20VC：我们如何促成 Fred Wilson、Benchmark 和 Index 投资 9400 万美元 | 为何 Robinhood 的战略是错误的 | 为何一对一沟通是扯淡，以及每位创始人对股权的误解 | 为何品味胜过 AI，但 AI 如何瓦解组织架构图——与 fomo 首席执行官 Paul Erlanger 对话

当你从零到一，再从一到十的时候，情况就完全不同了。

It changes when you're scaling from zero to one to one to 10.

在这一阶段——也就是我们最近这轮融资之前——从一到十的过程中，你最大的经验和反思是什么？

In that scale phase that pre this latest round, which we'll get to, but in that one to 10, what are your biggest lessons and reflections on that?

我们非常警惕招人太快这个问题。

Hiring too fast is something we're very, very vigilant of.

我看太多人在从一到十扩张时犯的最大错误，就是开始收购公司。

Some of the biggest mistakes I've seen in other people scaling from one to 10 is they start to acquire businesses.

一旦收购公司，你就没法对所有被带进来的人做面试，结果一下子给公司添了大量冗余。

And when you acquire businesses, you're not interviewing all the people that you're bringing over, so you end up just adding tons of bloat all of a sudden to your business.

我觉得这可能会带来大问题。

And I think that could be a huge issue.

这真有意思。

So It's so funny.

我觉得我们现在正处在最矛盾的时刻：大家都在用代币取代人。

I think we're in the biggest paradox moment ever where we're replacing everyone with tokens.

我们根本不需要再用工程师来取代所有人了，现在靠令牌就够了。

We don't need engineers anymore replacing everyone with tokens.

你去问每一个创始人，他们最大的问题是什么？

And then you speak to every single founder and you ask what's your biggest problem?

他们都会说：‘招人难啊。’

And they're like, Oh, hiring.

我就想问：到底哪个才是真问题？

I'm like, Which one is it?

我觉得关键在于，现在顶尖人才的价值比以往高太多了。

Well, I think how you reconcile it is the best people are just so much more valuable now.

就像你可以用ChatGPT画画，但还得有人把握整体创意方向。

Because it's like, you could use ChatGPT to make art, but you need to have like the creative direction behind it.

你的软件工程师现在是架构师，他们在做很了不起的事，只不过能用AI来处理那些原本由中级或初级工程师做的基础工作了。

Your software engineers are your architects, and they're doing amazing things, but now they can use AI to do the lower level things maybe a B tier or a lower level engineer would do.

那我们的团队是不是会变得小得多？

So do we just have dramatically smaller teams?

是的

Yeah.

我也这么觉得

I think so.

所以早期多给些股权也没关系

And which is why it's okay to give more equity early.

我们当时就是这么看的

That's kind of how we saw things.

那我们该怎么规划未来的团队结构呢？

How do we think about structuring the teams of the future then?

目前FOMO的组织结构非常扁平

So currently FOMO is extremely horizontal.

我们不开例会，也不做单独沟通

We don't have meetings, one on ones.

我们基本上没有层级

We don't really have any hierarchy.

大家现在都是自我汇报，我觉得等规模扩大到一定人数时，这种方式就得改改了，你

Everyone is kind of self reporting, and I think as we scale to a certain number that will have to change, you

觉得现在大概多少人？

will What number are you at today?

17人。

17.

到时候你打算

What will you

一年之后呢？

be in a year's time?

希望能在25以下。

Hopefully below 25.

好的。

Okay.

我们真的不会大规模扩充人手。

We are really not scaling head count.

听我说，也许情况会变，但目前我们真没觉得有必要。

Listen, maybe things change, but currently, we really don't see a need.

我们之前确实

We did have

工程这边有过瓶颈。

a bottleneck on our engineering side.

我们刚聘了两三个特别出色的工程师。

We just hired two to three incredible engineers.

优步和微软都对AI工具在工程中带来的效率提升提出了疑问，表示持怀疑态度。

Uber and Microsoft have both put question marks around the productivity gains that come from AI tooling and and engineering, saying they are questioning it.

你觉得这很蠢吗？还是你始终坚信不疑？

Do you think that's moronic and you unwaveringly see it?

还是你其实也觉得，虽然产出的代码多了，但速度并没加快？

Or do you actually say, Yeah, we get a load more code, but we're not faster?

我觉得效率确实提升了。

I think it's definitely faster.

这不仅仅是代码量变多了。

It's not just a lot more code.

就连彻底审查代码，也比写代码快得多。

It's a lot faster to thoroughly review even than write.

比如蒂娜，她是我们的资深前端工程师。

For example, Tina, she's a staff level front engineer of ours.

