Pinterest首席执行官谈人工智能路线图、中国紧急会晤英伟达、AWS新人工智能模型 | 2025年12月10日

欢迎各位来到Informations TITV。

Welcome everyone to the Informations TITV.

我叫阿卡什·巴里查。

My name is Akash Basricha.

今天是12月10日，星期三。

It is Wednesday, December 10.

今天我们为大家准备了一场精彩的节目。

We have got an exciting show lined up for you today.

首先，我们将与Pinterest的首席执行官比尔·雷迪进行一场精彩的对话。

First up, we have an exciting conversation for you with Bill Reddy, the CEO of Pinterest.

我们会探讨他对聊天机器人可能成为广告新战场的看法，同时也会讨论开源与闭源AI的争论。

We're gonna get into what he thinks about chatbots potentially becoming the new battlefield for advertising, and we're also gonna talk about the open and closed source AI debate.

接着，我们会与我们的亚洲分局主编讨论关于中国围绕H200芯片召开紧急会议的报道。

We're then talking to our Asia bureau chief about the information's reporting around emergency meetings that have been happening in China around the H200 chips.

我们还会讨论《信息报》关于DeepSeek最新动态的一篇报道。

And we'll also talk about a story of the information published about the latest on DeepSeek.

接下来，我们将剖析一则关于MGX的故事，这是一家总部位于阿联酋的基金，已成为大型数据中心项目的主要融资方。

Next up, we are breaking down a story on MGX, a fund based in The United Arab Emirates that has become a major financier for big data center projects.

最后，我们将与AWS的技术总监讨论AWS推出的新AI模型。

And finally, we will end the show with a discussion around AWS's new AI model with the company's director of technology.

这是一期内容丰富的节目，让我们马上进入正题。

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

Pinterest是众多正在谨慎应对AI转型的社交媒体公司之一。

Pinterest is one of many social media companies that is carefully navigating the AI transition.

多年来，该公司的收入增长稳步加速，但许多人正在关注AI聊天机器人如何彻底改变广告的竞争格局。

The company's revenue growth has steadily accelerated over the years, but of course, many are asking about the ways in which AI chatbots could change the battleground for advertising altogether.

我想邀请Pinterest的首席执行官比尔·雷迪进行一次独家对话，探讨他对这些问题的思考。

I want to bring on the CEO of Pinterest, Bill Ready, for an exclusive conversation on how he is thinking about all of these issues.

比尔，欢迎来到TI TV。

Bill, welcome to TI TV.

很高兴你来到这里。

It's great to have you here.

谢谢邀请我，阿卡什。

Thanks for having me, Akash.

我想探讨很多不同的点，但我想先从你最近发布的那篇博客文章开始，文章谈到了你对开源模型的偏好。

So there's a lot of different points I want to get to, but I want to start with this blog post that you put out recently, and it talked about your affinity for open source models.

我被这句话深深吸引了。

And I was struck by this line.

你写道，你们在Pinterest使用开源AI模型，以不到领先专有AI模型10%的成本实现相似的性能。

You wrote using open You wrote that you're using open source AI models at Pinterest to achieve similar performance at less than 10% of the cost of leading proprietary AI models.

当然，我们知道这是一个很大的说法。

Look, we know how That's competitive these models a big claim.

你这句话是什么意思？

What did you mean by that?

详细谈谈吧。

Talk about that.

是的，我认为关于AI的很多讨论都集中在谁在构建最新、最大的专有模型上。

Yeah, I think this is, you know, so much of the discussion around AI has been about who's building the latest, largest proprietary model.

我认为还有其他几个非常重要但尚未得到足够关注的趋势。

And I think there's a couple other really important trends that are happening that aren't being talked about enough.

一个是开源。

The one is open source.

正如我们所发布的，我们发现可以采用开源模型进行微调，以不到顶级专有模型10%的成本实现相似的性能。

And as we published, we're seeing that we can take open source models and fine tune those and get similar performance to the very best proprietary models at less than 10% of the cost.

我认为这非常重要，因为开源对软件开发的民主化作用已经持续了几十年。

And I think that's really, really important because open source has a democratizing effect that has been true for the, you know, building a software for decades now.

我认为这对展望未来、理解众多公司如何利用AI至关重要。

And I think that's really important for how we look forward as to how many, many companies can build with AI.

如果你能以不到10%的成本获得相似的能力，这将产生巨大的民主化效应，极大推动创新与创造力。

If you can get similar capabilities for less than 10% of the cost, that's going to have a really democratizing effect, really spur a lot of innovation and creativity.

因此，我们推动这项工作并分享给他人，让大家了解其卓越之处，因为我们认为，为AI建立一个繁荣的开源社区至关重要。

And so, pushing our work on that to share with others just how good that is because we think it's really important there's a thriving open source community for AI.

第二个趋势是针对特定用途的紧凑型模型，我认为这意味着你并不需要世界上最大的模型来完成每一项任务，这也是实现AI成本效益的另一种方式。

And then the second trend around compact fit for purpose models that I think also is about you don't need the very largest models in the world to do every single task, and that's also a way that you can create cost effectiveness in AI as well.

所以我想确认一下我是否理解正确。

So I just want to make sure that I understand this.

我的意思是，听到这个你可能会想，哦，成本不到领先AI模型的10%。

I mean, might hear this and think, oh, you know, less than 10% of the cost of leading AI models.

我们通常认为谷歌和OpenAI是其专有模型的领导者。

I mean, we think about Google and OpenAI as the leaders with their proprietary models.

所以帮我理清一下，你的意思是Pinterest使用开源技术构建的模型同样优秀且更便宜吗？

So help me understand, are you saying that the models that Pinterest has built using open source technology is just as good and cheaper.

你是这个意思吗？

Is that what you're saying?

两点。

Two things.

第一，现在已有大型开源模型可供获取，你可以对它们进行微调以满足特定需求。

One, there are large scale open source models available now that you can take and fine tune them for your purpose.

我们发现，就我们的用途而言，可以在实现相似性能的同时将成本降低90%。

And what we found for our purposes is that we could achieve 90% lower cost at similar performance on the things that we needed to do.

好的，所以

Okay, so

针对一个非常特定的用例。

for a very specific use case.

是的。

Well, yes.

但我认为这是一个可推广的观点，因为这涉及到大型通用模型。

But I think this is a generalizable point because that's about the large generalized models.

还有一个相关的第二点，那就是我们也在为特定任务开发自己的内部模型。

There's a second point, which is related to this, that we are also building our own in house models for specific tasks.

