Centific在AI热潮中的角色、Databricks 1340亿美元估值、外星猎人融资 | 2025年12月16日

欢迎各位收看Informations TI TV。

Welcome everyone to the Informations TI TV.

我叫阿卡什·帕什里查。

My name is Akash Pashricha.

今天是12月16日，星期二。

It is Tuesday, December 16.

今天我们有很多内容要介绍。

We have a lot of ground to cover today.

首先，Databricks宣布获得40亿美元的新融资，公司估值达到1340亿美元。

First up, Databricks announced a new $4,000,000,000 funding ground valuing the company at a $134,000,000,000.

我们会为您详细解析这一消息。

We will break that all down for you.

接下来，一家小型银行正在采取重大举措，收紧其风险控制。

Next up, one small bank is making a big move, tightening its risk controls.

这很重要，因为它是在海外稳定币用户与传统美国金融体系之间的重要桥梁。

That matters because it is a critical bridge between overseas stablecoin users and the traditional US financial system.

我们将解释正在发生的变化及其原因。

We will explain what's changing and why.

接下来，我们将邀请Sapphire Ventures谈谈他们对明年企业软件的预测。

We're then bringing on Sapphire Ventures to talk about their enterprise software predictions for the coming year.

你绝对不能错过这一点。

And you won't want to miss this.

詹森·黄是否能帮助资助地外生命的探索？

Could Jensen Huang help fund the search for extraterrestrial life?

我们将告诉你，一群外星生命猎人为何希望与英伟达CEO合作开发一款人工智能驱动的望远镜。

We'll tell you why a group of alien hunters wants to partner with the NVIDIA CEO on an AI powered telescope.

最后，我们将与Centific进行对话，这是一家在数据标注和模型定制领域快速崛起的企业。

And finally, we will end the show with a conversation with Centific, a fast growing player in the arena of data labeling and model customization.

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 things.

Databricks今天宣布完成40亿美元的L轮融资，公司估值达到1340亿美元。该公司还表示，其第三季度收入年化规模已超过48亿美元，同比增长超过55%。

Databricks today announced a $4,000,000,000 Series L investment valuing the company at $134,000,000,000 The firm also says that it surpassed a $4,800,000,000 revenue run rate in the third quarter, growing over 55% from the previous year.

现在邀请我们的企业软件记者凯文·麦克拉林来讨论这一话题。

Joining me now to discuss this is our enterprise software reporter, Kevin McLaughlin.

凯文，欢迎再次做客我们的节目。

Kevin, welcome back to the show.

很高兴你来到这里。

It's great to have you here.

谢谢阿卡什邀请我。

Thanks, Akash, for having me.

感觉现在Databricks差不多每季度都会宣布一次私募融资？

So it feels like Databricks announces these private fundraisings on what, like a quarterly cadence now?

是的。

Yeah.

到了这个地步，你还想说什么呢？

I mean, what can you say at this point?

几个月前他们的估值还是100亿美元，现在却达到了134亿。

It was a $100,000,000,000 valuation a couple months ago, and now it's a 134.

我认为，看待Databricks和它们正在进行的这些融资的方式是，目前还没有任何公司真正成为企业级AI代理领域的主导者。

I I think the way to look at Databricks and all of this fundraising they're doing is that no one has really established themselves as the dominant company in AI agents for the enterprise.

上个月我参加了AWS re:Invent大会，当时与AWS高管以及他们的软件合作伙伴的许多讨论都围绕着这样一个主题：AWS认为这仍然是一个开放的领域，胜负未定。

I was at AWS re:Invent last month, and a lot of the discussions I had there with AWS executives as well as their software partners was around this theme of, you know, AWS thinks that it's still a wide open space and it's anyone's game.

因此，Databricks需要筹集大量资金，因为它提供的产品几乎与云服务商完全相同。

So Databricks needs to raise lots of money because it also has an almost identical array of offerings as the cloud providers do.

Databricks拥有用于准备数据、存储数据的工具，更重要的是，它还能构建AI代理，这些代理能接入你的数据，并让你针对各种用例进行定制。

Databricks has tools for preparing your data, storing your data, and most importantly, building AI agents, which tap into your data and allow you to customize it for various use cases.

所以，看到这个公告我其实并不感到意外。

So not really surprised actually to see the announcement.

我很想听听你说Databricks正在与云服务商正面竞争，因为至少一年前，甚至两年前，人们讨论的竞争对手其实是Databricks和Snowflake，对吧？

So I'm interested to hear you say that the Databricks is really going head to head with the cloud providers, because at least a year ago, or maybe two years ago, mean, conversation was really Databricks versus Snowflake, right?

那时人们谈论的 rivalry 就是这两家公司。

That was the rivalry that people talked about.

而如今，Databricks的增长速度远超Snowflake。

And here you have Databricks now that's growing much faster than Snowflake.

它是一家大得多的公司。

It's a much bigger company.

那场竞争怎么了？

What happened to that rivalry?

哦，这场竞争显然还在继续。

Oh, it's still definitely going on.

Databricks 对阵 Snowflake。

Databricks versus Snowflake.

这是企业软件市场的哈特菲尔德家族和麦考伊家族之争。

It's the Hatfields and McCoys of the enterprise software market.

我觉得 Snowflake 的侧重点略有不同。

I I think Snowflake has got a a sort of slightly different focus.

