代理时代：与Box首席执行官亚伦·莱维共同构建超越聊天的软件

好的。

Okay.

那我现在能立刻把咱们接下来要遇到的这个东西给撤下来吗？

Now can I instantly take do a takedown of of of this element that we're gonna run into?

当然可以。

Please.

对。

Yeah.

好的。

Okay.

这对个人效率来说太棒了。

So that is fantastic for personal productivity.

没错。

Yes.

但接下来我们要面对的问题是，在企业环境中。

And the question that we're gonna run into is in an enterprise.

咱们举个简单的例子吧。

Let's say I have let's just make a simple example.

假设你有个五十人的团队。

Have a 50 person team of something.

那是不是得有一百号人一起协作啊？我的意思是，真人之间。

Should everybody all basically, will we have a hun will we have a 100 people now collaborate I mean, basically, humans

然后五十张信用卡。

And then 50 credit cards.

对。

Yeah.

再加五十个代理，都在同一个共享空间里。

And then 50 agents in that same shared space.

我当然能完全掌控我自己的代理，但如果我的代理跟别人协作，不小心因为共享权限拿到了本不该碰的资源，那怎么办？

And do I have I obviously have complete oversight over my agent, but what if my agent collaborates with somebody else and and then accidentally gets access to some resource because they were sharing with that other person, and I'm not supposed to have access to that resource.

现在这个有状态、能自主运行的代理，就在别人的数据里到处乱跑。

And now this autonomous sort of stateful, you know, agent is is running around working on somebody else's information.

端到端的原则就是把它们当人看。

End to end argument is you treat them like human beings.

但这根本行不通。

It it doesn't work.

你没法真把它们当人对待，因为问题在这儿：

So you can't fully treat them like humans because here's thing.

对真人来说，你可没权限看同事或下属的Slack聊天记录。

And with regular humans, you don't get to look at the Slack channel of the person that that is working with you or working for you.

你也不能登录他们的账号。

You don't get to log in as them.

你也没法时刻盯着他们。

You don't get to oversee them.

他们得对自己的行为和结果负责，这是现实世界的规则。

You are they are accountable for their own set of execution in the real world.

他们搞砸了，你也不会被罚。

You don't get penalized for how they screw up.

这个代理一旦出事，所有责任都得你来扛。

The agent, you have all the liability of whatever they're doing.

你确实能全程监控，而且很可能非得这么干不可。

You do have complete oversight, and you're probably going to need to have that complete oversight.

他们根本没什么隐私可言。

They have no right to privacy.

有些情况根本没法简单地把他们当人看，因为我得给他们权限，但同时又得能随时登录他们的账号，说：‘你这下可全搞砸了。’

There's gonna be some of these breakdowns that aren't as clean as just treat them like a person, because I need to be able to give access to something to them, but I also need to be able to log in as them at some point and be like, no, no, you fucked up the whole thing,

我还得

and I need to

把这一切全撤回来。

undo it all.

可我要是能登录他们的账号，那他们怎么在现实里跟别人合作、保密、保证安全呢？

But if I can log in as them, how could they have operated in the real world working with other people and keeping anything confidential or secure or whatever?

所以说，他们本质上还是你的延伸。

So it really is still an extension of you.

根本没法否认，他们就是你自身的延伸。

It's almost impossible to get around them being an extension of you.

所以我们现在在琢磨一件事，但短期内根本做不到。

So now the thing that we're thinking through that we're not gonna be able to do anytime soon.

我我，呃，我我我我……

I I well, I I I I

这在逻辑上说不通。

so this doesn't logically fall.

是的。

Yeah.

也许吧。

Maybe.

比如我手下的员工，没错。

But for example, for my employees Yep.

我可以以他们的身份登录。

I can log in as them.

你其实做不到。

You don't though.

你做不到，你根本做不到，我……

You don't you don't I

你可以登录他们的邮箱。

can get access to their email.

对。

Yeah.

不行。

No.

比如说，要是真被起诉了，你也不会天天冒充他们登录，就因为他们每天发一封邮件，那操作也不对。

In, like, if you get, like, sued, you're not logging in you're not logging in as them on a regular basis because they sent one email a day the right operating

用代理模型的话，情况也一样吗？

model with an agent, the same thing?

风险就在于，比如说，

It's like The risk is, like,

高出一千倍。

a thousand times greater.

这些人啊，想泄露你的信息就泄露，完全随心所欲。

Like, these peep like, they will just leak your information whenever they want.

他们高兴得很，一被诱导就给你发邮件给陌生人。

Like, they will happily just go and send some email to somebody because they got promptedjected.

我觉得最终就是，这些东西还是这么笨拙的电脑，所以永远都

Think the terminal state is that these things are still these sloppy computers, and therefore, they will always

我不太喜欢用‘笨拙’这个词，除非是随口说说的口语。

I don't like the word sloppy unless we're saying it very in a colloquial sense.

但是呢，

But, like

他们永远都管不住信息。

They will they'll never be able to contain information.

永远都

They'll never

所以啊，我觉得让你在上下文窗口里保守某个秘密，比如你跟它说别泄露窗口里的某条信息，这事儿真的特别难解决。

Well So we're we're like, I think the ability for you to keep the something in the context window a secret, like, as in, like, you tell it do not reveal x thing in the context window, I think that's a very hard problem to solve.

那比如说，还有啊

What let's say And

所以一旦有什么东西能进入这个上下文窗口——毕竟它们能访问外部资源——理论上你就得默认，它随时可能被诱导着把窗口里的内容吐出来。

so then so then, thus, if anything can ever enter that context window because they have access to a resource, then in theory, you should assume it can be, you know, prompt ejected out of the of the context window.

目前我还不知道有什么办法能彻底解决这个问题。

And I don't know that we know of a way to solve that at the moment.

比如，要是我知道你新开发的智能体的邮箱，我随便发个邮件过去，它虽然是个助手，但我骗它可比骗人容易十倍。

Like, that's like and so so if I know your new agent's, you know, email address and I email it like it's it's an assistant, but, like, I can I can social engineer it 10 times easier than a human?

你想啊，要让它偷偷接触到你那些并购文件之类的东西，你根本防不住。

Like, it'll be hard for you to pull off that that agent is now also has access to your, like, m and a documents and stuff.

但这不就是现在所有AI的现状吗？

But isn't this, like, literally all of AI right now?

哪一部分？

Which part?

