The Information's TITV - 英伟达与思考机器达成交易,腾讯加入中国AI代理竞赛,Vibe编码范式转变 封面

英伟达与思考机器达成交易,腾讯加入中国AI代理竞赛,Vibe编码范式转变

Nvidia–Thinking Machines Deal, Tencent Enters China AI Agent Race, Vibe Coding Paradigm Shift

本集简介

Menlo Ventures的Venky Ganesan与TITV主持人Akash Pasricha讨论了英伟达与Vera Rubin芯片的交易及其对Thinking Machines Lab的投资。我们还与《The Information》的Aaron Holmes探讨了微软新推出的Office + Copilot捆绑产品及其反垄断风险,并与财经编辑Ken Brown讨论了亚马逊为资助AI基础设施而进行的420亿美元债券发行。随后,我们与Tencent的Juro Osawa和Jing Yang探讨了微信AI代理,以及与South Park Commons普通合伙人Aditya Agarwal讨论了“氛围编码”的范式转变。 本集讨论的文章: https://www.theinformation.com/articles/tencent-joins-chinas-ai-agent-race-top-secret-wechat-project https://www.theinformation.com/articles/org-chart-microsoft-legal-staff-girding-cloud-bundling-suits https://www.theinformation.com/newsletters/applied-ai/microsoft-doubles-seat-based-pricing-ai https://www.theinformation.com/briefings/amazon-raising-42-billion-bonds 订阅: YouTube:https://www.youtube.com/@theinformation The Information:https://www.theinformation.com/subscribe_h 订阅AI议程简报:https://www.theinformation.com/features/ai-agenda TITV每周一至周五太平洋时间上午10点/东部时间下午1点在YouTube、X和LinkedIn播出。您也可以在您收听播客的平台找到我们。 关注我们: X:https://x.com/theinformation IG:https://www.instagram.com/theinformation/ TikTok:https://www.tiktok.com/@titv.theinformation LinkedIn:https://www.linkedin.com/company/theinformation/

双语字幕

仅展示文本字幕,不包含中文音频;想边听边看,请使用 Bayt 播客 App。

Speaker 0

欢迎各位收看Information的TITV。

Welcome everyone to the Information's TITV.

Speaker 0

我叫阿卡什·帕什里卡。

My name is Akash Pasricha.

Speaker 0

今天是3月10日,星期二。

It is Tuesday, March 10.

Speaker 0

今天首先报道的是,英伟达正与思维机器实验室合作。

First up today, Nvidia is teaming up with Thinking Machines Lab.

Speaker 0

该公司宣布与TML达成一项长期战略合作,计划使用英伟达即将推出的维拉·鲁宾芯片部署至少一吉瓦的算力。

The company has announced a long term strategic partnership with TML, planning to deploy at least one gigawatt of computing power using Nvidia's forthcoming Vera Rubin chips.

Speaker 0

我们将邀请来自Menlo Ventures的合作伙伴解析这笔交易,同时也会讨论另外两笔引发关注的重大交易。

We'll break down that deal with a partner from Menlo Ventures, and we'll also talk about two other big deals making news.

Speaker 0

Axios报道称,Meta收购了MoltBook,此外还有OpenAI收购PromptFu的消息。

Axios is reporting that Meta has acquired MoltBook, and we also have OpenAI's PromptFu acquisition.

Speaker 0

接下来,我们的微软记者将加入我们,讨论该公司在本周推出全新Office 365产品组合之际的战略调整。

Next up, our Microsoft reporter joins us to discuss the company's evolving strategy as it launches a new bundle of Office three sixty five products this week.

Speaker 0

我们还将与我们的财经编辑深入探讨亚马逊大规模的新债券发行。

We'll also dig into Amazon's massive new bond sale with our finance editor.

Speaker 0

最后,我们将转向一篇观点文章。

And finally, we will turn to an opinion piece.

Speaker 0

南公园社区的一位顶级投资者最近在《The Information》上发表文章,讲述了他个人如何应对‘氛围编程’这一范式转变。

A top investor at South Park Commons recently published in the information about how he is personally grappling with the vibe coding paradigm shift.

Speaker 0

这是一期内容丰富的节目,让我们马上开始。

It is a big show, so let's get right on into it.

Speaker 0

Mirror Moradis Thinking Machines Lab与英伟达签署了一项十亿瓦的协议,将使用其即将推出的Vera Rubin芯片用于其AI产品,这是又一项具有里程碑意义的芯片交易。

Mirror Moradis Thinking Machines Lab has signed a one gigawatt deal with Nvidia to use its upcoming Vera Rubin chip for its AI products in another landmark chip deal.

Speaker 0

我想邀请Menlo Ventures的合伙人Venky Ganesan来帮我们解析这笔交易。

I want to bring on Menlo Ventures partner Venky Ganesan to help us break down this deal.

Speaker 0

Venky,今天早上宣布这笔交易时,你对此有何看法?

Venky, what did you make of this deal this morning that was announced?

Speaker 1

嘿,Akash,很高兴能上你的节目。

Hey, Akash, great to be on your show.

Speaker 1

一个老观众,第一次成为粉丝。

A longtime viewer, first time fan.

Speaker 1

所以我想说,Thinking Labs能与Nvidia合作,获得接近十亿美元的投资,这真是一个令人兴奋的进展——他们通常说一吉瓦的算力相当于十亿美元的投资。

So I wanted to just say, I think it's a very exciting development for Thinking Labs to get Nvidia on board, to get close to a billion dollars of they usually say a gigawatt of capacity is almost like a billion dollar investment.

Speaker 1

这对Menlo和Moradis来说是一种认可,也证实了他们所做事情的价值。

So it's a validation for Menlo and Merardi, it's a validation of what they're doing.

Speaker 1

而最终,这既是资金支持,也是算力支持。

And ultimately, it's both funding and compute.

Speaker 1

我认为对Nvidia来说,这简直是不二之选。

I think for Nvidia, it's a no brainer.

Speaker 1

他们锁定了另一个新的实验室,将其纳入自己的架构体系。

They're locking up another new lab to be in the architecture.

Speaker 0

让我觉得特别突出的是,Thinking Machines Lab显然还是一家较早期、较年轻的初创公司。

The thing that sort of stood out to me is that Thinking Machines Lab, obviously it's an earlier, a younger startup.

Speaker 0

他们在新闻稿中表示,目标是在明年实现部署。

They said in a press release that they're targeting deployment next year.

Speaker 0

当然,我们说的是维拉·鲁宾望远镜还没有正式发射。

We of course mean, the Vera Rubin hasn't even been launched yet officially.

Speaker 0

我们正在等待下周的GTC大会,看它如何亮相。

We're waiting for GTC next week to see it being unveiled.

Speaker 0

我的意思是,这留下了一个空白期,我们还需要很多时间来弄清楚如何部署它,甚至招聘人员来让它运转起来。

I mean, it sort of left this runway of we still need a lot of time to figure out how we're going to deploy this and even hire the people to make it work.

Speaker 1

我认为有些人可能没有意识到,当一款新芯片发布时,需要几个月的时间才能让它稳定下来。

And the thing that I think some people may not realize is when a new chip comes out, it takes a few months for it to get smoothened out.

Speaker 1

大多数实验室都不会。

Most labs will not.

Speaker 1

大多数人不会立即采用新芯片。

Most people will not go on the new chip right away.

Speaker 1

他们希望有一些时间。

They want some time.

Speaker 1

因此,对英伟达来说,好处是他们锁定了一个高知名度的NeoLab作为最早的客户。

So the benefit for Nvidia is they're locking in a high profile NeoLab to be the earliest customer.

Speaker 1

对Thinking Machines Labs来说,好处是他们获得了算力和计算资源,我相信还有来自英伟达的投资。

The benefit for Thinking Machines Labs is that they're getting capacity and compute, and I'm sure investment from Nvidia.

Speaker 1

所以这感觉像是双赢。

So it feels like a win win.

Speaker 1

双方都能从中获益。

Both parties get something out of it.

Speaker 1

对英伟达而言,这是一个不错的时机来推出Vera Rubin,因为大家都知道,在任务初期的几个月里肯定会遇到一些问题。

And for Nvidia, it's a nice way to launch the Vera Rubin, because everybody knows they'll be teething troubles in the first few months of the trip.

Speaker 0

我问你一下,风投和初创圈里对Thinking Machines Labs正在构建的东西有多少讨论?

Let me ask you, how much chatter is there in the venture and start startup community about what Thinking Machines Lab is building?

Speaker 0

我们当然知道他们以100亿美元的估值融资了20亿美元。

We of course know that they raised $2,000,000,000 at a $10,000,000,000 valuation.

Speaker 0

我们还报道过,他们计划在未来某个时候锁定500亿美元的估值。

We've also reported they're looking to lock in a $50,000,000,000 valuation at some point.

Speaker 0

我的意思是,这是一家极其耀眼的公司。

I mean, this is as flashy a company as it gets.

Speaker 0

你关注这家公司实际生产出什么产品呢?

What are you watching for in terms of what this company actually produces in terms of product?

Speaker 1

稍微背景介绍一下,Menlo Ventures 全力投入人工智能领域。

A little bit of context, Menlo Ventures is all in on AI.

Speaker 1

我们一直在关注并投资了不少神经科技公司,包括像 Acxiom、MAT 这样的公司,它们涉及飞行器领域。

We have been looking and have invested a bunch of neolabs, including companies like Acxiom, MAT, which is in the and flapping airplanes.

Speaker 1

所以对我来说,真正有趣的是,最近关于 Thinking Machines 有一些消息,比如团队是否有人员流动,到底发生了什么?

And so to me, what is really interesting is that there's been a little bit of news around Thinking Machines, have they had some team turnover, what's going on?

Speaker 1

他们刚刚推出了 Tinker。

They just launched Tinker.

Speaker 1

事实上,很少有人知道他们内部真实的情况和长期计划,但这对公司来说是一个非常重要的里程碑。

The reality is very few people know what's really going on and what their long term plans are, but this is a very meaningful milestone for the company.

Speaker 1

能获得 Nvidia 和 Jensen 这样的承诺,让他们投入新芯片并提供投资,我认为这是一个重要的里程碑。

To get this kind of commitment from Nvidia and Jensen for them to put their new chips and to lay some investment, think it's a milestone.