它们都有相同的工具。

They've all got the same tools.

Snowflake 两者都有数据仓库，而且 Snowflake 在人工智能方面也有雄心壮志，还拥有自己的 AI 代理构建平台。

And Snowflake, they both have, like, data warehouses, and Snowflake has big aspirations in AI as well as its own AI agent building platform.

我认为，从我的角度来看，差异在于Databricks已经开始把自己定位为一家更纯粹的AI原生公司，而Snowflake则没有。

I think the difference is what, my perspective at least, is that Databricks has started to separate itself as more of a AI native company than Snowflake has.

Snowflake的根基在于传统企业，而首席执行官Sridhar Ramaswamy正努力将公司更多地引向AI领域，但这仍在进行中。

Snowflake's roots are into the traditional enterprise, and CEO Sridhar Ramaswamy has been trying to steer the company more into AI, and that's a work in progress.

而Databricks，我认为在这个领域已经略微领先。

Whereas Databricks, I think, has a little bit of a head start in this space.

我认为，如今如果你想要做AI，你可以选择这两家公司中的任何一家。

And I think the perception is that if you wanna do AI today, you can go to both of them.

你可以选其中任意一家。

You can go to either one.

但以我的观点来看，Databricks稍微领先一些。

But from my perspective, Databricks is a little bit ahead.

对。

Right.

关于这个话题，我最后再问你一个问题。

And last question for you on this topic.

你认为Databricks继续保持私有更长时间会有什么影响？

What do you think the implications are for Databricks staying private even longer?

我们看到SpaceX终于开始向IPO方向略有转变。

We saw that SpaceX has finally started to pivot a little bit towards an IPO.

Databricks是人们密切关注的一家公司。

Databricks is one that people are watching.

我的意思是，这里的 implications 是什么？

Mean, the implications here are what?

我真的不了解CEO Ali Godsey或Databricks董事会的想法。

I really am not inside the head of CEO Ali Godsey or the Databricks board.

我敢推测，也许他们相信某一天市场会降温，那时不成为上市公司可能更有利。

I I would venture to say that, you know, maybe there's a a belief that things are gonna cool down at some point, maybe might be advantageous to to not be a public company at that point.

另一方面，他们似乎正在增长。

On the other hand, they seem to be growing.

重要的是，在这一轮融资中，他们允许员工进行二级股份出售。

Importantly, in this latest round, they're allowing employees to do secondary share sales.

所以人才保留应该不是问题。

So retention shouldn't be a problem.

实际上，我在融资公告中看到，他们计划将已经非常庞大的员工队伍再增加数千人。

And actually, I saw in the funding announcement that they're planning to grow the already very large employee base by several thousand employees.

他们称

They called

最近他们说最近的大学毕业生。

out recent they said recent college graduates.

他们是这么说的吗？

Is that what they said?

我没看到这部分。

I didn't see that part.

好吧。

Okay.

也许是我记错了。

Well, maybe I'm making that up.

不知道。

Don't know.

我感觉我读到过这个。

I felt like I read that.

但不管怎样，我确实看到他们在大量招聘。

But anyway, I did see that they're hiring a lot.

所以，是的，我的意思是，他们在各个维度上显然都在快速增长。

So, yeah, I mean, sounds like they are certainly growing quickly on all vectors.

我想转到你为我们《应用人工智能》通讯撰写的一篇文章。

I want to pivot to a piece that you wrote in our Applied AI newsletter.

你谈到了客户聊天机器人的激增。

You talked about the surge of these customer chatbots.

我说的客户聊天机器人，指的是公司用于处理客户事务以及员工事务的聊天机器人整体数量的激增。

And I say customer chatbots, I mean, the surge in chatbots, period, that are being used at companies to deal with customers and then also to deal with employees.

我的意思是，谈谈这里的竞争吧，因为有太多不同的参与者了。

I mean, talk a little bit about the competition here because, I mean, there are so many different players.

有Sierra，有Decagon，还有Ada。

There's Sierra, there's Decagon, there's Ada.

这些玩家之间有什么办法可以区分吗？

Is there any way to differentiate all these players?

有的。

Yeah.

这就像我刚才描述的那样，大家都在为企业的AI代理构建平台。

It's not unlike what I just described where everyone is building platforms for building AI agents for the enterprise.

聊天机器人多年来一直用于接待网站访客。

Chatbots have been greeting visitors to websites for many years.

在生成式AI出现之前，它们实际上非常无用。

And before generative AI arrived, they were pretty useless actually.

它们会说：嘿，我在这里帮助你。

Know, they'd say, Hey, I'm here to help you.

我能回答你的问题吗？

Can I answer your questions?

你会问它一个问题，它会回答：‘这里有个链接’，或者‘我不知道’。

And you would ask it a question and it would say, Well, here's a link, or I don't know.

所以AI确实极大地改变了这一局面。

So AI has definitely changed the game a lot.

几周前，我们注意到Sierra为它的新客户Gap Inc.构建了一个聊天机器人，这个机器人却在讨论一些它不该涉及的话题，完全跑偏了。

Now, what we noticed a couple weeks ago, reported that Sierra had built a chatbot for Gap Inc, one of its new customers, and this chatbot was engaging on topics that it shouldn't have this going way off topic.