她开发了我们的信息流。

She built our feed.

她现在正在为我们新产品做滑块组件。

She is building sliders for our new product.

前端有好多这类小功能，本来得花不少时间去实现和熟悉。

There's like all these like small things on the front end that probably would take a while to implement and learn.

你得去YouTube上看视频，或者到处搜库，虽然她对很多东西都有经验，但有些小组件还是新的。

You have to go watch YouTube videos or go search to find libraries, and she has experience with a lot of this, but some of like the small components are new.

但如果你让AI来帮忙，它会给你一个大概的搭建思路。

But if you could ask AI to do it, they'll kind of give you an overview of how to build this thing.

它们甚至能帮你写好代码，然后Tina再回头重新整理和重写大部分内容，但有了对如何编写和思考的整体理解，哪怕是最厉害的工程师，学习速度也会快很多。

They'll even write the code for you, and then Tina will go back through and even restructure and rewrite most of the code, but having the framework of understanding how to write it, think just speeds up the learning process significantly, even for the best engineers.

产品推进速度，我跟你说过，我们刚把所有事情都停下，就为了赶在下周上线这个新产品。

The product velocity, like I told you that we just dropped everything to ship this new product we're shipping next week.

我们只用三周就做出了这个产品，而这个产品所包含的功能，相当于其他整个应用的全部内容。

We built this product in three weeks, and this product is basically what entire other apps, their entire product is.

我们的网页应用，一个月就做出来了。

We built our web app in one month.

有意思的是，保罗·格雷厄姆昨晚说，他现在问所有YC创业团队的一个新问题是：怎么用AI为你的产品增加护城河？

So funny, Paul Graham said last night, the new question that he asked all YC batch members is, How do we AI protectify your product?

怎么加入一些非AI的功能，来构建真正的竞争优势？

How do we put in non AI features that build defensibility?

我觉得，对你来说，社交关系网就是最稳固的那一个。

I think the social graph for you is unwaveringly one of those

没错。

Exactly.

我觉得这特别有意思。

Which I think is really interesting.

当我们回头看看之前提到的那些AI工具带来的赋能作用时，现在团队到底在用哪些工具呢？

When we go back to the enabling powers of AI that come from some of the tooling that we've mentioned there, What are the team using today?

这些都是核心代码吗？

Is this all core code?

是用Cursor吗？

Is this cursor?

是用Codex吗？

Is this codex?

我对这些工具的使用分布特别感兴趣。

I'm just fascinated by distribution of tooling.

是的。

Yeah.

我们内部有一项AI使用政策，确保只使用企业账户，防止敏感信息被上传等等。

So we have an internal AI policy to make sure that we're only using enterprise account, that there's not sensitive things being uploaded, etcetera.

我觉得这一点非常重要。

I think that's really important.

在这些规定下，我们的工程师大多在用云端代码辅助工具。

And then within those guidelines, most of our engineers are using cloud coding codecs.

这方面有变化吗？

Has that changed over time?

对我们来说，我觉得没什么变化。

I don't think so for us.

我们确实遇到一些问题，比如模型性能下降，最近云端代码的使用成本也涨得厉害，确实有不少压力。

I think we have seen some frictions, like these models degrade, and then the cloud code credits got really expensive recently, so there's definitely some frictions there.

你们对价格敏感吗？

How price sensitive are you?

完全不敏感。

Not at all.

因为我们工程师人数还不够多，目前还没对成本造成太大影响。

We don't have enough engineers that it's really, really hurting our bottom line yet.

等它开始大规模使用时，也许我们会设个使用限额，但现在不会。

Once it starts to do, maybe we'll have some kind of quota there, but no.

你觉得将来会到那个阶段吗？

Do We think encourage there's a time when it will?

会的。

Yeah.

我觉得这取决于我们规模扩得多大，还有……

I think depending on how big we get and how

目前来看，限制使用根本毫无意义，得不偿失。

much we use it, Currently, the trade off is it just makes no sense to limit it.

坦白说，AI这个行业的核心问题就一个：开发者薪资里会有多少比例花在令牌费用上？

For me, the core question on on AI bluntly is an industry, and it is determined by one question, which is, like, what percent of developer salaries will we see spent on tokens?

就拿马克·贝尼奥夫来说，他说他们去年在Anthropic上花了三亿美元。

Right now, you know, if you look at Mark Benioff, he said they spent 300,000,000 on Anthropic.

这差不多占了开发者薪资总额的3.8%。

That's about 3.8% of developer salaries spent on tokens.

如果一直维持在当前水平，花一万亿美金投资OpenAI和Anthropic就太贵了。

If it stays there, paying a trillion dollars for OpenAI and Anthropic is grossly overvalued.

