埃隆·马斯克诉OpenAI庭审、OpenAI 500亿美元员工股票池与电商现实 | 2026年1月8日

欢迎各位收看Information的TI TV。

Welcome everyone to the Information's TI TV.

我叫阿卡什·佩斯里查。

My name is Akash Pesritcha.

今天是1月8日，星期四。

It is Thursday, January 8.

今天在节目中，我们将与本周在法院跟进埃隆·马斯克对OpenAI诉讼的记者进行对话。

Today on the show, we are talking with our reporter who was at the courthouse this week following the Elon Musk OpenAI lawsuit.

我们还独家报道了OpenAI为员工设立的500亿美元股票激励计划。

We also have some exclusive reporting around OpenAI's $50,000,000,000 stock grant pool for employees.

此外，我们将深入挖掘我们获得的有关OpenAI电商野心及其在实际中进展的最新内部报道。

Separately, we are digging into some new inside reporting we have about OpenAI's ecommerce ambitions and how those are going on the ground.

随后，我们将与网络安全公司Netskope的首席执行官对话，探讨人工智能时代这一行业的情况，并以一场关于健康的对话作为结尾。

And then talking with the CEO of cybersecurity company Netskope about that sector in the age of AI, and we will wrap with a conversation about health.

我们将邀请一位医生和一位资深企业家，深入剖析我们对OpenAI最新健康功能的看法。

We are bringing on a doctor and a longtime entrepreneur to break down what we think of OpenAI's latest health features.

这将是一场精彩的节目，让我们马上进入正题。

It's going to be a fun show, so let's get right on into things.

加利福尼亚的一名法官裁定，埃隆·马斯克对OpenAI的诉讼将由陪审团进行审判。

A judge in California decided that Elon Musk's lawsuit against OpenAI will go to a trial with a jury.

我的同事罗克特·德鲁在法庭上旁听了整个庭审过程。

My colleague, Rocket Drew, was in the courtroom watching the proceedings play out.

我想请他来和我们分享他所看到的情况。

I want to bring him on to talk to us all about what he saw.

罗克特，欢迎再次回到节目。

Rocket, welcome back to the show.

很高兴你来到这里。

It's great to have you here.

谢谢，阿卡什。

Thanks Akash.

很高兴能来这里。

Great to be here.

所以我想确认一下，我理解得对不对？

So I just want to make sure I had this right, okay?

因为你已经给我解释过好几次了。

Because you've explained this to me a couple times.

现在我要向你解释一下，我理解这个诉讼是怎么回事。

I'm going to now explain to you what I understand this lawsuit to be about.

这个诉讼是埃隆·马斯克对OpenAI说：你们曾承诺要保持非营利性质。

So this is the lawsuit where Elon Musk said to OpenAI, You said you were going to remain a nonprofit.

但你们没有信守承诺。

You didn't keep your word.

他因此起诉了他们。

He sued them.

OpenAI确实变成了营利性机构，但埃隆·马斯克仍然想从OpenAI拿钱，因为他认为你们违背了承诺。

OpenAI still became a for profit, but Elon Musk still wants money from OpenAI because he says that you didn't keep your word.

没错。

That's right.

没错，就是这样。

That's exactly right.

我的意思是，他对OpenAI有一长串的不满，认为OpenAI在早期曾向他许下过许多承诺，但都没有兑现。

I mean, he has a long list of grievances against OpenAI, a number of ways that he feels that OpenAI hasn't kept its promises to him that were made in the early days.

非营利结构是其中之一，但他还觉得OpenAI背离了其最初的慈善使命，比如没有开源技术。

The nonprofit structure was one of those, but there are other ways he feels they've departed from their original charitable mission, including not open sourcing technology.

他认为安全性已经被边缘化了。

He feels that safety has become deprioritized.

所以他还有其他不满，但完全转变为营利性结构无疑是首要问题

So he has other grievances as well, but the full transition to a for profit structure is definitely top

之一。

of mind.

那么昨天在法庭上发生了什么？

And so what happened at the courthouse yesterday?

OpenAI当时向法院提出了特定的要求。

So OpenAI was making a particular ask to the court.

他们说，我们已经弄清楚了案件的许多事实。

They were saying, We've figured out a lot of the facts of the case already.

我们已经让许多与OpenAI关系密切的人接受了质询，包括萨姆·阿尔特曼、格雷格·布罗克曼和萨钦·恩德拉。

We've made a bunch of people who are close to OpenAI sit for depositions, including Sam Altman, Greg Brockman, Sachin Endela.

我们掌握了事实。

We know the facts.

OpenAI表示，即使你以对埃隆·马斯克最有利的方式解读这些事实，从法律上讲，你的结论仍然是他的案件毫无依据。

OpenAI is saying even if you interpret those facts in a way that's very generous to Elon Musk, still as a legal matter, you're going to find that his case doesn't have any merit

的。

to it.

所以你可以

So you can

现在就做出裁决。

just make a call right now.

省去我们的麻烦。

Spare us the trouble.

别让陪审团来操心。

Spare a jury.

直接驳回吧。

Just toss Exactly.

直接扔掉吧。

Out the Exactly.

法官最终得出相反的结论，她认为埃隆·马斯克仍保留的四项主张在当前审判阶段仍然有效。

The judge ended up finding to the contrary, she feels that all of the claims there are four claims that Elon Musk still has that are sort of live at this stage in the trial.