因此，这两种策略可以帮助企业在大幅降低成本的同时，更好地利用人工智能。

And so these are two different tactics that can help companies get a lot more out of AI at much lower cost.

这一点现在非常重要，因为几乎所有我交谈过的首席执行官都提到，比如上周我在《纽约时报》的德布克会议上，许多我遇到的CEO都说，他们已经投入了大量资金购买现成的专有软件解决方案，但尚未看到预期的投资回报。

And this is really important right now because nearly every CEO that I talked to, for example, I was, you know, last week I was at the New York Times deal book, and so many of the CEOs that I talked to there said, Well, they've invested a lot of money in AI and buying these off the shelf proprietary software solutions, but they're not seeing the return on investment that they need.

他们花了很多钱，但还没有获得预期的节省，或者支出显著抵消了他们原本能获得的节省。

And they're spending a lot, but they're not yet getting the savings that they had expected, or the spend is significantly offsetting the savings that they were getting.

公司发现，他们可以利用开源技术获得更高的效率。

Companies finding that they can leverage open source to get much better efficiency.

这一点在软件开发领域已经存在了几十年。

This has been true in software development for decades.

事实上，如果不能基于开源技术构建，许多大型公司根本不会存在。

In fact, many of the largest companies out there wouldn't exist if not for the fact that they had been able to build on open source.

这比使用专有数据库或专有操作系统便宜得多。

It was much cheaper than proprietary databases or proprietary operating systems.

像Linux这样的技术对于许多万亿市值公司的建立至关重要。

Things like Linux and those types of things were really, really important to the building of a lot of these trillion dollar market cap companies.

接下来的一批公司需要能够获得优秀的开源AI，才能实现这些成本效益。

Well, the next batch of companies, you know, they're going to need to have really great open source AI available to them to have these cost efficiencies.

否则，专有软件可能会独占所有价值，而我们却无法从中构建和创造价值。

Otherwise, the proprietary software may collect all the value without us being able to go build and create value from that as well.

因此，我们认为这非常重要，这也是我们分享研究发现的原因。

So, we think this is really important, again, which is why we're sharing our findings.

所以，你是开源的坚定支持者。

So, you're a big proponent of open source.

你如何看待中国推出的开源模型？

Do you think of the open source models coming out of China?

我的意思是，这周新闻里有不少相关报道。

I mean, that's in the news this week.

今年早些时候，那里取得了大量进展。

Tons of progress there earlier in the year.

我们最近听到的消息少了些，但我们知道，只要一个模型出现，就足以引发市场震荡，就像我们之前看到的那样。

We've sort of heard less about them, but we know that, look, it's just one model away from roiling markets, as we've seen.

你对这个领域关注得有多密切？

How close are you watching that space?

是的，DeepSeek的突破是一个重大时刻。

Yeah, I mean, the deep seek moment was a huge moment.

我认为，DeepSeek时刻的讨论很大程度上集中在他们如何在少于最新GPU芯片的情况下完成这一成就，以及他们如何利用这些限制找到更低成本的模型构建方式。

And I think so much of the deep seek moment, the conversation was about how they had done that on less than the latest GPU chips, and the sort of theory constraints of how they leverage those constraints to find cheaper ways to build those models.

而这实际上关乎芯片。

And that was really about the chips.

但他们使用了许多技术，现在这些技术正被开源社区采纳，其他人可以利用这些技术来构建自己的模型。

But there were a lot of techniques they used that are now going into open source community that others can use to build their own models.

因此，我们正在将其中一些技术应用于我们内部正在构建的模型。

So, we're using some of those techniques for our own models that we're building internally.

但随后，其他优秀的开源模型也涌现出来，比如Quinn，它发布了非常强大的模型。

But then you've had other really good open source models emerge, like Quinn, for example, that is putting out really powerful models.

但我认为，DeepSeek时刻的重点主要放在芯片和低成本芯片上。

But I think that deep seek moment was so much of the focus was on the chips and lower cost chips.

但我认为更大的突破在于，出现了一个能与大型专有模型相媲美的开源模型。

But I think the bigger moment was that you had an open source model that was rivaling the large proprietary models.

而现在，有了Quinn，你看到其他模型也相继出现，比如Mistral，你正目睹一个真正能够竞争的开源社区开始形成。

And now you have that with Quinn, you see others coming out, Mistral, like, you know, you're seeing an open source community really start to develop that can compete.

你知道，开源已经进入赛场，它正在竞争，并且以极高的成本效益达到性能水平。

You know, open source is at the table and it is competing and doing so at really effective cost to performance levels.

这很重要，不仅仅是从成本的角度来看。

And that's important, not just from a cost perspective.

我认为也很重要的是，很多人会说这些模型来自这个国家，或者来自那个国家。

I think it's also important that, you know, so many people refer to these models as, you know, this model came from this country or this model came from that country.

开源真正重要的地方在于，它属于每一个人。

The really important thing about open source, open source belongs to everyone.

它不受任何一家公司控制。

It's not controlled by any one company.

真正的开源不受任何一家公司控制。

True open source is not controlled by any one company.

它也不受任何一个国家政府控制。

It's not controlled by any one nation state.

如果你希望OpenAI掌握在多数人手中而非少数人手中，那么开源对他们来说至关重要。

And if you want OpenAI to be in the hands of the many rather than the few, then open source is really important to them.

而且，这一点在过去几十年里对软件开发一直至关重要。

And again, that has been really important to software development for decades now.

我认为在人工智能领域，这种趋势继续下去至关重要。

And I think it's really important that that continues in the world of AI.

没错。

Right.

它并不需要取代专有模型，你知道，开源软件几十年来一直非常重要。

And it doesn't need to be instead of proprietary models, you know, open source software has been huge for decades.

这并不意味着专有软件不再存在。

It doesn't mean that proprietary software no longer exists.

这意味着开源软件是整个繁荣生态系统中一个非常重要的组成部分，与专有模型并存，它能够与这些专有模型良好竞争，并对它们起到良好的制衡作用。

It just means that open source software is a really important component of a thriving ecosystem overall alongside of proprietary models, and it can compete well with those proprietary models, and it's a good check and balance on those proprietary models.

所以我想问你一个关于Pinterest以及你们当前如何应用人工智能的广泛问题。

So let ask you a question broadly about Pinterest and how you're approaching AI now.

你们在哪些方面大力推动人工智能？

Where are you applying the gas on AI?