如果涨到20%，也就是很多人预测的开发者薪资中有20%花在令牌上，那它们就值五万亿美元了。

If it goes to 20%, which is what many think is, 20% of dev salaries go stones, they're $5,000,000,000,000 companies.

你能想象自己把开发者薪资的20%花在令牌上吗？

Can you feasibly see yourself spending 20 percent of dev salaries on tokens?

当然能。

Definitely.

这得看令牌的具体价格。

I mean, depends on the price of the tokens.

我希望看到价格不断下降，主流模型能变得像大宗商品一样，越来越便宜，别出现价格串通的情况。

I hope that there's a race to the bottom and these major models are kind of commoditized and they get cheaper and there's not like price collusion.

是的。

Yeah.

所以希望随着时间推移，价格能越来越低。

So hopefully they get cheaper with time.

能源成本下降，计算资源也越来越便宜，这些东西自然就便宜了。

Energy gets cheaper, compute gets cheaper, and then these things get cheaper.

但就目前来看，绝对是这样。

But at the current state, absolutely.

我觉得20%完全在合理范围内。

I think 20% is definitely within reason.

我们刚才提到，真正厉害的人才是关键。

Going back to what we said there about like, hey, you just have the really great people.

在新的环境下，设计和工程人员的比例会有变化吗？

In terms of design to eng ratios, does that change in this new world?

现在我们只有一个设计师。

Right now we only have one designer.

但我确实觉得，设计会变得越来越重要，尤其是对那些拥有大量中级工程师、而这些工程师做的工作AI又很容易接手的大公司来说。

But yeah, I do think that design becomes more and more important, especially for like some of these bigger businesses that do have a lot more mid level engineers and doing tasks that AI can kind of take over pretty easily.

如果设计越来越重要，我们是不是该全力押注Figma？毕竟它才是艺术和创意真正发生的地方。

If design becomes more and more important, do we double down on Figma, and that is the stage where art and creativity is fundamentally performed?

或者像你之前说的，我们会进入一个追求速度和快速迭代的世界，直接做原型，用 Rapplet、Lovable 这类工具，快速把产品推出来？

Or to your point earlier, do we move to a world of speed and iterations where we just prototype it and we use other tooling, Rapplet and Lovable and you name it, to get fast product out the door?

我觉得是两者之间，但更偏向后者。

I think it's somewhere in between, but mostly the latter actually.

Figma 在这方面优势很大，因为人还是想要掌控感。

Figma has a huge advantage here because humans want some control.

比如在 FOMO 这个项目里，我们完全可以只靠一个大模型自动执行交易之类的操作。

So for example, on FOMO, we could just be an LLM, you execute trades, etc.

也许未来会有个界面让你这么做，但我还是想打开 Harry 的个人主页，看到他所有交易记录的清晰可视化展示，能通过图形界面追踪这些数据。

Maybe in the future we have an interface that allows you to do that, but I still want to go to Harry's profile, see in this beautiful view everywhere you've trade, and be able to track that through a graphical user interface.

有人在 Figma 上会说：我想要这个设计。

Someone on Figma will be like, I want this design.

它自动生成矢量文件，但你依然可以自己调整，想怎么改都行。

It generates you the vector file, and then you could still manipulate it and do whatever you want.

我觉得这种混合模式特别重要，因为人终究还是希望觉得自己能掌控一切。

And I think that's really important to have the hybrid, because humans still want to feel like they're in control.

至于Lovable，就难多了，因为他们没做出以用户为中心的软件。

So with Lovable, it's much harder because they haven't built the human centric software.

我觉得在上面叠加一个大语言模型要容易得多，尤其是现在这些主流模型都已经把它变得很普及了。

I think it's much easier to add the LLM on top, especially as it becomes commoditized by all these major models.

为什么自从Snap之后，就再没出现过大的社交公司呢？

Why have we not had a big social company since Snap?

这真的很难。

It's really hard.

消费类产品太难做了。

Consumer is so difficult.

一点小失误都可能带来致命后果。

Some small mistakes could be pretty existential.

比如Clubhouse。

Like for example, Clubhouse.

它当初可是开始火起来了。

That started to take off.

那时候发展得特别好。

That was doing really well.

疫情期间人人都在用，核心用户群特别喜欢它。

Everyone was using it during COVID, and they had a very core user base that loved them.

但后来他们请来一堆明星，反而把真正喜欢这个产品的核心用户给淹没了，那些明星的粉丝根本不在乎产品本身。

But then they started bringing on all these celebrities, and it overshadowed the core user base that actually would love the product with people that don't really care about the product.