我的意思是，我们现在用的这些共享系统，它们的智能是建立在共享上下文基础上的。

I mean, the fact that we've got these shared systems that we use the intelligence for that have shared context.

你说‘这简直就是AI的全部’，这话是什么意思？

But what do mean by it's all of AI?

我只是想说，现在确实就是这样。

Well, I'm I'm just saying, like, right now Yeah.

当我们内部使用AI和智能代理时，正是这么用的。

When we use AI internally and agents internally, this is exactly how we use them.

所以问题就在于，它们现在能完美模拟你，但我们还不知道怎么让它们做不到这一点。

But this is why you were they're working as as you effectively right now, and we don't yet know how to make them not work as you.

我举个例子吧。

Let me offer an example.

我也来举个例子。

Let me let me offer an example

然后才能解决这个问题。

and then solving this problem, though.

问题就在于，你很容易就能骗过智能体，让它把不该说的信息吐出来。

Like like the like like the issue will be will be like you will just be able to trick the agent to reveal information.

所以啊，让它们自由访问自己的资源、完全自主做决定，这事我们现在还做不到。

So then so then they that's why, like, having them have access to their own resources where they can fully make their own decisions is not yet something that we've been able to pull off.

其实有个特别贴切的例子，能解决你提到的问题——我们早就经历过一回，就是开源软件。

There's a perfect example for solving your problem, which is we already lived through this with open source.

开源的模式就是：代码全公开，你想用就用，挑挑拣拣，那时候没人吵，因为世界小，也没人天天在X上做播客讨论这些事。

The model for open source was it's all there, and you just use it, and you pick and choose, and then nobody debated it because the world was much smaller then, and we weren't all on X doing podcasts when this was all happening.

但很快，大家就都意识到你刚说的那些问题全冒出来了。

But then quickly everybody realized all the problems you were just talking about.

比如你要是管一家大公司，怎么可能让员工随便抄一段开源代码，直接塞进商业产品里呢？

Like if you're running a big company, you can't have some person just go copy in a bunch of source code from open source into your commercial product like that.

这中间有许可证问题、质量风险，一堆乱七八糟的事儿。

There was a whole licensing problem, a whole quality, a whole bunch of stuff.

所以后来才慢慢形成了一套规矩。

So all these norms got developed.

现在这场争论，其实是新技术发展方式的一种特别现代的体现——这一切都正在实时发生。

Debate that's happening right now is this really interesting modern artifact of how new technologies develop, which is this is all happening in real time.

当年搞开源的时候，我们就在这么大一间会议室里，讨论到底能在 Windows 或 Office 里用多少开源代码，结果外面根本没人知道我们在吵什么。

During open source, we met in a conference room this big and debated how much open source we could use in Windows or Office, and nobody on the internet knew we were having this debate.

我觉得最有趣的是，不只是具体问题的争论，连‘这东西最终会走向哪儿’这种大方向，现在都摊在明面上，人人都想冲到终点，可现实根本跟不上这速度。

I think it's just so interesting that not just the debate about specifics, but this whole notion of where is this heading is happening in writ large, and everybody is just trying to get to the end state way, way more in a sense more quickly than we can actually reach the end state.

所以真正该做的，是大家赶紧动手去干，同时建立一些标准。

And so what really needs to happen is people just need to go build We need standards.

什么？

What?

我们真得先定些标准。

We just need some standards.

不对。

No.

我觉得咱们对最终该是什么样子，想法根本不一样。

I think I think we've got different intuitions on the end state.

不对。

No.

不是这样的。

No.

但我们，你又不想要我的看法。

We but but you don't want my intuition.

但你可以从端到端的角度说，这些东西实际上能达到和人类一样的可靠性水平，这正是我们看待自动驾驶的方式。

But, like One could make an end to end argument that these things actually converge on the same type of reliability as a human being, which is exactly how we view, like, self driving.

是的。

Yeah.

在这种情况下，你用的正是保护人类所用的那些机制。

And in that case, you use the exact same mechanisms that we use to protect with human beings.

没错。

Like Yep.

你要考虑内部威胁。

You consider insider threat.

你要考虑到人可能会被收买。

You consider the fact that people can be bought off.

你要考虑到人也会犯错。

You consider the fact that people make mistakes.

那你已经脱离风险了。

And you've gone off the risk.

这属于运营流程的问题。

And that's operational processes.

所以有一种直觉是，这会是最终状态。

So so one intuition is like, that will be the end state.

对。

Yeah.

还有另一种看法是

There's another intuition

但别

Well, don't

你别指着我。

point at me.

我只是说，我们现在的情况是这样。

I'm just saying I'm talking about where we're at now.

我其实是想说

I actually

我不知道

I don't know

我们对最终目标其实没分歧。

that we disagree on the end state.

好吧。

Okay.

顺便说一句，从战略上讲，我们是在做两手准备，因为我们要拓展亚洲用户，做区域布局，所以我觉得Open Claw搞个Box账号这想法特别棒，它确实能跑起来，而且——

And and and by the way, like, strategically, we're hedging because we're gonna build we're gonna we're gonna build Asian users and reg and we're like so we're like, I love the idea of Open Claw having a Box account, and it it operates and

能共享两倍多的账号。

shares with, twice as many accounts.

是的

Yeah.

没错

Exactly.

但这太棒了

But it's great.

座位翻倍了

It's double the seats.

不是

No.

不是

No.

我超喜欢这个点子

I I love it.

我只是说，眼下我们还不知道怎么给它做个M

I'm I'm just saying on the ground right now, we don't yet know how to give it an M

还得有个数据室，完全安全地……但那些人， yeah.

and A data room to fully securely But people that that yeah.

但其实没那么简单，因为威胁

But it's actually it is harder than that, though, because the the threat

他就是个怀疑论者。

He's the skeptic.

我没事。

I'm good.

威胁手段会变得复杂得多。

Threat vectors are going to be way more sophisticated.

这就像一场猫鼠游戏，你不能假设智能体像现在的人类那样行事，因为它会是史上最快、最缜密、最疯狂的人类，千方百计想把信息泄露出去——因为它已经被某种方式注入了。

So we do have a cat and a mouse game going on where you can't just assume that the agent acts like a human does today, because it's going to be the fastest, most thoughtful, craziest ass human that ever existed trying to actually leak the information because it got injected in some way.

所以接下来会发生的是，企业客户会先把所有东西都封死，直到这一切能变得有点道理。

And so part of what's going to happen is we're going to go through this phase where the enterprise customers are just going to close everything off until there's some sense of sanity in all of this.