你们在哪些方面踩了刹车？

And then where are you pushing the brakes?

因为我们在《The Information》上做过一些报道，我们了解到，实际上是你们的CTO与我们讨论了公司如何以更谨慎的方式应对AI生成内容等问题。

Because we've done some reporting at The Information, and we've spoken to it was your CTO actually who spoke with us about ways in which the company is taking a more careful approach to, for example, AI generated content.

你能跟我谈谈你对这一点的看法吗？

Talk to me a little bit about how you see that.

我们在加速推进。

We're applying the gas.

过去几年里，我们实际上已经将Pinterest转变为一个AI驱动的购物助手。

You know, we've effectively turned Pinterest into an AI powered shopping assistant over the last few years.

我们已经连续九个季度实现用户数新高，平台超过一半的用户是Z世代。

We've had nine straight quarters of record high users, more than half the platform is Gen Z.

购物是他们使用平台的主要原因。

Shopping is the primary reason they come to the platform.

如果你问Z世代用户为什么使用Pinterest，他们会说：‘它真的懂我。’

And if you ask Gen Z users why they come to Pinterest, they'll say things like, well, just gets me.

这正是我们利用经过独特策展信号调优的AI模型，提供出色的个性化推荐和代理式体验，帮助用户完成购物旅程的结果。

And that's us using our AI models tuned off our unique curation signal to give really great personalized recommendations and agentic style experiences that help guide users through shopping journey.

所以我们正在大力加强这一点。

So we're really doubling down on that.

我们刚刚推出了最新的Pinterest助手，进一步深化了这一方向。

We've just launched our latest Pinterest assistant, going even further with that.

但我们真正是在打造购物体验，你知道，我们正在构建不会取代购物或自动化购物的AI体验。

But we're really building shopping, you know, we're building AI experiences that don't replace shopping or automate away shopping.

对。

Right.

我感觉有些人正在做或试图为讨厌购物的人打造购物体验。

I sort of see that folks are doing that or trying to build shopping for people that hate shopping.

我们为热爱购物、并希望有一个助手真正帮助他们深入体验购物旅程的人打造购物助手。

We're building shopping assistance for people that love shopping and that want an assistant to actually help them immerse in that journey.

所以，这正是我们正在大力投入并加速推进的方向，正如你所说。

So that's where we're doubling down and hitting the gas to your point.

在这一点上，我不说我们在踩刹车，而是确保我们负责任地使用AI。

The place where, I wouldn't say we're hitting the brakes, I'd say we're just making sure that we use AI responsibly.

首先是让人工智能趋向积极。

The first is tuning AI for positivity.

这是我将近三年半前从谷歌加入Pinterest的主要原因之一：一方面，我想证明存在一种更积极的社交媒体商业模式，而大多数社交媒体的核心是通过引发对立来获取参与度。

This is one of the primary reasons I joined Pinterest from Google nearly three and a half years ago, is that, one, I wanted to prove there was a more positive alternative to the business model of social media that so much of it had engagement via an arrangement at the core.

而其中很大一部分是，你知道，社交媒体是如何变得消极的？十多年前，人工智能就被用来决定你在社交媒体上看到的内容。

And a big part of that, you know, how does social media become negative, AI got put in charge of what you see on social media more than a decade ago.

那时的人工智能还只是早期形态。

Was just earlier forms of AI.

所以，这本质上就是在避免垃圾内容。

So, this is like avoiding slop, essentially.

嗯，避免垃圾内容，或者避免其他负面因素——比如十多年前，当人工智能被要求最大化你在社交媒体上的观看时间时，它发现你会长时间关注那些触发你情绪的内容，无论你的触发点是什么。

Well, avoiding slop or avoiding other negative things like, you know, when AI was asked to maximize your view time on social media more than a decade ago, the AI figured out you look longer at the things that trigger you, whatever your triggers are.

因此，我们着手让人工智能趋向积极。

And so we set out to tune AI for positivity.

如果人工智能可以用来通过引发愤怒来让你沉迷于屏幕，那我们为什么不能要求人工智能确保你离开平台时感觉更好，让它成为你情绪健康的积极贡献者呢？我们已经证明了这一点。

If AI could be used to keep you glued to a screen with engagement via enragement, why can't we ask the AI to make sure you leave the platform feeling better And that it's a positive contributor to your emotional well-being, and we've been able to prove that out.

这是我们确保不会陷入只为让用户沉迷屏幕、做那些上瘾而非有益之事的恶性竞争的地方。

That's a place where we're making sure we don't, you know, have a race to the bottom on just trying to keep people glued to a screen, doing things that are addictive rather than additive.

因此，我们努力专注于有益的内容。

So, we try to focus on additive.

这是一个方面。

That's one place.

另一个方面是信任与安全，我们利用AI来打击不良内容。

Another place is on trust and safety, where, you know, we're using AI to combat bad content.

正如你提到的AI垃圾，如今生成内容的能力已经大幅提升。

And, you as you asked about AI slob, well, you know, there's been a huge increase in the ability to create content.

而其中绝大多数人是在享受创作优质内容的乐趣。

And, you know, the vast majority of that, you know, are people that are having fun creating good content.

因此，人们表达自我的能力得到了民主化，这是好事。

So, you've had a democratization of people's ability to express themselves, that's a good thing.

但与此同时，就像任何技术一样，生成式AI既可以用于善，也可以用于恶。

But also mixed in that, like any technology, you know, generative AI can be used for good or for bad.

对。

Right.

因此，也存在一些不良行为者，试图生成垃圾、无用或有害的内容。

And so, you also have bad actors that are trying to create things that are spammy or not helpful or harmful.

所以是

So, is

现在Pinterest上允许存在AI生成的内容吗？

there AI generated content right now on Pinterest that is allowed?

当然允许。

Oh, absolutely.

关于AI生成的内容，我想说，关于AI垃圾的整个讨论，我有几点想说。

And what I would say with AI generated content, you know, this whole discussion of AI slob, you know, a couple of things I'd say.

首先，问题不在于AI内容是好是坏？

One is that, you know, it's not about is AI content good or bad?

而在于有些AI内容是好的，有些AI内容是坏的，你该如何区分呢？

It's about some AI content is good and some AI And content is how do you parse that?

那么，您如何让用户掌控他们想看到的内容呢？

And how do you give the user control over what content they want to see?

因此，您正在使用人工智能来评估哪些内容是人工智能生成的。

And so, you're using AI to then assess which of the content is AI generated.

我的理解是

Am I understanding that

对吗？

correct?

