投资者如何通过AlphaSense的瑞安·芬内蒂提升专家电话和人工智能能力

最终，我们会共同决定向您推荐的人选。

And then, ultimately, it's a joint decision on we'll recommend to you.

我们认为这个人是可信的。

We think this person's credible.

我们以前和他们合作过。

We've worked with them before.

如果是新招募的，我们是否收到信号表明这个人是在装样子，我们不建议您选择他们，或者他们通过了初步筛选？

Or if they're freshly sourced, are we getting signals that this person is kind of faking it and we wouldn't recommend you take them or they pass their screen?

这可能和我们之前讨论的一些内容有关，但我还是想再强调一下。

This might apply to some stuff we've already talked about, but I do wanna hit again.

您可以进行两种类型的访谈，显然它们的范围更广。

There's two types of calls you can do, and obviously they're broader.

在我看来，两种类型的访谈是：我需要实时信息，对吧？

The two types of calls in my mind are, I want real time information, right?

我们不是在寻找新产品发布信息，也不是在寻找季度数据。

We're not looking for an NPI, we're not looking for quarters.

但你和我今天是2月9日录音，你可能会想和那些担任公司CIO的人聊聊SaaS。

But you and I are, again, recording February 9, there's the SaaS You might want to talk to people who are the CIO for a company.

你可能会问：‘你们现在正在重新评估软件预算、SaaS预算、Percy预算吗？’

And you might want to say, hey, how much are you reevaluating your software budget, your SaaS budget, your Percy budget right now, right?

这是一种实时的温度检测，不同于长期趋势。

That's a real time temperature check versus the longer term.

你想要和CIO聊聊，问他们：‘从长远来看，你们是怎么看待Zoom和Microsoft Teams的？’

You want to talk to the CIO and say, hey, how are you thinking about Zoom versus Microsoft Teams in the long term?

或者这是一个更具体的例子，但你可能还想看看整个行业格局。

Or that's a very specific example that's more But you might wanna look at the overall industry landscape.

你可能会说：‘你负责Duolingo，对吧？’

You might wanna say, Hey, you run Duolingo.

你们是怎么规划未来五年的评估和发展的？

How are you guys thinking about the five year evaluation progression?

比如，Duolingo还可以在哪些方面拓展？

Like where else can you expand the Duolingo?

你以前在学习领域，现在转到象棋了。

You were in learning, now you're in chess.

你能把它应用到四年级的数学上吗？

Can you apply it to fourth grade math?

你能把它应用到学习打篮球上吗？

Can you apply it to learning how to play basketball?

我不知道。

I don't know.

但这是一个长期的问题，而不是即时性的问题。

But that's a longer term thing versus more in the moment thing.

你觉得专家访谈真正擅长的领域在哪里？

Where do you think expert calls like really excel?

你觉得它们在哪些方面特别出色？

Where do you do you think they excel both?

你认为人们会觉得其中一个比另一个更好吗？

Do you think people see one is better than the other?

你认为人们该如何最好地利用这些呢？

How do you think people can use these the best?

我觉得他们可以两者兼顾。

I think they can do both.

我认为，越来越可行的事情打开了许多过去难以实现的机会。

And I think what's increasingly possible opens up a lot of opportunities that were harder to get to.

所以我会分别谈谈这两方面。

So I'll I'll speak to both.

所以，总的来说，正如你所指出的，关于理解商业模式、驱动因素等，存在更深层次的问题。

So, generally, as you laid it out, there are deeper questions around understanding business models, drivers, etcetera.

而对于基本面投资者来说，当这些对话处理得当时，通常会带来更大的满足感，因为这类对话本身就适合这样进行。

And those, think, generally for a fundamental investor, have been more satisfying in large end counts when done properly because those conversations lend themselves that way.

你所描述的前者——实时市场影响、正在发生的一切——确实非常重要，但这也是人们寻求实时洞察以获得市场视角的地方。

What you're describing on the former sort of real time market impact, what's all what's happening here, like, is absolutely something, but I that is absolutely a place where people go for real time insight to get perspective on the market.

这非常重要。

That's very important.

它们总会存在的。

They'll always be there.

我想你没有直接这么说，但你暗示了另一种形式，显然是通过调查和渠道检查。

I think you I don't know if you said it directly, but you're alluding to another form, obviously, is, you know, in surveys and channel checks.

越来越多地将这些对话视为收集趋势和具体数据点信号的途径。

Increasingly treating these conversations as places to collect signal on trend, specific data points.

我认为，对于越来越多的人来说，除非他们有非常复杂的内部系统来完成这项工作，否则他们会发现专家令人沮丧或不可靠。

And that I think is where more and more people, unless they have really sophisticated internal setups to do that, have found experts frustrating or unreliable.

我要说的是，发生变化的是，首先，它们曾经极其昂贵。

And what I will say is what has changed well, one, first, they were just incredibly cost prohibitive.

因此，为专家网络运营这些服务的成本，与单次专家通话的成本并没有太大区别。

So the the cost to operationalize those for an expert network didn't look that different from an individual expert call.

你不会为一次调查花费十万美元，但你可能会为一次专家通话花费两千美元。

You're not gonna spend a 100,000 for a single survey, whereas you could spend it on 2,000 for, know, for an expert call.

但人工智能实际上是其中一个最重要的领域，虽然我们还处于早期阶段，但我预计这将对你的问题——即专家通话在哪些方面最具优势——产生重大影响。

But AI is actually one of the biggest places where, like, we're early, but I expect this to have a big impact on, you know, your question there of, like, where expert calls may be most powerful.

我实际上认为，像调查和渠道检查这样的洞察，AI彻底改变了这一领域的成本模型和运营模式，这与我们行业以往的任何经验都大不相同。

I actually think on the things like survey like and channel check like insights, AI makes the entire cost model and the operating model behind that, like vastly different from what we've ever experienced in the industry.

所以现在，你可以通过访谈者来实现，他们不需要是人类来安排时间。

So now you could get, know, with interviewers, they're not a human that has to arrange time.

你可以让他们，比如说，去与十位首席信息官交谈，并且在他们自己的时间里完成。

You could have them, you know, go talk to 10 CIOs and do it on their clock.

因此，这让他们能够实时快速地获得关于这类问题的洞察。

So it's their availability to get to get really quick insight on a question like that in real time.

而就在六个月前，这种事还很难实现规模化操作。

And that was just stuff that was really hard to operationalize even six months ago.

这很有趣，因为我安排这次访谈、整理笔记的方式是：前半部分讲专家访谈，后半部分讲AI。

It's so funny because the way I've structured this interview, structured my notes is expert calls front half, AI second half.

而这已经是第五次我们谈到终点了，我心想：哦，我应该谈谈AI将如何推动这一领域的发展之类的内容。

And like for this is like the fifth point where we've hit the end state and I'm like, oh, I should talk about how AI is gonna evolve this thing and stuff.

但即使只是进行这次访谈，你也能看到AI如何悄然渗透到这些方方面面，不过我会把这些留到后面再讲。

But even just like doing this interview, you can see how AI is creeping into a lot of these things, but I'll save it for them.

让我记一下笔记，好吗？

Let me just note taking, okay?

我最近刚做了一次专家访谈，我想你和我周三还做过一次预筛选通话，我那时刚从一次专家访谈回来，对吧？

I just did an expert call last I think you and I did a pre screening call on Wednesday, I was literally coming from an expert call, right?

去做一次专家访谈。

Do an expert call.

我读一份专家访谈记录，不管是什么。

I read an expert call, whatever it is.

我个人觉得比较难的一点是，如何跟踪和记录这些专家访谈的内容，对吧？

One of the tough things I personally find is keeping track or note taking on these expert calls, right?

我会在Tegus或AlphaSense应用里做标注。

I'll highlight it in the Tegus or AlphaSense app.

我会记下笔记，但有时候挺难的。

I'll write down notes, but it can be hard.

如果你在六个月里读了四份关于XYZ公司的专家访谈，要记住这些内容确实不容易。

You read four expert interviews over six months on company XYZ, and it can be hard to remember these things.

要记住你读过的关于一家公司的信息很难，尤其是专家访谈，内容很容易混在一起。

And it's hard to remember anything you read about a company, but especially an expert call, it can kind of blend into it.

等人工智能发展起来后，可能会有所帮助。

AI, when we get there, will probably help a little bit.

但你怎么才能找到最合适的人选，尤其是在实时进行访谈时？

But how do you find the best people, especially in real time when they're doing the interviews?

他们是怎么做笔记的？

How are they taking notes?

他们关注什么，才能记住并真正内化这些专家访谈中学到的内容？

What are they focused on so that they remember and kind of ingrain whatever learnings they're having of these expert calls?