Speaker 1

我认为这要归功于 Vera Rubin 让这一切成为可能。

I think it's a kudos to Vera Rubin for making this happen.

Speaker 0

你下周会去参加GTC大会吗?

Are you going to be at GTC next week?

Speaker 1

当然会去。

Absolutely.

Speaker 1

嗯,我本来打算在那里吃晚饭。

Well I'm supposed to get dinner there.

Speaker 0

这就对了。

There you go.

Speaker 0

好吧,我们以后得再请你回来聊聊这个大会。

Okay, well, we'll have to have you back on to talk about that conference.

Speaker 0

我想转到谈谈你一直关注的一个领域——网络安全。

I want to pivot to talking about a space that you spend a lot of time in, cybersecurity.

Speaker 0

我们这周看到OpenAI收购了一家名为PromptFu的公司。

We saw this week that OpenAI is acquiring a company called PromptFu.

Speaker 0

这家公司在你的关注范围内吗?

Is this a company that you have followed?

Speaker 1

是的,我们对红色主题漏洞领域进行了深入跟踪。

Yes, we follow the red theming vulnerability space very intensely.

Speaker 1

网络安全是我关注的重点领域。

Cyber's a big area of focus for me.

Speaker 1

我很幸运曾投资过Palo Alto Networks,并在那里担任董事会成员六年。

Was fortunate to have been an investor in Palo Alto Networks and been on the board there for six years.

Speaker 1

我要告诉你们的是,发令枪已经响起,比赛已经开始了。

What I'll tell you is the bell has rung and the race has begun.

Speaker 1

人工智能和网络安全正走向正面碰撞,在我看来,OpenAI收购PromptFu应该让所有网络安全公司和其他AI模型都感到不寒而栗,因为他们表明自己是来认真参赛并志在必得的。

AI and cybersecurity are in a collision course, and to me, OpenAI buying prompt food should send shivers down the spine of all cyber companies and other AI models, because they are showing they are here to play and they're here to win.

Speaker 1

我认为这是OpenAI的一个绝佳举措。

Think it's a great move by OpenAI.

Speaker 0

让他们不寒而栗。

Shivers down their spine.

Speaker 0

再多说说这一点。

Say more about that.

Speaker 0

你是什么意思?

What do you mean?

Speaker 1

我只是觉得,当Claude几周前发布他们的产品时,你看到网络安全公司股价下跌了。

Well, I just think that when Claude released their product a few weeks ago, you saw cybersecurity companies take a hit.

Speaker 1

那时更多是说,这是Claude代码可能的应用方式,但他们并没有真正提供更多细节进行发布。

At that point, it was much more sort of a, hey, this is how Claude code could be applied, but they didn't really have they didn't actually launch out with more detail.

Speaker 1

OpenAI正在真正投身于网络安全领域,并且表明他们打算进行收购。

What OpenAI is doing is they are really committing to the cybersecurity journey, and they're showing that they're going to acquire.

Speaker 1

我实际上认为,这意味着AI公司将会开始收购网络安全公司。

I actually think that it means that AI companies are gonna start acquiring cyber.

Speaker 1

所以,网络安全和AI是相辅相成的。

So cyber and AI is a pad trade.

Speaker 1

每次AI发展,都会扩大攻击面,从而扩大网络安全的市场。

Every time AI happens, you increase the surface area and you increase the market for cyber.

Speaker 1

而且,坏人利用AI的速度比好人更快。

Also, bad guys can use AI faster than the good guys can.

Speaker 1

所以对于这两点,我认为人工智能对网络安全是巨大的助力。

So for both things, I think AI is a huge tailwind for cyber.

Speaker 1

但现在OpenAI所表达的意思是,我们不仅要自己构建,还可能收购你。

But now what OpenAI is saying is not only are we gonna build it, but we're also gonna potentially acquire you.

Speaker 1

所以我确实认为这场竞赛已经开始了,接下来将非常有趣,看看网络安全公司如何应对这一人工智能威胁。

So I I do think that this the race has begun, and it's going to be a very interesting time to see how cyber companies deal with this AI threat.

Speaker 0

你能从很高的层面给我们解释一下,OpenAI在其前沿产品中试图构建什么,以及这如何融入其中吗?

Can you explain to us at a very high level what OpenAI is trying to build with its frontier offering, and then how this fits into that?

Speaker 1

我得先说明我的偏见。

So I have to say my bias.

Speaker 1

我们是Anthropic最大的投资者之一。

We are one of the largest investors in Anthropic.

Speaker 1

我们是一家Anthropic公司,所以我花了很多

We are an Anthropic company, so I spent a lot

Speaker 0

嗯,你为了

of Well, you for

Speaker 1

在推进这件事。

making it.

Speaker 1

所以我认为OpenAI正在迎头赶上。

So my guess is OpenAI is catching up.

Speaker 1

我的意思是,OpenAI没有意识到编程会变得如此重要。

I mean, OpenAI didn't realize how big coding was going to be.

Speaker 1

云平台很早就占据了领先地位。

Cloud took a very early lead.

Speaker 1

现在OpenAI正通过Codex努力追赶。

OpenAI is now going with Codex and trying to catch up.

Speaker 1

我认为他们意识到的一件事是,网络安全将会变得至关重要,尤其是云平台——他们就在一周前,可能两天前,刚发布过关于在浏览器中发现的所有漏洞的内容,对吧?

And I think one of the things they've realized is that Cyber is going to be and especially, you know, Cloud has been they just published a week ago, maybe two days ago, about all the vulnerabilities they found in browsers, right?

Speaker 1

所以OpenAI正在试图对此做出回应,我认为他们的对策是收购PromptFlu,以便将网络安全功能内置到他们的模型中。

And so OpenAI is trying to have an answer to that, and I think their answer is to acquire PromptFlu so they can start having that cyber functionality built into their model.

Speaker 0

我很感兴趣,你是Anthropic的投资人,能帮我整体了解一下AI网络安全领域的状况吗?

And I am curious, you are an investor in Anthropics, so just help me understand the AI cybersecurity landscape at large.

Speaker 0

你提到你在关注更多的收购行为。

You mentioned you're watching to see more acquisitions.

Speaker 0

我们应该关注网络安全领域的哪些方面?

What pockets of cybersecurity should we be watching for?

Speaker 1

我认为人工智能真正擅长的是识别漏洞。

So I think what AI is going to be really good at is in identifying vulnerabilities.

Speaker 1

本质上,Claude 已经展示了这一点,我认为 OpenAI 也正在展示这一点。

So in essence, already Claude has shown that, I think OpenAI is showing that.

Speaker 1

因此,网络安全公司的机会在于,一旦人工智能识别出漏洞,你该如何修补和修复它们?

And so the opportunity for cyber companies is that once AI identifies vulnerabilities, how can you patch them and remediate them?

Speaker 1

我们一直花很多时间思考的一个领域是如何实现自动修复。

One of the areas we've been spending a lot of time in is just thinking about how do you do automatic remediation.

Speaker 1

我认为网络安全的终极目标是能否在零天内自动修复漏洞。

I think the holy grail for cyber is can you automatically remediate your vulnerabilities in zero days.

Speaker 1

我们非常幸运地投资了一些杰出的公司,包括 Zaffron,它们利用人工智能来自动修复你的漏洞。

We've been very fortunate to back some amazing companies, including Zaffron, that actually help in using AI to automatically remediate your vulnerabilities.

Speaker 0

我想转到我们本周关注的第三笔交易。

I want to turn to a third deal that we are watching this week.

Speaker 1

有一个

There's a

Speaker 0

今天新闻真多,我知道。

lot of news today, I know.

Speaker 0

我知道。

I know.

Speaker 0

我们今天的交易都很契合,Venky,感谢你配合我们。

Are our deals correspondent today, Venky, so thank you for playing ball with us.

Speaker 0

今天早上我们看到了Axios的报道,称Meta收购了MoltBook,并且吸纳了MoltBook的创始团队。

So we had the Axios report this morning that Meta has acquired MoltBook and that they're bringing on the founders that worked on MoltBook.

Speaker 0

MoltBook当然就是那个基于代理的社交网络,曾经在硅谷引起轰动。

MoltBook, of course, was that agent based social network that sort of took Silicon Valley by storm.

Speaker 0

当然,我们也看到OpenClaw的Peter Steinberger被OpenAI聘用了。

We of course saw that Peter Steinberger from OpenClaw was hired by OpenAI.

Speaker 0

我的意思是,只是想知道你对此有什么看法。

I mean, just wondered what you thought here.

Speaker 0

Meta,我们还在等待他们的超级智能实验室会推出什么成果。

Meta, we're still waiting to see what comes out of their superintelligence lab.

Speaker 0

我们看到OpenAI正在开发Sora和他们的社交网络。

We saw OpenAI working on Sora and their social network.

Speaker 0

你是怎么看待这个问题的?

How are you thinking about this?

Speaker 1

我想到的是,2026年将会是智能代理的元年,对吧?

What I'm thinking about is that 2026 is going to be the year of agents, right?

Speaker 1

我的意思是,如果你还记得今年早些时候,Meta还收购了Manus,这是一家总部位于新加坡的中新加坡公司,专注于多种代理型软件。

I mean, if you know, remember earlier this year, Meta also bought Manus, which is a sort of Chinese Singaporean company based out of Singapore that does a bunch of agentic software.

Speaker 1

通过这次收购,我认为Meta正在全力投入。

With this acquisition, I think Meta is going all in.

Speaker 1

我现在可能会想,马克是否像亚历山大·王在安全超级智能方面那样开心呢?

Now I might wonder, like, is Mark quite as happy as Alexander Wong on safe superintelligence?

Speaker 1

看起来,你知道的,他是在传递一个信号,表明他想分散自己的投资,但他

Seems like, you know, he's sending a signal that he wants to diversify his bets, but he's

Speaker 0

不过他发了一张新照片。

He selling a new posted that photo, though.

Speaker 0

我们别忘了。

Let's not forget.

Speaker 0

你看到Threads上的那张照片了吗?

Did you see the photo on threads?

Speaker 1

没有,我没看到。

No, I didn't.

Speaker 1

好的。

Okay.

Speaker 1

我是Twitter用户,所以好吧。

So I'm a Twitter guy, so Okay.

Speaker 0

好的。

Okay.