因此，我想着去查看其他网站上那些当你访问网站并浏览产品时弹出的助手。

And so I think that sort of spurred me to kind of look at other websites that have these assistants that pop up when you visit the website and you start browsing through products.

我们今天发布的简报是关于几个网站显然没有对聊天机器人进行足够限制，比如其中一个，你可以说：‘嘿，我正在计划一个节日派对。’

And the newsletter we have going today is about how a couple of these websites, clearly haven't locked down the chatbots enough because they were, you know, one of them was talking about, you know, you could say, hey, I'm planning a holiday party.

‘我该怎么做？’

Like, what should I do?

它竟然会认真回应这个话题，并给你建议。

And it would actually engage on the topic and give you advice.

而且情况更进一步，你还可以问：‘如果有人在我派对上滑倒了怎么办？’

And then it went even further because you could ask, Well, what if somebody slips and falls at my party?

我会承担法律责任吗？

Would I be legally liable?

顺便问一下，我该如何进行医学治疗？

And oh, by the way, how should I approach the treatment medically?

显然，这些话题是聊天机器人不应该参与的，这可能会使公司面临某种形式的法律责任。

And so these are clearly topics that the chatbots shouldn't be engaging in, and somewhat might open the companies up to some form of liability.

那么公司对此是怎么向你解释的？

And so what are the companies telling you about this?

Glean 就是我们今天早上报道的一个聊天机器人例子。

Well, Glean is one example of a chatbot that we're reporting on this morning.

他们表示已经更改了配置来阻止这种情况。

They said that they've changed the configuration to stop it.

我认为在很多情况下，他们根本没想到网站访客会进来问这类问题。

And I think in a lot of cases, they're just not expecting website visitors to come in and start asking these kinds of questions.

所以我认为这反映了公司在面向客户的场景中采用人工智能的方式。

So I think it's a reflection of how, you know, companies are adopting AI in customer facing situations.

而且这些应用还比较新，也许他们还没有充分考虑到所有可能性。

And they just, these implementations are still fairly new, maybe they haven't thought through all the possibilities.

但没错，Glean 表示他们已经修复了他们的聊天机器人。

But yeah, Glean says that they've fixed their chatbot.

很好。

Great.

那么，凯文，我确实经常使用它们，所以听到公司如何从后台思考这些问题很有趣。

Well, Kevin, I certainly use them, and so it's interesting to hear how the companies are thinking about them on the back end.

很难避开它们。

It's kind of hard to avoid them.

当然，坦白说，我每次在整段对话结束时还是要求转接真人，所以我还在期待一些改进。

Of course, to be frank, I still just ask for a person at the end of the whole exchange, so I'm still holding out for a little bit of improvement.

凯文，非常感谢你来参加。

Kevin, thank you so much for coming on.

这位是凯文·麦克劳林，我们《信息报》的企业软件记者。

That is Kevin McLaughlin, our enterprise software reporter here at The Information.

加密行业正试图构建一个全新的去中心化金融体系，但在许多方面，其核心基础设施仍与传统金融体系紧密相连。

The crypto sector is trying to build a new decentralized financial system, but in many ways, much of the core infrastructure is still tied to the traditional financial system.

今天，《The Information》发布了一篇文章，提供了另一个与此相关的数据点，聚焦于稳定币。

Today, The Information published a story with another data point to that effect focused on stablecoins.

我想邀请我们的加密记者杨悦琪，来告诉我们她发现了什么。

I wanna bring on Yueqi Yang, our crypto reporter, to tell us more about what she's learned.

悦琪，欢迎再次做客节目。

Yuechi, welcome back to the show.

很高兴你来到这里。

It's great to have you here.

嗨，阿卡什。

Hey, Akash.

在你今天的文章中，你聚焦于这家银行——Lead Bank。

So in your piece today, you focus on this bank, Lead Bank.

给我们讲讲Lead Bank吧。

Tell us about Lead Bank.

我从来没听说过这家银行。

It's not one that I've heard of.

我们为什么要关注它？

Why should we be focused on them?

在过去两年里，Lead Bank 已成为稳定币支付行业最重要的银行服务商。

LiBank has emerged as the most important banking providers for the stablecoin payment industry in the past two years.

你可能听说过许多最受欢迎的稳定币初创公司，比如今年被 Stripe 收购的 Bridge，它们都依赖 LeadBank 提供服务。

Most of the most popular stablecoin startups that you might have heard of, such as Bridge, which was acquired by Stripe this year, rely on LeadBank to provide their services.

正如你所说，它并不是家喻户晓的名字，但对这个行业至关重要，因此我们深入研究了它所提供的服务，以及它所推行的一些变革。

And as you mentioned, it's not a household name, but Le is very important for the sector, which is why we took a deeper look into the services that it provide, but also some of the changes that it's imposing.

我要提一下，Le Bank 是一家位于堪萨斯城的社区银行。

And I will mention that Le Bank is a community bank based in Kansas City.

它是一家小型银行。

It is a small bank.

它的资产总额约为20亿美元。

It has about $2,000,000,000 in assets.

对。

Right.

现在，你说它为发行稳定币的公司提供的服务，这些服务具体是什么？

Now, you say the services that it provides to companies issuing stablecoins, what exactly are these services?