是的。

Yeah.

我认为，最佳做法显然是在访谈后留出足够的时间进行整理和总结。

I think I mean, a best practice is obviously to book enough time right after to go synthesize and take stock.

但我认为，这种技能和自律，真希望我们早就不再需要它了。

But I I think that skill set and that discipline, I I wish it was all it it was obsolete with us already.

但你看，所有的发展路线都指向同一个方向：当你通过Tigus进行专家洞察或专家访谈时，能够记录、即时转录并发送给你，这已经是基本要求了，而我们今天已经能做到这一点。

But look, all road maps are leading in this direction where you do an expert insight, you do an expert call through Tigus, it's table stakes that that should be able to be something that's recorded, instantly transcribed, sent to you, which we do today.

但更重要的是，还需要一个AI摘要和整合功能，能够反映你希望如何组织这些笔记的方式。

But more importantly, there's an AI summary and synthesis that mirrors the way you wanna organize your note taking on that.

我认为，我们还没有达到这个水平，但用不了几个月，大多数人就会朝这个方向发展。

Like, I think the fact that we're not there is I mean, within months, think, like, most people are gonna be moving in that direction.

但要回答你的问题，传统上，那些做得很好且系统化的企业，都会形成一种纪律：一旦访谈结束，立即整理笔记，并存入我们所有人都能访问的内部驱动器中。

But to answer your question, like, traditionally, I think the the funds that have done this really well and systematically have a discipline around as soon as we're done, we take the notes, it goes into internal drive that we can all extract from.

另外我想说的是，这在传统观念中是一个非常重要的部分，我们稍后会谈到下一个话题：人们过去认为，专家就是专家。

And then the other thing I'll say that is a really big part of, you know, we'll get to the next conversation, is traditionally people thought of like, there's expert.

还有许多提供专有研究和投资研究服务的机构。

There's all these services for proprietary research and doing investment research.

还有各种传统工具，比如数据源和其他提供商。

And there's all these, like, tools, the traditional, you know, data data feeds and other providers.

然后就是AI工具，以及我们内部的内容。

And then there's AI tools, and then there's our internal content.

而且，越来越多的情况是，整个公司都在不断进行专家访谈。

And increasingly, what's happening is you're doing expert calls as a firm all the time.

你有投资备忘录。

You have investment memos.

然后还有外部数据提供商，将所有这些整合起来，并利用人工智能提取这些洞察。

And then there's external data providers and plugging all that in and using AI to extract those insights.

这最终是事情发展的方向。

That's ultimately where things are going.

当我们进入人工智能的话题时，我会分享一些我看到的非常有趣的使用案例，以及这些洞察是如何产生的。

And then we'll get as we get to the AI conversation, I'll talk through some use cases that I'm seeing that are really interesting and how insights are coming out of that.

但归根结底，我认为，作为通话后一名专业笔记记录者的角色，其生命周期将非常短暂。

But, ultimately, I think the world of having to be a really expert notetaker on the back of your call has a very short half life.

人工智能本应为你做到的一件事，就是不必让你的日常工作大量耗费在记笔记上。

And, like, one of the things AI should be able to do for you is not make that a huge part of your routine.

它应该让你能够即时收到一份准确、符合你需求的洞察摘要。

It's you being able to have that and immediately send you a summary of exactly the insights and the structure you want.

技术可以做到这一点。

That's the technology can do that.

对吧？

Right?

我们总是谈到人工智能。

Well, we we keep coming on to AI.

那我们开始转向人工智能吧。

So let's start transitioning to AI.

在我看来，关于人工智能的讨论几乎分为两个部分。

And I will say in my head, the AI discussion has almost two parts to it.

一是普遍使用人工智能工具。

There's just using AI tools in general.

二是由于我们是从专家访谈开始的，所以要讨论人工智能工具如何塑造和改变专家访谈——这显然是其中的一个子部分，但我认为它属于这个范畴。

And then because we started with expert calls, there's how AI tools are shaping and evolving expert calls, which is obviously a sub piece of that, but I think it kind of fits into this.

所以，让我从问专家访谈的同一个问题开始。

So let me start with the same question I did for expert calls.

如果我是一个听众，无论是关注专家通话中的AI，还是AI本身，如果我听完这场对话后只能带走一件事，关于如何用AI成为一名更好的投资者，你会怎么回答呢？

If I'm a listener and whether choosing AI on expert calls or AI in general, if I'm a listener and I'm gonna walk away from this conversation with one thing about how I can use AI to be a better investor, what would your kinda how would you kinda answer that?

是的。

Yeah.

让我告诉你，在面向公开市场的投资者中，我们看到的最主要应用在哪里。

Look, I'll tell you where we're seeing all the action for public markets focused investors.

对吧？

Right?

其中一个能让你立即获得优势、让生活立刻变更好的使用场景，就是财报分析。

And that is, like, one use case where you can immediately start getting leverage and making your life instantly better is around earnings.

对吧？

Right?

所以，你给我的建议和表达方式是，你最需要做的就是找一个你花大量时间进行手动整理、综合多个数据源，并在时间压力下形成观点的地方——而这正是AI最擅长的领域。

So the, you know, the advice and the way you'd kind of phrase it to me, was like, the number one thing you need to do is pick a place where you find yourself spending a huge amount of time doing hand to hand comment on synthesis and taking multiple data sources and forming a view under time pressure that is ultimately where out where AI is strongest.

因此，在公开市场中，财报季正是我们看到这种应用的典型场景。

And so earning season is where we're seeing that in public markets quite a bit.

我给你举几个例子。

I'll give you some examples.

有些事情，人们过去常常需要在背后处理，比如我的投资组合里有一个名字。

There are things that habitually people would have to do on the back of like, I've got a name in my portfolio.

这其实是一次真正的投资者对话。

Like, is a real investor conversation.

我看了，我已经读过了。

I'm looking I have read it.

他们刚刚发布了财报。

They just published earnings.

管理层的指引非常积极。

Management guidance was very positive.

现在我必须基本上更新一下投资逻辑，看看我们是否还想继续持有这只股票，以及周围发生了什么。

Now I've gotta go basically update the thesis on whether or not, you know, we want to stay in the stock and what's happening around us.

过去你必须亲自逐项处理的事情，现在几个小时内就能完成。

The things that you used to have to do very hand to hand, you can do now within hours.

因此，这个人为此设置的一个提示是：好的，这是管理层的指引。

And so one of the prompts that this individual has set up is, okay, here's management guidance.

我想让你将CEO的说法与我告诉你的已经发布财报的最近五家可比公司实际的现金流报表进行对比。

I want you to compare what this what the CEO is saying to the actual cash flow statements of the last five comps that I tell you that have already reported.

这使得人们能够迅速得出结论：尽管这位发言者态度积极，但现金流报表显示其他公司的情况却普遍消极。

And what that's allowing people to do very quickly is say, okay, this individual speaking positively, but the cash flow statements show that there's a lot of negativity on everyone else.

这能说明什么？

So what does that tell you?

两点。

Two things.

第一，Reddit是个例外，整体情况非常积极，为什么？

One, Reddit's an outlier and things are going really positively and why?

或者第二，管理层过于自信。

Or two, management's overconfident.

对吧？

Right?

我们已经在为那里设置一个疑问号了。

And we're already setting up for a question mark there.

这些正是AI真正擅长的事情——从多个来源综合洞察，并建立人类难以快速发现的联系。

These are the types of things that are happening right around, you know, what AI is really good at is synthesizing insight from multiple sources and drawing connections that are very hard for a human to do quickly.

这可能是我认为公开市场投资者最应该关注的首要领域，因为人们现在一直在做各种事情。

And that's probably the number one place I would say public markets investors, there's multiple things that people are doing right now all the time.

这太有趣了。

So that's super interesting.

如果我稍微反驳一下你。

If I could just push back on you slightly.

当然。

Sure.

这是《又一个价值》播客。

This is Yet Another Value podcast.

我的常规播客是邀请一位嘉宾，花一小时讨论一只股票。

My average podcast is a guest comes on and we talk about one stock for an hour.

这是一种深入研究、通常高度集中的投资。

It's a deeply researched, generally concentrated investment.

当你提到盈利和需要快速完成的事情时，我脑海中第一反应是，他指的是那些交易季度财报、追逐内幕消息的播客团队，对吧？

When you say earnings and things that need to be done quickly, I just know in my mind, my first thought process was he is talking to pod shops who are trading quarters and whisper numbers and all this sort of stuff, right?

让我重新表述一下这个问题。

So let me just reframe the question.

如果我忽略短期因素，一个只持有五只股票的长期投资者会如何利用人工智能来优化他们的投资流程？

If I was ignoring immediate term stuff, how would someone who's five stocks concentrated long term investor, how are they using AI to evolve their process?