我记得有次马克·安德森周日晚上在上面分享见解。

I remember when it was like Marc Andreessen just sharing wisdom on a Sunday evening.

没错。

Exactly.

那种幕后内容特别精彩，让人学到很多，完全自发，特别有魅力。

And it was the most amazing behind the scenes, but fascinating lesson from the back, and it was so spontaneous and cool.

从这件事里你能学到什么？

What do you learn from that?

关键是要找到平台本身天然的创作者。

That it's very important to find native creators to your platform.

所以他们没必要到处去拉其他平台的创作者，这些人本来就已经在那儿了。

So instead of going out and bringing on all these creators from other platforms, they already established that.

比如洛根·保罗，他最早是在Vine上火起来的。

Like Logan Paul, he got big on Vine for the first time.

我觉得每当出现一个新的社交平台，早期入驻的创作者会拥有巨大的战略优势，因为他们会成为这个平台的代表性人物。

And I think when there is a new social platform, there is an outsized strategic advantage for creators to build an audience on that platform early, because they'll be known as the creator of that platform.

我不是想让勒布朗·詹姆斯靠怕错过的心态来凑热闹。

So I'm not trying to get LeBron James to trade on FOMO.

我想要的是这些原生创作者。

I want these native creators.

听你这么说真有意思。

It's so interesting to hear.

就像查理·达梅利奥，你知道的，很明显

It's like Charlie D'Amelio, I think, you know, obviously

在TikTok上。

On TikTok.

在抖音上，关键是要培养内部支持者，不能把一个Instagram明星直接搬过来，就说‘嘿，这是Paul Erlanger的粉丝群’。

Be on TikTok where it's like the lesson there is you have to make internal champions and you can't bring an Instagram star to TikTok and say like, hey, Paul Erlanger audience.

真的特别有意思。

Really, really interesting.

没错。

Exactly.

B Real也是我参与过的一个例子。

B Real was another one that I was in.

从B Real身上，你学到什么经验了吗？

Is there any lessons for you from B Real?

因为那个平台也经历过Clubhouse那样的热潮，但没能持续下去。

Because that too had the clubhouse hype cycle that didn't sustain.

我觉得B Real缺少反馈机制。

I think B Real didn't have the feedback loop.

它要求用户每天都要参与，但没人愿意每天都得做点什么。

It required people to do something every day, and people don't want to have to do something every single day.

一旦失去这一点，势头就会迅速下滑。

And I think as soon as you lose that, you lose momentum very quickly.

有没有办法在用户旅程中人为地制造出这种势头？

Is there a way to synthetically create momentum within a user journey?

当然有。

Absolutely.

比如在FOMO上，最重要的功能之一就是分享卡片。

So for example, one of the most important things on FOMO are the share cards.

当我查看哈利的持仓时，我能看到你所有的持仓，还能把那些精美的分享卡片发到其他社交平台，或者直接发给朋友。

So if I go to Harry's positions, I can see all of your positions and share any of your positions of these beautiful share cards on any other social media platform or your fumbles.

假设你卖得太早，结果价格突然暴涨，我就能看到你错失了多少利润。这样一来，我就能在其他平台上公开分享你的操作，大家想实时看到这些动态，就会主动来FOMO，这就形成了一个内部的增长闭环——每当有人做出亮眼的交易，比如那个叫Iceman的，一夜之间把一万美元翻到了两百五十万美元。

So let's say you sold too early, and then the price rockets, then I could see how much you missed out on, and what this does is it creates this feedback loop where I can fully publicly share your things on other platforms, and then people want to see that in real time, so then they come to FOMO, and then you're building this growth feedback mechanism within the app, that every single time there's a top person that's having a top trade, whether it be this guy Iceman who turned he turned 10 k to $2,500,000 overnight in one night on FOMO.

还有个叫Remiss的，一个月就把三百美元做到了一百五十万美元。这些成绩都被公开分享到其他社交平台，自然就带来了大量关注。

There's another guy, Remiss, he turned I think it was $300 into 1,500,000 in a month, and then these are being publicly shared on other social media platforms, and then it's driving attention to our platform.

那你们怎么不干脆做一家媒体公司呢？

Why aren't you also a media company?

我这么说是因为，如果你为每个人制作精彩的短视频，他们可能并不希望这样，还想保持匿名。

And and the reason I say that is because if you were to do amazing Shorts with each of maybe they don't want it and they wanna stay anonymized.

但如果你做些关于300美元变成百万美元的爆款短视频，这绝对是TikTok上最炸裂的内容。

But if you were to do amazing Shorts on turning $300 into a million dollars, I mean, is the most viral crack content for TikTok.