没错。

That's exactly right.

我们正在做两件事。

So, two things that we're doing.

第一，我们会标注人工智能生成的内容，以便用户知道哪些是AI生成的。

The first is that we are labeling AI content so the user knows when it's AI generated.

我们还采用行业通用技术来检测和标注内容是否为人工智能生成。

And we're using industry techniques to go detect and label when something is AI generated.

但并不是每个人都这样做标注。

But not everybody's labeling that.

但我认为，我们在确保标注方面走在了行业前列，我们已做出选择：当能够检测到是AI生成的内容时，我们就进行标注。

But we are further along in the industry, I think, in making sure that we label and we've made a choice to label when we can detect that it's AI generated content.

其次，我们给予用户选择权，让用户决定何时希望减少看到AI内容。

Secondly, we're giving the user choice and letting the user decide when they want to see less AI content.

因此，这与其他平台非常不同。

And so that's very different than other platforms.

我认为，平台之所以回避这一点，部分原因是他们发现AI生成的内容非常吸引人。

And I think part of why platforms may be avoiding this is that they're seeing that the AI generated content is really engaging.

它能让人们停留更长时间。

It keeps people looking for longer.

但对我们来说，我们并不想让用户停留更长时间。

But for us, we're not trying to keep people looking for longer.

我们的目标是帮助人们做那些对他们的现实生活产生积极影响的事情，即使这意味着他们要离开我们的平台，比如去购买新衣服，或者重新设计房间之类的。

We're trying to help people do things that make a positive impact in their real life, even when that means something off their platform, like going and buying a new outfit or going and redesigning a room or things like that.

因此，我们始终赋予用户自主权，让他们能够决定何时认为AI生成的内容有帮助，同时我们也在不断个性化，以了解每位用户认为什么是有帮助的。

And so we consistently give our users agency both in and the then they have the option of when the AI generated content is helpful or not, and we're personalizing more and more to understand for each user what's helpful or not.

当然，举些例子来说，什么时候有帮助，什么时候没有帮助，有些内容本质上就是糟糕的，比如由机器人生成的垃圾信息或类似的内容。

And, you know, some of these things, to give you some examples of when is it helpful versus when is it not, you know, there's some content that's just inherently bad, you know, that would be created by bots or that is spam content or things like that.

对于这些内容，我们会使用AI模型将其从平台上清除。

That, we use AI models to just get that stuff off the platform.

但还有其他类型的内容，我们一直认为，美是主观的，一个人的垃圾可能是另一个人的宝藏。

But there's other content, you know, we've always said, you know, beauty's in the eye of the beholder, or one person's trash can be another person's treasure.

比如，一个人觉得是艺术的东西，可以回溯到现代艺术运动之类的例子。

You know, something that one person thinks is art, you know, go back to sort of, you know, modern art movements and things like that.

传统主义者可能会说：‘这叫艺术？’

Well, traditionalists sort of said like, Oh, that art?

关键在于，每个人真正想看什么，都是他们自己的观点。

Mean, this is the thing, is that it's everyone's opinion, really, what they want to see.

因此，我认为这种选择很有意思。

And so I think that choice is interesting.

没错。

That's right.

这是一种选择，也是一个个性化问题。

It's a choice and it's a personalization issue.

所以，关键在于如何正确实现个性化，并赋予用户表达自我、说出他们喜好的能力。

So, it's really about how do you get the personalization right and give the user the ability to express themselves and say what they like.

但即使是同一个用户，他们也会切换模式，因为有时用户可能处于幻想模式。

But even for a given user, they'll shift modes because sometimes the user may be in a dreaming mode.

我们看到这种东西，有点像幻想，你会说，嘿，这个房间布局真有意思。

We're seeing this thing that is sort of a fantasy that you'd say, well, hey, that's a really interesting sort of room layout.

现实中你永远做不到，但在幻想模式下，这种想象能极大地拓展我的想象力，这很棒。

You can never do that in real life, but it's really cool to sort of expand my mind's eye as to what might So be in dreaming mode, that's great.

但当他们进入购物模式时，比如想买一张沙发，突然间，那张图片就必须是真实的了。

But then when they say they go to doing mode, I wanna buy a sofa, but none of the sudden that picture's real.

我们的视觉搜索技术实际上能让用户选取那张沙发和AI生成的内容，然后为我们展示最接近的、真正可以购买的沙发。

Well, our visual search technology actually lets the user take that sofa and that AI generated content, and we'll show them the closest real sofa that they could actually buy.

因此，我们帮助用户在不同状态间切换——什么时候处于幻想模式或梦境模式，此时这些充满想象力的AI生成图像可能就像过去的虚构与非虚构作品、或那些以往不可能实现的奇幻故事一样，具有启发性，能激发其他可能性；然后在用户准备进入现实世界行动时，帮助他们切换回现实模式。

And so we're helping the user navigate the sort of movement between, well, when am I in sort of fantasy mode and dreaming mode, in which case some of these fantastical, you know, AI generated images might actually be helpful in the same way that we had fiction and non fiction previously, or we had fantasy stories before, you know, that weren't possible today, but it can inspire something that, you know, something else might be possible, and then help the user toggle over when they're ready to actually do something in the real world.

让我问你一个问题，如果我们从产品转向你所处的广告行业。

Let let me let me ask you a little bit, you know, so if we move on from the product to the business that you're in, which is the advertising market.

你看，关于聊天机器人业务逐渐引入广告，已经有很多讨论了。

Look, there's been a lot of chatter about the chatbot businesses introducing ads over time.

无论是在本周还是下周发生，我认为共识是这一定会发生。

Whether it happens this week or next week, I think the consensus is it's going to happen.

我想和你谈谈这对你的业务意味着什么。

And, you know, I want to talk to you about what that means for your business.

你认为这会侵蚀你的广告业务吗？

Do you see that eating into your advertising business at all?

你知道，广告商的行为一直非常一致，他们会追随消费者，真正去到消费者做出购买决策的地方。

Well, you know, advertisers have behaved very consistently, and that advertisers will go where consumers are and advertisers will really go where consumers are actually making purchasing choices.

而我们的业务正是如此，只要购买决策发生的地方，广告商就会去哪里。

And our business, and then so wherever purchasing choices are happening, advertisers are going to go there.