Speaker 0

好吧,网上之前有一些关于马克和亚历山大·王之间关系破裂的传闻,但他发了一张照片在Threads上,笑容满面地和亚历山大·王在一起,还说:‘别担心,朋友们。’

Well, let's just so there were some rumors online of a rift between Mark and Alexander Wang, he posted a photo on threads, him smiling as wide as could be with Alexander Wang saying, Don't worry, guys.

Speaker 0

一切都好。

Everything's okay.

Speaker 0

I

Speaker 1

我确信亚历山大在达成这笔交易后肯定会笑得很开心,所以看到马克也笑得很开心,我很高兴。

definitely know Alexander is going be smiling after that deal, so I'm glad Mark is also smiling.

Speaker 1

但回到Mold Book的话题,我认为Meta想传达的是:今年是智能代理的一年。

But coming back to mold book, I do think that what Meta is telling you is this is the year of agents.

Speaker 1

本质上,智能代理框架将占据主导地位,他正试图——我的意思是,你看一下。

Essentially, agentic frameworks is gonna take over, and he is trying to I mean, I give look.

Speaker 1

对我来说,马克和Meta从来都没有退出过。

I for me, Mark and Meta, they're never out.

Speaker 1

我认为他是认真的,一定会让AI在Meta取得成功。

I think he is committed, and he's going to make AI work at Meta.

Speaker 1

在某些方面,他既有计算能力,也有资金去尝试实现这一点。

And in some ways, he's got both the capacity to compute and the capital to go out and try to do this.

Speaker 1

我认为这是他为Meta构建AI武器库的又一步骤。

And I think this is another step in his direction of building that AI arsenal at Meta.

Speaker 0

对。

Right.

Speaker 0

很好。

Great.

Speaker 0

好了,文基,感谢你前来做客。

Well, Venky, I want to thank you for coming on.

Speaker 0

这位是来自Menlo Ventures的合伙人文基·加内桑,正在TITV节目中。

That is Venky Ganesan, a partner at Menlo Ventures here on TITV.

Speaker 0

微软本周推出了一套新的Office 365产品组合。

Microsoft launched a new bundle of Office three sixty five products this week.

Speaker 0

这是该公司试图从现有客户群中获得更多收入的一部分战略。

It is part of the company's ambition to try to get more revenue out of its existing customer base.

Speaker 0

为了帮助我们理解这些内容,我想邀请我们的微软记者亚伦·霍尔姆斯。

To help us understand what is what, I want to bring on Aaron Holmes, our Microsoft reporter.

Speaker 0

亚伦,帮我们解释一下这个新套餐是什么。

Aaron, help us understand what this new bundle is.

Speaker 2

本质上,这是微软首次将Office 365与Copilot 365结合在一起,后者是一款由OpenAI和Anthropic模型驱动的AI产品,此前已作为独立附加服务销售了大约三年。

So, essentially, this bundle is for the first time Microsoft combining Office three sixty five with, three sixty five Copilot, which is the AI product powered by OpenAI and Anthropic Models that they have been selling as a separate add on for about three years now.

Speaker 2

具体来说,他们不仅将这两款产品整合在一起,还加入了一些网络安全工具,以及一款专为管理AI代理设计的新软件。

And, specifically, you know, they this combines those two products as well as some cybersecurity tools and a new piece of software that's kind of geared towards managing AI agents.

Speaker 2

而且它的价格有点高,每人每月99美元。

And it's a bit steep at 99 per user per month.

Speaker 2

但这是微软首次尝试将AI与Office打包成一个整体产品提供给客户,目的是让那些已经为其中一项服务付费的Office用户愿意花更多的钱。

But this is essentially the first time that we're seeing Microsoft try to present, you know, AI plus Office as one single bundle to customers in an attempt to get, you know, Office subscribers that are already paying for one of those things to spend more money.

Speaker 0

所以我想确认一下。

So I just want to understand.

Speaker 0

在推出这个套餐之前,微软提供的是一个Office套件产品,包括Word、PowerPoint、Excel等。

So before this bundle, so there was one offering for the Office suite of products, which was Microsoft Word, PowerPoint, Excel, etcetera.

Speaker 0

然后之前还有一个独立的Copilot订阅,现在它们被合并在一起了。

And then there was a separate subscription for Copilot, and now they've joined them together.

Speaker 2

没错。

That's right.

Speaker 2

再加上一些额外的网络安全产品和这个新的代理工具。

Plus some additional cybersecurity products and this new kind of agent product.

Speaker 0

明白了。

Okay.

Speaker 0

那么在合并之前,这两个订阅分别多少钱?现在新的套餐又是多少?

And so how much were the two subscriptions separately before, and then how much is the new bundle?

Speaker 2

之前最贵的Office版本,也包含一些网络安全软件,大约是每月每用户60美元。

So the the kind of, most expensive version of Office that you could get before, which also had some cybersecurity software, that was about $60 per seat per month.

Speaker 2

而Copilot,也就是AI工具,是每月每用户30美元的附加服务。

And then Copilot, the the AI tools was a $30 per seat add on.

Speaker 2

所以两者加起来大约是90美元。

So those combined would have been about $90.

Speaker 2

这个新套餐比那还要贵。

This new bundle is even more expensive than that.

Speaker 2

每月99美元,因为他们增加了Agent 365产品和更多的网络安全软件。

It's $99 per month because they're adding this Agent three sixty five product and even more cybersecurity software.

Speaker 0

好的。

Okay.

Speaker 0

但这里的关键是,你现在无法单独购买Office订阅了。

But the whole point here is that you can't get the Office subscription on its own now.

Speaker 0

必须支付Copilot套餐的费用。

Have to pay up for the Copilot bundle.

Speaker 2

如果你愿意,还是可以单独购买的。

You can get it on its own if you want to.

Speaker 2

据我从客户那里听到的消息,很多人仍然计划这样做。

And I think from what I've heard from customers, a lot of people are still planning to do that.

Speaker 2

这本质上就像是一个诱饵,让客户同时购买这两者:通过整合一些额外的产品和这种新的代理软件,而我听说,对于那些考虑部署自己AI代理的客户来说,这可能会很有吸引力,他们希望确保自己从微软其他产品中已获得的网络安全和合规保护,也能适用于他们的AI代理。

Essentially, this is just kind of a carrot for customers to buy both of them together by, you know, rolling in some additional products and and this new agent software, which I'm hearing could potentially be of interest to customers who are thinking about, you know, deploying their own AI agents and wanna make sure that they can get the, you know, cybersecurity and compliance protections that they might already be getting from Microsoft for other types of products to apply to their AI agents.

Speaker 0

我想回到你报道中探讨过的一个问题,就是人们到底从Copilot中获得了多少实际用途。

And I wanna return to a question that you've explored in your reporting, the question around how much use people are getting out of Copilot.

Speaker 0

关于人们是否真的从每月30美元的订阅中获得了价值,目前情况如何?

Where are we on that issue here of people actually getting value for that $30 a month subscription they were paying for?

Speaker 2

所以这是一个尚无定论的问题。

So it's kind of an open question.

Speaker 2

我们所知道的是,截至一月,微软表示,只有大约3%的Office 365客户同时购买了Copilot附加服务。

What we do know is that as of January, Microsoft said that around 3% of the total number of customers that have Office three sixty five are also paying for the Copilot add on.

Speaker 2

因此,真正已经为Copilot付费的用户比例非常小。

So that's a pretty small slice who are actually already paying for Copilot.

Speaker 2

我从与客户的交流中也了解到,有时候人们会担心,比如为公司员工购买了数百个Copilot席位后,却发现员工并没有真正使用它们。

I've also heard just from talking to customers that, you know, there is sometimes a concern that you might buy, you know, however many 100 seats of Copilot for staff at your company only to find out that people aren't necessarily using them.

Speaker 2

据我了解,推动Copilot更广泛地使用,也是微软内部的一个关键重点。

And, you know, it's my understanding that that also has been a key focus inside Microsoft is driving even more usage of Copilot.

Speaker 2

有趣的是,微软一直在追赶Anthropic,后者今年早些时候因其‘同事’产品获得了大量关注。

And, you know, one thing that's interesting there is Microsoft, has been kind of playing catch up with Anthropic, which has gotten a lot of hype earlier this year for their coworker product.

Speaker 2

我们实际上看到微软昨天宣布,将在其Office 365 Copilot中内置Anthropic模型,推出自己的Anthropic Co-Work版本,我认为这是为了让Copilot对用户更有用,并希望追上Anthropic在该产品上获得的热度。

We actually saw Microsoft just yesterday announced that it would have its own version of Anthropic co work powered by Anthropic models inside of its Office three sixty five Copilot, which I think is, you know, an effort to make Copilot even more useful to people and hopefully catch up on some of the hype that Anthropic has gotten for that product.

Speaker 0

所以为了确认一下,现在这个新套餐包含了Office产品、Copilot、安全产品,同时还加入了这个Co-Work产品,它本质上就是Anthropic Co-Work的原版。

And so just to confirm here, so now so this new bundle now, it it has the Office products, it has the Copilot, it has the security products, and then it also has this co work product, which is it's literally just what Anthropic Co work had.

Speaker 0

只是打上了微软的标志。

It's just under a Microsoft logo.

Speaker 0

是这个意思吗?

Is that the idea?

Speaker 2

是的。

Yeah.

Speaker 2

实际上,它比Anthropic推出的产品要简单一些。

It's actually a little bit less advanced than what anthrop Anthropic has put out.

Speaker 2

Anthropic的产品旨在自动化你在电脑上能做的任何工作,能像人类一样控制鼠标和键盘,而这个产品仅限于在Office 365软件内执行高级任务,比如撰写研究报告,或根据你邮箱里的所有笔记自动生成会议议程等。

Anthropic's product is meant to kind of automate any work that you can do on a computer and and, you know, controls the mouse and keyboard similar to a human would, whereas this is just something that can do advanced tasks inside of Office three sixty five software, like, you know, writing research reports or compiling, you know, meeting agendas based on all of the, you know, notes that you have in your email, for example.

Speaker 2

此外,微软还想通过另一个卖点来推动人们购买这个套餐,那就是全新的Agent 365产品,它功能更强大。

And then additionally, you know, another selling point that Microsoft is trying to get people to buy this bundle with is this new agent three sixty five product, which is Even more.