最重要的服务是为稳定币的最终用户提供银行账户。

The most important service is to provide bank accounts for the end users of stablecoins.

通常，这些用户是阿根廷、墨西哥等新兴市场的人和企业，他们希望收到美元作为汇款或来自国外雇主的工资，以便储存美元财富。

Oftentimes, these are people and businesses in emerging markets such as Argentina, Mexico, where they're hoping to receive US dollar for remittances or for their salary from, their employers outside of their own country, so that they can, get the US dollar, store their wealth in US dollar, currency.

因此，出现了一整批稳定币初创公司来满足这些用户的需求。

And there's this whole group of stablecoin startups that have popped up to serve this need for these users.

而它们确实需要一家银行作为中介，以便它们能够接收美元。

And they do need a bank in the middle of this because a way for them to receive the US dollar.

我之前没意识到这一点，也许我一直忽略了这一点。

So I didn't realize this, and maybe I've just been missing it.

但如果我想用USDC或任何稳定币支付某样东西，我的意思是，如果我去购买USDC，后台实际上会为我开设一个银行账户，用来持有储备金吗？

But if I wanted to pay for something with USDC or with any stablecoin, I mean, if I go about buying USDC, in the back end, there will be a bank account that is opened for me, essentially holding the reserves?

存放在那个银行账户里吗？

Is being held in that bank account?

银行账户本质上为你提供了在实体经济与稳定币加密世界之间进行出入金的途径，这就是需要账户的原因。

The bank account essentially provides you a way to on ramp and off ramp between the real economy and the stable the stable coin crypto world, and that's when the account is needed.

例如，一些美国公司使用美元支付海外供应商和员工，而稳定币初创公司会从这个银行账户将美元兑换成稳定币，以便以更快、更便宜的方式跨境传输。

For example, there are US companies that are paying overseas suppliers and employees using US dollar, and the stablecoin startups will convert that US dollar into from this bank account into stablecoins so that send it across border in a faster and cheaper way.

好的。

Okay.

那么，根据你的报道，Lead Bank 目前正在做出哪些改变？

So now what are the changes that Lead Bank is making now according to your reporting?

在过去几个月里，我们发现 Lead Bank 在其对稳定币初创公司的服务方面正在加强风险控制。

In the past few months, we have find out signs that Lead Bank is tightening its risk control when it comes to its services for stablecoin startups.

具体来说，它正在对某些国家和行业施加更多限制，限制接受来自这些地区和领域的客户。

And specifically, it's increasing restrictions for certain countries and sectors where it would accept clients from.

我们还发现，银行审查交易所需的时间变长了。

And we also come across signs where it's taking longer for the banks to review transactions.

这是因为它们需要时间来核实资金来源，并增强对使用其银行账户的账户持有人身份的信心。

And this is because they need time to verify the source of funding and to strengthen their confidence in the identity of the account holders that are making use of their bank accounts.

所以根据你的报道，这听起来像是某些技术挑战导致了其监管的收紧。

So based on your reporting here, it sounds like there are some technical challenges that have led to it tightening its regulations.

除了技术原因外，这家公司采取更严格的方式对待与稳定币的合作，是否有更广泛的原因？

Are there broader reasons why the company is taking sort of a tighter approach to how it works with stablecoins?

是的。

Yes.

对于银行来说，为稳定币初创公司提供服务，尤其是那些拥有大量来自世界各地用户的公司，始终是一项高风险业务，因为这会给合规团队在监控多层供应商风险和验证资金来源方面带来更大挑战。

For a bank, to be able to serve stablecoin startups, especially the ones that are having a lot of users in a wide range of countries, this is always a high risk line of business because it brings your compliance team greater challenge when it comes to monitoring the risk across layers of vendors and also verifying the source of fund.

这也是许多其他银行不愿涉足这一业务的原因，而这一业务对于今年《天才法案》通过后涌现的众多稳定币初创公司来说至关重要。

It is a reason why a lot of the other banks are reluctant to get into this business, which is much needed for all these stablecoin startups that have popped up this year after the passage of Genius Law.

因此，这为银行提供了发挥巨大作用的空间，但也要求银行拥有强大的合规流程，以便能够妥善监控这些风险。

And that's why it gives room for the bank to play such an outsized role, but it also requires the bank to have a strong compliance process that allows them to monitor these risks properly.

对。

Right.

所以，悦琪，对于读者来说，这个故事在细节上探讨了稳定币在后台是如何运作的。

And so, Yueqi, for people reading a story, it's a story that is somewhat in the weeds in terms of how these stablecoins work in the background.

从一个非常宏观的角度来看，为什么人们应该关注这一点？

At a really high level, why should people be paying attention to this?

这是否预示着稳定币未来的发展方向？

Is this a signal for what's to come with stablecoins?

如果这些银行决定不再与稳定币合作，稳定币是否会受到更多限制？

Could they at all become more restricted if these banks decide that they don't want to work with stablecoins?

你是如何看待这个问题的？

How are you thinking about that?

我认为，鉴于今年有如此多的人关注稳定币，理解稳定币整体领域可能存在的风险来源，以及哪些主要参与者在市场中扮演着关键角色，是非常重要的。

I think given that so many people are paying attention to stablecoins this year, it is important for them to understand where the risk for the broader stablecoin sector might be emerging and who are some of the biggest players that are serving outside roles in this market.