是的。

Yeah.

我认为另一个领域是，当你打算投资一家公司时，最终需要做大量工作，去分析这家企业的基本驱动因素，以及我能否获得与市场共识不同的观点？

So I think there's another area is when you are going to, you know, take a position at a company, I think there's ultimately a heavy heavy amount of work and what are the fundamental drivers of this business and can I get a differentiated view versus consensus?

对。

Yes.

对。

Yes.

是的

Yep.

我很喜欢你提到的‘差异化观点’。

I love that you said differentiated view there.

是的

Yep.

对

Yeah.

我认为，其中一些非常有趣的用例是，共识是在多个层面形成的。

And I think ultimately, some of the really interesting use cases there are, you know, like consensus is formed across multiple layers.

对吧？

Right?

如果这是一个被广泛覆盖的公司，卖方的研究观点是什么？卖方在讨论哪些关键争议点？

What is like what are sells if it's a, you know, widely covered name, what are what are the key debates on the sell side and what they're saying about it?

大家都在说什么？那些专家对这些关键驱动因素又怎么说？

What are all the people saying what are all the experts saying on this on the key drivers that matter?

那我们对这些观点的内部看法是什么？

And then what is our internal view on those?

你可以进行交叉验证。

And you can triangulate.

你可以比较这些不同的观点。

You can compare those perspectives.

你之前问我一个问题，关于AI真正独特且超越普通投资者能力的地方是什么？

One thing that AI I think you had asked me a question coming into this is, like, what is AI actually really good at uniquely that surpasses the ability of the average investor?

对吧？

Right?

而它只是达到你引入基金的初级分析师所能做的水平。

Versus where it's merely coming up to the ability to do what an analyst that you you know, junior analyst you bring into the fund can do.

我想说的是，它能够以网格化的方式，综合并比较来自大量不同来源的观点。

And one thing I will say, it is the ability to go synthesize and compare perspectives across tons of different sources in a grid like format.

我认为我们看到投资者更多从基本面角度使用AI的一个例子是，你可以查看众多不同的组成部分，比较管理层的指引、卖方的观点、我们进行的专家访谈，它们与外界说法有何不同，以及交易记录库中的专家访谈内容是什么。

And so one of these things that I think we have seen investors using, more of the fundamental first, is that you can look at so many different components and compare what is management guidance saying on this, what is the sell side saying on this, what are the expert calls we're doing, how do they compare to what's being said, what are the expert calls in the transcript library?

我认为这让人能够意识到，这些才是真正关于该股票、关乎价值创造核心的争论。

And I think that's allowing people to say, hey, these are the real debates on this name that are really fundamental to the value creation story.

而这就是我们将要投入更多精力的地方。

And that's where we're gonna do a lot more work.

我认为，像这种深度的分析，你根本不会想到要去这么做。

And I think that's the kind of stuff that like, you just wouldn't know to do that level.

我是说，像这种80乘80的网格，对多个数据源的输入进行对比。

Like, I'm talking about, like, an 80 by 80 grid comparison of inputs across multiple data sources.

人类分析师根本不可能完成这样的工作。

It's just not feasible that a human analyst would do that.

但这种分析能揭示出投资者值得深入挖掘的真正有价值的方向。

But that reveals really insight insightful places to go and dig deeper for investors.

我们可能会回头再讨论这个。

We'll probably come back to this.

让我突然想到的一点是，Tigas上有些股票一年有80次专家访谈，对吧？

Like one thing that just jumps out to me is there are some names on Tigas where there's 80 expert calls a year, right?

我的意思是，也许吧。

There's no I mean, maybe.

但如果你说你要跟踪30家公司，那你根本不可能去阅读这30家公司的80场专家电话会议。

But if you're saying, hey, I'm gonna follow 30 companies, there's no effing chance you're gonna read 80 expert calls on 30 different companies.

AI可以在半秒钟内完成并为你总结，对吧？

AI can do it in half a second and summarize it for you, right?

所以我想就这个问题问两个问题。

So I wanna ask two questions on that.

第一个问题是，我知道我不只是一个人这样。

The first question, I know I'm not alone in this.

现在有很多工具可以自动为你构建财务模型并进行推算。

There are lots of tools that will automatically build financial models for you and extrapolate them.

比如康卡斯特公布第三季度财报，系统就会自动将其录入，更新整个模型。

Comcast reports Q3 earnings, they'll automatically put it in, update the model and everything.

但我所有模型都是自己手工搭建的，尤其是当我快要做出投资决策时，因为亲手去做这件事能让我学到东西，让我深入思考，诸如此类。

I of, I build all my models by hand, especially as I get like close to making an investment Because there's something about just going and doing it that makes me learn and makes me think and all that sort of stuff.

但如果我只是被直接提供的话。

Whereas if I just had presented to me.

使用AI工具时，我有点担心这一点，对吧？

With AI tools, I kind of worry about that, right?

如果我只是让AI总结ADX的访谈内容，而不是现在亲自去阅读，这其中确实有区别——获得摘要可能让我无法完全理解或深入内化。

If I just have AI summarize ADX for calls versus now, it is a lot, Going and reading, there is something about getting the summary that maybe I don't quite understand it or internalize a lot.

所以当你与公司交流时，尤其是与投资组合经理级别的人交谈时，他们是如何谈论这种权衡的：我根本不可能读完80场专家访谈，尤其是在涉及30家公司的情况下， versus 如果我只是得到80份摘要，我就无法内化。

So when you talk to firms, especially portfolio manager level people, how were they talking about that trade off of I could never read 80 expert interviews, especially across 30 names versus, hey, if I just get 80 summarized for me, I don't internalize.

我没有深入思考。

I don't think it through as much.

我正在失去那种优势和洞察力，我只是把一切都外包给了AI。

I'm kind of losing that edge, that insight, I'm just outsourcing to all the AI.

你听到人们是如何谈论这种权衡的吗？

How are you hearing people talk about that trade off?

是的，我认为这是一种合理的权衡。

Yeah, look, I think it's a fair trade off.

这是一种非常容易理解的情绪反应。

And it's a very, very understandable emotional reaction.

我的确也经历过这种感受。

I mean, I've had it myself.

我曾经亲历过构建公司模型的过程。

I went through the experience of building company models.

我知道，就像你描述的那样，通过自己亲手搭建这些驱动因素并运行各种敏感性分析和情景模拟，才能真正理解其中的逻辑。

And I know that, like, what you're describing that, like, it clicking the drivers and the sensitivities by actually actually building the drivers myself and running the sensitivities and the scenarios through it.

不过，我想说的是。

Here's what I'd say, though.

我发自内心地相信，到2030年，下一代的顶尖投资经理中，会有很多人从未经历过这些繁琐步骤，却依然表现卓越。

I think ultimately, like, I I believe in my bones that, like, by 2030, they're gonna be really high performing portfolio managers to this next generation coming up who, like, never who absolutely never had to go through that.

他们从未亲手搭建过极其复杂的并购模型，但依然擅长利用这些工具获取洞见，交叉验证真正重要的信息，从而取得良好的投资成果。

Like, they you know, they've never built a super detailed m and a model, and yet they're pretty good at leveraging this stuff to to get to insights and triangulate on what really matters and get good investment outcomes.

因此，我们行业内的争论，恰恰如你所说：在完全信任这些工具之前，它们仍然容易出错，存在判断偏差，以及我无法信赖或认同的数据。

And so the debates we're having in the industry are more about exactly what you said, which is like, until I can fully trust this stuff, it's still prone to like errors and judgment, data that just like I don't trust or believe in.

我认为，我们构建AlphaSense的理念中非常重要的一点是，所有内容都能完全追溯到原始来源。

And I think so a huge part of, you know, like, to name our philosophy for how we've built is that everything in AlphaSense is fully fully traceable down to the source.

这一点非常关键，因为即使在我自己进行市场推广研究时，我也需要立即知道每一个洞察的来源。

And that's really important because like when I go through workflows, even for my own research for like go to market, I need to see instantly where that insight is coming from.

否则，它就会打断我的工作流程。

Otherwise, it just were it just interrupts my workflow.

我不希望花了两个小时后，突然发现所有结论都建立在脆弱的基础上。

I don't wanna get two hours in and then suddenly have it all be on a shaky foundation.

所以我觉得这一点非常重要。

So I think that's really important.

其次我想说的是，AI非常容易受提示方式的影响，如果你以某种方式提问，它就会非常坚定地给出结论。

And then the second thing I'll say is, like, look, AI is very prone to if you prompt it a certain way, it'll pound the table.