Speaker 2

更多。

Even more.

Speaker 2

是的。

Yeah.

Speaker 2

它本质上是一种工具,旨在帮助企业追踪理论上日益增多的代理,并确保这些代理拥有正确的权限,以及像人类一样拥有登录权限。

It it's essentially like a tool that's sort of meant for companies to keep track of, in theory, the growing number of agents that they're going to have and make sure that those agents, you know, have the right permissions and have, I guess, you know, like, logins in the way that humans do.

Speaker 2

我认为对大多数客户来说,这在很大程度上仍属于假设性概念,但微软认为这很快就会变得重要。

I think for the most part, that's still somewhat hypothetical for a lot of customers, but it is something that Microsoft thinks will be important soon.

Speaker 0

那么,现在所有这些功能都包含在这个套件中,从你接触的客户来看,他们对这个套件的兴趣如何?

So now with all of these features included in the bundle, from the customers you're talking to, is their appetite for this bundle?

Speaker 0

你认为它会如他们所期望的那样流行起来吗?

Do you think it's gonna catch on the way they hope?

Speaker 2

你知道,这正是一个关键问题。

You know, that's that's kind of the the big question.

Speaker 2

据我所知,许多客户仍不确定他们究竟需要多少像Agent 365这样的产品,也不确定是否每个员工都应该拥有访问权限,还是仅限于IT部门等需要密切监控AI代理在系统内行为的人员。

I mean, from what I've heard, a lot of customers are still not entirely sure how much they're going to need something like agent three sixty five and whether that is a product that every employee should have access to or if you just need to limit it to, for example, you know, people working in your IT department who might need to keep a close eye on on what AI agents are doing inside your systems.

Speaker 2

与此同时,我也与一些参与软件合同谈判的人交谈过,他们表示,客户或许更明智的做法是,先逐个试用这个捆绑包中的某一项,比如先试用Agent 365订阅,而不是一开始就全盘接受整个捆绑包。

At the same time, you know, I also have spoken to folks involved in negotiating software contracts who say it might be smarter for a customer to, you know, ask to just try out one piece of this bundle at a time, such as the agent three sixty five subscription before they go all in on the bundle.

Speaker 2

这样做的原因是,购买软件的一个普遍规则是:一旦你购买了捆绑产品,就很难再拆分它们,这基本上可能导致你被锁定在大量订阅中,无论你是否真的在使用它们。

And the reason for that is that, you know, that the rule of buying software is that once you've bought bundled products, it's pretty difficult to unbundle them, which, you know, basically can just lead to you essentially getting locked into a large number of subscriptions, whether or not you're actually using them.

Speaker 2

没错。

Right.

Speaker 0

我快速切换一下话题,想聊聊你今天发表的关于微软的另一篇报道。

Very quickly, I want to pivot to a story that you published today, also on Microsoft.

Speaker 0

你写到了公司里那些帮助微软应对反垄断问题的法律专家。

You wrote about the legal experts at the company that are helping them navigate their antitrust issues.

Speaker 0

在深入这些专家之前,我们先回顾一下这里所指的反垄断问题是什么。

Before we get to the experts, remind us about what antitrust issues we're talking about here.

Speaker 2

是的,这其实是个很好的过渡,因为微软目前正受到欧洲、英国、澳大利亚以及美国联邦贸易委员会(FTC)等多个反垄断监管机构的调查。

Yeah, so this is actually a great segue because Microsoft is being investigated by several antitrust regulators in Europe, in The UK, Australia, as well as the FTC here in The US.

Speaker 2

具体来说,这些监管机构正在审视微软捆绑软件和定价的方式,有竞争对手指控微软利用其庞大的Windows和Office用户基础——数亿付费用户——通过将AI、安全工具或Azure云服务等产品与用户已购买的Office和Windows产品捆绑销售,并大幅降低捆绑价格,从而将客户锁定在这些附加产品上。

And specifically, those regulators are actually looking at the ways that Microsoft bundles its software and and prices them with the allegation that competitors have made being that Microsoft essentially, you know, uses its existing base of Windows and Office subscribers, which, you know, number in the hundreds of millions of paying users and ropes those customers into using other products like AI or security tools or, you know, services on its Azure cloud by making it much cheaper to, you know, buy the bundle of those tools in tandem with the Office and Windows products that they already pay for.

Speaker 2

有趣的是,我们正看到微软创建一个包含人工智能的新捆绑包,据我了解,这正是FTC目前正在调查的内容,以决定是否对微软提起反垄断诉讼。

So it's interesting that we're seeing Microsoft, essentially, create a new bundle that includes AI since, from my understanding, this is essentially what the FTC is currently probing as it decides whether or not to pursue an antitrust action against Microsoft.

Speaker 0

那么到目前为止,这个案件进展到什么阶段了?

And where where are we in that that case so far?

Speaker 0

时间表是怎样的?

What's the timeline?

Speaker 2

是的。

Yeah.

Speaker 2

我今天报道说,过去一年里,微软一直在向竞争对手询问有关其涉嫌反竞争行为的问题,尤其是最近几个月,我听说FTC实际上在告诉竞争对手,确保他们提交给该机构的材料必须完整、清晰地整理成一份详细的文件包。

So I I reported today that, you know, Microsoft has been asking competitors about Microsoft's alleged anticompetitive behavior in the last year, but especially, you know, in the last few months, I have heard that the FTC is essentially telling competitors to, make sure that whatever they're submitting to the agency is, like, fully, you know, written out in in a very clear packet.

Speaker 2

一些与FTC接触的人认为,这表明该机构可能在接下来的几个月内即将结束调查。

And some people who are talking to the FTC interpret that to mean that the agency could be close to wrapping up its probe in the coming months.

Speaker 2

之后的下一步将是FTC的委员们投票决定是否正式对微软提起反垄断诉讼。

The next step after that would be for the FTC to vote, that the commissioners will vote on whether or not to actually pursue an antitrust lawsuit against Microsoft.

Speaker 2

目前来看,结果仍有可能朝任何方向发展。

It could go either way at this point.

Speaker 2

我们不知道这场诉讼是否会成真,但人们认为这可能在接下来的几个月内发生。

We don't know whether that lawsuit will will come to pass, but people think that that could happen in the next few months.

Speaker 0

你今天在组织架构图中指出了微软内部领导这场对抗FTC调查的核心人员,还画了一个小型的组织架构图。

And so the people that you identified today in the org chart you put a little mini org chart together of the people at Microsoft that are leading the charge against the FTC in this probe.

Speaker 0

我想知道,所有这些被点名的人之间有什么共同点?你从他们组建团队的方式中学到了什么?

I wonder what was the common thread between all people identified, and what did you learn about how they're structuring their team?

Speaker 2

是的。

Yeah.

Speaker 2

微软非常擅长聘请具有政府工作经验,尤其是曾在微软所面对的监管机构任职的政策专家。

So Microsoft is very good at, you know, hiring policy experts who have experience in the government, often in the regulatory bodies that Microsoft is contending with.

Speaker 2

例如,领导微软反垄断部门的尼克,此前曾是欧洲顶尖的科技监管者之一,曾主导欧盟对多家科技公司的反垄断调查。

So for example, Nick who leads the antitrust org at Microsoft was previously one of the top tech regulators in Europe, who actually led some of the European Union's investigations, antitrust investigations into tech companies.

Speaker 2

他们还聘请了曾任职于FTC或司法部、处理科技反垄断案件的律师。

They also have, you know, lawyers who previously worked at the FTC, or at the DOJ doing tech antitrust action.

Speaker 2

通常,这些人都对这类调查的运作方式有深刻理解,知道微软该如何最好地构建论点,以应对反竞争行为的指控。

And, you know, typically, those people obviously bring a deep understanding of how these investigations tend to play out and how, you know, Microsoft could best structure its arguments against allegations of anti competitive behavior.

Speaker 2

因此,我们看到,拥有这些背景的人正在被提升到公司内部的高级法律职位。

And so we're seeing, you know, people with those backgrounds being elevated to to top legal roles within the company.

Speaker 0

亚伦,我只是想确认一下,这次调查涉及的核心问题是什么?

Aaron, I just want to make sure I understand here, what is at stake here with this probe?

Speaker 0

我们刚才谈到了他们在AI方面进行的捆绑销售,这显然是他们推动人们采用Copilot并真正认识这款软件用途的一种方式。

We talked just now about the bundling that they are doing with AI, and this is a clear way that they're going to try to encourage people to adopt Copilot and really see the use of this software.

Speaker 0

我的意思是,这项调查仍需要数月,甚至数年才能见分晓。

I mean, this probe will still take months, maybe years to play out.

Speaker 0

我的意思是,这里到底有什么风险?

I mean, what is at stake here?

Speaker 0

一切都会被拆分吗?

Would everything become unbundled?

Speaker 0

我们可能会看到什么?

What could we see?

Speaker 2

我的意思是,这确实是有可能的。

I mean, it is definitely possible.

Speaker 2

我们过去看到欧洲监管机构在

We have seen in the past European regulators a

Speaker 3

几年前

couple

Speaker 2

迫使微软停止将Teams消息应用与Office捆绑销售,因为以前只要你购买Office,就会免费获得Teams。

of years ago forced Microsoft to stop bundling its Teams messaging app with Office because previously, you know, whenever you bought Office, you would get Teams for free.

Speaker 2

Slack(现已被Salesforce收购)在欧洲对此提出了投诉,欧洲监管机构最终与微软达成和解,要求其停止捆绑Teams。

Slack, which is now owned by Salesforce, lodged a complaint about that in Europe, and European regulators actually reached a settlement with Microsoft to stop bundling Teams.

Speaker 2

所以我认为,你知道,在反垄断诉讼提起之前,我们不太可能看到FTC采取如此激烈的措施。

So I think, you know, we we probably wouldn't see anything that drastic come out of the FTC until after an antitrust lawsuit is filed.

Speaker 2

与此同时,你知道,我理解FTC正在调查其他问题,比如微软如何销售安全软件。

At at the same time, you know, I understand the FTC is looking into other questions like how Microsoft sells security software.

Speaker 2

因此,他们可能会寻求一些补救措施,比如影响这类产品的定价。

So, you know, they they may potentially pursue remedies and, like, that would dictate the pricing for that type of product.