你很有洞察力。

You have great intuition.

我们在FOMO正大力打造一个庞大的媒体部门。

We're building a huge media arm at FOMO.

目前它还是独立于产品之外的。

Right now it's external to the product.

我们现在正全力扶持内容创作者，跟很多主播合作，大量制作精彩片段之类的。

So we're doubling down on content creators, we're doing tons of partnerships with streamers, we're trying to do like a lot of the clipping content, etcetera.

我们希望成为全球科技公司里最大的媒体业务之一。

And we want to become one of the largest media businesses for a tech company in the world.

你刚才提到剪辑的时候，我们正在聊媒体的事。

You mentioned clippers there when we were talking about media.

这是一种全新的媒体形式。

It is a new form of media.

确实如此。

It is

一种占主导地位的媒体形式。

a dominant form of media.

你们最初是怎么入手的？

How have you approached that first?

现在社交媒体上的注意力机制就是这样，你不得不跟着这套规则走。

It's kind of a game you have to play because of how attention works on these social media platforms now.

你们怎么分配预算？

How do you budget for it?

你们怎么跟用户生成内容合作？

How do you work with UGC?

具体是怎么操作的？

What does that actually look like?

所以我们整个团队都内部运作。

So we actually have this all in house.

我们有专门的创作者经理。

We have these creator managers.

他们全都是公司内部员工，负责管理大约三四十位创作者。

They're fully in house, and we manage a group of 30 to 40 creators.

我们一直在淘汰表现差的，不断引入新人，同时重点扶持那些有潜力的。

We're constantly getting rid of the bad ones, adding new ones, and doubling down on bad

那什么样的创作者才算优秀？

creator, what makes a good creator?

说白了，就是个数字游戏。

It's honestly numbers game.

完全看他们的曝光量。

It's just based on their impressions.

就像做产品一样，很大程度上靠的是直觉，猜用户会喜欢什么。

So like when you're building a product, it's a lot about intuition, what your users will like.

说到增长，尤其是在这些平台上，完全看数据指标，比如他们带来的曝光量、转化量这些。

When it comes to like growth, especially on these platforms, it's just based on metrics, like how many impressions they're driving, how many conversions they're driving, etcetera.

那些表现不达标的，不管是按千次展示成本还是获客成本对比用户终身价值，只要比例不对，我们就会直接淘汰掉。

And the ones that whatever, the CPM or the CAC based on the acquisition cost versus the lifetime value of the user, if that ratio isn't right, then you'll just kinda churn out that How

你们怎么计算获客成本？

do you determine acquisition cost?

是按每次下载算，还是按用户充值并入金来算？

Is it on a per download basis, or is it on a per funds and deposited basis?

对我们来说，就是看收入。

It's revenue to us, right?

必须是那些充值并且交易过的人才算。

So it's someone who has to deposit and trade.

充值本身是免费的。

So depositing is free.

我们看的是，每个月从这些渠道直接带来的用户里，有多少人进行了交易，我们从中赚了多少钱，再对比我们为获取这些用户花了多少成本。

If you trade, and then we take the total amount that we earn on any given month, from I people who trade from those channels that are directly attributed, and then how much it costs us to get those users.

你有没有发现，哪些类型的创作者能成功，哪些不行？

Do you see commonalities in talent that works and that doesn't?

当然有，这是我学到的最重要的一课。

Absolutely, and this is the most important lesson that I've learned.

刚开始的时候，你做出一种内容形式，或者找到一个已经有效的创作者，然后你就想：好了，接下来找下一个能火的。

Early on, you create a form of content, or you find a creator that has a form of content, and they're working, and you're like, Okay, now figure out the next thing that works.

这种想法完全错了。

That is completely wrong.

你应该持续优化现有的内容，让它变得越来越好，然后复制这种模式，让这类内容不断重复出现。

What you want to do is continue to iterate on that and make until it works better and better and better, and then replicate, and just have that type of content being replicated.

我们还在培养这方面的能力。

This is something that we're still building up the muscle for.

用户更容易被转化，如果字体是这种风格，颜色是这种搭配，位置是这样安排，或者是由这个人出镜而不是那个人——你一旦摸清了这些规律，就该全力放大那些有效的做法。

Users are more likely to convert if this is the type of font, if this is the color of the font, if this is the placement of the font, if it's this person talking versus this person talking, and you figure out these things, and you kind of just double down on what works.

现在从用户生成内容的角度来搭建这个剪辑内容管理系统，你有什么反思可以分享给其他正在考虑做类似系统的创始人吗？

Any reflections now from UGC building this Clipper content management system that other founders should know if they're thinking about it?

所以我觉得这里有两点。

So So I guess there's two things here.