我有过这样的经历：我要求它，比如，以CRO向董事会汇报的视角，为AlphaSense制定市场推广计划。

And I had that experience where I say, like, you know, like, build my go to market plan for AlphaSense in the lens of a CRO reporting board.

那个提示会让我对某些观点产生极强的信服感。

Like, the prompt like, the conviction it will give me in certain things.

我觉得这毫无道理。

And I go, like, that makes no sense.

我的判断告诉我，也许这是对的。

Like, my judgment suggests that, like, well, that might be true.

我们有一系列与客户的真实通话，明确提到x是真实的。

There's a verbatim series of calls we had with customers saying x was true.

我了解 enough，那个细分市场的总潜在市场规模（TAM）根本不符合这个建议。

I know enough that, the TAM of of, like, that segment doesn't make any sense for that recommendation.

所以我认为，对于投资者来说，判断力、对市场结构和商业模式的理解，其价值更高，但很多像我们这样留在行业里的人——我离开了这个行业。

So I think that's ultimately, I'd say for the investor, like the value that comes from judgment, and understanding market structure and business models, I think, like goes higher, but a lot of the like, like, you know, a lot of us in the, you know, for those who stayed in the industry, I left the industry.

但对于那些留下的人，你们作为投资者的自我认同，很大程度上来自于你们的技术能力和分析能力。

But for those who stayed a lot of you, like your, your sense of self as an investor is your technical prowess and your in your analytical skills.

我认为，随着时间推移，这些能力正在变得商品化。

I think those over time are getting commoditized.

而更重要的是你们的模式识别能力、判断力，以及推动这些事情的能力。

And what's much more important is your pattern recognition, judgments, ability to push on these things.

你看，你刚才说的每一件事，尤其是最后部分，完全契合我的世界观。

Look, everything you just said, especially towards the end, just like matches my worldview.

那我问你一个问题。

So let me ask this.

你提到，如果我没记错的话，在做投资时要有独特的观点。

You mentioned, if I'm quoting, having a differentiated view when you're making investment.

对吧？

Right?

这正是你在做集中型长期投资时所寻找的东西。

That's kind of what you're looking for when you're making, especially a concentrated long term investment.

如果每个人都使用AlphaSense和AI来总结相同的AlphaSense专家库，这就是我不看卖方报告的原因。

If everyone is using AlphaSense and AI to summarize the same AlphaSense expert library, this is why I don't read sell side reports.

因为如果你读了所有卖方报告，然后基于这些得出结论，那你只是得到了市场观点或卖方的分析。

Because if you read all the sell sides and then you make your conclusions based on that, then you've just got the market view or you've got that sell side review.

如果每个人都用AI来总结一切，那人们该怎么思考呢？嘿，这不过是基本门槛，对吧？

If everyone's using AI to summarize everything, how are people thinking about, hey, that's the table stakes, right?

我需要这个。

I need that.

我需要这种基本的思考方式：人们是怎么想的？我如何获得一个与众不同的观点？我的独特优势在哪里？当别人都在用同样的AI来总结相同的专家通话时，我如何做到与众不同？

I need that basic How are people thinking about, hey, how do I get a differentiated viewpoint or where's my special sauce where I'm going to kind of have a differentiated viewpoint than everyone else is using the same AI to summarize the same expert calls?

是的，我认为在很多

Yeah, I think with a lot of

这些技术革新只是改变了基准线。

these like technology innovations, it just shifts the baseline.

所以，你可以想想，在个人电脑和Excel出现之前做财务分析，对吧？那时候，拥有非常复杂的分析方法确实很特别。

So, you know, I think like the like, you know, you can think of like doing financial analysis before the PC and Excel like, right, like that no longer was it like having these really sophisticated ways of doing that.

但后来这变成了基本要求，如果你不做财务分析，那就完全落伍了。我认为我们现在正走向一个阶段：一直以来，关键都是信息的获取，以及你能否建立一套能产生他人无法达成的投资成果的流程。

Like that became the baseline if you weren't doing financial and outright like and so I think where we're getting to is like, it's always been about access to information and then your ability to have an investment process that yields results that others can't get to.

我认为AI正在做的，我们一直都在说市场是有效的。

And I think what AI is doing, you know, we've always talked about like markets are efficient.

每个人都拥有X和X，但其实我们知道这并不真实。

Everyone has x and x, but like, we know that's not true.

这就像我们当初都被训练去仔细研读笔记，深入分析10-K和10-Q文件，真正整合这些分散的信息，从而形成一种独特的见解，甚至在我们讨论通过另类数据集获取优势之前就已经如此。

It's like why we were all trained to like, sweat the notes and go deep into the 10 k's and the 10 q's and like, really synthesize all these disparate things and get to something that was differentiated even before we talk about getting an edge through like alternative data sets.

我认为，AI的发展使得技术带来的任何优势都变得越来越难以获得。

I think what's happened with AI is just the the the technology is so powerful that any gains from that are getting harder to come by.

因此，真正的超额收益其实来自于我们一直讨论的那些相同因素。

And so really, the the alpha comes from, I think, some of the same things we've always talked about.

关键在于让这些系统为你服务，从而让你能更迅速、更有信心地做出决策。

It's like the ability then have these systems work for you so that can make decisions much faster with conviction.

我给你举个例子。

I'll give you an example.

在私募市场中，过去十二个月里，我亲眼见证了这一点的显著体现。

In private markets, it's it's like, I've seen this really, like, come to play in the last twelve months.

这与我们长期讨论的、专注于少数重仓股的公开市场投资者非常相似，私募股权基金也同样是通过少数几笔集中投资来运作的。

And, like, you know, this is parallels that we talked about for, like, a long term concentrated public market investor, there are very big parallels to, you know, PE fund that makes a couple concentrated bets here.

是的。

Yep.

是的。

Yes.

当我问他们的时候，我会说：嘿，这对你有什么影响？

And like, when I've asked them, I'm like, hey, how's this impacting you?

你们是在看更多的公司、更多的机会吗？

Are you look are are you looking at more more names, more opportunities?

是的。

Yes.

你们每年做的投资更多了吗？

Are you making more investments per year?

没有。

No.

这不是我们的策略。

That's not our strategy.

我们仍然只会投资三到五笔。

We're still gonna only make three to five.

但由于我们看到了更大的可能性，我们对这三到五个投资项目的信心更强了。

But we are much, much more convicted on those three to five as a result of what's possible.

过去，我们的尽职调查流程从A点到C点需要很长时间，但现在从A点到B点的时间已经从几周压缩到了一天。

So the due diligence we used to do that would get us to point from investment process from point a to point c, like, the time to get through point a to b in our process has compressed to a day from weeks.

因此，我们把更多的时间和精力投入到B点和C点，这通常是投资委员会讨论的关键环节——价值创造来自哪里、业务的驱动因素是什么，以及我们对此的独特见解——这才是真正下功夫的地方。

Therefore, the amount of time and energy we spend are really diligent in b and c, which is usually the key debates in the investment committee around where the value creation comes from, where what are the drivers of the business and our differentiated view on that, that's where all the real work is going.

你觉得

Do you think

他们应该这么做吗？你说过三到五个，但他们说，我们更有信心了。

they should be going so you said three to five, and they say, hey, we're more convicted.

是的。

Yep.

我觉得你一开始提到过，既然他们能更快地从A点到B点，那他们是不是应该把投资数量从三到五个增加到八到十个？

I think you suggested at the beginning, hey, because they can go from point A to point B faster, should they instead of doing three to five, should it be eight to 10?

还是应该反过来？

Should it be the other way?

如果他们更有信心，并且能够更深入地进入B2C领域——这可能是他们解决真正细分案例和实现真正差异化的地方——那么与其做三到五个，是不是应该说，我们更有信心，所以应该更集中？

Like if they're getting more convicted and they're able to go deeper into B2C, which is probably where they're addressing the real niche cases and their real differentiation, instead of three to five, should it be, Hey, we're more convicted, so we should be more concentrated.

我们应该做一到三个，而不是三到五个。

We should be doing one to three instead of three to five.

你觉得这才是正确的答案吗？

Do you think that should be the right answer?

是的。

Yeah.

我不确定。

I don't know.

因为我觉得确实有一些基金说过，是的。

Because I do think there are some funds who've said, yeah.

这实际上增加了我们一年内所投资的项目数量。

It actually has increased the amount of things we'll do in a year.

对吧？

Right?

还有其他人表示，这根本不是我们的运营理念，我们通常只会做三到五个项目。

And there are others who are saying that's just not our operating philosophy and will only be, you know, three to five that we usually do.

是的，当然。

And, yeah, sure.

也许有些基金的想法是，我们现在信心更强了，所以要把基金押注在一两个想法上。

Maybe some it's been like, we're gonna go we have even higher conviction now, so we're gonna bet the fund on one or two ideas.