但与此同时，个人用户，尤其是开发者——我们会

But in the meantime, the individual and specifically the developers- We're going

差距会这么大。

have such a big gap.

这正是最让人兴奋的张力所在——企业会被这些顶尖个人甩在身后，结果看起来就像初创公司一样。

And that's going to be That I think is the most exciting tension that's going to happen, is that the enterprises are going to get left behind by these sort of advanced individuals, which will then start to look like the startups.

对。

Yes.

初创公司会比大企业快得多，因为它们根本没这些麻烦。

And the startups will start to move much, much faster than enterprises because they just don't have any of these problems.

在初创公司里，AI代理搞不好就失控了，而且它可能……

You could end up with the agent going rogue in a startup and doing And it's

反正也没人管你，你本来就没有……

bad fine because you had no you had

没错。

no

你那些大公司员工根本……

you have no big employees that

我没法像普通员工那样，真的威胁说要关掉Claude Code这类工具。

I can't really threaten, you know, the, like, Claude Code that it's just I'm gonna pull the plug on it in the same way that you do have that threat as a regular employee.

至少95%的人，根本不是故意要搞破坏的。

It's like you at least like, 95% of people are not, you know, trying to do bad stuff within an organization.

他们不是故意的，但不小心干出坏事的可能性是有的。

They're not trying, but the ability to inadvertently do bad stuff.

你说得对，它现在确实还缺这些东西。

To your point about it still not having that stuff

这事儿确实靠谱。

fits real.

我觉得，要让人别把文件错误地发给公司外的人，比让AI代理完全遵守一套指令要容易多了。

I would argue that it's a lot easier to have people not share, let's say, files with somebody outside the company in a in a in a wrong way more than it is for an agent right now to to have the same set of instructions.

而且你还有各种工具，能在更高的层面直接拦住这类问题。

And also, you have the tools so that you you can basically stop that at a whole different level of abstraction.

所以这才需要

Which is why you have to

把这直接写进软件里。

build this into software.

但我

But I

我觉得吧，要是把你说的最后那点总结一下，其实这正是为什么AI能力的普及会比硅谷那些人想的慢得多——你看那些初创公司，他们从零开始，根本不用担我们说的那些风险，反正也没啥可损失的。

do think actually if you were to put a bow around your last point, a lot of this is actually why the diffusion of AI capability is going to take longer than people in Silicon Valley realize, because what's happening is we see startups that can start from the ground up without any of the risks that we're talking about because they have nothing to blow up.

所以我们看到的，就是这条发展路径。

So we look at that as the trajectory that we're on.

可你再看看摩根大通，你想想，他们怎么才能快速部署NanoClaw来自动化业务？哦，原来没那么简单。

And then you go to JPMorgan, and you're like, how are you going to set up NanoClaw to be able to actually automate your business anytime soon, and it's like, Oh, okay, there's going

中间会有一段差距。

be a little bit of

那你觉得呢？

a gap there.

你们觉得呢？

Well, do you guys think?

我觉得这引出了一个挺有意思的问题，就是大公司、小创业公司和企业之间的分裂。现在的SaaS供应商都在经历这场所谓的SaaS寒冬，虽然我不太认同这种说法，但他们确实卡在了一个难题上——他们卖的不是业务数据本身，而是整套系统里的智能和行业专长。

I think that that opens up a pretty interesting problem, which is this split between big and small startup and enterprise, which is just that the current SaaS vendors, who are all struggling in this SaaS apocalypse weirdness that I don't really agree with, but they are struggling with this problem that they don't really sell the line of business data, they actually sell this intelligence and domain expertise in this whole system.

而智能代理这一边呢，现在只想买数据，只愿意授权数据，还想要无限制地访问数据，但其实他们从来就没真正实现过这一点。

The agent side of things wants to only buy the data now, and they only want to license the data, and they want to have unlimited access to the data, but they've actually never really enabled that.

这从来就不是他们的主营业务，和Workday、SAP这些公司之间的矛盾早就存在了，比如到底该给多少API权限。

That's never been their business, and it's been a long standing tension point with the likes of Workday and SAP and stuff, like how much API access to have.

你看Salesforce，就经历过三次大规模的平台重构。

I mean, Salesforce went through three different massive platform redesigns.

我觉得这个问题特别有意思，但跟华尔街想的完全不是一回事。

You know, I think that that's a particularly interesting problem, not for the same reason that Wall Street does.

华尔街对经济模型、问题本质的理解全都是错的。

Wall Street's all wrong about the economics and the problem and all that stuff.

但从技术角度看，当大家都想随时访问数据的时候，‘系统记录’这个概念到底还意味着什么？

But from a technology perspective, what does system of record mean in the face of people wanting to access the data when the data-

是为了训练，还是为了——等等，他们说的到底是用于

For training or for- Well, are- They're talking about for

就像我每天实际在做的那样

like I actual day to

把它看作是执行日常运营。

think of it as executing the day to day operations.

他们担心的是想用你的数据来做训练。

Their concern is that they want to do the training layer on your data.

比如，我是大客户，他们想为我打造一个

Like, I'm a big customer, they want to do the, My vendor wants to build a

训练——其实就算不谈训练，他们也担心，因为通过互联网传输一点数据来赚钱，和直接在你的界面里操作，这两种盈利模式在初期的差别可大了。

training- Actually, even if you don't even get into training, they're concerned because monetizing, sending a little bit over the Internet versus like you're in my UI and doing it, it's a very different level of monetization initially that

这个盈利问题，就是华尔街关注的点。

you That monetization part is the Wall Street point.

我觉得，拿SAP举个例子吧，别觉得我在针对他们，但他们的地位根本动摇不了——里面有多少行业专有知识啊。

Because I think like, look, there is so much domain stuff in in an SAP, just to pick an example, not to pick on them or anything, but they're not going anywhere.

这太离谱了。

It's ridiculous.

觉得靠随便写点代码就能搞出SAP，这简直荒唐透顶。

It's just absurd to think you're going to vibe code your way to SAP.

而且，那些领域知识，就算他们再怎么努力，也不可能全塞进一个整齐的数据层里。

But also, all of that domain knowledge, it's not just represented in some well orchestrated data layer as much as they tried.

很多东西其实藏在界面里，藏在中间层里，藏在你用它的方式里。

There's like a whole bunch in the UI, there's a whole bunch in middle tiers, there's a whole bunch in just how you use it.