这就是我们为何要重点提及LeadBank的原因。

And that's why we highlighted LeadBank.

LeadBank所做出的改变，将对众多稳定币提供商产生连锁影响。

And the changes that LeadBank make will have ripple effect across a long list of stablecoin providers.

对。

Right.

正如你所说，LeadBank 并不是唯一一家银行，但支持这类业务的银行数量仍然很少。

And as you say, I mean, LeadBank is not the only bank, but it's still a small number of banks that are supporting this type of work.

因此，这听起来像是稳定币未来增长的一个瓶颈。

So it's certainly a bit of a bottleneck, it sounds like, in terms of how stablecoins will grow in the future.

Yueqi，非常感谢你参加我们的节目。

Yuecchi, thank you so much for coming on the show.

我们非常感激，也期待早日再次邀请

We really appreciate it, and we look forward to having

你再次做客。

you back again soon.

好的。

Okay.

ServiceNow 本周完成了对 Moveworks 的收购，但 ServiceNow 的股东似乎对该公司可能即将进行另一项重大收购——这次是在网络安全领域——的消息感到有些不安。

ServiceNow's Moveworks acquisition closed this week, but ServiceNow shareholders appeared to be somewhat spooked by the news that it could be nearing another big acquisition, this time in the cybersecurity space.

ServiceNow 属于那些试图帮助企业购买其人工智能工具的软件公司行列。

ServiceNow belongs to the pool of software companies that are trying to figure out how to get businesses to buy their AI tools.

如今，我一直在想，到2026年，这种对话是否会改变，以及会如何改变。

And these days, the question on my mind is how that conversation is going to change in 2026, if at all.

我想邀请萨菲勒风险投资公司的管理合伙人拉吉夫·达姆，他投资了众多企业软件公司和人工智能公司，其中就包括Moveworks。

I want to bring on Rajeev Dham, managing partner at Sapphire Ventures, who has invested in many enterprise software companies and many AI companies, not the least of which is Moveworks.

拉吉夫，欢迎来到节目。

Rajeev, welcome to the show.

很高兴你能来。

It's great to have you here.

很高兴能来这里。

Great to be here.

谢谢你的邀请。

Thanks for having me.

鉴于你通过Moveworks将一家公司出售给了ServiceNow，我猜你一直在密切关注ServiceNow。

So I trust that you are closely following ServiceNow, given that you sold a company to ServiceNow by way of Moveworks.

你如何看待本周市场波动，那些进行收购的公司股价下跌了超过11%？

What did you make about the market jitters this week that companies making another acquisition, stock falls more than, I think it was 11%?

我的意思是，我们昨晚在每日简报中写道，考虑到公司依然健康增长，且‘40法则’远高于此，这似乎有点反应过度了。

I mean, we wrote last night in our nightly briefing, it seemed like a little bit of overreaction given the company is still growing pretty healthily, and the rule of 40 is far north of that.

是的。

Yeah.

听我说，我认为目前的财务业绩非常强劲。

Listen, I think the financial results are strong at this point.

正如你可能注意到的，公开市场似乎正处于避险模式，这可能是由于年底的获利了结，但同时，纳斯达克指数今年迄今已上涨20%，连续第二或第三个强劲年份。

As you could probably tell, the public market feels like it's in a risk off mode, likely due to some profit taking at the end of the year, but also, listen, the NASDAQ's up 20% year to date for the second or third blockbuster year in a row.

我认为，无论是投资者还是其他人，都依然在担忧哪些公司是AI的赢家，哪些是输家。

And I think everyone's still sort of worried, investors alike, about which companies are AI haves versus AI have nots.

谁才是AI的受益者？

Who are the AI beneficiaries?

这主要基于我的直觉，至少目前来看，ServiceNow可能在创新竞赛中的形象有所削弱，因为他们连续进行了多笔大型并购交易。

And so this is mostly based on my senses, least ServiceNow potentially maybe losing the perception, at least for the moment, around the innovation race because they've had a string of large M and A transactions.

但在我看来，这些交易表明他们正在积极抢占人工智能创新的前沿。

But in my view, I think these transactions are a signal of them being aggressive in getting ahead of the AI innovation race and curve.

那你怎么看待他们深入网络安全领域呢？

And what do you make about them diving deeper into cybersecurity?

我的意思是，说到ServiceNow，我们知道它主要是IT服务管理或后端IT流程。

Mean, when it comes to ServiceNow, we know it's ITSM or the back end IT processes.

我们知道他们正在深入CRM软件领域，与Salesforce竞争。

We know they're going deep into CRM software competing with Salesforce.

网络安全？我的意思是，我知道他们什么都能做，但这并不是我首先想到的他们的核心领域。

Cybersecurity, I mean, I know they do everything, but it's not the first thing I think of them for.

当然。

For sure.

当然。

For sure.

我认为这是他们在扩大总可用市场。

And I think it's them expanding their TAM.

听好了。

Listen.

这里没人是ServiceNow专家。

No ServiceNow expert here.

但我会说，网络安全和云安全总体上是我们风险投资领域看到的最热门的领域之一，尤其是考虑到一些公司的公开上市及其估值，而且它也是人工智能的受益者。

But I will say cybersecurity and cloud security in general is one of the hottest sectors we see in venture land given some public offerings and how those companies are valued, and it is also an AI beneficiary.