我还没怎么看到这种情况。

I haven't seen that as much.

但我认为，总体原则是，大家都意识到，估值现在很高。

I think just the the general principle, though, is I think everyone recognizes, like, valuations are elevated.

竞争更激烈了。

It's more competitive.

需要配置的资金更多了。

There's more to put to work.

因此，当我们出手时，必须更有信心，才能竞标这些优质资产。

And so when we go, like, we have to be much more convicted to go bid for these good assets.

这就是稀缺性。

Like, that's the scarcity.

因此，我们工作的很大一部分就是确保我们有一个可信的故事，说明我们如何创造价值并实现真正退出。

And therefore, so much more of the work is making sure that we have a credible story for how we're gonna have value creation and a and a real exit.

在过去两年里，这个门槛已经发生了巨大变化。

And that bar has just shifted dramatically over the last two years.

并不是我们主动选择这样，而是我们能感受到周围的人如何迅速带着坚定信念行动。

Not because we chose it to, but be we can feel it around us, like how quickly people are moving on opportunities with conviction.

我们必须跟上步伐。

We have to we have to stay in line.

我认为这最终就是正在发生的事情。

I think that's ultimately what's happening.

不。

No.

我只是问一下，因为你刚才说的正是我想说的。

I I just asked because exactly what you're saying.

我有一些朋友，以前每年会做五笔投资。

I have some friends who used to do, let's say, five investments for a year.

但现在他们说，因为有了AI工具，我可以更快地完成，所以现在做十笔。

And now they're like, hey, because of the AI tools that I can get up to these faster, I do 10.

还有些朋友说，我以前做五笔，现在还是做五笔，但更有信心了。

And then I have friends who say I did five, but now I do five with a lot more conviction.

但我还没遇到任何人说，因为我更有信心了，所以从五笔减到三笔，你知道吧？

But I haven't had anyone be like, because I have more conviction, I do three instead of five, you know?

所以我只是想说

So I was just

没有，我还没听说过这种情况。

No, I haven't heard that yet.

是的。

Yeah.

不过我也想强调一下，虽然最终进入筛选池的结果可能还是三到五笔，但在到达这一步之前，被审视的项目数量已经大大增加了。

I do wanna understate too though that like the but I didn't mean to say that while the end result in the funnel might result in like the same three to five, the amount of things that get looked at before they even get to that has expanded.

我认为，归根结底，你要想的是，如果这个范围已经显著扩大，那么你能够考察多少可能最终达成目标的资产呢。

I think that's, you know, ultimately you think of like how many assets can you look at that might get there if that universe has expanded sometimes materially.

有些人说，我现在看的东西是以前的两倍，因为你可以获得一个SIM。

Like, some have said, I look at twice as many things now because, you know, you get a SIM.

你可以用AI和我们内部的所有工具瞬间分析这个SIM，并快速得出绿色、黄色或红色的结论，而以前这需要耗费数周的分析师人力。

You can analyze that SIM instantly with AI with all of our internal stuff and get a green, yellow, red in a way that, like, took weeks of analyst capacity.

所以我认为这带来了巨大的变化。

And so I think that's been a huge difference.

你之前提到，几年前，财务分析实际上就是用Excel表格。

So earlier you were talking about, hey, years ago, financial analysis, it was literally like Excel spreadsheets.

因为在把数据输入电脑之前，你真的要先在纸质表格上手工构建所有内容。

Was because before you put it into the computer, there was literally a physical piece of paper, a spreadsheet that you would build everything out.

最终，这些工作都转移到了线上，并变得商品化。

So eventually that goes online, that gets commoditized.

现在，有各种工具可以自动基于Excel模型生成内容。

Now there's all sorts of stuff that will automatically build off the Excel model.

所以我认为，六十年前，你仅凭头脑中的量化分析就能赚钱，对吧？

So I would posit to you that sixty years ago, you could make money with quant in your head, right?

如果你是一个非常出色的纯粹的财务分析师，对吧？

If you were a really good literal financial analyst, right?

你可以通过建模来赚钱。

You could make money by modeling.

想想本·格雷厄姆，他只是计算净营运资本。

Think about Ben Graham, just calculating net working capital.

完全正确。

Totally.

我认为，也许十年前到十五年前，你很可能在定性分析方面赚到更多钱，对吧？

I would posit that maybe ten to fifteen years ago, you could make a lot of money probably better on the qualitative side, right?

财务分析师已经变得商品化了。

The financial analysts got commoditized.

赚钱的关键在于定性分析。

The qualitative is where all the money made.

我想说的是，看看过去十五年。

And I would just say like, look at the past fifteen years.

如果你仔细想想谷歌、Facebook、亚马逊，无论哪一个，它们都是有史以来最棒的商业模式。

If you thought it through Google, Facebook, Amazon, whichever one, these are the best business ever.

世界的发展趋势正是如此。

The world was trending that way.

互联网带来了无限的资本回报和规模效应，诸如此类的东西。

The internet infinite returns to capital scale, all this sort of stuff.

如果你能看懂这一点，那可不是靠电子表格算出来的数字。

If you could figure that out, that was not a spreadsheet number.

真正让你成功的，是定性的判断。

That was qualitative that got you there.

AI方面，我并没有看到它取代了定性分析，但AI确实极大地提升了质量标准。

AI is kind of I'm not seeing replacement qualitative, but AI actually really raised the bar in quality.

你觉得接下来能创造超额收益的技能会是什么？

What do you think the next skills are that kind of generate alpha?

如果财务分析已经下降，很多那种定性因素也会随之下降。

If financial analysis has already come down, a lot of that qualitative comes down.

必须有一些技能被提升起来，因为六年前，如果你定性能力很强但财务能力很差，你是无法成功的。

There has to be some skills that get elevated, whereas you know, six years ago, if you were a great qualitative and you were terrible financial, you couldn't make it work.

但当财务技能变得商品化后，定性能力如果也在下降，那你认为下一个关键技能会是什么？

But then when the financial gets commoditized, qualitative makes if that's coming down, what's the next skill set do you think?

是的。

Yeah.

我来分享一下我的观点，当然，这里还有很多需要验证的地方。

I'll give you my thesis, and, you know, there's a lot to to be proven out here.

我想先谈私募市场，再谈公开市场，因为我觉得两者有一些相似之处，但也会有所不同。

I think so I'll talk private markets first, and I'll talk public markets because I think there's some parallels, but they're gonna be different.

我认为在私募市场方面，真正优秀的人在财务结构和交易撮合方面获得的回报正在下降，这一点在业内已经广泛讨论。

I think on the private market side, I think what's been happening is, like, the returns from being really good at, you know, financial structure and deal making have been going to I think that's, like, widely discussed in the industry.

因此，真正重要的是，AI可能会帮助我们更快地应对更广泛的机会，并在符合我们策略时赢得更多交易。我认为，那些专注于投资后组合价值创造的公司会受益。

And so, really, it's about, like, what AI will probably help is facilitating the ability to act really quickly on a much bigger opportunity set and and win more deals when they fit in your I think, like, firms that focus on portfolio value creation post.

对吧？

Right?

我觉得有很多机会。

I think there's a lot of opportunity.

我们没谈到那里。

We didn't go there.

这里更多是从AI和投资流程的角度来看。

Here, this is more of a lens on, like, AI and the investment process.

但我去年参加了一个中型私募股权会议，当时房间里大家都在热议如何利用AI赋能被投企业，推动价值创造的故事。

But I think one of the other things I was at a mid market PE conference last year, and, like, all the buzz in the room was, like, the thing people could do taking AI to portfolio companies to drive value creation stories.

所以我认为，这很可能是AI带来重大收益的领域之一——通过高效运用AI，拓展我们能关注的范围，提升对交易的判断信心，正如我们之前讨论的，但我也认为这会大量转化为对被投企业的价值创造。

So I think that is a likely place where I think some of the big gains will come from, you know, using AI, like, really effectively to drive more places where you can look and get higher conviction on the deals in the way we discussed, but I think a lot of it will also translate to portfolio value creation.

而且我认为这是因为AI确实有一系列真正基础性的应用场景，正在改变运营模式和成本结构，而这在私募领域是切实可行的。

And I think that because I actually think AI has a real fundamental set of use cases where it's changing operating models and cost structures that make sense in that world.

在私募领域，这很难，而在我专注的公开市场方面则不同。

On the the private front, this is hard on the public side, which is where I'm focused.