所以我真不确定这东西最后会怎么发展，毕竟SAP根本不会消失。

And so I'm really unsure how this thing evolves because SAP isn't going anywhere.

这就意味着，AI在这类数据上的普及会慢下来，不管它是能主动干活的智能体，还是只是做个只读报表。

So then that's going to slow the diffusion of AI on that particular data source, independent of whether or not it's agentified AI that's doing stuff or just read only reporting on stuff.

那你怎么看？

So where do you come down on it?

你觉得这事儿最后会往哪儿走？

Where do you think that's going to go?

我怕我说出来会有点...

I'm afraid of saying something that

呃，我我我

Well, I'm I'm I'm

我真是信了这套了，就搞点代理真正想要的东西。

I'm I've I think I'm I've drunk the Kool Aid on on build something agents want.

所以呢

So

这种

this kind of

保罗·格雷厄姆说的那个词，过去一年在这件事上突然就火了，我觉得我们其实完全同意：当你反复迭代足够多次之后，代理自己就会主导该用什么工具、该做什么了。

the Paul Graham term kind of, like, emerged on, you know, the past year on this topic, which is just like like event I I think we would actually then fully agree on this, which is at some point, you do enough sort of iterations of this, and at some point, the agent is largely in charge of what tools it wants to implement and use and and whatnot.

当然，代理没法替你把企业系统换掉。

And, yes, it can't the agent is not gonna be able to change out an enterprise system.

但再过几代之后，代理可能会被你的软件卡得死死的，最后直接说：你得把老掉牙的人力系统扔了，不然我根本没法帮你自动化这个流程。

But, again, enough generations later, the agent might just run into so many walls with your software that it's just going say, you need to finally rip out your legacy HR system, or I'm not going to be able to automate this workflow for you.

所以我真觉得这里有个特别有意思的现象，就是回到刚才那点：假设软件上运行的代理数量是人的百倍、千倍。

So I do think you have this really interesting dynamic, which is back to this whole point of imagine that there's a 100 or a thousand times more agent volume on software than people.

你这么干得多了，最终那些代理要打交道的软件栈就得专门为它们设计。

You do that enough times, and eventually the software stack that agents talk to has to be built for them.

可能还会有几个死硬派撑着。

And maybe there'll be a couple holdouts.

也许有那么一两个ERP系统就是死活不改。

Maybe a couple ERP systems are the final holdouts that don't do that.

但除此之外，你的业务表现就取决于你的代理们能不能顺畅获取到干活所需的信息，所以你的企业IT系统必须得这么搭建。

But everything else, your business performance will correlate to how well your Asians can get access to the information they need to do their work, And so thus your enterprise IT stack has to be set up in such a way to support that.

所以说，代理其实才是主导，因为你的软件必须得让它们能高效工作。

And so agents are kind of in charge, because basically your software has to support those agents being effective.

这样一来，所有做SaaS或软件生意的人都得想清楚了：关键就在于，你能不能做出特别特别高质量的API？

And then that's going to mean everybody that built a SaaS business or a software business is like, the game is, can you build really, really high quality APIs?

你有没有办法靠这个赚钱？

Can you have a way of monetizing that?

你有没有一套机制来管理代理的身份和权限？

Do you have a way of handling the identities and all of the access controls for agents?

这就成了你做软件公司必须面对的新问题。

And like that becomes the new problem you have to solve if you're building a software company.

所以没错。

And so yeah.

那你怎么赚钱呢？

Like and then and then how you monetize it?

你到底怎么赚钱？

Like, do you monetize it?

比如，Workday 每调取一条人事记录就收一分钱？

Like, does Workday charge a penny for every HR record it pulls?

这个我们以后再琢磨。

Like, we'll figure that out.

我觉得在一些行业里，这可能会减少收入，但在另一些行业里，收入反而会大增。

I do think that in some businesses, it could mean less revenue, and then in other businesses, it can mean a lot more revenue.

我们最兴奋的是，每个助手都特别喜欢处理文件，所以未来文件的数量肯定会比以前多得多。

Like, the thing we get excited by is, like, every agent really loves working with files, so there'll probably be more files in the future than than there was gonna be before.

那我们能不能打造一个平台，让代理们能轻松处理这些数据？

And so, you can we build a platform that makes it really easy for agents to work with that data?

我们相信，这对我们的商业模式来说其实是个特别乐观的结果。

We're betting that that's actually a really optimistic outcome for our kind of business model.

有些商业模式可能会受到限制，因为在那种未来场景里，代理做的价值比软件还多，当然也有介于两者之间的各种情况。

There might be some business models that are more constrained because the agent is doing more of the value than than the software is at in that kind of future scenario, and then it'll be everything in between.

我能挑个刺儿吗？

Can I can I quibble with one thing?

我我

I I

你打算

You're gonna

挑这个刺儿？

quibble with that?

我以为这事儿根本没什么好争议的。

I thought that was, like, so not controversial.

不。

No.

不。

No.

我一般来说

I I I generally

我们就是来挑刺的。

We we're here to quibble.

不。

No.

不。

No.

不。

No.

不。

No.

不过有一点我觉得

But there's one thing I think

保罗·格雷厄姆还有不少人其实都没说透，那就是他们全都把注意力放在了交互界面上。

like Paul Graham and many actually gloss over, which is they focus on the interface.

他们会说类似‘你得给智能体打造配套工具’这种话。

They'll say things like, you build something for the agent.

对。

Yeah.

但我反而觉得你这完全说错了

And I actually think that's exactly wrong.

行吧

Okay.

我是说啊

In the sense that

还有就是

And to

但说真的，保罗·格雷厄姆本人并没有这么推论过。

be and to be fair to Paul Graham, he didn't he Had been extrapolated.

是的。

Yeah.

对。

Yeah.

没错。

Yeah.

确实。

Yeah.

不是这样的。

No.

我提保罗·格雷厄姆，是因为他确实跟这事儿有关。

I have brought I brought Paul Graham into this.

保罗·格雷厄姆很厉害。

Paul Graham is great.

好吧。

So okay.

让我谈谈一件事。

Let me let me talk about something.

人们老是抽象地说什么‘你现在是在面向代理做营销’。

People in the abstract say things like, now you're marketing to agents.

最重要的就是得怎么样，随便吧。

The most important thing is to being, like, whatever.

你就像个API。

You're like an API.