所有这些主张都基于需要由陪审团裁定的事实问题。

All of them rest on questions of fact that are matters for a jury to decide.

因此，这四项主张都将进入审判。

So all four of them are set to go to a trial.

好的。

Okay.

所以这件事要上法庭了。

So this is going to trial.

我想问问你，除了富豪之间的争斗本身很有趣之外，普通人为什么应该关注这个故事？

I want to ask you about why the average person should be paying attention to this story outside of the reasons that billionaires fighting each other is kind of fun to watch play out.

作为记者，你是怎么看待这个问题的？

As a reporter, how do you think about this?

我的意思是，最终是不是因为我们能随着证据的披露，更多地了解双方的情况？

I mean, is this ultimately the idea that we could just learn a lot more about both sides of the story as you know, evidence comes out?

你已经做了一些报道，关于实际可能涉及的资金数额，我认为实际金额并没有人们想象的那么大。

I mean, you've done some reporting about how much money could actually change hands, and I think the answer is it's not as big as people thought.

那我们为什么应该关注这件事呢？

So why should we be paying attention to this?

是的。

Yeah.

没错。

That's right.

我的意思是，即使在最极端的情况下，马斯克取得压倒性胜利，OpenAI可能需要支付的金额达到数十亿美元，但这对OpenAI来说也不会伤筋动骨。

I mean, even in the most extreme situations where you get sort of a slam dunk win for Musk, maybe in that case, the amount that OpenAI has to pay could it could reach into the billions, but that's not going to break the bank for OpenAI.

而且，目前法院不太可能介入并试图逆转OpenAI已经完成的重组，因为加利福尼亚州和特拉华州的总检察长已经批准了这一重组。

It's also unlikely at this point that the court would intervene and try to rewind the restructuring that OpenAI has already completed because the attorneys general for California and Delaware have already signed off on that.

因此，目前来看，此案的结果对双方都不太可能具有生存性影响。

So it's true that at this point, the outcomes from the case are unlikely to be sort of existential for either party.

但我认为，正如你所说，此案中涌现出的大量证据非常有趣。

But I think it's interesting, first of all, like you said, because of the treasure trove of evidence that's coming out in the case.

我的意思是，那些参与了2023年导致萨姆·阿尔特曼被OpenAI解雇、后又最终复职的谈判的人，提供了非常真实的第一手叙述。

I mean, really interesting first person accounts from people who were involved in the negotiations that led Sam Altman to be fired from OpenAI in 2023 and then eventually reinstated.

这让我们得以一窥那些对话实际发生时的场景。

So it's a real look inside the room where those conversations were happening.

这同时也是衡量公众意见的一个有趣指标。

It's also an interesting gauge of public opinion.

我的意思是，加州奥克兰的陪审团会对埃隆·马斯克和萨姆·阿尔特曼持什么看法？

I mean, what is a jury in Oakland, California going to think about Elon Musk versus Sam Altman?

我的意思是，陪审团更可能同情哪一方？

I mean, which of those sides is a jury going to be more sympathetic to?

我觉得很难说。

I I think it's hard to say.

所以当审判进行时，将会很有趣。

So it'll be interesting when a trial plays out.

这将为不可避免的电视剧提供绝佳素材。

And it'll give great fodder for the TV show that inevitably There will.

我觉得某天它终会问世。

Comes of it at some point, I think.

好吧，火箭，我肯定你很兴奋，因为如果你昨天在法院的话，既然要开庭了，我想你会在现场报道这场审判，对吧？

Well, Rocket, I'm sure you are excited because if you were at the courthouse yesterday, if it's going to trial, I I take it you will be covering the trial on the ground then whenever Yeah.

这太棒了。

It does Great.

这将会是一场精彩的对决。

Well, it'll be a fun one to play out.

火箭，感谢你前来做客。

Rocket, I wanna thank you for coming on.

这位是火箭，我们《信息》杂志的AI与机器人记者。

That is Rocket True, our AI and robotics reporter here at The Information.

好的。

Okay.

硅谷当前热议的一个话题是AI人才大战，以及各公司为吸引顶尖人才开出的巨额薪酬包。

One big conversation across Silicon Valley is the AI talent war and the massive pay packages that companies are offering to secure top talent.

同样地，《信息》独家报道了OpenAI庞大的员工股票激励池。

In that same vein, The Information has exclusive reporting about OpenAI's big employee stock grant pool.

现在加入我们的是这篇文章的作者、我们的OpenAI和Anthropic记者谢丽·穆皮蒂。

Joining me now is the author of the piece, Shree Mupiti, our OpenAI and Anthropic Reporter.

谢丽，欢迎再次做客我们的节目。

Shree, welcome back to the show.

很高兴你来到这里。

It's great to have you here.

很高兴能来这里。

Great to be here.

所以我来读一下你撰写和报道的文章的第一句话。

So I'm going to read the first sentence of your story that you wrote and reported.

你说，去年秋天，OpenAI 设立了一个价值占公司10%的员工股票奖励池，而该公司在10月的估值为5000亿美元。我不想假设大家都知道这些股票奖励池是如何运作的。

You said OpenAI last fall set aside an employee stock grant pool worth 10% of the company, which was valued in October at $500,000,000,000 So don't want to assume that people know how these stock grant pools actually work.

帮我们理解一下你发现了什么，以及这意味什么。

Help us understand what you found and what it means.

完全正确。

Totally.

