two arguments that we hear on this.

所以这就是为什么Palantir看起来会用它来支持大量美国政府机构。

And so this is why it looks like Palantir is is gonna be using this to run a lot of US government agencies.

抛开所有这些争议，我觉得只要能保住机密数据，不用传到外部网络，很多机构都会愿意用。

And beyond all of the drama, I just think for keeping classified data without having to send that to an outside network, I think a lot of agencies will wanna use that.

而且机构还能保留自己训练的定制模型权重，这基本上解决了过去阻碍安全AI广泛应用的核心问题。

So agencies also get to keep the customized model weights that they train, which I think solves basically the core problem that's kept a lot of secure AI adoption stuck, and and being very slow in the past.

特别是Nemotron，这是一个开源模型，会在Palantir的主权AI操作系统上运行。

And in particular with Nemotron, this is an open model, and it's gonna run inside of Palantir's Sovereign AI operating system.

整个系统都会基于NVIDIA的加速硬件运行。

It's gonna be all run on NVIDIA's accelerated hardware.

系统会处理数据授权问题。

There's gonna be handling data authorization.

会有隔离机制，所有审计都会在一个封闭环路内完成。

There's gonna be isolation, and all of the auditing is all gonna be in a closed loop.

我也要说，这其实并不算离谱。

I'll also say this isn't too crazy.

目前三分之二的公司已经在使用开源模型了，我觉得很多人这么做主要是为了降低成本。

Like, two thirds of companies right now already use open models, and I think a lot of them say they're doing this because of cost efficiency.

NVIDIA 把这当作在受监管、敏感领域——比如金融、医疗和政府——扩展 AI 的标准模式。

NVIDIA frames this as kind of the template for scaling AI in regulated environments, sensitive environments, so like finance and healthcare and government, obviously.

在 AI 监管方面，伊丽莎白·沃伦和斯坎隆正在重启他们的健康与位置数据保护法案。

In AI regulation, Elizabeth Warren and Scanlon are reviving their health and location data protection act.

他们在这项法案里加了不少新内容。

They're adding a bunch of new things to it.

新法案明确禁止 OpenAI、Anthropic 和 xAI 这类公司，把用户输入到 ChatGPT、Claude 和 Grok 中的健康和位置数据拿去售卖。

It now explicitly bans AI companies like OpenAI, Anthropic, and xAI from selling health and location data users enter into Chatuchify, Claude, and Grok.

这项法案将在十年内拨款十亿美元给联邦贸易委员会来执行，本质上是在填补一个监管空白，因为很多 AI 实验室都在拼命拉用户。

The bill is gonna allocate $1,000,000,000 to the FTC over ten years to enforce this, and, basically, this is filling a regulatory gap because a lot of these AI labs are racing to get users.

说真的，就连我自己，也在用 ChatGPT。

And, I mean, even myself, you have ChatGPT.

他们刚推出了 ChatGPT Finance，当然，我老婆特别讨厌这种东西。

They just added ChatGPT Finance, which, of course, I, my wife like hates this kind of thing.

她不愿意把数据交给任何 AI 模型，这其实也有道理。

She hates giving data to AI models and it's probably for good reason.

但我自己呢，已经把所有的信用卡和银行账户都连到了我的 ChatGPT 上。

Myself though, I went and added basically all of my credit cards and bank accounts to my ChatGPT.

它是通过 Plaid 做的集成，但真的特别有用，能直接问它我们钱花在哪了、固定开销是多少、生意收入怎么样。

It does the integration through Plaid, but it's been super useful to be able to ask it questions about where we're spending, what our recurring expenses are, income from businesses.

说实话，跟它聊一聊，能挖出好多有意思的数据。

I mean, there's just so much interesting data you can get just chatting with this thing.

所以我特别喜欢这个功能。

So I love it.

我知道很多人也用它来管理健康信息，确实有不少好处。

I know a lot of people use this for health things as well, and there's a lot of benefits.

但很多人还是担心，把这么重要的财务和健康数据交给这些公司不太安全。

But a lot of people are concerned about, you know, giving all of these financial and health data to these companies.