但在私募领域，我认为这将会是这样：如果AI是一个每个人都能使用的工具，就像老沃伦·巴菲特说的，你在游行队伍里踮起脚尖，就能看得更清楚，但当所有人都踮起脚尖时，情况就变了。

But on the private front, I I actually think it's going to be you know, if AI is a tool that everyone can use, so you have the old Warren Buffett, you know, if you're at a parade and you stand on your tiptoes, you get a better view, but then everyone stands their parade.

所以没人真正占到便宜。

So no one's better off.

实际上，每个人反而都稍微更糟了。

Actually, everyone's a little bit worse, better off.

我真觉得金融分析会被AI商品化，而很多定性工作反而会被货币化。

I actually think it's going to be financial analysis commoditized AI and a lot of qualitative gets monetized.

我认为，那些在人力资源和人际交往方面最出色的人，将成为私募领域真正被提升价值的技能群体。

I think the people who are best at human resources and people are actually going to be the people who I think that's going to be a skill set that gets elevated on the private side.

但你知道，我不抽烟，也许我只是在抽什么奇怪的东西，或者想法太天马行空了，简直是银河系大脑。

But you know, I don't smoke, but maybe I'm just smoking something or you know, just too far out there galaxy brain.

那在公开市场方面呢？

What about on the public side?

你认为哪些技能会得到提升？

What do you think is skill sets get elevated?

是的，你知道，我认为这在每个行业都是一个常见的讨论，即这个答案存在一个长期问题。

Yeah, you know, I think this is like the common discussion in every industry, which is like like there's a there's a long term problem with this answer.

但我确实认为，我们之前讨论过的一点是，从分析师技能向架构师技能的转变。

But I do think, like, know, one thing that we've talked about is, like, this shift from the analyst skill set to the architect's skill set.

所以那些擅长利用这些工具在投资过程中创造杠杆的人，比如对于专注三到五个长期持仓的投资者来说，这可能不太相关，但我认为这将影响公开市场。

So people who are really adept at using these things to to create leverage in the investment process from, like, portfolio I know, like, for a concentrated three to five name long term investor, this is probably less resonant, but I do think this will impact public markets.

我认为你会看到更多人使用人工智能在基本面投资工作中进行组合监控、创意生成，更快地分析更多数据。

I think you'll see a lot more people using AI in fundamental still fundamentally fundamental investment work to to do portfolio monitoring, idea generation, just like look at a lot more a lot more quickly.

我认为这将改变你所说的那种基金团队的行为方式，我确实认为这种压力会蔓延到行业的更多领域。

I think that is going to change like, the stuff you said, like, how pod shops behave, I actually think that pressure is gonna move into more places in the industry.

从长远来看，我认为真正的问题是，你所描述的基本面投资会如何演变？

And then long term, I think ultimately, like, the real question mark is, like, what happens to fundamental investing in the way you described?

最终，我们这一代人，像你我这样成长起来、深谙此道、通过多年投资积累了模式识别能力的人，会不会被新一代取代？

Like, ultimately, do the do we have this, like, cohort of people who grew up in the world that you and I grew up in and are deep experts in it and understand through years of investing pattern recognition and we lose that with another group?

还是说，这个新进入的群体能够跨越式发展，开始关注那些我们一直信以为真、但在数据中并不相关、实际上并不重要的陈规？

Or does this new group that come in leapfrog that somehow and start looking at, you know, like truisms that we've all lived with that like are uncorrelated in the data and actually don't matter.

然后还有一种完全不同的版本。

And then like there's a whole different version.

这不是量化投资，也不是基本面投资，而是介于两者之间的东西。

It's not quant investing, it's not fundamental, but it's something in between.

对吧？

Right?

这是我非常担心的其中之一。

It's one of I'm very worried.

我已经是个恐龙了。

Am already a dinosaur.

让我回到我们之前讨论的偏见问题。

Let me go back to our bias discussion.

这是我经常思考的事情。

Is something I think about a lot.

实际上，在此之前，关于公开市场方面，我出于好奇想问一下。

Actually, before we go there, just on the public market side, I have to ask for my own curiosity.

在投资组合监控方面，您如何看待专注的基本面投资者使用人工智能进行投资组合监控？

The portfolio monitoring side, how are you seeing concentrated fundamental investors use AI for portfolio monitoring?

是的，我认为真正大的例子是，比如我会给你极端情况和日常情况。

Yeah, I think like really big examples would be, you know, I think there's, like, these I'll give you, like, the extreme scenarios, and then there's, like, day to day scenarios.

所以极端情况就是去年的解放日。

So the extreme scenarios was, like, liberation day last year.

对吧？

Right?

比如，我们看到一些人拥有这些投资组合，瞬间就想知道自己的风险敞口在哪里，有什么建议，或者应该去深入研究哪十到十五只股票。

Like, we saw people who had these portfolios and, like, were instantly, like, what is my exposure, and what are the recommendations, or where should I go dig across, you know, like, ten, fifteen names.

对吧？

Right?

人工智能在这些紧急情况下非常出色，能在几小时内就指出所有与他们观点不同或一致的研究内容，以及他们的风险敞口所在，而真正的工作就是在此基础上完成的。

And what AI was very good at was, like, in those kind of fire drill moments, like, within hours, right, had kinda indicated all the places, all the different research that was, like, different from their view and aligned to their view and where their exposure was, and then that's where the work was done.

我认为这是一个极端的例子，但我们也最近再次看到了类似情况，正如你所描述的，围绕SaaS和人工智能敞口的悲观情绪。

I think that is, like, an extreme example, but we also saw that again around, actually, as you described, more recently around all this, like, bearishness around SaaS and AI exposure.

人们一直在使用人工智能快速理解这类问题。

Like, people have been using AI to very quickly get their head around things like that.

从持续的投资组合监控角度来看，要让这一切有效运作，关键在于你能接触到多少市场数据集。

From an ongoing portfolio monitoring perspective, ultimately, what really matters, for this to work well is, like, it's only as good as the number of datasets you have access to in the market.

但最终，我认为市场已经从帮助我寻找问题答案的工具，转变为帮助我自动化日常流程的工具，现在更是发展为能够自主运行报告的定制化系统，仿佛有一位分析师在为我工作。

But, ultimately, like, I think the market has shifted from things that help me go find answers to questions I'm looking for to things that help me produce, like, kind of these workflows I'm constantly doing, to now custom autonomous things that run reports as if I had an analyst working on it.

因此，人们使用投资组合监控工具，要求每个周五生成一份特定格式的报告，告诉我这些参数相对于我的投资组合的趋势和转折点。

So people are using portfolio monitoring to say every Friday, I wanna report in this format that tells me trends and inflections on these parameters against my portfolio.

对吧？

Right?

而这些正是投资组合监控以持续、定制化方式发挥作用的典型例子。

And that's like those are the types of things where portfolio monitoring is just like always on custom way.

想象一下，如果你拥有无限的分析师资源，你会希望有哪些可选的、能让你对投资组合更有掌控感的事项？

Just imagine if you had infinite analyst resources, where where would be something like nice to have discretionary things you'd ask for that would make you feel more in command of your portfolio.

而这些正是人工智能擅长做的事情。

And that's the kind of stuff that AI does pretty well.

是的。

Right.

我先快速说一下偏见问题。

Let me go to bias real quick.

我觉得AI里有三种偏见。

So there's three types of bias I could see in AI.

对吧？

Right?

如果我为AI设计提示词，那我自己就可能存在偏见，对吧？

If I'm crafting prompts for AI, there's bias in myself, right?

如果我针对一只看涨的公司设计提示词，我就会在提示词中带入自己的偏见。

If I'm crafting a prompt on a company I'm bullish on, I can bias myself in the prompt.

公司方面也存在偏见，对吧？

There's bias in terms of the company side, right?

如果让AI阅读公司所有的投资者日活动和所有财报电话会议，管理层通常对自己都极其乐观，他们的表述本身就充满偏见；而分析师也不敢当场激烈质疑公司，否则就会被切断联系，再也无法接触到公司管理层。

If I have AI read every Investor Day and every earnings call a company's ever done, Management teams are generally pretty darn bullish on themselves and they've got a lot of bias in the way they present and analysts aren't exactly going to get on and scream at the company because then they'll get cut off and they'll never get to talk to the company again.

所以当我提问时，我会担心自己存在的偏见。

So I worry about bias for myself when I ask.

我担心如果让AI训练于一家公司的数据集，会带来公司方面的偏见。

I worry about if I have AI train on a company's dataset, bias on the company side.

至于专家方面，如果我们让AI阅读大量专家电话会议内容，正如我们之前讨论的，我认为专家们往往带有一定程度的负面偏见。

And then on the expert side, if I have AI read a bunch of expert calls, as we talked about with expert calls, experts in my opinion, to be a little bit more negatively biased.