Speaker 2

可能采取多种形式,但过去拆分确实一直是考虑的选项之一。

It could take a lot of different forms, but unbundling in the past has definitely been on the table.

Speaker 0

很好。

Great.

Speaker 0

非常感谢你前来做客。

Well, I want to thank you for coming on.

Speaker 0

这位是我们在《The Information》的微软记者亚伦·霍尔姆斯。

That is Aaron Holmes, our Microsoft reporter, here at The Information.

Speaker 0

据彭博社报道,亚马逊正在筹集高达420亿美元的公司债券。

Amazon is raising up to $42,000,000,000 in corporate bonds, according to Bloomberg.

Speaker 0

这些债券被分为最多11个期限,从两年到五十年不等。

They are being marketed in up to 11 tranches that range from two years to fifty years.

Speaker 0

在过去的几个月里,我们看到许多像亚马逊这样的大型科技公司发行了这种大规模的债券,因为它们都在为人工智能扩张筹集资金。

We have seen many big tech companies like Amazon issue these types of massive bond sales over the past few months as they are all looking to fund their AI expansions.

Speaker 0

我想请我们的财经编辑肯·布朗来帮我们解读这条新闻。

I want to bring on Ken Brown, our finance editor, to help us break down the news.

Speaker 0

肯,你对今天早上看到的这期债券发行有什么看法?

Ken, what did you make of this Spond offering that we saw this morning?

Speaker 4

嗨,阿卡什。

Hi, Akash.

Speaker 4

这是一大笔钱,但亚马逊有很多钱。

Well, it's a lot of money, but Amazon has a lot of money.

Speaker 4

这只是一个趋势,这些大型AI开发者正在为建设获得资金。

It's just more of a trend where these big AI developers are getting funding for the build out.

Speaker 4

其中很多资金来自他们自己的现金,但今年他们都加大了借款力度。

And a lot of it has been funded with their own cash, but they're all stepping up and borrowing too this year.

Speaker 0

债券投资者对这些大型科技公司发行债券的兴趣有多大?

How much appetite is there amongst bond investors for these big tech company issuances?

Speaker 4

需求非常旺盛。

There's plenty of appetite.

Speaker 4

亚马逊长期以来一直发行债务,因为他们有云业务和仓储业务,拥有所有仓库,但像Meta或微软这样的公司并不常来市场融资。

Amazon has long sold debt because they have a cloud business and a warehouse, they have all the warehouses, but a lot of these companies don't come to market too often like Meta or Microsoft.

Speaker 4

但无论如何,这些公司信用评级都很高,能产生大量现金。

But either way, these guys have high credit rating, they generate a lot of cash.

Speaker 4

这正是债券投资者想要的。

It's exactly what bond investors want.

Speaker 4

所以他们毫无疑虑地大量购入这些债券。

So they've been gobbling this stuff up with little doubt.

Speaker 0

这与债务市场的其他部分有何不同?

How does that contrast with other parts of the debt market?

Speaker 0

我的意思是,你在每个不同的资产类别中看到的都是同样的情况吗?

I mean, is this uniform story in every different asset class you look at?

Speaker 4

股市一直不太稳定。

Well, the stock market's been jittery.

Speaker 4

我们现在正处于战争中。

We're in a war right now.

Speaker 4

因此,我认为投资者将债券市场视为一种避险渠道。

And so I guess investors are seeing the bond market a little bit as a safe haven.

Speaker 4

那里也出现了一些波动,但总体上仍是一个避险渠道。

There's been some jitters there, but generally a safe haven.

Speaker 4

但股市一直相当动荡。

But the stock market's been pretty volatile.

Speaker 4

商品市场也十分波动。

The commodities market's been very volatile.

Speaker 4

另一点是,过去几个月私人信贷市场一直有些动荡。

The other thing is the private credit market has been in a bit of turmoil in the last few months.

Speaker 4

因此,这也是企业为人工智能建设融资的另一个渠道。

And so that has been another place where companies have borrowed money for the AI build out.

Speaker 4

因此,私人信贷市场遭遇了大量赎回。

And so the private credit market has been hit by a lot of redemptions.

Speaker 4

人们想要回他们的钱。

People want their money back.

Speaker 4

他们担心未来会发生什么。

They're worried about what's gonna happen in the future.

Speaker 4

所以,虽然亚马逊并不需要求助于私人信贷市场,因为那主要是给信用评级较低的借款者用的,但这也暴露了人工智能建设融资模式中的一丝裂痕。

And so, not that Amazon would need to go to the private credit market because that's more for lower rated borrowers, but that has been a bit of a crack in the AI build out, the funding scheme.

Speaker 0

因此,对于那些不太关注债务市场的人来说,这里的层级是:状况较好的公司会向债券投资者融资,它们可以自由地买卖这些债券。

And so for people who don't follow the debt markets as closely, so the hierarchy here is that the companies that are better off, they go to bond investors and they have the luxury of basically trading in and out of these bonds.

Speaker 0

私人信贷则更加定制化。

Private credit is a little more bespoke.

Speaker 0

你必须真正地谈判这些交易。

You have to really negotiate these deals.

Speaker 0

这就是你的意思吗?

Is that the idea?

Speaker 4

实际上,这些是提供给信用评级较低公司的贷款或信贷。

Well, it's more like these are loans that go or credit that go to lower rated companies.

Speaker 4

你也可以去债券市场融资,但对许多公司来说,私人信贷市场提供了更好的条件。

The bond market tends You can go to the bond market too, but the private credit market has been a better deal for a lot of these companies.

Speaker 4

亚马逊是一家AA级公司,仅比美国政府低一级。

Amazon is a AA rated company, which is just one notch below the US government.

Speaker 4

因此,它们不会遇到融资困难。

And so they don't have trouble.

Speaker 4

毫无疑问,他们会还钱的。

There's no doubt they're gonna pay the money back.

Speaker 4

所以,债券投资者都争相抢购这些东西。

So, the bond investors just glom onto all this stuff.

Speaker 4

我敢肯定,当今天这笔交易完成时,我们会发现它被超额认购了数倍,人们想要的远多于他们能获得的。

I'm sure when the sale is done, which will be today, we'll know that it was multiple times oversubscribed and people wanted a lot more than they could get.

Speaker 0

现在,我们已经看到其他大型科技公司走上了这条路。

Now, we've seen other big tech companies go this route.

Speaker 0

Meta 已经发行过,我想甲骨文也是。

Meta has done an issuance, I think Oracle, of course.

Speaker 0

你还在关注哪些其他公司?你觉得还有哪些公司可能会进行这类发行?

Who are the other names that you're watching, other names that you think could do these types of issuances?

Speaker 4

正如我之前所说,人工智能建设的大部分资金都来自于这些大公司产生的现金流。

Well, one of the things, as I said before, the bulk of the AI build out has been funded by the cash that these big companies generate.

Speaker 4

所以,现在许多公司表示:好吧,我们不想把所有现金都用掉。

So a lot of them now have said, Okay, well, we don't wanna use all our cash.

Speaker 4

我们打算借一些钱。

We're gonna borrow some.

Speaker 4

而且他们的借款利率很低,这很合理。

And they get good rates and it makes sense.

Speaker 4

有很多公司其实可以借更多钱。

There's a bunch of companies that could borrow more.

Speaker 4

我的意思是,Meta可以借更多,微软也可以借款。

I mean, Meta could borrow more, Microsoft could borrow.

Speaker 4

市场对它们的需求绝对不缺。

There's no shortage of demand for them.

Speaker 4

这仅仅取决于它们如何平衡自身的公司财务状况。

And it's just a matter of them balancing their own corporate financial situation.

Speaker 4

英伟达一直是借款方。

Nvidia has been a borrower.

Speaker 4

它们还可以再次借款。

They could borrow again.

Speaker 4

所有这些公司。

All these guys.

Speaker 4

我的意思是,最有趣的是甲骨文,今年早些时候它出来借款了。

I mean, the most interesting one was Oracle, which came out and borrowed earlier this year.

Speaker 4

人们曾担心它们花费过多,但它们的做法是借了250亿美元,同时也出售了250亿美元的股票。

There've been worries about them spending too much, but what they did was they borrowed 25,000,000,000, but they also sold 25,000,000,000 in stock.

Speaker 4

这些大公司通常不会出售股票,因为它们的股票价值非常高。

And none of these big guys usually sell stock because their stock is so valuable.

Speaker 4

所以借款更便宜。

So it's cheaper to borrow.

Speaker 4

但甲骨文对债券市场明确表示:听着,我们不会来找你们借这么多钱。

But Oracle told the bond market essentially, Listen, we're not gonna go and hit you up for all this money.

Speaker 4

我们要保持信用评级稳固,并且通过稀释股东权益来获得这笔资金。

We're gonna keep our credit rating solid and we're gonna actually dilute stock investors to get this money.

Speaker 4

所以这才是有趣的策略。

So that was the interesting play.

Speaker 4

其他大公司都不需要做类似的事情。

None of the other big guys are gonna have to do anything like that.

Speaker 0

在你看来,这是一招聪明的策略吗?

Was smart it play in your mind, in your opinion?

Speaker 4

这确实是一招精明的财务操作。

Well, it was a smart financial play.

Speaker 4

他们花在基础设施建设上的资金有多少,是否承担了巨大风险,这是另一个问题。

The underlying issue of how much they're spending on the build out and are they taking on a huge amount of risk is another question.

Speaker 4

你的财务实力远不如这些大公司,与他们正面抗衡可能会有问题。

And you have much less financial capability than these big guys, and sort of go toe to toe with them could be a problem.

Speaker 4

因此,他们在人工智能基础设施建设中被视为风险较高的投资之一。

So they're seen as one of the riskier plays in the AI build out.

Speaker 4

所以,这将成为关键问题。

And it's So that's gonna be the question.

Speaker 4

我认为他们有能力偿还债务,这一点毋庸置疑。

I don't think there's a doubt that they can pay down the debt.

Speaker 4

我的意思是,他们有一个庞大的软件业务和数据库业务,但后者就更有争议了。

I mean, they have a huge software business, a database business, but that's a more questionable one.

Speaker 4

其他巨头们可以随意借贷。

The rest of the giants there, they could borrow all they want.

Speaker 0

很好。

Great.