他们拿这些数据干嘛？

What are they gonna do with it?

如果这项法案通过了，联邦贸易委员会必须在一百八十天内制定实施细则。

So the FTC has to write implementation rules within a hundred and eighty days of this bill being passed, if it is passed.

联邦机构、各州检察长以及个人都可以对任何违规行为提起诉讼。

The agency state and attorney general and individuals can all sue for any violations on this.

这项法案还有一些共同提案人。

There's some cosponsors on this.

包括参议员罗恩·温登和伯尼·桑德斯。

We have senator Ron Winden, Bernie Sanders.

原始法案是2022年6月提出的，但当时只针对数据经纪商。

The original bill is from June 2022, but it only covered data brokers.

并没有涵盖AI公司如何从上游收集数据的问题。

It didn't have AI companies and kind of how they're collecting data upstream.

聊天机器人里的健康对话内容特别丰富，因为用户会上传化验报告、扫描影像，还会不断追问各种细节。

Chatbot health transcripts are unusually rich, right, because users are gonna be pasting in their lab results, or they're gonna upload scans, they're gonna ask follow ups, all of that.

所有这些信息加起来，能形成非常完整的医疗档案，其他传统的健康类应用根本做不到这一点。

And all of that can create a lot of really complete medical profiles that no other health, you know, traditional app would actually be able to capture.

对吧？

Right?

人们上传的不只是X光片，还有更多种类的资料。

People are uploading pictures of their X rays and and so much more.

所以呢，我真的很支持这种数据保护，不让这些信息被拿来出售。

So anyways, I I do love all of this kind of data protection, not being able to sell that.

这对我来说很有道理，尽管我对伯尼·桑德斯在AI问题上的某些立场，比如禁止AI数据中心在纽约设立，并不完全认同。

That makes a lot of sense to me even though I don't agree with everything that Bernie Sanders does or says when it comes to AI in particular, banning AI data centers from New York.

你想想，这其实和科技进步以及AI能给人们带来的好处背道而驰。

I mean, just think that that's kind of counterintuitive to a lot of what progress and the benefit that these data centers and, well, the benefit that a lot of this AI can have on people.

不过撇开这些不谈，就这项法案本身而言，我确实支持伊丽莎白·沃伦和伯尼·桑德斯的立场。

But, anyways, all of that aside, this was one where I would definitely say I'm agreeing with Elizabeth Warren and Bernie Sanders on the face of this bill.

我现在也不清楚里面有没有什么隐藏的猫腻，但总的来说，别把我给你的健康数据卖给外部公司。

Now I don't know if there's secret sneaky things stuck in there, but overall, don't sell my health data to outside companies that I give you.

我完全同意这一点。

I am, definitely in agreeance with that one altogether.

好的。

Okay.

我们来聊聊 Flexion Robotics 这家公司的情况吧。

Let's talk about what's going on with Flexion Robotics.

他们正在训练人形机器人去跑办公室的杂务。

They are training humanoid robots to run office errands.

他们让这些机器人能完全自主地完成这些任务。

They're helping them do this all autonomously.

这是一家瑞士初创公司。

This is a Swiss startup.

创始人是前 NVIDIA 的研究人员，他们开发了一套软件系统，让这些人形机器人能够执行一系列复杂的办公室任务。

It was founded by ex NVIDIA researchers that built a software stack, essentially that's letting these humanoid robots do multistep office tasks.

比如送文件、取快递之类的——对了，你只要跟它们说话就能指挥它们做这些事。

So things like and, oh, and by the way, you can do this all by just talking to them.

我平时见得比较多的是波士顿动力那些狗形机器人，通常还得用遥控器来操作。

So I think I'm really used to seeing those, like, dog robots from Boston Dynamics, and there's, like, remote controls where you can kind of control them.

我在各种展会上也见过它们。

I've seen them at, like, conferences and stuff.

但这个就不一样了。

This will be different.

你是直接走到人形机器人面前，跟它说：嘿。

This is you're literally going to humanoid robot and you're like, hey.