因此，如果让AI基于专家电话会议进行训练，我会担心训练数据中存在负面偏见。

So if I have the AI trained on expert calls, I worry about negative bias in the training data on AI.

那么，投资者在使用AI工具时，是如何看待这三种偏见的呢？

So how are investors thinking about kind of those three biases when they're using an AI tool?

是的，我认为这其实回到了上一个问题：投资者的技能如何随着时间推移而显现差异？

Yeah, I think this goes back probably the last question of like, where is the skill of an investor become differentiate over time?

实际上，我认为几乎每一个用于评估投资的数据集都天然带有偏见。

And I actually think like, look, bias is inherent in almost every dataset you can look at to evaluate an investment.

现在不同的是，你可以以过去难以全面做到的方式，综合 triangulate（交叉验证）这些带有各自偏见的多个信息源。

What's different now is you can triangulate multiple of these sources with their their biases in a way that was really hard to do as comprehensively.

所以我认为投资者的做法并不是试图完全消除所有偏见，而是意识到这些偏见普遍存在，他们正在采用越来越复杂的方式比较和对比不同观点，以找出争议所在，进而形成自己独立的判断。

And so I think what investors are doing is like rather than trying to oversaw for how to eliminate all bias, I think there's a recognition that they all have that, and they're doing increasingly sophisticated ways of comparing and contrasting sentiment perspective to see where the debates are and then forming their own independent view.

谁是对的，谁是错的？

Like, who's wrong and who's right?

管理层显然带有某种偏见。

Like, management obviously have a certain bias.

正如你所说，这些专家可能非常消极。

And as you said, these experts might be really negative.

但真相究竟在哪里？

But where do we think the the truth lies?

如果我们仅凭目前所看到的信息，该如何变得更聪明？

And how can we get smarter on that if we can't based on what we're looking at here?

我不确定这看起来怎么样。

I don't if that look.

这并不是一个直接的问题，但我觉得这就是我所看到的实际情况。

That wasn't a super direct answer question, but, like, I I think that's just what I see happening.

在过去，你能评估的信息来源更少，而且这些来源都有偏见。

It's like in a prior world, you had fewer sources you could evaluate on the time you had, and they had bias.

现在你可以评估更多的信息来源，而它们全都带有偏见。

Now you have more sources you can evaluate, all of whom are biased.

对吧？

Right?

但你在这不同来源和观点之间进行交叉验证的能力，远超以往任何时代。

But the triangulation you can do across these different sources and perspectives is infinitely higher than you could look for.

而我认为，这最终就是投资的本质。

And that's ultimately, I think, investment.

没有完美的答案。

There's no perfect answer.

让我最后问一下关于AI和专家电话的问题。

Hear let me let me end by asking AI and expert calls.

对吧？

Right?

AI极大地改变了专家访谈库尤其是专家访谈的使用场景。

AI really shifts the use case for especially expert call libraries, but also expert calls.

我们之前略过了两个方面，我来简单总结一下。

And the two which we've kind of hidden at, I'll just summarize.

第一，如果我想对一家公司进行100次专家访谈，深入挖掘，我是做不到的。

One, if I wanted to do 100 expert calls on a company, I just want to dive really deep, I can't.

我受限于自己的时间。

I'm limited by my own time.

我可以让一个AI代理担任提问者，完成这100次访谈。

I could have an AI agent serve as the questioner for 100 things and do that.

理论上，我完全可以实现这一点。

And I theoretically could have that happen.

这是第一点。

That's number one.

第二点，我无法阅读80份关于80家不同公司的专家访谈记录。

Number two, I can't read 80 expert call transcripts on 80 different companies.

AI可以。

AI can.

所以现在有两种方式，你可以更充分地利用专家通话库的转录内容，也可能根据需要获取更多专家通话。

So there's two ways that fundamentally now you can use more expert call library transcripts and maybe you can get more expert calls if you want to.

你如何看待AI与专家通话的共同演进？

How are you seeing AI and expert calls evolve together?

你如何看待那些在AI和专家通话使用上处于最前沿的客户？

And how are you seeing your customers who are at the far, far tail end of using AI and expert calls?

他们是如何将这两者结合的？

How are they marrying the two?

是的，我认为正如我们在对话开头提到的，专家通话始终是一种独特的情报来源，因为它们来自人类，视角多样，不会给你简单的是非答案。

Yeah, I think So ultimately, expert calls have always as we kind of in the first part of our conversation, they're a really unique source of insight because they're humans and they're varied and they're not gonna give you yes or no answers.

对吧？

Right?

你可以从他们那里挖掘出很多信息。

You can tease out a lot from them.

所以我认为，第一点，当我们构建业务和专家通话时，我们提到，这可能是市场上最具独特性的数据资产之一。

So I think the first thing, you know, when we talked at as we're building, like, the business and the expert called, what we were saying, like, this is probably one of the most unique data assets in the market.

你可以想象，所有可投资市场、研究标的和公司中所蕴含的专家知识量，这些知识都存在于平台之外。

It's like like, you can think about the amount of expert knowledge that sits out there on all the investable markets, research names, companies, and it's off platform.

它无法被任何其他方式提取出来。

It doesn't it's not you can't extract it anywhere.

因此，我认为AI基本上带来了这些商业模式的创新，扩大了可以进行的专家通话数量，以及能够捕获和搜索的信息量。

And so I think AI, basically, you know, we had these business model innovations that opened up the amount of expert calls that could be done and how much it could be captured and searched.

这可以说是TIGUS模型相对于传统模式的第一版。

That was, like, version one point o of, like, the TIGUS model versus traditional.

然后第二点，AI实际上是在这一趋势上更进一步，现在你可以拥有大量的AI。

Then two, what AI is basically doing is playing, like, further on that trend, which is, like, now you can have, you know, tons of AI.

如果下一个瓶颈是投资者亲自进行这些通话的时间，那么这个限制已经不复存在了。

If the the next gate was investor time to actually conduct those calls, like, that's no longer constraint.

对吧？

Right?

所以归根结底，这取决于有多少资源可以投入到这些事情上。

So ultimately, it's just the amount of resourcing available to go run at all these things.

所以，抽象地回答你的问题，我认为真正有趣的一点是，如今能够被捕捉、查询、纵向分析并跨时间比较的专家见解数量。

So I think to answer your question a little bit abstractly, I think one thing that's really interesting is the amount of expert insight out there that can be captured, queried, looked at longitudinally, and compared and contrasted over time.

这是一个每天都在不断积累的实时数据资产。

That is like a real time data asset that's building every day.

而几年前这还做不到。

And that wasn't true in a few years ago.

因此，我认为这就是一些投资者，尤其是大型投资者，正在真正意识到并认可我们的地方——他们自己也拥有这样的能力。

And so I think that's where I think investors some investors are really recognizing that and recognizing us, especially really large ones, that they have their own.

他们进行海量的专家访谈。

They do massive amounts of expert calls.

其中一些是横跨公开和私募市场的基金，他们正在将这些见解相互对比。

Some of them are crossover funds, public and private, and they're comparing insight against those.

于是，你现在突然拥有了两组不同的数据集。

So now you suddenly have two different datasets.

私募市场和公开市场中正在发生什么我们能看到的变化？

What's happening in private markets that we can see and public?

这正是帮助我们形成对不同标的信心的原因。

And that's how that helped shape our conviction on different names.

我认为，人工智能正在加速一个原本就在专家网络领域悄然兴起的趋势。

I think that's where AI is a accelerant of a trend that was already happening in the market around expert networks.

我认为投资者正将此视为最具独特性的数据资产之一，并希望从中分得一杯羹。

And I think investors are really seeing this as, like, one of the more unique data assets that is being built, and they want a stake in it.

他们还将其自身的专有信息融入其中，而这些是其他人无法接触到的。

And they also bring their own proprietary stuff that others can't see to it.

因此，我们看到的一个最大趋势是，许多投资者最初只是想找一个专家访谈文稿库和搜索工具。

So I think one of the biggest things we're seeing is a lot of investors initially were just, like, looking for an expert transcript library and AI tooling to search it.

但越来越多的投资者也开始将内部内容一并纳入。

Increasingly, they're also bringing their internal content alongside.

而这使得这一趋势对那些资源雄厚、开展大量工作的大型基金尤为重要，这也是市场中的一个重要趋势。

And that's where, you know, this is much more relevant, think, for larger well resourced funds that do a ton of work, but that's a big trend in the market.

他们能够看到别人看不到的东西，因为他们 across 各个不同团队在市场中进行了大量的研究。

Like, they're able to see things that others can't because of all the things that all the research that they're doing in the market across disparate teams.

这是一个比较轻松的问题，然后我还有两个问题。

Softball ish question, and then I'm I have two more questions.