Speaker 0

好了,肯,感谢你今天来做客。

Well, Ken, I wanna to thank you for coming on.

Speaker 0

我预计我们会看到更多这样的情况,因此我期待未来能多次邀请你来做客。

I imagine we'll see more of these, and so I anticipate we'll have you on more and more.

Speaker 0

这位是肯·布朗,我们《The Information》的财经编辑。

That is Ken Brown, our Finance Editor here at The Information.

Speaker 0

AI代理的研发竞赛正在升温。

The race to develop AI agents is heating up.

Speaker 0

从硅谷到中国,各大公司正在竞相推出能够为员工和消费者完成任务的AI助手。

From Silicon Valley to China, companies are competing to launch AI assistants that can complete tasks for both workers and consumers.

Speaker 0

我方《信息》杂志亚洲局的同事刘chaner和小川今天早上独家报道,腾讯也在悄然加入这场竞赛,为其广受欢迎的微信应用开发一款新的AI代理。

My colleagues in the Information's Asia Bureau, Channer Liu and Osawa, reported exclusively this morning that Tencent is also looking to join that race, quietly building a new AI agent for its popular WeChat messaging app.

Speaker 0

我想播放一段我们亚洲局局长杨静与朱罗就其报道进行的对话。

I want to play for you a conversation that our Asia Bureau Chief, Jing Yang, had with Juro about his reporting.

Speaker 0

以下是这段对话。

Here is that conversation.

Speaker 5

朱罗,你能为我们详细说明一下,微信正在尝试构建什么样的AI代理吗?

Juro, could you break down for us what kind of AI agent is WeChat trying to build here?

Speaker 6

我的同事chaner和我独家报道称,腾讯正在开发一款能够处理各种任务的AI代理,包括现实生活中的任务,比如购物或预订旅行。

So my colleague, Chenner, and I reported exclusively that Tencent's working on an AI agent that handles all kinds of tasks, including, like, real real life tasks, like, grocery shopping or booking trips, for example.

Speaker 6

为了实现这一点,腾讯计划将这个AI代理与微信的小程序连接起来。

And in order to do this, Tencent's planning to connect this agent with, WeChat's mini programs.

Speaker 6

小程序就像是运行在微信内部的众多应用。

Mini Programs are like many apps that run inside WeChat.

Speaker 6

在中国,有数百万家企业运营着这些微信小程序,微信用户无需离开微信即可访问所有这些服务。

And in China, there are millions of businesses that operate those WeChat Mini Programs, and WeChat users can access all of those services without leaving WeChat.

Speaker 6

所以,这个想法是,如果智能体能够访问所有这些不同的小程序,它们就能为用户执行各种任务,现实生活中的任务。

So the idea is that if the agent can access all these different mini programs, they'll be able to perform all kinds of tasks, real life tasks for the users.

Speaker 5

这听起来确实会具有很多

That does sound like it will have a lot

Speaker 0

of

Speaker 5

潜力。

potential.

Speaker 5

不过,我记得小程序并不是什么新鲜事物。

However, I remember mini programs is not a new thing.

Speaker 5

微信大约在2017年就推出了小程序。

WeChat launched mini programs around 2017.

Speaker 5

而且在过去至少十二到十八个月里,AI代理一直是中国和美国许多科技公司追逐的目标。

And Aditya has been the pursuit of many tech companies, both in China and in The US in at least the last twelve to eighteen months.

Speaker 5

那为什么微信直到现在才着手做这件事呢?

So why is WeChat only doing this right now?

展开剩余字幕(还有 157 条)
Speaker 6

我认为有几个不同的原因,但我认为公司在策略上一直相当保守。

I think there are several different reasons, but I think the company has been quite conservative in its approach.

Speaker 6

这是因为微信对他们来说太重要了。

And that's because WeChat is so important to them.

Speaker 6

这是一个拥有超过14亿用户的平台。

And this is like a platform with more than 1,400,000,000 people.

Speaker 6

这意味着几乎每个中国人都在使用它,而且每天都用。

So that means pretty much everybody in China uses this and, you know, uses this every day.

Speaker 6

所以如果你推出一个不成熟、半成品的东西,可能会产生很大的负面影响。

So if you launch something that's not, you know, that's half baked, that could have a big impact, negative impact too.

Speaker 6

对吧?

Right?

Speaker 6

所以如果你要为微信推出像AI这样重大的功能,它必须做到完美。

So if you if you're going to launch something with AI for WeChat, you know, something this big, it has to be perfect.

Speaker 6

这确实需要很长时间。

So that really takes time.

Speaker 6

此外,腾讯一直在开发自己的AI模型,但人们认为这些模型相比字节跳动和阿里巴巴仍不够有竞争力,而这两家公司是中国AI开发领域的领军者。

And also, Tencent's been developing their own AI models, but that, you know, people think that's still not very competitive, you know, compared to ByteDance, Alibaba, which are, you know, two of the leaders in AI development in China.

Speaker 6

所以腾讯正在努力追赶。

So Tencent's trying to catch up.

Speaker 6

他们一直在为团队招聘人才。

They've been hiring, for the team.

Speaker 6

但问题是,他们到底会为这个AI助手使用哪家的AI模型?是使用内部模型,还是采用其他公司的AI模型?

And but, you know, there's a question of whether you know, whose AI model they're going to use for this AI agent and whether they're going to use the in house models or they will use other companies' AI models.

Speaker 5

是的。

Right.

Speaker 5

但如果他们使用外部模型,可能会带来一些不必要的复杂性。

Then that's But if they I guess if they use external models, maybe there are some like unnecessary complication.

Speaker 5

在数据迁移、与第三方模型进行验证等方面,可能会变得更加困难,我觉得是这样?

It will be much harder when it comes to like, in my mind, like data migration, validate it with those third party models, that kind of issues?

Speaker 6

是的,有可能。

Yes, potentially.

Speaker 6

所以我们报告过,他们不仅测试了内部模型,还测试了第三方模型,除了这款名为混元HY的腾讯模型外,他们还有微信项目。

And so we've reported that they have tested, you know, third party models as well as in house including, so apart from this Tencent model known as Honyan HY, they also have WeChat project.

Speaker 6

微信也在独立开发自己的AI模型。

WeChat is separately developing its own AI models as well.

Speaker 6

因此有多种选择,我们报告过,到目前为止他们已经测试了不同的方案。

And so there are different options, we've So we reported that they have tested different options so far.

Speaker 5

他们测试了哪些方案?

What are the options that they have tested?

Speaker 6

例如,他们测试了来自聚脉的开源模型,聚脉是最近在香港上市的一家北京新兴AI公司,还有阿里巴巴的模型、DeepSeek模型,以及微信的内部模型。

So for example, they have tested open source models from Jupu, which is one of the upstart AI developers from Beijing recently went public in Hong Kong and also Alibaba model and DeepSeek model and also the WeChat in house model.

Speaker 6

是的。

And so yes.

Speaker 6

这个项目仍在进行中。

And this project is still ongoing.

Speaker 6

所以,你最终会决定使用哪个模型,以及最终产品会是什么样子。

So, you know, what which model do you, you know, decide to use in the end and what exactly the end product will look like.

Speaker 6

这仍然需要时间。

It still takes time.

Speaker 5

有趣的是,阿里巴巴和腾讯长期以来的竞争对手,实际上可能将在这项工作中合作,或者更准确地说,腾讯竟然会考虑使用其最大竞争对手之一开发的技术。

It's quite interesting that Alibaba and Tencent long time nemesis is actually potentially we'll be working together on this or maybe more accurately that Tencent wouldn't be considering using the technology developed by one of its biggest rivals.

Speaker 5

我认为这或许恰恰说明了HY或Honya模型相比之下并不怎么样。

And I think that really maybe is a comment on sort of how, you know, how not great the HY or Honya models are by comparison.

Speaker 6

我们还报道过,腾讯一直在努力为其自身的AI开发组建更强大的团队。

Well, we have reported also that Tencent has been really trying to, you know, build a stronger team for its own AI development.

Speaker 6

因此,去年九月,他们聘请了一位来自OpenAI的研究员,他现在正领导他们的内部AI开发工作。

And so last year in September, they hired a researcher from OpenAI, and, he's now leading their, in house AI development effort.

Speaker 6

所以他们显然也在努力提升这一领域的实力。

And so they definitely are trying to step up in that area as well.

Speaker 5

是的。

Right.

Speaker 5

我觉得这将会变得非常激烈。

I think it's gonna get really heated.

Speaker 5

但说到阿里巴巴,我们再稍微回溯一下这一点。

But speaking of Alibaba, just circle back on that part a little bit.

Speaker 5

你最近也报道过,阿里巴巴的夸克应用——不要和夸克系列模型混淆——已经开始开发许多身份商业功能,整合了电商、旅行预订、支付、导航等阿里巴巴旗下各种业务单元。

You also reported recently that Quinn, Alibaba's Quinn app, not to be mixed up with the Quinn series models, but the Quinn app has also started developing a lot of identity commerce features, integrating ways like e commerce to travel booking, to payment, to navigation, all kinds of business units within Alibaba's umbrella.

Speaker 5

你认为这一举措,结合微信正在做的事情,将如何塑造未来中国消费者AI代理的竞争格局?

What do you think this move coupled with what WeChat is doing is going to shape the competition landscape going forward for AI agents for consumers in China?

Speaker 6

我认为,今年AI代理领域的竞争将成为中国科技行业最关键的战役。

Well, I think this year, the battle over AI agents will be the most critical battle in China's you know, tech industry.

Speaker 6

正如你提到的,阿里巴巴在这方面的投入确实非常多。

And, as you mentioned, Alibaba has been really doing a lot in this, area.

Speaker 6

他们将把自家的AI应用打造成一个完整的AI代理,视为首要任务。

And they, you know, they consider their and turn their AI app into a full on AI agent a top priority.

Speaker 6

因此,他们把包括电商在内的大量核心服务都整合进了他们的AI应用中。

So that's why they integrated all you know, a lot of the main services into know, including ecommerce into their AI app.

Speaker 6

所以他们确实在大力推广,尤其是在上个月的中国新年假期期间。

So they are definitely pitching, you know, especially during this Chinese New Year holiday last month.