这意味着，总共有价值500亿美元的潜在股票奖励将分配给员工，或作为现有员工的追加奖励。

So what it means is essentially there's $50,000,000,000 worth of potential stock grants that would go to employees or refreshes to existing employees.

这些奖励预计将在未来五年内发放，充分体现了OpenAI在吸引、奖励和留住人才方面的重视程度。

And so that's supposed to last over the next five years and really shows the extent at which OpenAI is prioritizing, rewarding, and retaining talent.

这一安排的由来是，去年秋天OpenAI进行重组，转型为一种允许其最终上市的公司结构，当时公司估值为5000亿美元，它从中划出了约1300亿美元的股份。

And so how that came about was as part of the restructuring last fall when OpenAI turned into a corporate structure that allows it to eventually go public, it set aside roughly a $130,000,000,000 worth of the company when it was valued at 500,000,000,000.

这大约占公司总股本的26%。

So that roughly equals out to 26% of the company.

在这26%中，已有800亿美元分配给了员工或前员工。

Of that 26%, 80,000,000,000 of that has already been allocated to employees or former employees.

这被称为已归属股权。

So this is called vested equity.

而剩余的500亿美元则用于这个池子，将分配给未来的员工，或作为对现有员工的奖励更新。

And then the 50 remaining 50,000,000,000 remaining is towards this pool that will go towards future employees or rewarding existing employees as part of these refreshes.

这是一笔巨大的金额，对于像OpenAI这样的公司来说非常令人振奋，因为人工智能人才的竞争正变得越来越激烈。

So that just is a huge amount and is really exciting for a company like OpenAI, just one of these AI talent awards are heating up even more and more.

所以这些股份实际上已经创建了。

So basically these shares have been created.

总共有1300亿美元的股份。

There's 130,000,000,000 worth of shares.

他们已经支付或分配了800亿美元，现在又发现了另外500亿美元，这些是他们预留出来，未来根据员工希望和谈判的股票数量进行发放的。

They've already paid out or allocated $80,000,000,000 and now there's $50,000,000,000 more that you've basically found that they've set aside to pay out at some point in the future to their employees, depending on how much stock they want and negotiate for.

所以我认为，你故事中更深层的问题是，这与其他公司如何分配股份相比如何。

So I think the bigger question that you kind of got at in your story is how does this compare to other companies and how they allocate

他们的股票？

their stock?

完全正确。

Totally.

对于像OpenAI这样的私营公司来说，这是一笔相当大的金额，但实际上与Meta这样的公司更为相似。

It's quite a large amount for a private company like OpenAI, but actually is more similar to a company like Meta.

例如，在过去五年中，Meta分配了大约660亿美元的股票补偿。

For example, over the last five years, Meta allocated roughly 66,000,000,000 of stock compensation.

因此，将此与OpenAI的500亿美元相比，例如，但Google的支出要多得多。

So comparing this to OpenAI is 50,000,000,000, for example, but Google, for example, spends a lot more.

他们在过去五年中花费了大约1000亿美元。

They spent roughly 100,000,000,000 over the last five years.

当然，这些是历史数据，而500亿美元是针对OpenAI未来五年的规划。

Of course, these are historicals versus the 50,000,000,000 is towards OpenAI's future five years.

因此，我们或许会看到Meta和Google的支出也进一步增加。

And so maybe we'll see Meta and Google spend also increase.

但真正突出的一点是，Meta和Google显然是规模大得多的公司。

But one thing that really stands out is that Meta and Google, of course, are much larger companies.

它们拥有巨大的市值，年收入高达数千亿美元。

They have huge market caps and they are in, like, hundreds of billions of revenue.

因此，当你将OpenAI去年的收入130亿美元与之对比时，就能明显看出两者之间的巨大差距。

And so when you're comparing against OpenAI that's at least ended last year in 13,000,000,000 in revenue, you really see sort of the sort of the jump in disconnect there.

但我认为这再次表明了OpenAI的远大抱负——试图与这类巨头竞争，因为如果你拥有顶尖的AI人才和研究人员，就必须能够跟上Google和Meta的步伐。

But I think that really shows again, opening eyes and ambitions to try to compete with the likes of this, just because if you are having these top AI talent AI researchers and you're competing against Google and Meta, you need to be able to keep up.

所以我认为这一点充分说明了这一点。

And so I think that that shows that.

我想强调的另一点是，目前在5000亿美元的估值下，员工股票授予池占公司总股本的10%。

The other thing I wanted to highlight too was that that the employee stock grant pool is 10% of the company at the current 500,000,000,000 valuation.

但当然，随着OpenAI筹集更多资金、估值上升，这个股票池的价值也会随之增加。

But of course, as OpenAI raises more money and the valuation increases, that stock pool worth actually goes up as well.

因此，你也能从他们去年夏天的预算中看到这种趋势，体现在他们分配的股票补偿费用上。

And so that kind of use, you see that as well with the stock compensation expenses that they've allocated from their, like last summer's projections as well.

投资者喜欢这样吗？

Do investors like this?

我认为投资者明白，这是私营公司未来的一部分，为了奖励人才和保持竞争力，这是必要的。

Investors, I think, understand that this is a part of a private company's future and, like, it's necessary to be able to reward talent and be able to compete.

因此，总体上他们理解这一点，但当涉及到像1300亿美元的员工股票期权这样的稀释时，尤其是这500亿美元的股票池，确实会让人感到痛苦。

And so overall, they understand this, but of course it bites when you have like dilution from, for example, dollars 130,000,000,000 worth of employee stock grants, especially this $50,000,000,000 pool.