这是一个比较轻松的问题，接下来我会提一个真正棘手的问题。

Softball ish question, and then I'm gonna end with a true knuckleball question.

这是一个比较轻松的问题。

Softball ish question.

我们重点关注了投资者以及公开与私募市场。

We focused on investors and public private.

AlphaSense 服务了大量公司。

AlphaSense does a lot of companies.

公司是否会进入专家访谈库，利用专家访谈来获取信息、思考并调整策略，或者仅仅通过观察投资者提出的问题来改变他们对投资者关系的回应方式？

Are companies going into the expert call library and using expert calls to source and think and change strategy or even just seeing the questions investors are asking to change kind of how they're responding to IR?

或者你也可以告诉我，伙计，那些公司本身就是专家。

Or you could also tell me, dude, the companies are the experts.

他们不需要去专家库。

They don't need to go to an expert library.

他们可以直接打电话给他们的供应链经理。

They can just call up their supply manager and have them.

所以我很好奇，你是否看到公司正在调整和适应专家通话和AI库的工作方式。

So I'm kinda curious if you're seeing companies kind of adjust and adapt to how both expert call and AI libraries work.

是的。

Yeah.

听好了，公司是我们业务的重要组成部分。

Look, companies are are like a big part of our business.

在AlphaSense，我们近40%的业务都来自企业并购、企业战略和投资者关系团队。

At AlphaSense, we almost 40% of our business is built on corp dev, corp strat, and IR teams.

因此，他们是与投资者所关注的相同洞察力的重度使用者。

And so they are huge consumers of the same, insight that investors look at.

他们的使用场景更加细致且略有不同，但我认为，这个行业已经从过去极度关注基本面投资者，转变为如今成为高端企业决策的核心组成部分。

Their use cases are nuanced and slightly different, but I think this went from an industry that was, you know, very focused on fundamental investors to now becoming very much a core part of how, you know, sophisticated corporate decision making is made.

是的

Yeah.

我只是想知道，他们是不是在用你们自己的系统。

I I I was just that's exactly what I was wondering if they're using your own.

好的。

Okay.

有个很冷门的问题，特别奇怪，但让我先说下背景。

Knuckleball question, super weird, but if I can give you the background.

在2011到2014年间，中国股市曾发生过大规模的反向并购欺诈事件。

In 2011 to 2014, there was this big Chinese reverse merger fraud in the stock market.

你会读这些公司的20F报告，上面写着：我们拥有六亿英亩的林地。

And it was you would read the 20F of these companies and they would say, hey, we have 600,000,000 acres of woodland in China.

人们就会想：一英亩林地值16亿美元，那总共就是600亿美元，这股票该买。

And people would say, well, an acre of woodlands worth $1.600000000 this is worth $600 it's a buy.

但事实上，中国根本不存在六亿英亩的林地。

Well, it turns out 600,000,000 acres of woodland doesn't even exist in China.

所有这些行为都是欺诈。

All these things were frauds.

我想知道，如果在人工智能时代，欺诈的规模和回报有所提升，会怎样？因为如果你是一家公司，在进行欺诈时，如果人工智能只是检测到你赚了1万美元，而没有人类去指出：‘天啊，你们的总部居然只是博卡莱顿的一个邮政信箱。’

I wonder if there is a return to If the scale and returns to fraud improves in an AI age, because if you're a company and you're running a fraud and you're getting to 10 ks and AI is just detecting it and they don't have that human who's going and saying, Dude, their headquarters is like a PO Box in Boca Raton.

现在，是的，这些信息确实写在10K报告的首页，但那个会亲自指出‘这支管理团队疯了’的人类，已经不存在了。

Now, yes, it's the front page of the 10 ks, but the human person who goes and says, This management team is out of their mind.

你认为人工智能会不会反而提高了欺诈的回报率，或者让那些真正恶劣的公司更容易得逞？

Do you think just AI kind of increases the return to fraud or the return to far left really nasty companies?

因为如果它们只是被归入这个庞大的量化人工智能池中，就更难被发现了。

Because if they just get bucketed into this big quantitative AI pool, it's kind of tougher for them to detect.

这说得通吗？

Does that make sense?

是的。

Yeah.

你能再说一遍，或者稍微重新表述一下吗？

Can you say it one more time or reframe just slightly for me?

我只是在想，我实际上是在考虑中国反向并购欺诈的问题。

I'm just wondering, I'm thinking about the Chinese reverse merger frauds is really what I'm thinking about.

如果我去AlphaSense，说：嘿，帮我找一下按资产价值来看被低估的公司。

If I went to AlphaSense and I said, hey, find undervalued companies on an asset value.

如果只是读取中国反向并购欺诈公司的数据，它会说这是股票中最有价值的。

If it was just reading the Chinese reverse merger fraud twentieth, it would say this is the best value in the stock.

其他所有类似公司，比如Woodland Acres，每英亩只卖1美元。

Every other peer with Woodland Acres trades for $1 per acre.

而这家公司的价格却是每英亩0.05美元。

This is trading for $05 per acre.

它还会建议我买入，对吧？

And it would be telling me buy, right?

当时市场上有很多这样的情况。

And there were a lot of these things out there.

我只是举了Woodland Acres的例子，但当时市场上有很多类似的情况。

I just use the woodland, but there were a lot of these things out there.

而且需要有人四处打听、暗中调查，许多投资者才意识到这些问题。

And it took somebody like kind of calling around going snooping and plenty of investors felt for these things.

但我好奇，四年之后，所有这些事情，如果由人来阅读，都会觉得这里有问题。

But I wonder if in four years, all these things that a human reading it would say, hey, there's something wrong here.

或者一个亲自飞过去查看的人会说，哦，这家价值40亿美元的公司，总部竟然在商场的三楼。

Or a human who like literally flies and says, oh, this $4,000,000,000 company, their headquarters is in the 3rd Floor of a mall.

这有点奇怪。

Is kind of weird.

而AI不会注意到这一点。

And AI wouldn't see that.

所以我怀疑，AI是否会提高欺诈的回报率，因为随着量化资金和量化手段的增多，这种人为的核查就会消失。

So I'm wondering if AI increases the returns to fraud because as you get more quantitative money and as you get more quantitative things, that kind of human check goes away.

是的。

Yeah.

关于这一点，我有两个想法。

So two thoughts on that.

有趣的问题。

Interesting question.

我认为，首先想到的是，讽刺的是，AI在反欺诈行业中被广泛应用。

I think, you know, first thing that comes to mind is, like, ironically, AI is being used a lot in the fraud detection industries.

对吧？

Right?

比如用来发现模式，识别出某些异常情况。

Like, to find patterns and and things that just indicate that something, like, is amiss.

因此，我认为AI本身并不必然成为加剧欺诈的推动力。

And so I'd say that I don't think there's anything inherent about AI that suggests it becomes an accelerant for the fraud that's possible.

因为我认为，如果使用得当，AI同样可以成为检测欺诈的有力工具，在其他行业中以非常独特而直接的方式发挥作用。

Like, because I think equally, it can be when used right, a pretty powerful weapon for detecting fraud and pretty and idiosyncratic and straightforward ways in other industries.

所以我认为，AI并没有什么限制，无法像你提到的例子那样，去查看视觉图像或卫星图像，这确实是个很好的观点。

So I don't think there's anything that, you know, like, stops it from looking at, you know, like, to your example there, like, go looking at visual imagery, satellite imagery of Great point.

对。

Yep.

对吧？

Right?

而且比如说，嘿。

And being like, hey.

有些东西与公司指引不一致。

There is something that is mismatched versus company guidance.

比如，我们可以从航运通道的图像中看到，实际交通量根本与公司指引不符。

Like, we can go see from imagery on shipping lanes that, like, the traffic is not even remotely what company guidance is.

我确实认为，这在帮助投资者分析那些过去必须靠实地考察或派人前往查看才能获取的信息方面，可能非常强大。

I actually think it could be powerful in helping investors parse those pieces that used to be required by having someone on the ground or go going and sending someone to go look.

另一方面，我要说的是，这其实回到了我们讨论的核心：就像AI最终在某些方面表现得超乎人类一样。

On the other hand, what I will say, and, like, this goes back to the core of what we're discussing is, like, just like AI ultimately does some things in a superhuman way.

我认为，AI最终擅长的是综合分析，发现那些极其细微的细节和关联点，这是人类无法复制的。

And I think ultimately that is synthesis and finding really discrete details and connection points in a way that's very like, humans cannot replicate what AI is able to do in that domain.

但另一方面，在任何类似投资建议的复杂决策上，AI绝对达不到产品经理或资深投资者的水平。

On the other hand, it is absolutely not at the level of a PM or sophisticated investor on anything that remotely looks like our investment recommendation.