Speaker 6

阿里巴巴投入了大量资金开展了一场大规模活动,发放电商优惠券,让用户尝试通过AI代理进行电商购物。

Alibaba ran a big campaign spending a lot of money giving away e commerce vouchers for people to try like AI e commerce through the AI agent.

Speaker 6

所以,他们显然在加大力度,在某些方面,他们已经领先于腾讯和字节跳动。

So, you know, they are definitely stepping up and they're, you know, in some ways they are ahead of Tencent and so is ByteDance.

Speaker 6

字节跳动也在他们的AI应用上做了很多工作。

ByteDance has also been doing a lot with their AI app.

Speaker 6

字节跳动的豆包AI应用是目前中国最受欢迎的AI应用,上个月用户数达到约3亿。

And ByteDance's Doubou AI app is the most popular AI app in China with about 300,000,000 users last month.

Speaker 5

这场竞争将会变得异常激烈。

It's gonna get really intense.

Speaker 5

我有这种预感。

Have a feeling.

Speaker 5

谢谢你,Juro。

Thank you, Juro.

Speaker 6

谢谢。

Thank you.

Speaker 0

这是《信息》亚洲局的朱罗·小川和杨静。

That was Juro Osawa and Jing Yang from the Information's Asia Bureau.

Speaker 0

振动编码已促使硅谷许多人重新思考他们进行软件开发的方式。

Vibe coding has caused many folks in Silicon Valley to rethink how they approach software development altogether.

Speaker 0

南公园共同体的普通合伙人阿迪亚·阿加瓦尔最近为《信息》撰写了一篇观点文章,谈到了这一转变对他个人的意义。

Aditya Agarwal, a general partner at South Park Commons, recently wrote an opinion piece for the information on what this shift means to him personally.

Speaker 0

阿迪亚曾是Dropbox的首席技术官,也是Facebook的早期工程师。

Aditya was previously the CTO of Dropbox and an early engineer at Facebook.

Speaker 0

他写道,自己最初对编码变革感到的悲伤,最近已转变为狂热的能量,最终转化为精通。

He wrote about how this initial sadness that he had about the transformation in coding has recently turned into reckless energy and eventually mastery.

Speaker 0

我想邀请阿迪亚来帮助我们理解他是如何思考工程领域的这一范式转变的。

I want to bring on Aditya to help us understand how he is thinking about this paradigm shift in engineering.

Speaker 0

阿迪亚,欢迎来到节目。

Aditya, welcome to the show.

Speaker 0

很高兴你来到这里。

It's great to have you here.

Speaker 3

嗨,阿卡什。

Hi, Akash.

Speaker 3

很高兴能来这里。

Pleasure to be here.

Speaker 3

谢谢您邀请我参加。

Thank you for having me on.

Speaker 0

那么告诉我,你为什么决定写这篇文章?

So tell me, why did you decide to write this piece?

Speaker 3

原因是我觉得,人们很容易形成一种观点,认为人工智能本质上是一种末日预言,或者人工智能会让一切变得完美,我们都会非常快乐和满足。

The reason is because I think it's very easy to have a point of view that AI is essentially like a doomer prophecy or AI is essentially a everything will be completely fine and we'll all just be very happy and satisfied.

Speaker 3

但没有足够多的人在探讨,当你非常擅长的事情如今能被人工智能轻易生成时,这种经历意味着什么。

But not enough people were writing about what it means to go through the process of having something that you are very good at to now kind of essentially be capable of being produced by AI.

Speaker 3

这其实是一种非常有趣的情感混合,我想说的是,我经历过的个人情绪和感受。

And it was a really interesting mix of, I would say, personal emotions and feelings that I went through.

Speaker 3

所以我只是写下了一些发自内心的东西,我认为它引起了很多人共鸣,我想我们很多人都在过去大约六到七周里经历了类似的过程。

So I just penned something that came from the heart, and I think it kind of struck a chord with many people who are I think a lot of us are going through that same kind of process over the last, call it, six to seven weeks.

Speaker 0

那你现在感受到的是什么?

So what is it that you're feeling?

Speaker 0

你整个职业生涯都是工程师。

You've been an engineer your whole career.

Speaker 0

你从零开始构建了这些程序。

You have built these programs from the ground up.

Speaker 0

你曾在这些科技公司的最高层工作过。

You've worked at the highest levels of these tech companies.

Speaker 0

看到这些智能代理和 vibe 编程工具替你完成工作,感觉如何?

What was it like for you to see these agents and these vibe coding tools do it for you?

Speaker 3

我觉得这其实是一种深刻的惊叹,因为我觉得 ChatGPT 大约三年半前才问世。

I think it's actually just a profound wonder to kind of start, because I think that ChatGPT came out roughly about three and a half years ago.

Speaker 3

编码代理大约在十八个月前才真正开始出现。

Coding agents came about seriously, maybe about eighteen months ago.

Speaker 3

在这段时间里,每次我尝试使用它们时,都会有一种心理上的反应:哦,这真酷。

And I think in that time period, every time I tried it out, I could use it and I would have this mental kind of reaction like, Oh, this is cool.

Speaker 3

它能完成所有简单的工作,但无法真正做出我擅长创造的那些东西。

It can do all the simple stuff, but it can't really make all of the things that I'm good at making.

Speaker 3

那种数据与计算的复杂交互,正确存储数据并使其高效运行的能力。

The kind of the complex interplay of data and compute and storing data correctly, making it efficient.

Speaker 3

从觉得AI是个不错的初级程序员、像实习生一样能辅助我,到如今能一次性端到端地生成高度复杂的作品,我只感到一种深深的惊叹——它竟然真的做到了。而一旦发生这种变化,我紧接着的第二种感受却是悲伤:我曾经精通的这项技能,我引以为傲、几乎成为自我身份一部分的东西——尽管如今我是投资者,不再写代码,但我内心深处仍觉得自己是个世界级工程师。

And to go from that sense of the AI is a good junior programmera good intern kind of programmer or somebody that can assist me, to now being able to one shot end to end produce things of high complexity, it was just a sense of profound wonder that it could actually And once that happens, the second reaction that I had was actually a sense of grief, which is that this thing that I was world class at, this thing that I took a lot of pride in essentially being a part of my identity, even though I'm an investor today, I don't write code, I think there was a part of me that always felt as though I'm still a world class engineer.

Speaker 3

我基本上能和最顶尖的人并肩而立。

I can kind of essentially hang it with the best.

Speaker 3

但看到AI现在几乎能和我做的一样好,我感到悲伤,感到失落。

But to see AI now being able to essentially be at the same level of kind of what I can do, there was grief, there was sadness.

Speaker 3

有种感觉,仿佛有什么东西正从你身上被夺走。

There is a sense that something is being taken away from you.

Speaker 3

那是最初的第一反应。

And that was the first initial reaction.

Speaker 3

我花了几天时间沉浸在这种悲伤中,直到突然意识到:如果AI能做我过去做的事,那我现在能做什么?

And I sat, I think, with that grief for a couple of days before you kind of realized that, Oh, if AI can do the things that I used to do, then what can I do now?

Speaker 3

而我认为,这种感觉被一种近乎狂热的能量所取代,那就是每当我没有实际利用代码生成成果来自动化我在South Park Commons工作中的部分内容、或自动化我个人生活的某些方面时,都会感觉几乎是浪费时间,因为这些工具是如此强大且能力非凡。

And I think that it's replaced by a sense of, Just a sense of like manic energy that every moment that I'm not actually essentially producing artifacts with code to automate parts of my job at South Park Commons, automate parts of my personal life, it feels like almost wasted time because these tools are so capable and powerful.

Speaker 0

你在文章中写了这样一句话。

You wrote this line in your piece.

Speaker 0

你说:‘如果说有什么的话,这种转变正在教会我重新成为人类是什么感觉。’

You said, If anything, this shift is teaching me what it is like to be human again.

Speaker 0

你这么说是什么意思?

What did you mean by that?

Speaker 3

我认为,特别是在硅谷,以及个人成长、职业发展和人生旅程的一部分,总会有这样的时刻让你意识到,你唯一的选择就是向上迈一步,重新学习,在某种程度上为下一阶段的人生做好准备。

I think that part of, I think, being in Silicon Valley in particular, and part of, I think, evolving your own, I would say, journey, your career, your life, is that I think there come these moments where you realize that the only option that you have is to essentially take a step up and essentially retrain and kind of in some ways, ready yourself for the next part of your life.

Speaker 3

我想,这种转变在我们人类身上相当常见,从高中到大学,从第一份工作到第二份工作,随着我们不断蜕变。

And I think this kind of change happens to us as humans pretty often as we go through high school to college, to your first job, to your second job, and as we all kind of transform.

Speaker 3

这让我有一种感觉,哇,人工智能在某种程度上迫使我不得不深入反思,对我来说,'我之为我'究竟意味着什么。

And it was just a sense of, wow, the AI is in some ways forcing my hand to really introspect about what it means for me to be me.

Speaker 3

我认为,一次又一次地寻找个人身份认同,是一种深刻的人类情感。

And I think that part of finding your own personal identity over and over again is a deeply human kind of emotion.

Speaker 3

很久以来,我再也没有如此强烈而深刻地感受到过这种情绪。

And I hadn't quite felt that in such a powerful way for a long time, powerful and visceral way for a long time.

Speaker 0

所以,我理解你所说的,如果我理解错了请纠正我,我感觉你像是职业生涯中非常资深的人。

So the way that I understood what you're saying here, and correct me if I'm wrong here, I sort of read it as someone who is so senior in their career progression.

Speaker 0

正如我们所说,你曾经是Dropbox的CTO。

And like we said, you were the CTO of Dropbox at one point.

Speaker 0

我认为,作为人意味着什么,其核心就是学习新技能,并思考你是如何学习的。看,今天我们谈论的是‘氛围编码’。

I mean, what it means to be human, I think, at its core is learning new skills and grappling with how you learn Look, we're talking about vibe coding today.

Speaker 0

几个月前,我又开始重新学习弹钢琴,对吧?

I was learning how to play the piano again a couple months ago, right?

Speaker 0

那种反复练习、感到沮丧、再尝试的过程。

And it's that repetitive motion of getting frustrated, trying it again.

Speaker 0

在我看来,我听懂了你关于‘做人的意义’的观点,但对我来说,这不过是学习一项新技能而已,哪怕你已经60岁了,现在也必须去做。

And it sounds like to me I mean, I hear your point around being human, but for me, it's just learning a new skill that you have to even if you're 60 years old, I mean, you have to do it now.