但我认为他们明白这是必要的。

But I think they understand it's necessary.

因此，他们某种程度上是接受这一点的。

And so it's something that they're sort of like accepting.

我认为，思考OpenAI、投资者和员工之间的权力动态也挺有意思的。

I think it also is a little bit interesting to think about the power dynamic here between OpenAI and investors and then the employees.

一方面，研究人员是人才争夺战中的核心产品，因此他们拥有一定的议价能力。

On one hand, you have the researchers who they are the product in the talent war that everyone is going after, so they have some leverage there.

另一方面，OpenAI可能在投资者和资金来源方面拥有选择权。

And then you have OpenAI, which probably has their choice of investors and who to take money from.

但如果我仔细想想，OpenAI 需要投资人的资金。

Although if I think about this, OpenAI needs the money from investors.

我想这一切最终的结果是，我不确定谁真正拥有权力，但看到这些连锁反应如何逐步展开，确实很有趣。

I guess where this all nets out is I'm not sure who has the power here, but it is interesting to see how all these ripples very much play out.

完全正确。

Totally.

我觉得我听到的是，员工们现在正在努力工作，同时他们也得到了很好的回报。

I think employees are working hard right now is what I'm hearing, but they're being rewarded well as well.

例如，我的同事罗克特最近发现的一份法庭文件显示，到2021年，OpenAI 的联合创始人伊利亚拥有的股权价值约为40亿美元。

For example, in a recent court filings that my colleague Rocket had found, they found that Ilya, one of the cofounders of OpenAI had like roughly 4,000,000,000 invested equity by 2021.

因此，你可以想象，即使他已离开公司，他现在可能拥有的财富更多了。

And so you can imagine how much more he potentially has now, even after having left the company.

所以我认为，这仅仅展示了员工们所获得的收益规模之大。

And so I think that just shows the size of the stakes that employees are getting.

没错。

Right.

嗯，我想

Well, I want to

感谢你前来参加。

thank you for coming on.

我相信还有更多资金有待发掘，我认为这是一次非常出色的报道。

I trust that there is much more money to uncover, and I think it was some great reporting.

这位是我们在《The Information》的OpenAI和Anthropic记者李·穆皮蒂。

That is Sri Mupiti, our OpenAI and anthropic reporter here at The Information.

OpenAI一直公开表示其将购物功能整合到平台上的雄心，但今天《The Information》发布的一篇深度报道却对这些努力在实际执行中的情况给出了复杂的图景，揭示了要真正实现这一切有多么困难。

OpenAI has been very public about its ambitions to integrate shopping into its platform, although a new in-depth story from the information today gives a mixed picture on how those efforts are actually going on the ground and how difficult it is to actually get all this to work.

我想请我们的电子商务记者安妮·吉安上场，她撰写了这篇报道，和我们谈谈她的发现。

I want to bring on Anne Gian, our ecommerce reporter who wrote that story to talk to us about what she found.

安妮，欢迎再次回到节目。

Anne, welcome back to the show.

很高兴你在这里。

It's great to have you here.

嘿，阿卡什。

Hey, Akash.

那么，OpenAI 的购物计划进展如何？

So how are OpenAI's shopping efforts going?

进展还不错。

They're going.

正在稳步推进，但可能比一些人预期的要慢一些。

They're moving along, but maybe potentially a bit slower than some people would have hoped.

在我今天的文章中，我考察了 OpenAI 首次于九月在 ChatGPT 上宣布该功能后的实际落地情况。

In my story today, I took a look at what the rollout OpenAI first announced this feature for ChatGPT in September, what that's been like.

要让这一切顺利运行，OpenAI 和所有合作公司都付出了大量努力。

And it's taken a lot of work on the part of both OpenAI as well as all of the other companies that they've partnered with to make this work.

其他关键合作伙伴包括 Shopify，它为 OpenAI 打开了通往其平台上数百万商户的大门，还有 Stripe。

So some other key partners include Shopify, which has given OpenAI an in with all of the millions of merchants that use their software to host their stores, as well as Stripe.

OpenAI 和 Stripe 共同创建了一种名为‘智能商业协议’的系统，旨在建立一套统一的规则和标准，让各方就智能代理处理交易的方式达成共识，设定一些基本规范——有人向我描述说，这就像一种通用语言，让所有人都朝着同一个目标努力，从而让整个系统能更顺畅地协同工作。

OpenAI and Stripe actually created together what's called the agentic commerce protocol, which they intend to be this set of rules and standards so that everyone can be on the same page about how these transactions that agents will be handling, how it's gonna work and just kind of like set some some ground rules and, you know, was described to me as kind of like a a lingua franca so that everyone, you know, is building towards the same thing, you know, so that everything can work together more smoothly.

那么，这件事到底难在哪里？

And so what is so difficult about this?

是Shopify这一边更难，因为设置所有零售系统才是真正的挑战，还是支付环节才是关键难点？

Is the Shopify side harder in that setting up all the retail systems is the real challenge or is the payment side really the challenge here?

我的意思是，这里有很多需要考虑的因素。

I mean, there's a lot of considerations here.

我认为这之所以如此复杂，是因为涉及许多不同的系统，它们都必须协同工作、相互沟通。

I think part of why this is so complex is it's a lot of different systems that all have to work together and coordinate and communicate.