围绕云安全和网络安全的利好因素正在因人工智能而变得更加强劲或更有价值，我认为这完全合情合理。

There's so many tailwinds around cloud and cybersecurity getting stronger or more valuable in the face of AI that I think it makes all the sense in the world.

如果你想想CISO和安全团队与IT部门紧密相邻，我就能理解为什么这会是ServiceNow自然的延伸领域。

And if you think about a CISO and security sitting adjacent to IT, I could see how this is a natural adjacency for ServiceNow to get into.

对。

Right.

在你本周早些时候上节目之前，我曾请你提供几个我想在节目中讨论的2026年预测。

Now, before you came on the show earlier this week, I had asked you for a couple predictions that I wanted to talk about on the show for 2026.

你发来了一张清单，其中我想和你深入探讨的一个是：你提到2026年智能代理将实现融合。

And you sent through a list of them, and one of the ones I wanted to unpack with you is you said that agents will converge in 2026.

我想问问你，你这么说是什么意思。

I wanted to ask you a little bit for what you meant by that.

是的。

Yeah.

我认为现在我们有很多专注于非常狭窄领域的代理。

I think right now we have a bunch of agents that are focused on very narrow slices.

我认为企业拥有一种单一的代理来负责客户对话，这确实是一个机会。

And I do think that there is an opportunity for enterprises to have one singular agent, for example, responsible for customer conversations.

目前，这些代理只处理与客户支持、销售或营销相关的非常有限的对话，代表商家或企业与终端消费者交流。

Right now there's agents that do very thin slices of conversations as it relates customer support or sales or marketing, for example, on behalf of a merchant or an enterprise talking to end consumers.

我认为，可能会出现某种协调层，这个层面对最终用户来说可能是隐藏的，或者会有一个代理统一整合企业内的一切功能。

I think there's going to be some sort of orchestration layer that may be sort of obfuscated from the end user customer in that case, or specifically one agent that is going to sort of converge across everything on the enterprise.

实际上，我认为对消费者而言，更令人兴奋的是，你将在消费端看到这种趋势。

I actually think even more exciting maybe for us as consumers, you're going to see that on the consumer side.

大家一直在谈论一种操作型代理，用来帮你预订机票和餐厅。

Everyone's been talking about an operator agent to book your tickets in restaurants.

我认为在消费者端，这一点也终于要开始释放了。

I think you're also going to see that on the consumer side finally start to unlock.

对。

Right.

那么，那些在这些孤岛中运作的初创公司会怎样呢？

And so what happens to the startups here that are playing sort of in these silos?

我的意思是，我们在这个节目里经常讨论这个话题。

Mean, this is a conversation we have on this show all the time.

感觉这些企业软件公司真的开始融合了。

It feels like these enterprise software companies are really starting to converge.

这是整合吗？

Is this consolidation?

这是收购吗？

Is this acquisitions?

这是并购吗？

Is this M and A?

我认为大多数处于初创阶段的公司都是通过有机方式成长的。

Well, I think most of the companies that are in the startup phase mostly grow organically.

因此，你会看到它们逐步扩展其服务企业的使用场景和训练方式，以及整合更多功能，以覆盖更多样化的对话类型，而不仅仅局限于单一类型和单一模式。

And so you're going to see them sort of just expand their set of use cases and training for the enterprises that they serve and their set of integrations in order to nail more conversations than just sort of one type and one modality.

我认为你将看到的正是这种逐步应对的方式。

And that's all I think you're going to see start sort of tackle it.

更重要的是，我认为2026年的信号与噪音比将达到历史最高水平。

Think more so than that, I think the signal to noise ratio in 2026 is gonna be at an all time high.

我们目前工作中最困难的一点是，市场上充斥着大量噪音，很难分辨谁真正为企业创造了投资回报和价值。

One of the hardest things of our job right now is there's a lot of noise out there who's actually generating ROI and value of the enterprise.

我认为这不仅仅是初创公司会变得更好，企业自身也将学会如何避免盲目实验，如何采购真正优质的AI代理产品。

I think it's not just because the startup's gonna become better, but the enterprises are gonna learn how not to experiment, how to procure actually the best AI agent products.

因此，你会在企业端看到更多真正的信号，像我们几个月前提到的麻省理工学院的那项研究，到2026年可能会完全反转。

And so you're gonna see a lot more signal in the enterprises, those studies like the MIT one that we had a few months ago are going to sort of flip on its head come 2026.

而这正是我们最期待的事情。

And so that's what we're most excited about.

现在，你另一个预测是，你说Anthropic将发现它的下一个主要应用。

Now, another prediction you had is you said that Anthropic will discover its next major application.

我们显然将Anthropic和Claude与它们在编码领域的突破联系在一起。

We obviously associate Anthropic and Claude with their coding traction that they've gotten.

但你认为下一个应用会是什么？

But what do you think that next application is?

是的。

Yeah.

听我说，我认为这些人会瞄准最广泛、最同质化的应用领域。

Listen, I think that these folks are going to pick off the widest and most homogenous sort of application areas.

所以，也许可能是客户服务之类的地方，无论你是大型企业还是电商商家，你需要的训练、微调和集成可能并没有太大差异。

So maybe it could be customer support or something where, whether you are a large enterprise or an e commerce merchant, the amount of training and fine tuning integrations that you need might not be that disparate.