Speaker 3

是的,你说得很好,这确实关乎学习新技能。

Yeah, it's a good I like the way you phrased it, which is it is about learning new skills.

Speaker 3

但我认为学习这项新技能的一部分,就是以某种方式直面现实——你必须重新回归初学者的心态。

But I think part of learning that new skill is confronting in a way that you might In some ways, you have to go back to a beginner's mind.

Speaker 3

你必须回头承认:我在硅谷已经待了二十三年。

You have to kind of go back and admit to yourself, I've been in Silicon Valley for twenty three years.

Speaker 3

那些东西全都不重要,对吧?

None of that shit matters, right?

Speaker 3

某种程度上,我可以依靠我的经验,但同时我也得从第一步重新开始。

Like in some ways I can lean on my experience, but I'm also starting at step one in some ways.

Speaker 3

所以我认为,这其实是一次对现实的直面:如果我想学这个新东西,就必须接受我的经验固然有价值,但它并不能立刻让我像过去那样游刃有余地完成所有事情。

So I think it is a reckoning with the fact that if I want to learn this new thing, I have to accept that my experience counts for a bunch, but it also doesn't essentially immediately allow me to do everything that I thought I could do in the past.

Speaker 0

这改变了你招聘的方式吗?

How is it changing the way that you're hiring?

Speaker 3

这是个很好的问题。

Oh, it's a great question.

Speaker 3

我想说,我最钟爱的——好吧,我这算是透露了,这算是我的面试问题——但今天最重要的问题是:你日常生活中是怎么使用人工智能的?

I'd say that my favorite you know, I'm giving this away, I guess, you know, this is kind of my interview questions, but my most important question today is how have you used AI in your daily life?

Speaker 3

这周你用什么方法让你的工作和生活变得更好了?

What did you use to essentially make your work life, your personal life better this week?

Speaker 3

你为自己创造了什么?

What have you built for yourself?

Speaker 3

而其中很多潜在的动机是,这有点奇怪。

And a lot of kind of the underlying motivation there is that it's kind of weird.

Speaker 3

我注意到,有些人正在积极拥抱这项新技术,接受改变,重新体会学习新事物、做回人类的感觉。

I've noticed that there are some people who are running towards this new technology in some ways embracing the change, embracing what it means to learn something new to be human again.

Speaker 3

而有些人则坚决抵制。

And some people are really digging in their heels.

Speaker 3

我最初的反应是,这里可能存在某种关联:比如,资历越深的人越固守成规,更倾向于继续沿用过去的方式。

And my initial reaction was that there would be some correlator there in terms of perhaps, hey, listen, the more senior people are more set in their way, so they want to do keep on doing things the way they have.

Speaker 3

相比之下,资历较浅的人反而更容易学习新事物。

The more junior people in some ways, they're free to learn new things.

Speaker 3

但我发现,经验和年龄在某种程度上是无关的变量。

But what I found is that experience and age, these are in some ways uncorrelated variables.

Speaker 3

所以我现在真正寻找的是一个人的舒适度、适应能力,以及他们是否愿意使用这些工具来重塑自我和工作内容。

So what I'm really looking for now is someone's comfort level and adaptability and kind of a desire to use these tools to essentially reshape who they are and what their job is.

Speaker 0

换句话说,简单地说,年轻一代,那些没有学过旧式软件工程方法的人,他们更有优势;而职业生涯后期的人则有更多惯性需要克服——这种说法过于简化了。

So, in other words, a bit of an oversimplification to say, Oh, the younger generation that is coming up, people who haven't learned the older ways of software engineering, it's an oversimplification to say they are at an advantage and that people who are later in their careers, they have a little bit more inertia to sort of fight up against.

Speaker 0

你所说的是,这是一种心态,你在寻找任何职业阶段的人,他们真正深入研究了这种工具,并且说:你知道吗?

What you're saying is it's a mindset and that you're looking for people at any point in their career who have really dug into this tool and say, You know what?

Speaker 0

我一直在硅谷待了二十三年。

I've been in Silicon Valley I'm for twenty three

Speaker 3

准备好了。

ready to go.

Speaker 3

我准备好了,对吧?

I'm ready to go, right?

Speaker 3

是的。

Yeah.

Speaker 0

但让我问你一个问题。

But let me ask you this then.

Speaker 0

那么,你觉得学校的课程设置与你二十五年前学习软件工程时相比如何?我的意思是,你有没有和一些教授讨论过,他们是如何调整课程,使其更偏向实践或体验式学习,而不仅仅是理论?

So how do you think that the curriculums in school, compare this to when you were learning software engineering twenty five years ago, I mean, have you talked to people, professors, about how they're adjusting their curriculums and making

Speaker 2

更偏向于

it a little bit more

Speaker 0

应用导向或体验导向,而不是仅仅停留在理论层面?

application based or experience based rather than just theoretical?

Speaker 3

这是个好问题。

It's a good question.

Speaker 3

实际上,我最近一直在思考这个问题。

I would actually say I've been thinking a lot about this.

Speaker 3

但我目前还没有一个非常明确的观点。

I don't have a very concrete point of view on this yet.

Speaker 3

我猜测这可能会有些反直觉。

I would venture that this might be counterintuitive.

Speaker 3

我认为,至少在我的专业——计算机科学中,学习理论是非常重要的。

I actually think that learning the theory, at least in my profession, computer science, right?

Speaker 3

我认为理解理论、理解这些技术背后数学原理仍然非常重要,因为它能让你推导出这些底层系统和模型的能力边界。

I think understanding the theory, the math behind how a lot of these things work is still very important because it allows you to reason through what these underlying systems and models are capable of.

Speaker 3

但与此同时,我也觉得我上学时做的那些小玩具项目—— basically 就是把东西连起来——其实很傻。

But at the same time, I also think a lot of the small toy projects that I used to do in school that are basically wiring things up, that's dumb.

Speaker 3

AI 可以做这些事,对吧?

The AI can do that, right?

Speaker 3

所以某种程度上,你希望采用一种杠铃式的方法:一方面,你要阅读经典著作,学习理论,理解数学,掌握编程语言、编译器、系统和数据库的理论基础。

So in some ways you want a barbell approach where I think you want to read in some ways the classics, like read the theory, understand the math, understand kind of like the theoretical underpinnings of programming languages, compilers, systems, databases.

Speaker 3

而在杠铃的另一端,你应该极其 prolific 地运用这些理论去产出实际成果。

And on the other side of the barbell, you should be incredibly, I would say, prolific in terms of what you are producing with that theory.

Speaker 3

对吧?

Right?

Speaker 3

我认为中间地带正是如今许多大学陷入的困境:他们教你如何做事,却没有给你扎实的理论基础。

I And think the middle ground is kind of where a lot of colleges are stuck today, which is that they're teaching you how to do things without a solid underpinning of the theory.

Speaker 3

他们教的是技艺,但这种五年前定义的技艺,如今的智能代理已经能完成了。

They're kind of teaching you the craft, but the the craft as defined five years ago, these agents can do today.

Speaker 3

所以我主张课程设置更偏向两端的杠铃模式。

So I push kind of the curriculum to be more barbell.

Speaker 3

我认为现在这么做的人还不够多。

I don't think enough people are doing it today.

Speaker 0

所以当人们说,‘今天你甚至都不用学编程了’,你并不完全认同这种说法,对吧?

So when people say that, oh, you don't even have to learn how to code today, you actually don't believe that entirely?

Speaker 3

我会重新表述一下。

Well, I would I would rephrase it.

Speaker 3

我不认为你真的需要去学所谓的‘编程’,但我确实认为你需要理解这些系统是如何运作的。

I don't think perhaps you need to learn, quote unquote, how to code, but I do think you need to learn how the systems work.

Speaker 3

对吧?

Right?

Speaker 3

如果你是一名优秀的工程师,即使AI已经完成了大部分代码的编写,你也应该能够准确描述数据是如何存储的、数据如何流动、计算从哪里来、瓶颈在哪里、限制因素是什么。

You should be able to if you are a great engineer, even if AI does most of the writing of the code itself, you should be able to describe exactly what's happening in terms of how the data is stored, how the data flows, where the compute comes, where the constraints are, where the long pole is.

Speaker 3

但某种程度上,你确实不需要亲自写代码了,对吧?

But you don't need to write the code, right, in some ways.

Speaker 3

而在某种程度上,理解理论和系统能让你成为更出色的指挥者,统领你的各个智能代理,就像指挥一支交响乐团一样。

And then in some ways, understanding the theory and the systems allows you to be a much better, I would say, conductor of essentially the orchestra of your agents, if you will.

Speaker 0

对。

Right.

Speaker 0

好了,阿迪亚,非常感谢你来参加节目。

Well, Aditya, I want to thank you for coming on.

Speaker 0

这是一次非常精彩的对话。

It was a great piece.

Speaker 0

这位是南公园共同体的普通合伙人阿迪亚·阿加瓦尔,正在TITV播出。

That is Aditya Agarwal, a general partner at South Park Commons here on TITV.

Speaker 0

今天的节目就到这里。

That does it for today's show.

Speaker 0

提醒一下,我们每周一至周五上午10点(太平洋时间),下午1点(东部时间)直播。

Reminder, we are on this stream Monday through Friday at 10AM Pacific, 1PM Eastern.

Speaker 0

感谢大家收看。

I want to thank you all for tuning in.

Speaker 0

我们非常感谢您的观看。

We really do appreciate your viewership.

Speaker 0

请务必在YouTube上订阅我们的频道,并在X、Instagram、TikTok上关注我们,同时在您收听播客的平台关注我们。

Make sure to subscribe to the information on YouTube and follow us on X, Instagram, TikTok, and check us out wherever you get your podcasts.

Speaker 0

我已经迫不及待期待明天的下一期节目了。

I am already excited for our next show tomorrow.

Speaker 0

祝您周二剩下的时间愉快。

Have a great rest of your Tuesday.

Speaker 0

暂时再见了。

Bye bye for now.

关于 Bayt 播客

Bayt 提供中文+原文双语音频和字幕,帮助你打破语言障碍,轻松听懂全球优质播客。

继续浏览更多播客