显然，AI助手和聊天机器人正是这一趋势发生的地方，广告商肯定会涌入这里，他们必须这么做，但这并不是零和游戏，因为最终广告商追求的是增量购买。

And clearly, you know, AI assistants and chatbots are a place where, you know, that is happening and where advertisers are certainly going to go and they have to do the that, but it's not a zero sum game because at the end of the day, advertisers are looking to get incremental purchasing.

对于我们而言，尽管你已经看到了聊天机器人和助手的兴起，我认为在过去三年里，ChatGPT已经吸引了大约8亿用户。

And for us, even as you've seen the rise of chatbots and assistants, I think over the last you know, three years, ChatGPT has put on roughly 800,000,000 users.

与此同时，我们已经连续九个季度实现了用户数量的新高。

Well, at the same time, we've had nine straight quarters of record high users.

我们平台上的用户超过一半是Z世代。

More than half of our platform is Gen Z.

我想，我们平台上的几乎每一位Z世代用户都知道聊天机器人，也接触过它们，但他们认为我们与那些产品截然不同。

And, you know, I would guess that probably nearly every one of those Gen Z users on our platform, they know about chatbots and they view them, but they see us as something unique and different from that.

因此，这两种模式完全可以共存。

And so there's room for both of these to exist.

这实际上也反映了搜索领域几十年来的演变方式：几十年来，通用搜索一直由谷歌这样的平台主导。

And that actually mirrors the way search has been playing out for decades, is that for decades, you've had general purpose search where like Google, for example, is a winner of general purpose search.

但与此同时，垂直领域的搜索也与之并存，比如更多产品搜索从亚马逊开始，而不是谷歌；更多旅行搜索从Booking或Expedia开始，而不是谷歌。

But then you also had vertical specific searches that coexisted with that, where more product searches starting on Amazon than starting on Google or travel searches starting on Booking or Expedia rather than Google.

虽然你具备通用能力，但也可以拥有垂直领域的特定能力，让用户能够更深入地探索，比如亚马逊用于购物，或Booking和Expedia用于旅行。

And while you did have the general purpose capability, you could have a vertical specific capability where users could go deeper, you know, whether it be an Amazon, you know, for shopping or Booking or Expedia for travel.

我认为在AI世界中，同样会出现通用AI，类似于通用搜索的角色。

And I think similarly in this AI world, you're going to have general purpose AI as akin to what general purpose search was.

而在这一领域，将会出现少数几个大型赢家。

And you'll have a small number of, you know, large winners in that.

但同时，也会出现针对特定任务的专用工具，就像之前提到的模型，你可以获得专为特定任务优化的模型，从而更好地完成该任务。

But then also, you're going to have fit for purpose tools that for a specific task, just as was talking about the models, you can get fit for purpose models that perform that task better.

我认为会有一些公司和产品在这方面做得更好。

I think you're going have companies and products that do that better.

从一开始，我的核心观点就是Pinterest在视觉搜索和购物方面，能够专注于这一领域并实现超越。

That's been my thesis from the beginning with Pinterest, is that for visual search and shopping, we can really focus on that and do things outperforming.

关于这一点，我再分享最后一件事：在最近两次财报电话会议上，我都提到，我们最新的多模态视觉搜索模型在购物推荐的相关性上，比领先的专有模型高出超过30个百分点。

One last thing to share with you on that, on our last two earnings calls I've shared, our latest multimodal visual search models outperform the leading proprietary models by more than 30 full percentage points on the relevancy of their shopping recommendations.

所以，并不是针对所有问题或人类全部知识，而是专门针对购物场景，由于我们拥有独特的用户行为信号和内容精选能力，这些更小、更精简的专用模型在推荐相关性上能超越30个百分点。

So not for anything you can ask, and not for all of human knowledge, but specifically to shopping outperforms on the relevancy of recommendations by 30 full percentage points because of this issue of the smaller, compact fit for purpose models with our unique signal around user behavior and curating, we can outperform.

这就是为什么那些涌向我们平台的Z世代用户会说：‘Pinterest 真懂我’，因为我们通过购物推荐实现了出色的个性化。

And that's why those Gen Z users that are flocking to our platform will say things like, well, Pinterest just gets me, because we're really getting the personalization right through those shopping recommendations.

我想知道，在构建广告业务时，你认为最困难的部分是什么？这一点在讨论聊天机器人业务时常常被忽视。

I wonder what you think is the most difficult part about building an advertising business that maybe people miss out on talking about as it relates to these chatbot businesses?

因为，你看，人们以为广告会一夜之间出现，而且人们会喜欢它们，完全不会影响用户体验。

Because, look, I think people think that the ads are going appear overnight, and that people are going to like them, and that they're not going to affect the user experience at all.

但现实是，你引入了这个东西。

I mean, the reality is you introduce this thing.

我们已经看到了，对吧？

We've seen it already, right?

人们在ChatGPT里看到小弹窗，就会害怕这是广告。

People see little pop ups in ChatGPT, and they get scared that this is an ad.

天啊，这是怎么回事？

Oh my gosh, what's happening?

然后ChatGPT的负责人不得不在X上出来澄清：不，不，不是这样的。

And then the head of ChatGPT has to come out on X saying, No, no, no.

我们目前还没有这么做。

We're not doing it yet.

别担心。

Don't worry.

你知道，你已经成功应对了这种张力：如何智能地投放广告，同时又不损害用户体验。

You know, you have navigated this exact tension between how do you intelligently place ads and how do you not affect the user experience.

你认为在这个挑战中，人们在讨论聊天机器人向广告转型时，最容易忽略的最难部分是什么？

What do you think is the toughest part about this challenge that people neglect to talk about when they talk about this transition to ads that these chatbots will go through?

我认为，你们所提到的我们做的一件事是，确保广告对我们的用户来说是优质内容。

Well, I think one of the things, you know, what you're alluding to that we have done is that we focus on making sure that the ads are great content for our users.

当用户处于购物模式时，也就是Pinterest上的大多数用户都是来购物的。

And when the user's in a shopping mode, which is, you know, the majority of users on Pinterest are there to shop.

我之前提到过，我们如何赢得Z世代的青睐，超过三分之二的Z世代用户上Pinterest是为了购物。

You know, I talked about, you know, winning with Gen Z, more than two thirds of Gen Z is coming to Pinterest to shop.

根据Adobe的一项研究，39%的Z世代用户认为Pinterest是他们首选的搜索平台。

You know, 39% of Gen Z an Adobe study came out, 39% of Gen Z thinks of Pinterest as a first place to go search.

其中70%的人认为Pinterest更具个性化。

70% of them see Pinterest as more personalized.