但根据我的报道，目前大部分时间和精力都集中在与已上线的商家紧密合作，帮助他们处理产品数据上。

But right now, according to my reporting, it seems like a lot of the time and energy is focused on working pretty closely with the merchants that are live and helping them with their product data.

每个商家对尺寸、定价和库存信息的处理方式都不同。

Every merchant handles sizing and pricing and inventory information differently.

所以，再次强调，就是要让所有人都达成一致，确保一切都能正确协同工作，这样当用户试图下单时，所有信息都能准确无误、顺畅运行。

So just again, that idea of trying to get everyone on the same page and making everything work together correctly so that when it does get to a user who's trying to make a purchase, that everything's accurate and works smoothly.

那么，其他那些表示对购物功能感兴趣的平台呢？

Now, what about the other platforms out there that have said that they are interested in shopping?

我认为，如果我没记错的话，Perplexity 在这个领域也有过一些动作，对吧？

I think, if I'm not mistaken, Perplexity had something in this arena, right?

他们的进展更靠前吗？

Is theirs further ahead?

是的，没错。

Yeah, that's correct.

所以 Perplexity 很早就推出了电商功能。

So Perplexity, they were very early to launch commerce features.

有趣的是，他们也遇到了这些问题——要整理商家提供的大量杂乱数据，不仅要确保这些信息能正确展示给用户，还要在后台真正促成交易时，让一切正常运作。

It's interesting because they also dealt with some of these issues, wrangling all of this very messy data from merchants and trying to get it to not only show up to users correctly, but then on the back end when you're actually facilitating a purchase, making everything work correctly.

这显然不是 OpenAI 独有的问题，但恰恰说明了将 AI 代理或 AI 应用接入电子商务有多么具有挑战性。

This definitely is not a problem that's unique to OpenAI, but I think just really speaks to just how challenging adding in AI agents or AI apps to e commerce has been.

这真的非常复杂，涉及众多公司，而且需要无数环节无缝协作才能达到完美的效果，关于如何

It's just really complicated and there's a lot of different companies at play and there's a lot of different things that all need to work smoothly for it to be perfect for How the

我们认为这对 OpenAI 来说是一个多大的机会？

big an opportunity do we think this is for OpenAI?

我的同事们报告说，OpenAI希望在本十年末从其免费用户那里获得约1100亿美元的收入。

Well, my colleagues have reported that OpenAI wants to make something on the order of $110,000,000,000 from its free users by the end of this decade.

所以

So

在

in

在未来四年里，我认为许多人预计电子商务将成为其中的重要组成部分。

the next four years, I think a lot of people expect that commerce will be a big piece of that.

因此，OpenAI表示，他们可能会从ChatGPT上完成的某些销售中抽取分成。

So OpenAI has said they might take a cut of some of the sales that are made on ChatGPT.

这是一种相当标准的平台模式。

It's a pretty standard kind of marketplace model.

所以我认为，随着规模扩大，如果他们开始从销售中抽取分成，肯定会看到来自这一渠道的更多收入。

So I think as this scales up, they certainly will be seeing more revenue coming from that if they do start taking a cut of the sales that they make.

此外，引入电子商务也带来了广告的潜在可能，这是我们已经多次讨论过的另一个话题。

And then as well as introducing commerce also brings in the potential for advertising, which is a whole other debate that we've written about a bunch.

但这无疑是OpenAI的巨大机遇。

But it's certainly, it's a huge opportunity for OpenAI.

所以我认为，他们缓慢推出这项功能并试图在全面开放之前把一切都做到完美，这是很合理的。

So I think it makes sense that they are rolling this out very slowly and trying to figure out exactly how to get this perfect before they flip the switch and open it up to everybody.

此外，你还报道过社交媒体平台的电子商务，比如TikTok如何更深地涉足这一领域。

Now, you've also covered e commerce with the social media platforms and how TikTok, for example, has waded deeper into that category.

我们对Instagram做过一些报道，我想我们已经看到了它在购物功能上反复进出的故事。

We've done some reporting about Instagram, and I guess we've seen that story play out in terms of it going in and out of shopping.

我不知道。

Don't know.

我不确定它现在的情况如何。

I'm not sure where it stands now.

感觉他们总想增加更多购物功能，但实际的购物行为却反而减少了。

It feels like they always want to do more shopping and then there's sort of less of it happening.

但除了难度之外（这一点我们已经理解），你认为OpenAI能从TikTok尝试整合购物的方式中学到哪些经验？

But what lessons do you think OpenAI can learn from the way that TikTok, for example, tried to integrate with shopping beyond just it being difficult, which I think we understand?

我想知道，客户在购物行为上是否有某些方面是他们正在学习的，比如人们实际上是如何进行购买的？

I wonder if there are behavioral considerations from customers that you think they are learning from insofar as how people actually do their buying?

当然。

For sure.

我认为这是他们非常关注的一点。

And I think that's something that they're extremely mindful of.

我认为这也与广告话题有关。

I think this also relates to the advertising conversation as well.

你知道，人们使用ChatGPT的用途非常广泛，但很多时候，你与ChatGPT的对话都涉及个人隐私和私密内容。

You know, people people use ChatGPT for a huge range of things, but a lot of times it's it's really personal information and personal conversations that you're having with ChatGPT.