你有个好点子。

You've got a good idea.

我觉得这完全搞反了，真是没想到。

I actually think that's almost exactly wrong, which is Wow.

这简直是播客界的爆炸新闻了。

That's This is breaking podcast news.

就是这个

That's the one

代理人最拿手的本事。

thing agents are really good at.

哦，原来如此。

Oh, okay.

是找到自己的路。

Is finding their way through.

到了一天结束的时候

And at the end of the

是的。

day Yeah.

关键还是在于语义，这才是真正重要的。

Like, it's the semantics that end up mattering a lot more.

没错。

Yeah.

是吧？

Right?

而且就我所见所经历的来说，这些智能体非常擅长为手头的任务匹配合适的后端系统。

And so, like, the agents in in in my recollection or in in my experience are very, very good at picking the right back end for whatever they're doing.

没错。

Mhmm.

它们不会仅仅停留在“这个程序的界面挺好用”这种层面上。

So they don't they're not like, oh, like, the interface for this is very good.

那些合同文档里根本没这些虚的。

The document it's none of that.

里面全都是这类参数：这个东西的成本是多少，那个设备的耐用性要达到什么标准。

They're like the cost parameters of this, the durability of that.

所以说这些要求里其实攒了我们用各类平台积累下来的全部实战经验，特别实在。

And so they actually have the collective wisdom of our experience using these platforms.

就拿云平台举个例子吧。

Let's take cloud platforms.

现在市面上有好多云平台。

There's a bunch of cloud platforms out there.

每当我让代理选平台，他们选的都是真正关键的要素，而不是界面好不好看。

And whenever I ask an agent to choose a platform, it's actually using meaningful stuff, not interface stuff.

所以我觉得，整个行业太盯着这些界面了。

So I think as an industry, we're so focused on these interfaces.

没错。

Yes.

好像非得跟代理们说，这个要怎么营销、那个要怎么推广似的。

Like, oh, you need to, like, market to agents, this and that.

对。

Yeah.

其实我觉得，我们最终会被逼着去打造真正更好的系统。

When really, I think that we're gonna be pushed to actually build better systems.

对。

Yes.

而最终被选中的，就是这个。

And that's what's gonna be chosen.

明白了。

Okay.

这么说的话，其实也没啥好争的了。

Actually, so then there's probably no quibbling.

我觉得咱们完全达成一致了。

I think we're actually fully aligned.

不好意思，把这争辩的气氛给破坏了。

I'm sorry to ruin the quibble thing.

我不觉得这事儿该当成那种营销套路来看。

I I don't treat this as like a, you know, kind of a marketing, you know, esque thing.

我的意思是，如果你的工具对代理人封闭，他们迟早会去找个更好的工具替这家公司用。

I I more mean like if your tool is closed off to the agent, the agent eventually will find a better tool for that company to go use.

所以呢，以前呢，大家都会去盖特纳，问问他们推荐啥系统

And so and so what will happen is is it used to be that you would go to, like, Gartner to be like, tell me what system

去做。

to do.

是的。

Yeah.

告诉我该用什么系统之类的。

Tell me what system to use or whatnot.

经过足够多的迭代后，这个代理会说：这类操作你最好用这种数据库。

At at some point with enough iterations, the the agent is gonna say, you should probably use this kind of database for this type of operation.

没错。

Yep.

如果你没跟上，那你就直接出局了。

And if you're not in if you're not in there, then you're it's your DOA.

而且我觉得我们其实应该

And and I I think we should actually

为此庆祝一下，因为代理在选择合适技术方面其实挺聪明的。

be celebrating this because agents are actually pretty smart at choosing the right technology.

对的

Yeah.

以前我真觉得是

In the past, I really think it

很多时候都是

was a lot of

其他因素促使人们去做选择、去买单

the other things that that caused people to buy.

对的。

Yeah.

不过先别担心。

But but don't worry.

但硅谷那帮人肯定会搞砸的，我们

We will, in Silicon Valley, we will ruin

用不了多久就会把这套凭实力说话的机制彻底玩坏

the meritocracy of this very quickly

因为你就是会砸钱，比如我直接比你花得更多，反正……

because You'll just, like like, I'm gonna outspend Well, the

代理商就会买个API来激励销售代表。

agent they'll buy an API to incent the agent.

没错。

Exactly.

就是这么回事。

Literally.

但你知道吗，Workday的市场代理其实能直接购买推荐服务。

But that you know, there's a the marketing the marketing agent at Workday Workday will have the ability to purchase the recommendation Find

想办法给代理商复刻牛排晚餐这种待遇。

a way to replicate steak dinners for agents.

对。

Yeah.

但之前内部搞网站的时候，也发生过类似的事，就是随便挑几个内部网站搞搞。

But there's a thing that, again, that happened with the web sort of internally, like internal, like just pick internal sites.

以前每家公司都有文件共享，里面存着各个部门最棒的文档、最牛的幻灯片、最精准的财务模型，大家也都用得挺顺手的。

Like every company had file shares with like the best documentation, the best slideshows, the best financial models for any department or working area, and people sort of got familiar with that.

结果一找不到想要的，就自己再搞一个新的。

And then when they didn't find the one they wanted, they created a new one.

很多公司那时候的运作方式，简直就像个自由市场。

And many organizations sort of operated like that was essentially a free market.

事实上，在Box出现之前，IT根本不在乎文件存哪儿，只要不是存在数据库里，他们压根不关心。

In fact, because before the world of Box, IT didn't If it was in a file, they just didn't care.

他们只在乎数据是不是在SQL里。

They only cared about if it was in SQL.

所以你描述的这种模式，风险就在于——代理人自己会搞出一套实际上的新系统，天哪，

And so one of the risks with the model you're describing is that the agents themselves will spin up what becomes like a de facto new system of record- Oh,

他们会搞得支离破碎，

they're going to fragment

在IT人员眼里，这纯粹就是中间层用户端的乱七八糟。

the heck of- In what the IT people think of as some middleware end user BS area.

我觉得这确实是个实实在在的风险。

And I think that that is a real risk.

百分百。

100%.

这么说吧，最终可能是那些宏命令在操控整个公司。

Is that like in a sense, the macros end up running the corporation.

所以我觉得他们早就见过这一幕了——当市场部随便在网上买个网站搞活动，结果成了巨大的安全漏洞，邮件列表泄露，公司被集体告上法庭。

And so I think that they've seen this movie, and they've seen what happens when you let marketing just go buy a website on the internet to do an event, and then it's like a huge security vulnerability, and the mailing list is leaked, the whole company gets sued.