去给我买杯咖啡。

Go get me a coffee.

当然，让机器人去买咖啡可能不是它最该干的事。

Although I'm sure there's, you know, better things than telling it to go get you a coffee.

它更可能是到处走动，帮忙传真、扫描文件，或者把东西送到不同人的办公桌上。

It'd probably be, like, walking around and and getting things faxed or scanned or dropping things off different people's desks.

我也说不准。

I don't know.

对吧？

Right?

这其实还挺有用的。

This could actually be quite useful.

我几乎能想象到这个场景。

I almost imagine this.

比如在办公室里，每个人桌上可能都有个按钮，一按就能叫来这个机器人，就像我用Claude时，点一下语音输入按钮那样。

Like, I was in an office setting, everyone would have, like, a little microphone button maybe that, like, summons this thing, and you could, like, think of how you kind of use, like, Claude, or I guess how I use Claude is I hit, like, my voice to text button.

我下达指令，它就去办了。

I I give it an instruction and it runs off doing it.

整个办公室可能就配一台人形机器人，没必要人手一台，大概每层楼放一台就够了。

Let's say there's one humanoid robot for the entire office, not everyone needs their own obviously, so you could have like one or two, maybe one per floor or something like that.

你按个按钮，告诉它要做什么，它就走到你桌前，拿走文件、纸张，或者你让它处理的任何东西，然后去执行任务。

You push a button, you give it its command, you tell it what you want it to do, it comes over to your desk, it grabs the document, grabs the paper, grabs the, you know, whatever you need it to move or do, and then it goes and executes it.

这就像是Claude在现实世界里的版本。

It's like a real world version of Claude.

如果你没看出来的话，我真的很喜欢这个。

I love this if you can't tell.

这本质上是一个经过改造的通用人形机器人，负责运行他们所有的系统。

Basically, is a modified unitary humanoid robot, and it's running all of their systems.

他们让它去取邮件之类的活儿。

It's, you know, they're getting it to, like, go retrieve, like, mail and stuff.

据说它能上下楼梯。

Apparently, it can go up and down stairs.

它还能坐电梯。

It can go in the elevator.

功能超强，他们最近发布的演示视频里，它甚至在摆货架。

It's super capable, and they have, it stocking a shelf in a recent demo that they released.

所以说，它能力真的很强。

So it's very capable.

目前，他们在整个系统的每一层都用上了强化学习。

Right now, they're using reinforcement learning at basically every layer of what they've built.

他们有一套完整的规划系统。

They have a master planning.

他们搭建了模拟环境。

They have simulation environments.

他们有自己的电机控制器。

They have motor controller.

如果看整个市场，ABI研究公司预测，到2036年，机器人基础模型市场的规模将达到1500亿美元。

If you look at the overall market, ABI Research is projecting that robot foundational model markets are gonna hit about a $150,000,000,000 by 2036.

Flexion的做法是把软件本身当作核心产品，他们说，硬件已经变成大宗商品了。

Flexion is basically trying to position some of the software as the product, and they're saying, look, the hardware's a commodity.

未来会冒出一大批公司，都在做这种人形机器人，比如特斯拉。

There's gonna be a ton of these companies that are building these humanoid robots, maybe Tesla.

我们不跟他们比硬件，我们比的是软件——能让你家、你办公室里的机器人学会适应你的环境、听懂你的指令的软件。

We're not gonna compete on there, but we're gonna compete on the software that helps train these robots to do things in your environment, in your office, and and how you guys actually control that.

说实话，我特别喜欢这个思路。

So honestly, I kind of love that.

这样一来，人形机器人就更像一个大语言模型了，而且围绕它的各种应用也在不断涌现。

It's moving it makes, you know, the humanoid robot feel more like an LLM, and there's all of the stuff getting built on top of it.

Flexion 的设计是不依赖特定硬件的，他们跟多家人形机器人厂商都有合作。

Flexicon is hardware agnostic, so they're saying they work with a whole bunch of different humanoid robot makers.

他们自己不会去造机器人。

They're not gonna build their own robot.