那么，如果此时你开始引导他们购买商品或展示广告，人们会有什么感受？

And so how are people gonna feel if then you start prompting them to buy something or show them an ad?

因此，我认为OpenAI至少尚未公开说明他们的策略是什么。

And so I think that's something that OpenAI, hasn't at least publicly kind of said what their strategy will be.

但我认为他们对此非常谨慎，因为如果你在用户没有预期或尚未准备好的地方引入商业内容，这会让人非常反感。

But I think that's something they're extremely mindful of because I think if you start putting commerce in places where users don't expect it or aren't ready for it, I think that's a huge, huge turn off.

如果消费者对此反应不佳，这可能会严重阻碍他们的雄心。

So that could be a pretty big dent to their ambitions if it doesn't go over well with consumers.

很好。

Great.

安妮，非常感谢你前来做客。

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

这位是安妮·基恩，我们《信息》杂志的电子商务记者。

That is Anne Keehan, our e commerce reporter here at The Information.

好的。

Okay.

我想稍微谈一谈网络安全。

I want to talk a little bit about cybersecurity.

人工智能正在迅速改变企业软件领域的另一个细分领域。

It is yet another pocket of the enterprise software sector that AI is changing quickly.

Netskope 是这一故事的核心公司之一。

Netskope is one company that sits at the center of that story.

该公司股价低于其IPO价格，但过去十二个月的现金流已经为正。

The company is trading below its IPO price, although it is now generating cash on a trailing twelve months basis.

现在加入我们的是Netskope的首席执行官桑杰·巴里。

Joining me now is Sanjay Barry, CEO of Netskope.

桑杰，欢迎来到节目。

Sanjay, welcome to the show.

很高兴你能来到这里。

It's great to have you here.

很高兴受到邀请。

Great to have you invited.

谢谢。

Appreciate it.

我想谈谈人工智能时代下的网络安全领域。

So I want to talk a little bit about the cybersecurity sector in the era of AI broadly.

我对你提出的基本问题是：人工智能对网络安全是利大于弊还是弊大于利？

And the basic question I have for you is AI net good or net bad for cybersecurity?

我认为是总体有利的。

Well, net good is my view.

事实上，攻击者肯定会使用AI，正如你所见，他们会尽其所能地利用AI。

And look, the reality is attackers are going to use AI, as you've seen, and they'll use AI in every way they can.

他们能拿到的任何工具，都会用来发动攻击。

Every tool they can get their hands on, they'll use to attack.

但现实是，AI也可以用于正面用途，比如防护、自动化，帮助中小型公司以及大型企业全面提升防御能力，同时提高生产力。

But the reality is that AI can be used for good, for protection, for automation, to help midsize companies and small companies and large across the board defend themselves, and of course, increase productivity while doing it.

那么请你谈谈Netskope在AI新时代的应对策略，我想知道你如何看待自己与那些团队更小、行动更快的新兴初创公司的差异化优势。

And so talk a little bit about how you are approaching this new era of AI at Net Scope, I wonder how you think about your differentiation against some of these newer and flashier startups that probably have smaller teams, that can move a little faster.

你如何看待与它们的竞争？

How do you think about competing with them?

是的，如果我们看一下我们公司，我们是目前增长最快的上市网络安全公司之一，对吧？

Yeah, so if you look at us, we're one of the highest growth public market cyber companies out there, right?

上个季度，我们的经常性收入（ARR）实现了34%的增长。

We grew at 34% from an AR perspective last quarter.

我们已经实现了规模化，但我们有一句口号：创新或死亡。

We're at scale, but we have a phrase: innovate or die.

因此，我们保持灵活，并在研发上大力投入。

So we're nimble and we invest heavily in our R and D.

我们已经建立了八年的AI实验室。

We've had AI labs for eight years.

我们拥有超过160个领域特定模型，用于保护和识别敏感数据、防范恶意软件、检测您的网络及其他方面的异常。

We have over 160 models, domain specific models to protect and identify sensitive data, protect against malware, detect issues in your network and beyond.

除了这些模型和庞大的AI实验室团队——顺便说一句，AI已应用于我们平台二十多个产品中——我们还专注于保护和赋能AI。

In addition to those models and that broad AI Labs team, which by the way, AI is used throughout all of our 20 plus products on the platform, these models, we also focus on protecting and enabling AI.

这就是其中的细微差别。

And this is the nuance.

你不会走进一家公司说：嘿，顺便说一下，我们不希望你们使用AI。

You don't go into a company and say, Hey, by the way, we don't want you using AI.

不，没人会这么做。

No, nobody does that.

你必须使用它，就像任何行业的任何公司一样。

You have to use it, as any company in any vertical.

我们的做法是让你可以选择使用它，同时为其设置防护措施。

What we do is we let you say yes to using it and put guardrails around it.

你可能会说：你可以使用Gemini，但只能使用企业版，而且我不希望你用敏感的医疗数据进行查询。

You want to say, Hey, you can use Gemini, but only the corporate instance, and I don't want you querying with sensitive healthcare data.

MedScope上的简单策略。

Simple policy on MedScope.

因此，我们的重点是防护型AI，促进使用，并帮助人们充分利用它。

And so for us, our focus is guardrail AI, enable usage, and enable people to leverage it.

但我正试图理清一点：对于纯粹的网络安全公司来说，一方面，你有那些更炫酷的初创公司，它们筹集了大量资金，产品也获得了市场认可。

But I'm trying to sort of understand, you know, for pure play cybersecurity companies, on one hand, from below, you have these flashier startups that raise a ton of funding and their products get traction.