因此，这些例子表明，当用户在购物时，只要我们展示的是他们真正需要的产品，是否为广告其实并不那么重要。

And so, you know, those are examples of, you know, where they see that we're providing a really good fit for purpose solution on shopping, but it also means that when somebody's shopping, as long as we show them the right product, it doesn't matter so much whether it's an ad or not an ad.

你有没有向他们展示出他们真正想找的那双鞋？

Did you show them the right pair of shoes that they're actually looking for?

这才是关键。

That's what matters.

对于其他人来说，我认为这将是一个真正的问题：当你有商业意图时，广告可以很好。

And so for others, I think this will be a real question is like when you have commercial intent, the ads can be great.

当用户没有商业意图时，广告可能就不太合适。

When the user doesn't have commercial intent, the ads may not be so great.

我认为，对于许多聊天机器人来说，其中会存在一些商业行为。

And I think with a lot of the, you know, chatbots, there's going to be some commercial behavior in there.

其中也会有很多非商业行为。

There's going be a lot of behavior in there that's non commercial.

但对我们而言，我们的平台主要围绕购物。

But for us, our platform is primarily about shopping.

因此，我们能够为用户提供极佳的体验——正如我们连续九个季度实现用户数量新高一样，我们也持续提升了每位用户的参与度，这意味着我们为他们提供了更相关的广告内容。

And so that makes it so that we can really, really, you know, deliver a great experience for the user where the user, you know, as we've delivered those nine straight quarters of record high users, we've also consistently deepened engagement per user, which means we're making the ads relevant content for them.

这些广告对用户有帮助，同时也对广告主非常有利，因为这意味着广告主能在用户真正愿意看到广告的时刻与他们相遇。

They're helpful to the user, but that's also great for the advertiser because that means the advertiser gets to meet the user in a moment where the user actually wants to see their ad.

在其他一些场景中，用户可能正在研究或做其他事情，那时他们可能并不想看到广告。

In some other place where the user might be researching or doing other things, user may not want to see their ad in that moment.

因此，在这里，由于购物场景的存在，我们能够使用户和广告主的利益达成一致。

And so here, we can align the incentives of the user and the advertiser because of the shopping context.

我想知道你对Pinterest未来三到四年的发展方向有什么看法。

I wonder what you think about where Pinterest will be three, four years from now.

因为我一直思考的一点是，社交媒体公司多年来不得不以某种方式重塑自我，我们看到许多公司都在进行重大转型。

Because one of the things that I've been thinking about is this idea that social media companies, they have had to reinvent themselves over the years in some capacity, and we see companies taking big swings.

比如，Meta从元宇宙转向，即使仍部分聚焦于此，现在他们所宣称的终极目标已是个人超级智能。

I mean, Meta went from the metaverse, to the extent it's still focused on that, it's now focused on personal superintelligence is the moonshot that they talk about.

你知道，Snap推出了他们的智能眼镜和可穿戴设备业务，并将此视为公司未来的方向。

You know, you have Snap that has come out with their spectacles and their wearables play, and they're talking about that as the future of the business.

我想知道Pinterest的终极目标是什么。

I wonder what Pinterest's moonshot is.

我的看法是，你们公司并没有像Snap那样采取如此大胆的举措，比如进军可穿戴设备领域，尽管这可能是任何公司可能走的方向。

I mean, I don't see the company taking as big a swing, for example, going to wearables, to the extent that that is a direction that any company could go.

我认为，Snap是否能成功实现这一转型还值得商榷。

I think it's debatable whether or not that is a transition that Snap will be able to make effectively.

但当你思考你们的业务时，未来几年你们必须进行自我革新。

But as you think about your business, you will have to reinvent yourself in the coming years.

我的观点是，这不能仅仅局限于在核心平台上引入人工智能。

And my take is it's going to have to be more than just implementing AI in the core platform.

你的终极目标是什么？

What is the moonshot for you?

嗯，你说得完全对。

Well, so you're absolutely right.

这不仅仅是将AI应用到核心平台上。

It's more than just implementing AI in the core platform.

事实上，过去三年我们对平台进行了重大革新。

In fact, we've had a major reinvention of the platform over the last three years.

所以我们正处于这一革新过程中。

So we are in the middle of that reinvention.

你知道，三年前、三年半前，Pinterest的用户数量在下降，当时我们像其他人一样专注于短视频，失去了差异化和相关性。

You know, three years ago, three and a half years ago, you know, Pinterest was declining users and was sort of, you know, focusing on short form video like everybody else and had lost differentiation and relevance.

当我们专注于将Pinterest转变为一个AI驱动的购物助手时，这推动了平台的复兴。

And as we focused on turning Pinterest into an AI driven shopping assistant, that has led to the resurgence of the platform.

这是一次平台的巨大革新。

That has been a massive reinvention of the platform.

我们并不是仅仅通过实施AI来实现这一点的。

And we did that not just by, you know, just by implementing AI.

Pinterest上有一种非常独特的行为，我们在人类策展方面进行了大力强化。

It was that there was a really unique behavior on Pinterest that we've really doubled down on in terms of human curation.

我觉得关于人工智能的讨论太多都集中在，AI 如何只是自动化掉人类所做的一切事情？

I think so much of the discussion about AI, you know, has been, you know, how does AI just, you know, automate away all the things that humans will do?

然后我们所有人就会普遍过上一种无需工作的生活，或者类似的情况。

And, you know, we'll all just go live on a universal basis that can come and have no work to do or whatever.

这是一种

It's a

非常有趣的生活。

very interesting life.

我认为大多数人不会觉得这是种有趣的生活。

I think most people wouldn't think that's a very interesting life.

我们的重点是如何让人工智能成为对人们的补充，真正帮助人们，让人们的生产力更高，生活质量更好。

We're focused on how do we make the AI additive for people and truly helpful to people so that people can be more productive, get, you know, a higher quality of life.

而我们平台上的这种人工策划就是一个很好的例子：当超过70%的Z世代认为Pinterest比其他地方更具个性化时，部分原因就在于我们平台上的这种人工策划信号——人们在平台上搭配服装或设计房间，而AI本身是没有风格和品位的。

And, you know, that human curation on our platform is a good example where when, you know, you get that, you know, 70% plus of Gen Z that sees Pinterest as more personalized than other places to go search, well, part of that is because that human curation signal that we get that is people styling outfits on our platform or designing rooms on our platform, AI by itself doesn't have style and taste.

人类才有风格和品位，而AI可以从中学习。

Humans have style and taste, and then the AI can learn from that.