第一，得确保用户的终身价值真的值得投入；第二，不一定要总挑最容易的客户，我们现在正走在这么一条路上：有人的终身价值可能只有30块，你花8毛钱就能搞定；但另一个人你得花3块钱，因为他得看十次广告，而不是两次，才会转化。

One, need to make sure that the lifetime value of the user is actually worth it, but two, you don't always have to go for the lowest hanging fruit, and this is a journey that we're on now, where you have someone whose lifetime value might be let's say, 30, and you're only spending 80¢ on them, but then there's another user that you actually need to spend $3 on, because they need to see it 10 times instead of two times to actually convert.

只要客户获取成本低于终身价值，你就该适当提高CAC，哪怕LTV不变，也要去覆盖更广的用户群体。

You should actually start to increase your CAC, even if the LTV stays the same, to capture a larger and larger audience, as long as the CAC is lower than the LTV.

客户获取成本是上升还是下降？

Does CAC go up or down over time?

有些人说是上升。

Some people Up.

我觉得这背后有两股相反的力量。

Well, I guess there's opposing forces, right?

一股让成本下降，是你越来越懂怎么玩这个游戏，不断优化；另一股让成本上升，是因为每次新增的用户，通常都是剩下最难搞定的那批人。

The force that makes it go down is that you get better at the game and you iterate, but the force that makes it go up is that each incremental user, usually you get the lowest hanging fruit, right, to convert.

所以每多一个新用户，转化起来就越难，成本也就越高。

So each incremental user is harder to convert, so they cost more to convert.

而且因为品牌越来越多，你的品牌一旦成了默认选择或者排名第一，人们就会自然而然地找你。我们Air Wallachs投资的这家企业就是做足球球衣的，正好属于这个领域。

And then also it goes down because of brand proliferation, which is when you just become the default provider or the number one, inherently you just get people because you are the number one, and we're an investor in this business at Air Wallachs, and they're on the side of shirts, football shirts.

所以品牌营销就变得很重要了。

And so brand marketing comes into play.

品牌营销其实特别难理解。

Brand marketing is a very difficult one to understand.

品牌营销确实是最难搞的之一，因为你根本看不到直接的效果。

Brand marketing is actually one of the hardest things because you don't see the direct benefit.

它特别重要，但你甚至都说不清自己的获客成本到底是多少。

It's so important, but it's not like you don't really even know what the CAC is.

对吧？

Right?

你可能砸进去无穷多的钱，却连转化都看不到，

And you could spend infinite amounts of money and not even see conversion and

你会通过分析本地搜索结果，或者按账户追踪转化情况，看看用户是不是在阿森纳比赛的特定时段，身处北伦敦附近。

You try attribution by looking at hyper local search results, by looking at conversions on a per account basis and whether that was in the vicinity of North London where Arsenal played at a certain time of the game.

但这真的特别难。

But it's really freaking hard.

我常说的品牌营销，就是要找那些能长久存在的资产。

The thing I say actually on brand marketing is look for immortal assets.

比如说你赞助一个播客，一定要确保这个赞助是永久性的。

And what I mean by immortal assets is like, if you sponsor a podcast, make sure that the podcast has it in perpetuity.

就像我们之前和比尔·格利做的那一期，虽然录于三年前，现在每个月还是有成千上万的播放量。

Like, if you sponsored an episode that we did with Bill Gurley, it still gets thousands and thousands of plays per month even though it was recorded three years ago.

这价值可不小。

That's quite valuable.

如果是广告牌，两周就拆了，那价值就有限了。

If it's a billboard and in two weeks it's gone, it's not that valuable.

还有哪些类似的例子？

What are some other examples?

足球球衣。

A football shirt.

全世界都有孩子穿着十年前的曼联球衣，上面还印着沃达丰的标志。

There are kids all around the world wearing Man U shirts from ten years ago with a Vodafone logo on it.

能让那么多人在十年后还穿着印有你巨大logo的衣服，这价值真的很大。

That's pretty valuable to have people still wearing your massive logo in the thousands and thousands from ten years ago.

你明白我的意思了吗？

Do you see what I mean?

明白。

Yeah.

还有很多类似的资产是长久有效的，而不是转瞬即逝的。

And there's a lot of assets like that which are immortal versus very transient.

我接触过的每一个大创始人，包括Revolut的Nick，都说他们早期在营销上最大的错误，就是没有足够重视品牌营销。

Every single big founder I've had, Nick at Revolut included, said the single biggest mistake he made around marketing was he did not appreciate brand marketing enough early enough.

对。

Yeah.