他们会让你用 Unitree、Figure、X1，还有其他好几个平台。

They're gonna be able to let you use Unitree, Figure, X1, and and a bunch of other platforms.

中国 CXMT 公司刚和腾讯签下了三十亿美元的内存供应合同。

China's CXMT has landed a $3,000,000,000 memory supply deal with Tencent.

这大概是国内 DRAM 厂商有史以来最大的一笔商业订单了。

This is the largest commercial contract, I think, in basically domestic DRAM makers in like their entire history of their company.

这笔交易说明，中国最大的云服务商现在愿意从本土供应商采购 AI 所需的内存，因为美国对先进芯片的出口管制越来越严了。

The deal right now I think is showing that China's largest cloud operators are now actually willing to source AI memory from local suppliers because a lot of The US export controls are tightening around basically all of the advanced chips.

对吧？

Right?

所以腾讯是个大客户。

So Tencent is a huge customer.

他们需要的数量太大了。

They need so much of this.

过去他们一直从美国公司采购，但现在因为出口管制，似乎能从中国公司买到这些产品了。

And traditionally, they have bought a lot of this from American companies, but it seems like they're able to get this from Chinese companies now because of these export controls.

特别有意思的是，CXMT 正从智能手机和 PC 厂商转向超大规模的 AI 客户，这对他们来说应该是头一回。

Something interesting about this company in particular is that CXMT is kinda moving from, like, smartphones and PC OEMs to now hyperscale AI customers, and I think that's kind of the first time for them.

这跟 SK 海力士当年的做法很像，他们之所以能占据主导地位，就是因为拿到了英伟达这样的客户。

It's very similar to what SK Hynix did, and they kind of, you know, hit this dominance path because they were getting customers like NVIDIA to to buy from them.

我很想知道 CXMT 能不能走出中国，真正实现全球主导，还是说因为云、AI 和数据中心建设这些敏感因素，只能局限在国内市场。

I'll be interested to see if CXMT is able to move outside of just China, and really get more global dominance, or if it's something that's just gonna be geographically landlocked because of kind of the sensitive nature of everything happening with cloud and AI and and all of the data center build outs.

今天这期播客真有意思。

This podcast was a lot of fun today.

我正坐在瑜伽球上，抱着我的孩子。

I'm actually sitting here on a yoga ball holding my, baby.

几周前我当爸爸了，希望你没怎么听到他哼哼的声音。

I got a new baby as of a couple weeks ago, and, hopefully, you weren't able to hear him grunting too much.

他现在睡着了，应该挺舒服的。

He's sleeping, so he should be pretty happy.

不过没关系，我会继续为你更新这些播客的。

But, anyways, I'm gonna keep getting these podcasts out for you.

AI领域的新闻真是太多了。

There's so much happening in AI News.

我特别喜欢这些东西。

I love this stuff.

我对这个领域充满热情。

I'm super passionate about it.

今天聊得很开心，但如果你听到背景里有点小动静，别介意，我们正在和Clay一起在录音棚里录呢。

But lots of fun going over, but excuse any, little sounds you might be hearing in the background, as, we're we're recording in the studio with Clay.

如果你不介意的话，给个评价吧，这对我们帮助特别大。

So, anyways, if you wouldn't mind leaving a review, it would help to show out a ton.

我都特别感谢。

I appreciate them all.

我都认真看了。

I read them all.

不过话说回来，非常感谢，如果你想体验各种AI模型，一定要去试试 aibox.ai，特别是那个 MCP 功能。

But anyways, thanks so much, and make sure to check out aibox.ai, the MCP in particular, if you wanna get access to all of the different AI models.

平台上一共有大约八十种模型，你也可以把它们全都集成到 Claude 里，这样就能在 Claude 里直接处理图片、音频和视频了。

There's about 80 of them in one platform, or you can build them all into Claude so you get image, audio, video, all inside of Claude.

好了。

Alright.

你们早就知道了。

You guys already know.

链接在简介里。

The link's in the description.

我们下期再见。

I'll catch you in the next episode.