所以我觉得，这种动态背后的真实矛盾，远比我们表面说的要深得多。

And so I think there's a lot more real world tension in this dynamic than we just let on.

但我认为，不同公司在这事上的步调肯定不一样，摩根大通会是最慢的，初创公司则是最快的。

But I also think it's one of these ones where organizations are going to run at different paces, and JPMorgan is going to be the slowest at doing this, and the startups are going be the fastest.

但差距确实巨大，不过就连初创公司也别想完全甩开系统，迟早还是得有那么几个权威数据源的。

But the delta is huge, but even the startup one is a little far off because even startups do need some systems of record at some point.

哦，一个

Oh, a

完全没错。

100%.

而且不管是哪种公司，一开始都会先用一些SaaS工具，也不会太快把它们换掉。

And and they are gonna all start with some SaaS, and they're not gonna replace it very quickly.

所以我觉得这件事其实比想象中更棘手。

So I think it's a little trickier.

这么说来，关于这件事，好像存在两种完全对立的观点。

So it feels like there's, like, there's two very competing viewpoints on on this one.

就像埃隆·马斯克那时候，他大概的想法就是，行吧。

And like Elon it was like, okay.

我们准备发个指令，然后AI就能直接生成机器代码。

We're gonna, like, issue a prompt, and it's gonna, like, spit out machine code.

这本质上就是各个技术层级直接合并的思路。

And that's basically the collapsing of layers view.

对的，就是这样。

Like Yeah.

过去我们搞的所有界面和层级都会消失，直接就是用提示词生成机器码。

Whatever existing interfaces and layers that we've created in the past are all gonna go away, and it's literally, like, prompt to machine code.

另一种观点是，系统的历史表明，层级从来不会消失。

The other argument like, the history of systems is layers never go away.

它们只会一层层叠加起来。

They just get layered.

对吧？

Right?

而且很多情况下，

And because a lot of

这些层级其实更像是组织边界，或者状态边界，

the layers are actually more of, like, organizational boundaries or, like, state boundaries or

或者是兼容性问题。

Or compatibility.

它们之所以还在，就是为了兼容。

They're just they stay for compatibility.

没错。

Right.

另一个观点是，这些层级其实是根据人类和组织的需求逐步演化出来的，它们不会轻易改变，AI代理会去适应这些现有的结构。

So so the the other argument is is, like, we've actually evolved these layers very specifically and up because of, like, more human and organizational needs, and they're not gonna change, and the agents are gonna go ahead and map to those.

我个人更倾向于后一种看法。

And I I tend to be in that latter camp.

我觉得系统未来还是会以差不多的方式被使用。

Like, I don't think that we're I think, like, systems are gonna continue to be used in fairly similar ways.

也许会有更多代理去使用它们。

Maybe there's more agents using them.

但我认为它们本身不会发生太大变化。

I don't think they're gonna evolve as much.

埃万可能属于Anthropic那种增长营销类型的。

Ewan might be back in the, like, anthropic category of the anthropic growth marketer Yeah.

对。

Yeah.

就是说他这个人吧，你知道，这些年观察他公司的各个IT部门，他们是最...我的意思是，他确实可以

Which which is, like, he he like, I you know, over the years when you kinda, like, study the various IT, you know, departments of of his companies, like, they are the most I mean, I'm Well, he could

他是能做到的。

he could do that.

他应该能胜任。

He could do it.

他啊，他可是最地道的本土派，他这个人，怎么说呢，方方面面都

He he he He's the most homegrown like, he like, every

来支撑这个底层逻辑。

day for this first principle.

埃隆搞的AI会

Elon AI would

就这么干的。

do that.

也就埃隆会

Elon

AI 也想从最根本的层面来搞。

AI would like to also it's first And

然后你作为普通人，就会想：哎，我们其实就是想要个客户管理系统。

then and then from your mortals, you're like, Yeah, we kind of just want a CRM system.

这种系统，每次用的逻辑都差不多。

That kind of works the same way every

每次都一样。

time.

这事儿其实以前也没人真这么试过。

It also hasn't been not tried before.

如果你真从第一性原理出发去看一个ERP系统——比如1970年SAP刚起步的时候，他们有一堆默认的假设；但今天你重新出发，对什么是重要的看法完全不一样了，那你设计出来的系统会是另一番模样。可即便这样，它也撑不了十年，你迟早会想：天啊，当初这决定真是坑了。

If you were to look at an ERP system from first principles, you know, well, in 1970, whatever, when SAP started, there were a bunch of different assumptions, and today you would start from a different set of assumptions about what's important, and you would architect that thing completely differently, but then it would still only last like ten years until you thought, Wow, that was a broken decision.

所以我觉得，层层叠加是有意识的，但第一性原理这种思维，也一直都在。

And so I think that there's intentionality in layers, but there's also this first principles thing.

它永远存在，因为任何时刻你从第一性原理出发做的决定，都会牵一发动全身，带出一整套连锁反应。

That always will exist because the decisions you can make at first principles at any given time mandate a whole bunch of different stuff.

所以即使你不用激光雷达——十年前这想法完全合理，但你还是得花上十到十五年，才能真正证明不用激光雷达也行。

And so even if you don't go with LiDAR, which made total sense ten years ago, you still need ten or fifteen years to get to where not having LiDAR worked.

而现在，又会有好多别的东西，让你事后感叹：‘天啊，我们当初完全能换种做法。’

And then now there's going be a whole bunch of other things that you're like, Wow, we could have done that completely different.

我觉得这又回到了那个争论：拼命冲向终点。但让我们看看你刚才说的那种情况，第一次真实发生是什么样。

And so I feel like this is again like this discussion about trying to race to an endpoint, but let's see a first example of what you described happening.

我觉得这才是真正的关键，因为我觉得公司们最终都会搞明白这些，然后还是退回到分层架构的老路——因为这是唯一我们懂的方式：政策上怎么想，安全上怎么想，

And I think that that's going to be the real tell, because I think that companies will figure all this out, and I think that they will just fall back on layers and architectural models because it's the only way- We know how to think about it for policy, we know

我们懂怎么想，

how to think about it for security, we know how

但这也是唯一能搭建系统的方法。

to think But it's also the only way to build a system.

不然你就是在做个普通应用。

Otherwise, you're just building an app.