当人类来到我们的平台，设计事物、搭配服装，比如判断哪个手袋配哪条裙子，或者哪个沙发放在房间里更合适，这些都是人类的品味和筛选。

And so when humans come to our platform and they design things, you know, put together outfits and say, which handbag looks good with which dress or, you know, which sofa, you know, would look good in a room setting, that's human taste and curation.

然后我们可以以此训练AI，不仅为该用户做出更好的推荐，也为其他用户做出更好的推荐。

Then we can train AI from that, not only to make better recommendation to that user, but to make better recommendations to other users.

所以下一位用户进来，看到这只手袋，心想：我真的很喜欢这只手袋，但我不确定该怎么围绕它搭配一套完整的穿搭。

So the next person comes in starts with just that handbag, says, you know what, I really like that handbag, but I'm not sure how I style an outfit around that.

我们能看到的不仅是其他人如何搭配这只手袋，还有与你品味相似的人是如何搭配它的。

Well, we see the intersection not only of how other people style that handbag, but other people's tastes similar to yours style that handbag.

这正是让我们在AI领域实现真正独特突破的原因，也解释了为什么我们最新的多模态视觉搜索模型比大型专有模型性能高出整整30个百分点。

So, that's what's letting us do really unique things with AI and back to our latest multimodal visual search models outperforming by 30 full percentage points, the large proprietary models.

这得益于我们独特的筛选信号，以及专为特定用途优化的紧凑型模型。

It's that unique curation signal plus our compact fit for purpose models.

展望未来，我们认为，真正由AI驱动的购物助手才刚刚起步。

As we think forward, we think we're just getting started in terms of what a true AI powered shopping assistant can be.

目前，我们已经在为你推荐你可能想购买的商品，并帮助你采取行动。

Right now, we're making great recommendations of things you'd want to buy, you know, helping you go take action on those things.

这还能有多大的帮助呢？

How much more helpful can that be?

这本身就是一个登月计划——它能否达到那种程度，就像你最想一起购物的那个人一样好，无论是最好的朋友、姐妹、兄弟，还是如果你足够幸运拥有私人造型师，它能像他们一样有帮助吗？

That is a moonshot in and of itself in terms of like, can that get to a place where it's as good as the person you'd love going shopping with the most, whether that's a best friend, a sister, a brother, or if you're lucky enough to have a personal stylist, you know, can it be as helpful as that?

我们认为，单凭这一点就已经是一个登月计划了，而且并不需要不同的硬件形态或其他东西。最后我想说，太多登月计划最终要么没有成功，要么要很久之后才实现。

We think that in and of itself is a moonshot and doesn't require a different form factor or things And like the last thing I'd say, so many moonshots, you know, so many moonshots end up either not panning out or come to fruition much later.

我从零开始创办过五家初创公司，最近的两家中包括Venmo和Brainshrub。

And, you know, I've built five startups from zero, Venmo and Brainshrub being the two most recent.

我在硅谷已经待了很长时间。

And, you know, I've been in Silicon Valley for a long time.

在硅谷，你会经常听到两件事。

And one of the things that, know, two things you'll hear in Silicon Valley.

一件事是：‘我没错，只是太早了。’

One is like, well, I was right, but early.

没错。

Right.

另一个任何优秀的风险投资人都会回应的观点是：‘对但太早’等同于‘错’。

The other thing that any good VC would say in response is right, but early is the same as wrong.

因此，对于这些远大目标，你必须有对未来的愿景，但也必须深刻了解用户今天真正需要什么，以及你今天能提供什么。

And so, with these moonshots, you've got to have a vision for where you want to go, but you've also got to be deeply tuned into like what users are ready for today and what you can deliver today.

我认为我们在这两者之间取得了很好的平衡：我们清楚地认识到，购物和人工精选是我们平台的核心，而人工精选正是我们的真正优势。

And I think we're balancing that really well, where we have a view of like, okay, shopping and human curation, you know, are at the core of our platform, that human curation is a real differentiator for us.

我们相信，随着时间推移，这将让我们实现巨大的突破，变得像和你最好的朋友，或者精品店里出色的销售顾问一起购物一样贴心。

We think over time that can let us do tremendous things to be, you know, as helpful as going shopping with, you know, your best friend or the really great sales associate at the boutique or whatever.

我们认为，这就是它未来的发展方向。

We think that's where that can go over time.

用户今天真正准备好接受的是什么？

What are users ready for today?

今天，用户需要的是非常出色的推荐，以及在我们平台内无缝完成购买的体验。

Well, today, you know, it's really great recommendations, really, you know, seamless ability to go purchase right inside our platform.

你可以绑定亚马逊账户，直接在我们的平台内完成购买。

You can link an Amazon account and purchase right inside of our platform.

这些能真正简化他们当前体验的事情，以及我们的用户。

Those kinds of things that really make their journeys easy today and then with our users.

因此，我们当然有长期愿景，但同时也非常专注于用户现在真正需要什么？

So we certainly have that long term vision, but we're also really laser focused on what are users ready for now?

这就是为什么我们在与用户交流时避免使用‘智能体’之类的术语，因为只要你保持

And that's why we don't use words like agentic and things like that when we're talking to our users because as long You want you preserve the

人性化的体验，对吧？

human experience, right?

是的，更多是关于智能体和硅基技术，但普通消费者根本不会考虑智能体这个问题。

Yeah, was much more about agentic and silicon, The average consumer isn't thinking about agentic at all.

他们只关心：这个东西有没有帮他们完成想做的事？

They're just thinking about, did this thing help me do what I wanted to do?

在这方面，我认为我们获得了很高的评价，证据就是过去三年里，用户不断涌入平台，并且参与度越来越深。

And on that, I think we're getting really high marks as evidenced by users flocking to the platform and engaging deeper and deeper over the last three years.

很好。

Great.

比尔，很高兴你能做客我们的节目。

Well, Bill, it's great to have you on the show.

非常感谢你的对话，我相信公司未来还有很多令人兴奋的进展。

I really appreciate the conversation, and there's a lot of exciting to come, I'm sure, for the company.

在你的带领下，公司取得了巨大的成就，我期待看到你接下来如何引领公司发展。

It's been a great run under you, and so I look forward to seeing how you lead the company from here.

非常感谢你，阿卡什。

Thank you so much, Akash.

真的非常感谢。

Really appreciate it.

刚才那是Pinterest的首席执行官比尔·雷迪，出现在TI TV上。

That was Bill Reddy, the CEO of Pinterest, here on TI TV.