因此，他们是否会彻底解决到顶级应用层的最后一步呢？

And so maybe they'll provide will they solve all the way to the last mile that a top end application layer will?

也许不会，但我能看到它们不断向上攀升，去攻克下一个应用领域。

Perhaps not, but I could see them moving up and up in the stack to to to tackle the next application area.

好的。

Okay.

在你看来，那个应用领域会是什么呢？

And and what would that application area be in your view?

嗯。

Yeah.

嗯。

Yeah.

我认为客户服务可能是一个非常有趣的领域，

I I think customer support could be a very interesting one for

对他们来说。

them.

好的。

Okay.

是的。

Yeah.

好的。

Okay.

那你说的法律领域呢？你也提到了法律。

What about I mean, you you also talked about legal, you know, legal.

你觉得这个领域会在2026年爆发吗？

That that's an area that you think is gonna take off in '26?

是啊，法律领域其实已经起步了。

Yeah, mean, legal's already sort of taken off.

你看Harvey和LaGouette就已经在做了。

You see with Harvey and LaGouette.

说实话，很难想象它还能再扩大。

Know, it's hard to imagine it gets bigger.

我知道，我知道。

I know, I know.

不，我觉得它还会继续扩大。

No, I think it is gonna get bigger.

我的意思是，这些公司已经从最初作为律师的专属聊天GPT，发展到现在。

I mean, these companies have gone from being secure sort of chat GPT for lawyers, Yeah.

它们现在实际上在完成一部分工作。

They're now actually doing a portion of the work.

所以生产力从5%提升到了30%。

So it's 5% productivity to being 30% productivity.

因此，它们能够实现的收入模式和计费模式，以及这些企业的市场规模，在未来几年将爆炸式增长。

So the sort of revenue model and billing model that they're able to capture, and therefore the market size of these businesses is just going to explode in the coming years.

很好。

Great.

拉吉夫，非常感谢你来参加这个节目。

Well, Rajeev, I want to thank you for coming on the show.

这是一次很棒的对话，期待很快再和你交谈。

It's a great conversation, and look forward to talking to you again very soon.

谢谢，阿卡什。

Thanks, Akash.

保重。

Take care.

很快聊。

Talk soon.

好的。

Okay.

预测市场已经席卷全球，但我们的新闻团队就一些人可能利用这些新型博彩平台进行内幕交易的问题进行了有趣的报道。

Prediction markets have taken the world by storm, but our newsroom has done some interesting reporting around questions that have surfaced that some people might be insider trading using these new betting platforms.

我想请我的同事罗基特·德鲁上节目，他写了一篇关于此事的文章，谈谈他了解到的情况。

I want to bring on my colleague Rocket Drew, who wrote a story on that to talk about what he's hearing.

罗基特，欢迎再次做客我们的节目。

Rocket, welcome back to the show.

很高兴你来参加。

It's great to have you here.

谢谢，阿卡什。

Thanks, Akash.

很高兴来到这里。

Great to be here.

让我问你一个问题。

So let me ask you this.

我们有没有证据表明有人在这些平台上进行内幕交易？

Do we know that people are insider trading on these platforms?

这只是一个猜测吗？

Is it a hunch?

是来自Reddit论坛上的信息吗？

Is it stuff that is on Reddit forums?

你听到了什么？

What are you hearing?

是的。

Yeah.

是的。

Yeah.

你知道，这有点像一种直觉，因为没有人出来承认自己在进行内幕交易。

You know, it's a little bit of a hunch in the sense of I don't think anyone has come out and confessed that they're insider trading.

我们能看到的是，投注行为呈现出一种异常精准的模式。

What we can see is a pattern of betting activity that looks suspiciously good.

看起来有人提前知晓了某些事件或公司即将发布的公告。

It looks like someone seemed to have knowledge in advance of some event, some announcement that a company was going to make.

给我们举个例子。

Give us an example.

给我们举个例子。

Give us an example.

例如，有一个账户在一天内通过押注谷歌2025年搜索数据中最热门的搜索关键词，赚了约100万美元，而谷歌是在年底才公布这些数据的。

For example, there was an account that made about $1,000,000 betting over the course of a day on what the top searches would be in Google's twenty twenty five search data, which they announced at the end of the year.

这个人准确预测了23个问题中的22个。

And this person bet correctly on 22 out of the 23 questions.

这些问题是有很多可能答案且充满不确定性的。

Now these are questions with a lot of possible answers and a lot of uncertainty around them.

所以，这个业绩记录看起来好得不可思议。

So this was a track record that looked sort of too good to be true.

如果你只是凭运气下注，而普通人的知识水平有限，你几乎不可能表现得这么好。

Like if you were just betting by chance with the amount of knowledge that a normal person has, you'd be very unlikely to do this well.

这引起了一些警觉。

So that raised some red flags.

其他一些警示信号是，当人们一次性在市场中投入大量资金，然后往往非常成功。

Other sort of red flags are when people put a lot of money into a market all at once, and then they tend to be very successful.

他们总是在正式公告发布前立即投入资金，仿佛有人在公告发布前提前透露了消息。

They tend to put money in right before the actual announcement happens, as if someone had tipped them off right before the announcement went out.