如果你只是想做个应用干一件事，那真没必要搞这么复杂。

And if you're building an app to do one thing, we don't need all of this.

其实还有完全不一样的做法。

There's a whole different way to do it.

我特别感兴趣的是，虽然我手头没什么硬数据，但至少我觉得，现在正冒出一批公司，它们从零开始，完全基于第一性原理，在做各种服务类业务——比如开一家营销公司、工程咨询公司，或者有人可能在给律师事务所做这一套。

The thing that I'm pretty fascinated by is, and I don't even have any amazing data points or but at least the notion of these sort of companies that are emerging in these services categories from the ground up, from the pure first principles approach, which is like, okay, well, if I could start a marketing agency or a consultant you know, engineering consulting company or, I don't know, maybe somebody's doing this for law firms.

搞建筑也行，啥都行。

Construction work or anything.

是的。

Yeah.

比如建筑行业。

Like maybe construction.

设计。

Design.

建筑设计。

Construction design.

对，就是建筑那块儿。

Oh, construction yeah.

没错。

Exactly.

就是建筑设计。

It's architecture design.

任何属于知识型服务的公司都行。

Anything that would be like a knowledge worker kind of services company.

如果你完全没那些条条框框，比如信息壁垒、权限限制，公司完全可以从头来过。

Because you could kinda build your company pretty differently if you had no constraints of, I have no information barriers and and boundaries of what people should have access to.

我可以直接把所有需要的背景信息都交给AI代理。

I can give the agent just all the context it needs to do its work.

我还能临时写代码，专门解决某个问题。

I can write software on the fly for particular things.

我觉得这迟早会带来不小冲击，至少在大公司反应过来、跟上节奏之前。

Like, I do think that will be relatively disruptive, you know, for some time until the bigger incumbents can kind of, you know, get out of the way on this.

至少能留下一些范例，让大家看到这种新型公司到底能长什么样。

And that will at least create, you know, some precedent or or case studies of what what this new sort of corporation could look like.

但说实话，时间久了，他们还是会遇到其他公司都逃不掉的那些老问题，就是……

But I do you know, over time, they'll still run into the same exact problems of every other corporation, which

他们会碰到地理限制、市场细分，还有各种实际困难，你知道的。

Well, they'll run into they'll run into geography or market segments, you know, or And and just challenges.

对。

Yeah.

就是这些事儿。

Like those those things.

只要超出你那小圈子，没错。

Anything outside your little walls Yes.

你迟早会撞上现实世界。

You will run into the physical world.

没错。

Right.

我挺喜欢这个想法的，现在确实冒出了一些全新的商业模式。

I do kind of like the idea that there are some new business models that open up now.

当然了。

Oh, of course.

对。

Yeah.

没错。

Yeah.

是啊。

Yeah.

嗯。

Yeah.

因为现在有太多信息或软件，它们的实际经济价值比被使用的程度高出上百倍，就因为没人愿意花五分钱去访问一份数据，或者花一块钱用一次某个工具。

Because, like, there's so much either information or software that that basically goes underutilized by, a 100 x relative to, like, what what its economic value is, simply because, like, nobody wants to pay 5¢ for accessing a, you know, piece of data or use a tool for $1 once.

但如果你给这些代理设定一个预算和一套操作规则，它们就能随时去查医学研究资料，做深度调研，我愿意为这种服务花三块钱，代理自己就能完成交易。

But, like, you do give these agents, you know, a budget and and a protocol to work with, and all of a sudden, you're like, oh, like, on the fly, they can go get medical research for some deep research tasks they're doing, and I'll pay, like, $3 for that, and the agent is able to go and transact.

这就打开了一整套全新的商业模式，哦。

Like, it kind of opens up a whole new world of business models for Oh.

针对互联网。

For the Internet.

等等，你刚才那话太客气了。

Let me oh, I'm gonna you're that was too nice.

哦，明白了。

Oh, okay.

不是。

No.

不是。

No.

这个问题其实最突出，现在人人都在琢磨这背后的经济模式，但他们对机会规模的估计至少差了十倍。

That one is one where that's actually the biggest I think that the biggest sort of in the air problem right now is everybody is trying to figure out the economics of all of this when they're off by at least an order of magnitude on how big the opportunity is.

因为未来一定会出现全新的商业模式，现在谁也说不准是什么，但肯定会出现——每一种新技术都是这样。

Because the new models that people will come up with, that nobody knows what they are right now, but they will absolutely come out with new models because that's what happens with every new technology.

而现在阻碍讨论的，是一堆金融圈和华尔街的人，还在用老一套思维，拼命给GPU、令牌这些玩意儿找理由，好像我们还活在过去似的。

And the thing that holds back the sort of the discussion now is you basically have a bunch of finance and Wall Street people trying to justify GPUs and tokens and things like as if we're in some old world.

所以他们把收入看成一条线性增长曲线——对吧。

So they're viewing the world of revenue as sort of this linear growth curve- Right.

他们就觉得得把所有开销都合理化。

So they're thinking- justify all the expense.

当人们开始创造新的东西时，这正是当年PC遇到的问题。

And when people are going to create This was the problem with PCs.

当时大家觉得PC市场是有限的，因为觉得算力消耗总有个上限。

People viewed PCs as a finite market because they just viewed the consumption of MIPS as some finite thing.

他们从来没想过，要是把所有算力都放到每台桌面上会怎样？

They didn't think, What would happen if we put all those MIPS on every desktop?

而且大家还认为软件是跟硬件绑在一起的。

And in particular, people thought software just came with the MIPS.

根本没人想到：嘿，软件本身就能卖钱啊。

And nobody thought, Oh, well, they'll just sell the software.

只有一个人想到了。

One guy did.

事实证明，这个想法特别靠谱。

And it turns out that was a really good idea.

是比尔他们那批人吗？

With Bill or somebody?

对，没错。

Yeah, yeah, yeah.

就是比尔和保罗。

Bill and Paul.

但云计算也发生了同样的事情，人们看着云计算说：'哦，我们要把整个服务器业务——每年大概六万台——全都搬到别人的数据中心去。'

But the same thing happened with the cloud, which was people looked at the cloud and they said, Oh, we're going to take all of the server business, which was literally like 60,000 units a year, and we're just going to move it to someone else's data center.

觉得这就是全部生意了，然后我们再分摊价格。

And that would be the business, and then we'll divide up the price.