当产品发展到一定阶段，从靠直觉做产品就变成了靠数据说话，因为你用户太多了。我特别喜欢Robinhood早期的一个故事：iOS用户的存款金额是安卓用户的两倍。

At a certain point, building a product transfers from a game of intuition to a numbers game, because you just have so many users that One of the early stories from Robinhood I love is that the deposit amounts from people with iOS were twice the ones from people with Android.

他们当时以为，只是因为用iPhone的人钱更多，所以存的钱也更多。

And their assumption was just, Oh, like people with iPhones just have more money, so they're depositing more money.

但通过数据他们发现，其实是安卓端的加载页面慢了整整一倍，导致用户等不及就流失了。

And what they realized through data is actually there was something on the loading screen where it took like twice as long to load for Android, so people were just churning off and not using it.

一旦解决了这个问题，两者的存款金额就拉平了。

And as soon as they fixed that, then the prices converged.

我觉得数据能揭示很多反直觉的现象，当用户规模大到一定程度，决策就真的得靠数据驱动了。

So I think there's a lot of unintuitive things that data can explain, and at a certain point when it becomes a numbers game, it really is data driven.

我们还没到那一步，但已经意识到这一点了。

And we haven't gotten there yet, but that's something we're aware of.

接下来有个令人振奋的消息，Index和USV领投了我们的B轮，这两位可是业内顶尖的投资人。

As we move forward, a very exciting announcement on Index and USV doing the Series B, two of the best investors in the game.

我得问一下，这事儿是怎么促成的？

I have to ask, how did that come about?

是的。

Yeah.

那时候我们其实并不急着融资，所以对这件事挺有选择性的。

So we weren't in a position where we necessarily need to raise capital, and we were pretty opportunistic about it.

A轮之后，我们收到一些主动联系，但没着急回应，先放了放。

I think after the A, we got some inbound and we took some time.

我们基本没怎么跟其他投资人接触。

We didn't really talk to investors.

那时候我们一心在做产品，只跟USV聊了挺长时间。

We were just building products, and we were talking with USV for actually a few months.

我们特别欣赏弗雷德。

We really liked Fred.

我们跟弗雷德聊了很多次。

We spent a lot of time with Fred.

他真的非常出色。

He's incredible.

很多人说风投都不懂产品，但弗雷德对产品有特别敏锐的直觉。

Talk about VCs not being focused on the product, but Fred actually has a really good product intuition.

我觉得对于风投来说，这种情况非常少见，和他交流的时候也一样

I think it's very rare for VC, and just the conversations with him and

这正好结合了他两大热情所在：去中心化网络和网络效应。

Well, this is like the combination of two of his biggest passions in decentralized networks and then network effects.

没错。

Exactly.

这简直正中靶心。

This is like, right in the mesh.

所以我觉得，除了那次聊天之外，这是另一个特别的时刻，但我们在A轮融资期间没能和Fred好好聊上。

So I think it was actually the one of the other times besides the chathing conversation where there was this moment, but we didn't get to speak with Fred during the Series A.

为什么没聊成？

Why not?

我也不太清楚具体原因。

I don't exactly know.

我猜Fred当时在出差，时间对不上，不过现在总算赶上了。

I think Fred was traveling, and the time just didn't match up, but but it worked out now.

对吧？

Right?

不对。

No.

不过我最讨厌我这份工作的一点就是这种假期。

This is what I fucking hate about my job though, which is like a holiday like that.

我不是说一定是Fred，但这种假期一来，动不动就可能让公司损失几亿美元。

I'm I'm not saying it was Fred, but, like, a holiday in general can lead to, like, hundreds of millions of dollars lost.

这都是机会成本。

It's all opportunity costs.

是的。

Yeah.

很难说清楚。

Hard to know.

我知道。

I know.

好吧。

Okay.

所以你见了他

And so you meet him

在这轮投资中。

for this round.

他来过公司。

So he comes to the office.

我们和他建立了关系。

We're building a relationship with him.

他当初作为投资人时，就帮了我们不少忙，我们真的很感激。

He was actually super helpful on a few things even when he was an investor, and we just really appreciate that.

后来也有一些其他投资人主动联系了我们，我还记得第一次和指数资本交谈的情景。

And then we got inbound from some other investors, and I remember the first time we talked with Index Ventures.

他们真的是特别棒的合作伙伴。

They were just amazing partners.

我觉得Benchmark的资源没那么多。

I think Benchmark doesn't really have as many resources.

这些多阶段基金给创始人提供的资源真的特别多。

These multi stage funds have so many resources for founders.

我们A轮融资的公告，所有公关都是自己做的。

The Series A announcement, we did all the PR in house.