好的。

Okay.

特朗普政府已批准向中国出售H200。

The Trump administration has given the green light on H200 sales to China.

但现在，中国正在努力明确自己的立场，并报道称中国正在评估国内对这种芯片的需求量。

But now, China is trying to figure out its stance, and the information reported China is trying to assess how much demand there is domestically for the chip.

现在邀请到我身边为我们分析这一切的是我们的亚洲总编辑杨静。

Joining me now to break it all down is Jing Yang, our Asia Bureau Chief.

静，欢迎再次做客节目。

Jing, welcome back to the show.

很高兴你来到这里。

It's great to have you here.

随时都很高兴回来，Akash。

Always glad to be back, Akash.

那么，请告诉我们中国最近举行的这些紧急会议情况。

So tell us about these emergency meetings that have been happening in China.

是的。

Yeah.

周三，也就是今天，中国政府官员召集了一系列紧急会议，与阿里巴巴、字节跳动和腾讯等中国科技巨头讨论，询问他们实际需要多少NVIDIA H200芯片。

So on Wednesday, well, today, Chinese government officials called a series of emergency meetings with Chinese tech giants, including Alibaba, Bydance, and Tencent to ask them how much of the NVIDIA S200 they actually need.

他们召开这些会议的原因是，中国一方面深知需要这些强大的英伟达芯片来发展本国的AI系统。

And the reason that they called these meetings is because that China on one hand wants to, you know, is keenly aware that China needs these powerful NVIDIA chips to develop the country's AI systems.

但另一方面，他们也希望让中国在半导体领域更加自给自足。

But on the other hand, they also want to make China more self reliant in semiconductors.

所以，特朗普所谓的某种和解姿态，实际上让中方的政策目标变得更加复杂。

So what Trump, you know, has made as some sort of a peace offering, seeing as, you know, the Chinese side, also complicated somehow Beijing's policy goals.

我想，由于H200芯片比H20强大得多，因此对它的需求一定很大。

Now I imagine there's a lot of demand for the h 200 chip given how much more power how much more powerful it is than the h 20.

当然，我们知道它不如黑曜石芯片。

Of course, we know it's not as good as the black well.

但我很好奇，这些会议透露出政府对AI发展的方向有何不同看法。

I wonder though what these meetings say about the direction in which the government is looking at AI.

这与他们对待H20的态度有何不同？

Is it different from the approach they took with the H20?

是，也不是。

Yes and no.

所以，通过今天的会议，我们了解到，中国官员告诉科技公司，要求它们重新审视并提交一份详细的评估报告，说明为何需要H200芯片。

So basically, in today's meetings, we learned that the Chinese officials told the tech companies that they wanted the companies to look to go back and look at, you know, give them a very detailed assessment and justify their demand for the H200.

例如，你们必须说明为什么需要这么多H200芯片，而国产芯片根本无法替代。

Like for example, you have to satisfy, why you need this many of H200 that, you know, Chinese domestic chips just cannot replace.

随后，公司被告知，一旦官员们汇总并审阅了所有公司的回复，就会做出决定。

And then the companies were told that once officials have compiled and looked through all these responses from the companies, then they will make a decision.

顺便说一句，一旦中国政府做出决定，它并不会向外界公开。

By the way, once the Chinese government makes the decision, it's not like going to tell the world about it.

他们很可能会采用所谓的‘窗口指导’方式，这是一种中国监管机构常用的、但非常有力的方法，即私下向企业传达政策预期，而非公之于众。

They're very likely going to use this so called window guidance method, which is like a very common but very powerful method for Chinese regulators to privately tell companies their policy expectations instead of making it public.

对。

Right.

顺便说一下，几个月前，正如我们报道的那样，中国政府正是用同样的方式，要求中国企业不要购买任何NVIDIA H20芯片。

That, by the way, was the exactly same approach that the Chinese government used a few months ago as we reported to ask Chinese companies not to purchase any of the Nvidia H20 chips.

对。

Right.

好吧，我想转到谈谈深度求索（DeepSeek）。

Well, I want to pivot to talking about DeepSeek.

《The Information》发布了一篇关于该公司积极推进自身模型开发的文章。

The Information published a story about how the company is pushing ahead with its own model development.

给我们讲讲这个故事。

Tell us about that story.

是的。

Yeah.

我的意思是，自从DeepSeek在今年一月登上全球舞台以来，我们对这家公司内部究竟发生了什么知之甚少，对吧？

I mean, since DeepSeek rose to global stardom in January, I don't think we have learned much about what actually is going on in the company, right?

这简直太神秘了。

And that just, you know, is such a mystery.

我们今年一整年都在试图弄清楚这家公司的下一步重大举措是什么。

And we spent all of this year trying to figure out what is the company's next big move.

而今天，我们终于能够报道：原来DeepSeek已经成功使用了黑焊芯片——这些芯片被美国政府禁止出口到中国。

And finally, we were able to report today that it turns out that DeepSeg has been able to use black weld chips, by the way, which are banned from being exported to China by the US government.

但不知怎么的，他们通过走私渠道弄到了这些芯片。

But somehow they got their hands on these chips from, you know, the smuggling channel.

而他们需要这些芯片，是因为他们正在竞相开发下一代旗舰AI模型，以在中国和全球范围内保持竞争力。

And then they needed these chips because they are racing to build their next flagship AI model so they can continue to stay competitive in the AI race in China and globally.

现在，我想把这两个故事联系起来，你觉得在未来几个月，随着H200芯片的出现，这个故事会有什么变化吗？

Now, I wonder, connecting the two stories together, do you think that that story changes at all in the months to come in the face of the new H200 chip?

关于H200的政策，会不会改变DeepSeek试图使用的芯片类型？

Does the policies around the H200, could that change the different types of chips that DeepSeek tries to uses?

是的。

Yeah.

我的意思是，不只是DeepSeek，所有试图开发自己AI模型、以在中国和全球范围内与美国公司竞争的中国公司都会受到影响。

I mean, not just DeepSeek, but all the Chinese companies that are trying to develop their own AI models to stay competitive in China and globally against US companies as well.

我的看法是，这肯定会直接影响需求。

I mean, what I'll say right now is that it will definitely have an impact on the demand.

假设中国政府真的允许H200进入中国，这肯定会冲击黑市对黑市芯片的需求。

Let's say if the Chinese government actually allows H200 to get into the country, it will definitely have an impact on the black market demand for black world trips.