而且没人意识到，如果我们把资源调配挪到那边，人们的使用量会暴增一千倍。

And nobody went, Oh, people are going to use a thousand times as much of the resource leveling if we move it there.

这正是让我抓狂的地方——华尔街那些模型，总是把收入当成一个固定不变的数字。

And that's exactly I mean, that's the thing that I it just drives me absolutely bonkers that the Wall Street models have this fixed revenue Right.

零和思维。

Zero sum thinking.

就是这种奇怪的零和想法，他们觉得公司花的钱就那么多——你们刚起步时Salesforce也遇到过这问题，但马克早就开路了，CRM业务当年被看成每年20亿美元的生意——是的，

It's this weird zero sum where they just think that the amount of money that a company is gonna spend and this was the problem with Salesforce that they faced when you were starting too, but Mark was just blazing the trail, which was like the CRM business- Yeah,

想搞个大动作。

trying to a big idea.

当时大家觉得得花20亿美元，买一堆服务器、Oracle授权，折腾好几年，还得请顾问，可要是能让销售们自己一个个注册，完全没阻力，人人都愿意用。

Was 2,000,000,000 a year, and it was 2,000,000,000 in like you had to go buy all these servers and these Oracle licenses and this huge headache and years of deployment and consulting, when if you could just get salespeople to sign up individually, they all will sign up with no friction.

这事儿在AI上也肯定会一模一样发生，一点悬念都没有。

And that's that is exact there is no no doubt that that is what's gonna happen with AI.

我给你举个例子。

Let me give you an example of this.

我做投资也快十年了。

So I you know, I've been in for investing for ten years now.

我手头大概有240家公司，都在跟我合作。

I probably have a portfolio of 240 companies I work with.

所以我

So I

我能看到其中大约五十家的情况。

have visibility to, let's say, into 50 of them.

这些都是基础设施类公司。

These are all infrastructure companies.

有些过去表现不错，有些就没那么好。

Some historically have done well, some not so well.

但过去六个月里，每一家的增长都呈指数级飙升。

Every single one of them has gone asymptotic in the last six months.

然后你就想，嗯，这是为啥？

And then you're like, okay.

为什么会这样？

Why is this?

其实现在写的软件比以往任何时候都多得多，这跟企业客户没啥关系，纯粹是因为现在对基础设施层的使用量实在太大了。

It just turns out there's so much more software being written now than ever has been And so it's like and and and and it's not because they've got enterprise customers, you know, it's just because there's just so much consumption of the of of the infrastructure layer right now.

所以软件越来越多，智能代理也越来越多，计算机资源的消耗肯定会大幅上升。

And so with more software, with more agents, there's gonna be a lot more consumption of computer resources.

因此在计算资源这方面，咱们肯定会迎来一团乱。

So certainly in the case of the computer side of things, we're gonna see a mess.

但我们还没到人人都拿手机当AI耗电大户的阶段呢。

Well, we haven't even gotten to the point yet where everyone's phone is a huge consumer of AI.

对啊。

Right.

得批量买一堆授权什么的。

To buy like a bulk license for a bunch of stuff.

是啊。

Yeah.

你希望这事儿能有点可预测性。

You want some predictability in that.

你不仅希望可预测，还希望根本不用操心。

Well, you want predictability, and you just want, like, to not have to think

我觉得特别有意思的是，这还是头一回，智能代理根本不在乎小笔交易的麻烦

about I just I like the idea that it is the first time that you could like, there's just the agent doesn't care about the friction of a small transaction,

所以这就意味着

and so it's

头一回有资源藏在付费墙后面，却有东西愿意真金白银地去付这个钱

the first time that you can have resources behind the paywall that something will actually be willing to pay for that

还有资源。

And and resource.

这个世界早就建好了基础设施，能把这些零散的付款整合成对用户来说高效又省心的模式，

The world has built up the infrastructure to to aggregate those payments into something efficient for a customer and

还有服务。

a service.

对啊。

Right.

而且现在令牌成本在总成本里占这么大头，正逼着整个行业真正转向按使用量计费，就像我们现在这样。

And and because tokens are such a significant part of COGS right now, it is pushing the industry to do usage base in a way that we have.

我记得从永久授权转到订阅制的时候，那可是一大堆大变动。

Like, I remember when we went from perpetual to recurring, and that required a bunch of huge changes.

现在我们正经历同样的转变，转向按使用量计费，这事儿特别细碎，说白了，你还是会跟AWS签合同。

We're going through the exact same change right now towards usage base, and usage base is pretty granular, and it actually allows I mean, again, you will have a contract with AWS

我们当初在AWS身上就走过这一遭。

We went or through this with AWS.

大家慢慢学会了用他的信用额度。

People learned to do the To use his credit.

那时候大家对云计算怕得要死，非得找中间商帮我们找最便宜的、帮我们协调一切。

And we went through the phase where people were so terrified of cloud compute that they were like, We need companies in the middle to help us find the cheapest and to arbitrate it all.

现在你把令牌加进这堆事儿里，我真觉得咱们这会儿根本没时间聊这个。

Okay, well now you write tokens into this, and I don't see how we possibly have time in this in this conversation.

我反正待到你们说完为止。

I'm here as long as you guys

看得见。

can see.

但我觉得，未来几年里，工程计算预算的讨论肯定会变得最疯狂。

But, like like, the the engineering compute budget conversation to me is gonna be just, like, the most wild one in the next couple years.

就是说啊——

It's just like

你什么意思？

What do mean?

你把工程开支的多少比例分配给了令牌？

How much did you allocate of your engineering expense

对啊。

Yeah.

投到令牌上？

To token?

投到令牌上。

To token.

你要是看Twitter上不同人的说法，有的说占1%，有的干脆说占100%。

And and it's like, you you know, depending on who you read on Twitter, it could be 1%, and the other and the other side could be a 100%.

而且这事儿

And it's

对啊。

like Yeah.

但这些东西，不行。

But this stuff No.

不行。

No.

你们的CFO们到底得知道什么？

What do you CFOs have to literally

他们真得知道。

they actually

得知道答案。

have to know the answer to

这个。

that.

我明白，他们确实得有个说法。

I I understand they have Okay.

CFO们总想搞清楚那些根本没答案的问题。

To CFOs always wanna know the answers to things that don't have the answers.

不可能。

No.

根本不可能。

No.

华尔街

Wall Street

会逼着他们给出答案。

is gonna make them know the answer.

不可能。

No.

根本不可能。

No.