在某些情况下，它们的团队规模较小。

And in some cases, they have smaller teams.

另一方面，你有像微软和ServiceNow这样的大公司，它们拥有非常广泛的产品套件，其价值主张是：不要只把你的网络安全业务交给我们，把整个套件都交给我们。

On the other hand, you have these bigger companies like Microsoft and ServiceNow that have these really broad suites of products where their value prop is, well, don't just give us your cybersecurity business, give us the whole suite.

那么在中间位置，你如何思考自己相对于这两类公司的价值主张？

How do you, in the middle there, how do you think about your value proposition against both of those categories

是的。

and Yeah.

很好的问题。

A great question.

我相信，安全和网络领域并不存在一个万能平台。

So I'm a believer there is not one platform for all of security and networking.

我记得曾坐在房间里问过100位首席信息官：你们希望在这一光谱上处于哪个位置？

I remember sitting in a room asking 100 CIOs, Where do you want to be on the scale?

你们是想要100个安全和网络产品，还是只想要一个？

You want 100 secondurity and networking products or you want one?

他们说：不，不，我们哪个都不想要。

And they said, No, no, we don't want any of those.

不想有100个，也不想只有一个。

Don't want a 100, and we don't want one.

我们希望拥有几个核心平台。

We want a few core platforms.

这些核心平台之一就是我们所做的一切，即数据和网络安全的演进成果——SASE，根据大多数分析师报告，我们是这一领域的领导者。

And one of those core platforms is what we do, the evolution of data and network security, called SASE, which we're the leader in per most analyst reports.

因此，我坚信这一点。

And so I'm the believer in that.

几个核心的安全与网络平台在生态系统中协同工作。

A few core security networking platforms working well together in an ecosystem.

我们被授予了微软年度ISV合作伙伴称号，对吧？

We were named Microsoft ISV Partner of the Year, right?

我们之所以能获得分析师和其他厂商的认可，是因为无论你使用什么，平均客户都会使用两个以上的公有云、一百多个SaaS应用、数十个生成式AI应用，并访问数十亿个网站。

And there's a reason that we get that recognition from analysts and other vendors, is no matter what you use, two plus public clouds, average customer uses 100 plus SaaS apps, they use tens of generative AI applications, they go to billions of websites.

不管你使用什么都没关系。

It doesn't matter what you use.

你需要一个在网络安全层面统一、一致的层，来管理和保护所有这些应用。

You have one common, consistent layer at that network and security level that can govern and secure all of those.

这正是客户真正想要的。

That's really what customers want.

他们不希望被锁定在某个系统上，比如‘只适用于这个应用’或‘只适用于这个解决方案’。

They don't want to be locked into, Hey, only works for this app, or, Only works for this solution.

但他们也不希望只覆盖方程式中的一部分，比如仅提供数据保护。

Yet they also don't want something that is only going to cover one part of that equation, like only data protection.

所以，这就是那个恰到好处的平衡：我们不会强迫你购买全套产品，但也不会只给你一个点解决方案，我想是这样。

So that Goldilocks of, We won't lock you into the whole suite of everything, but we'll also give you more than just the point solution, I guess.

我想和你聊聊，2026年运营一家软件公司是什么样的体验。

I want to talk to you a little bit about just broadly what it's like to run a software company in 2026.

你已经经营这家公司大约十三年了，对吧？

You've been building the company now for, I think, thirteen years now, think, Yes.

如果我没记错的话，如今运营一家SaaS公司或广义上的软件公司，和十年前有什么不同吗？

If I'm not is running a SaaS company or a software company broadly different today than it was ten years ago?

你在招聘人才方面有什么变化吗？

Is there a difference in the talent that you have to go after?

人才更年轻了吗？

Is the talent younger?

他们更技术导向了吗？

Is it more technical?

你看到这种变化了吗？

How have you seen that change?

这是个好问题。

Yeah, it's good question.

我们一直重视技术人才和开发。

Look, we've always believed in technical talent and development.

我们从创新角度投入了大量资源在研发上。

We invest a significant amount in our R and D from an innovation perspective.

我会投资那些两三年后才会显现成果的东西。

I invest in things that are two, three years out.

八年前我们就设立了AI实验室。

Think about AI labs for eight years.

我会告诉你，过去五年里没人关心。

I would tell you nobody cared for five years.

我向客户谈论人工智能时，他们会说：‘你在说什么啊，桑杰？’

I talk about AI in front of customers and be like, What are talking about, Sanjay?

当然，现在他们关心了。

Now of course, they do.

所以我们有个说法，至少在我长大的加拿大，要滑向冰球将要去的地方。

And so we have a saying, at least where I grew up in Canada, skate to where the puck's going.

所以我觉得这就是...

And so I feel like that's

作为一位加拿大同胞，我喜欢这句话。

I like it as a fellow Canadian.

我喜欢这句话。

I like it.

我也喜欢。

I like it.

但我觉得不同的是，现在它得到了更多的认可，对吧？

But I feel like what's different is it's recognized more, right?

当你成为一家上市公司时，一个微妙之处在于，很多人都是按季度或按年度来管理公司。

One of the nuances of when you go and become a public market company, a lot of people, they manage their company quarter by quarter or year by year.

但这不是我的做法。

That's not how I do it.

我会在这里待上十年以上。

I'm going be here in ten plus years.