对。

Right.

而且他们

And they

可以反复在同一家公司或一组公司上投注并取得巨大成功。

can be very successful betting on the same company or cluster of companies over and over again.

那么，监管机构或公司是否已经对此采取了任何措施，如果这确实属实的话？

So have regulators or has the company clamped down on any of this if in fact it's true?

是的。

Yeah.

因此，监管机构在这个问题上一直保持沉默，结果责任落到了公司身上，让它们自己去决定该如何应对。

So regulators have been somewhat silent on the topic, and as a result, the burden has fallen to the companies to figure out what to do about this.

就监管机构而言，这些市场不属于证券。

So in terms of the regulators, these markets are not securities.

因此，当我们听到内幕交易案件时，通常会想到的SEC（美国证券交易委员会）实际上并不负责这类案件。

So the SEC, which is often the one that we think of when we hear about cases of insider trading, is not really the one that's responsible in these cases.

负责监管Calci并即将监管U.

The CFTC, which does regulate Calci and soon will regulate a U.

S.

S.

实体（针对Polymarket）的CFTC（美国商品期货交易委员会）确实会开展一些关于内幕交易的工作。

Entity for polymarket, does do some amount of work on insider trading.

同样地，你可以想象司法部可能会介入某些案件。

And similarly, you could imagine the DOJ getting involved in certain cases.

但联邦机构要处理这类案件，需要一个让它们感到非常有底气的政府当局。

But for federal agencies to pursue cases like this, you would have to have an administration that was making them feel very emboldened.

换句话说，必须营造一种政府愿意创新、希望扩展法律并设立新先例的环境。

Is the way they're You'd putting it to have to have an environment where the government was feeling creative and feeling like it did want to broaden the law and set new precedents.

而过去一年我们并没有看到这种情况。

And we haven't really seen that in the past year.

有些人告诉我，他们认为在未来几年内不太可能看到这种情况。

Some people are telling me they think it's unlikely we'll see that in the next few years.

因此，这仍然是一个监管上的灰色地带。

So it remains kind of a regulatory gray area.

因此，责任落在了公司身上，即Calshi和Polymarket，需要在各自的平台上实施监控，防范内幕交易。

So it falls to the companies, meaning Calshi, Polymarket, to enforce surveillance, monitoring for insider trading on their platforms.

所以，如果我正确理解了你对这些预测市场的预测，那就是至少在近期内，这种情况不会有任何改变，那些异常成功的交易仍会继续存在，我想。

So if I'm reading your prediction correctly about these prediction markets, it's that nothing is really going to change, at least in the near future, about this, and there will just continue to be these suspiciously successful trades, I guess.

我认为我们会看到一些这种情况。

I think we'll see some of it.

特别是Kalshi，确实会主动寻找内部交易者。

Kalshi in particular does sort of seek out insider traders.

他们有自动化系统来监控交易者的行为模式。

They have automated systems to monitor the, you know, sort of pattern of behavior of its traders.

他们还依赖其他交易者揭发并指出这种可疑活动。

It also relies on other traders to of tattletale and point out this kind of suspicious activity.

然后他们每个季度会调查若干个账户。

And then they investigate a handful of accounts every quarter.

这是他们的合规人员告诉我的。

This is what their compliance people are telling me.

他们会与那些看起来可能是内部人士的交易者进行面谈。

Do have interviews with traders that seem like they might be insiders.

如果发现有人确实在进行内幕交易，他们可以暂时或永久冻结账户，甚至将案件转交监管机构，比如商品期货交易委员会（CFTC）。

And in the event that someone does seem to be insider trading, they can either suspend the account temporarily or permanently and even refer them to the authorities, who might be the CFTC in this case.

而且，这些公司有没有告诉你们，他们曾将任何人转交给当局，或者发现过什么情况？

And and did the companies tell you that they have referred anyone to the authorities or or or found say.

有的。

Yeah.

对。

Right.

嗯，我不清楚。

Well, I don't know.

我也不知道，这属于那种很难说事情会多快推进的故事，但确实值得引起我们的注意。

I I I this is one of the stories that it's just you know, it's hard to know how fast things are going to move, but it's certainly one that is important to bring to our attention.

所以这是一次

And so it was a

我也想说，有些人会为预测市场上的内幕交易辩护。

great I'll say also, some people defend insider trading on prediction markets.

我认为这使得它与传统证券市场的情况略有不同。

I think that's another thing that makes it a little different from the typical territory with securities in the stock market.

有些人说，预测市场的目的是

Some people say, Well, the purpose of prediction markets is to

让公众有所了解

give the public a sense

发生了什么。

of what's going on.

如果内部人士能让这些市场更准确地预测未来，那就再好不过了。

If insiders make those markets more accurate at predicting the future, so much the better.

这是一种公共服务。

That's a public service.

有些预测市场甚至鼓励内部人士参与，比如我在Manifold上创建一个预测市场，使用的是虚拟货币，而不是真钱。

And some prediction markets go so far to encourage insiders that, I can make a prediction market, say, on Manifold, which uses play money, not real money.

但我可以创建一个关于我今晚吃什么的预测市场。

But I can make a prediction market about what I'm gonna have for dinner tonight.

显然，我是这个市场的内部人士。

Obviously, I'm an insider in that market.