因此，我经营我们公司的方式，以及我认为你在开发技术核心平台时应该采取的方式，就是必须为未来投资。

And so the way I run our company, and the way I think you should when you're developing core platforms in tech, is you've got to invest for the future.

我认为这一点在许多组织中都被忽视了。

And I think that's lost in a lot of times in organizations.

但现在我认为它已经得到了很好的认可，对吧？

But I think it's well recognized now, right?

意思是，等等，你正在关注下一波浪潮——人工智能。

In the sense of, wait a minute, you're looking at the next wave, AI.

你正在关注物理AI。

You're looking at physical AI.

你正在关注这一点，所以我认为这已经改变了。

You're looking at and so I think that's changed.

第二部分是当今的技术人才。

The second part is technical talent nowadays.

你必须将AI作为技术人才的一部分加以利用，对吧？

You have to leverage AI as part of that technical talent, right?

所以你需要那些愿意学习的人。

And so you want people who want to learn.

你希望人们像十年前没有人醒来就说：嘿，我是使用这些工具的专家那样。

You want people, like no one woke up ten years ago saying, Hey, I'm an expert in using these tools.

这些工具当时还不存在。

They didn't exist.

因此，我们一直寻找那些更具创业精神、更愿意学习、能够掌握新技术并快速行动的人，对吧？

And so we've always gone for people who are a little more entrepreneurial, a little more, they're willing to learn, to grasp new technology, to move quick, right?

他们不会满足于只是坐在那里使用二十年前的技术。

They're not satisfied by just sitting there and leveraging something twenty years ago.

因此，这会引导你走向一种略有不同的 talent 和人的思维模式。

And so that leads you to a little different talent, I would say, and mindset of people.

我们想知道，你们是在九月份上市的吗？

We go, so you went public in September.

我想了解你们当时决定上市的考量是什么。

I wanted to understand what the calculus was to go public when you did.

我们经常谈论这个 IPO 窗口，这有点像一场舞蹈，对吧？

We talk about this IPO window, and it's a bit of a dance, right?

市场将走向何方？

Where are the markets going?

什么时候才是合适的时机？

What's the right time?

你们为什么选择九月？

Why did you decide September?

对于目前正在观望这一窗口的后期初创企业创始人，你有什么建议？

And what advice do you have to late stage founders who are looking at the window right now?

是的，我们选择上市的主要原因是提升知名度。

Yeah, so look, the reason that we went IPO was for awareness.

你知道，现实情况是，如果我们能达成一个概念验证（POC），我们有超过80%的概率会成功。

And you know, the reality is we went over 80% of the time if we can get to a POC, proof of concept, I.

E.

E.

我们的理想状态就是让客户亲自试用一下我们的平台，对吧？

Our nirvana, is just try the platform, right?

这就是我们的理想状态。

That's our nirvana.

因此，当你想到这一点时，我们实际上是在与那些享有上市后自然带来的知名度和优势的上市公司竞争。

And so when you think about that, we really compete against public market companies who have the benefit of the awareness and the natural this that you get when you're public.

所以，上市的目的不仅仅是融资？

So it wasn't just to raise money?

不，事实上，每个人都知道在私募市场也能融资，对吧？

No, the reality is everyone knows you can raise money in the private markets too, right?

我的意思是，这没什么秘密可言。

I mean, there's no secret to that.

所以对我们来说，关键是利用上市公司带来的知名度。

And so for us, it was, look, we can grab the awareness from being a public company.

我们内部长期以来一直这样运作，对吧？

We've operated like that internally for a long time, right?

纪律、基础设施、委员会等等。

The discipline, the infrastructure, the committees, the so on.

第三，它是一种货币，对吧？

And then third, it is a currency, right?

当你考虑招聘时，当你想到，坦白说，我们通常都是有机增长的。

When you think about recruiting, when you think about, you know, we generally build organically, to be frank.

看看我们的平台，我们是有机地构建起来的。

Like look at our platform, we built it organically.

所以所有这些因素。

And so all of those things.

另一点是，这是我们公司的第一局。

And the other thing too is, look, this is the first inning of our company.

我还会在这里待上十年以上。

I'll be here in ten plus years.

如果你在公司里，每次会议、每次全员大会，大家都知道，好吧，那是第一局。

If you're in the company, you know, every meeting we have and every all hands, everyone knows, okay, that was the first inning.

我们现在正处于公司的第二局。

We're now in the second inning of the company.

所以这就是我们的心态。

And so that's the mindset.

如果你当时为了提升知名度而上市，现在你们作为上市公司已经几个月了，而之前一直是私营公司，当然，两者各有优劣。

If you went public then for awareness, now that you've been a public company for a couple months and you've certainly been private, I mean, look, there's pros and cons of both.

作为上市公司，义务更多。

There's more obligations as a public company.

那么，你对打算让公司上市的创始人有什么建议？

What then is your advice to founders looking to take their company public?

嗯。

Yeah.

首先，这不适合每个人。

So one, look, it's not for everybody.

而且这也不一定是每个人的最佳选择。

And it isn't necessarily the right answer for everybody.

所以你必须做好准备。

So you have to be ready.

首先，在财务上你必须做好准备。

So one, financially you have to be ready.

我们已经实现了正向现金流。

Look, we're cash flow positive.

我们的纯订阅收入增长率超过30%，对吧？

We grow at over 30% pure subscription, right?