那时候我们没人帮忙，什么事都得自己来，但我知道Index在Robinhood身上特别投入，USV在Coinbase身上也一样，更别说Julia和Jan两位合伙人简直太棒了，Fred在USV那边也是。

We didn't have anyone, any help with anything, and having know the history of Index being very involved in Robinhood, and USB being very involved in Coinbase, and then even aside from the fact, just Julia and Jan being such amazing partners, and same with Fred at USV.

我们觉得时机到了，就顺势把这轮融资做了。

We were opportunistic and thought it was time to do the rounds.

对我们来说，这时间点真是再好不过了。

It was really great timing for us.

我觉得这笔资金会让我们事半功倍，

I think the capital is gonna be high leverage for us to

这轮融资规模有多大？

How how big is the round?

我们这次融资七千五百万美元。

It's we're we're raising 75,000,000.

七千五百万美元。

75,000,000.

Index那次融了多少？

How much did Index do?

Index融了五千五百万。

Index did 55.

五千五百万。

55.

USV融了一千五百万。

And then USV did 15.

一千五百万。

15.

那估值是多少？

And the price?

估值是五亿五千万。

The price is 550,000,000

明白了。

Okay.

投后估值。

Post money.

你们是直接跟他们提了融资规模和估值吗？

Did you come to them with like, this is the round size and this is the price?

你们是主动提出

Did you kind of come

一起敲定的吗？

to it together?

在条款书出来之前很久，我们就已经讨论过什么估值区间对我们来说比较合适。

There was an early conversation far before the term sheet where we discussed what price range would make sense for us.

其实我们原本没打算做这一轮。

I think we weren't looking to do a round.

所以他们问，什么样的价格对你们来说比较有吸引力？

So they were like, what price would be interesting to you?

我们定了一个数字，我觉得这为对话提供了一个很好的起点。

And we set a number, and I think that kind of helped anchor the conversation.

当他们愿意接受这个价格时，我觉得这让整个讨论有了明确的方向，之后的交流也就特别顺畅。

And, yeah, when they were open ears to that, I think that, yeah, that anchored the conversation in a way where we had a great conversation from there.

在这样的融资轮次里，创始人常常会套现一部分股份。

In rounds like this, often, founders take secondaries.

你们有做二级市场转让吗？

Did you take secondaries?

有的。

Yeah.

我其实看过一集播客，讲的是谁来着？

I think I've watched the podcast actually with who was it?

是埃文·斯皮格尔。

It was Evan Spiegel.

他上过《CEO日记》这个播客，说过一句话，我们的目标是为未来十年制造一种错失恐惧感。

He was on the Diary of CEO, and one thing he said is that our goal is to build FOMO for the next decade.

他还提到，自己早期套现一部分股份，对他后来拒绝Meta的收购至关重要。

And he mentioned that him taking some secondaries early on was really important for him saying no to Meta, right, when they came to acquire him.

所以我觉得创始人套现是个好主意，只要别太过分就行。

So I think founder secondary is a good idea as long as it's not extravagant.

我们确实套了一点钱出来，但数额非常小。

We took a little bit off the table, but it very little amount.

根本没改变我们的生活方式。

Nothing that would really change our lifestyle in any way.

你们账上躺着七十五亿八千万美元，现在有哪些以前做不到的事可以做了？

When you have $7,580,000,000 bucks in the bank following around like this, what can you do now that you couldn't do before?

是的。

Yeah.

其实有几件事。

So there are a few things.

第一点，这显然进一步强化了我们之前谈到的关于市场周期的看法。

One, it obviously kind of helps us even more in the former point about market cycles.

我们的团队正在扩大规模。

Our team is scaling.

我的意思是，我们确实在赚钱。

I mean, we're making money.

我们账上的钱比过去融过的所有资金加起来还要多。

We have much more money in the bank than we ever raised.

其实没那么难。

It's that difficult actually.

Coinbase的布莱恩曾经提到过，

Brian at Coinbase has

在和我聊天时，他说过，他的业务面临的挑战就是波动性，它会影响团队文化和士气，尤其是在加密市场低迷的时候，真的很难熬。

said before the chat, I think he said on the show with me, is like the challenge of his business is just like volatility and how it impacts culture and morale in some cases where it's like, it's just hard when it's a fucking depressed crypto period.

没错。

Exactly.

你就说，没错。

And you're like Exactly.

所以啊，正因为这样，放手给团队这么多股权才重要，但同时有充足的资金，才能在低谷期保持士气——因为我们知道，不管多难，我们都有能力撑过去，不是在勉强维持，对吧？

And that's why, I mean, that's why giving away so much ownership is important, but also having this capital really keeps motivation high during those periods, because we know no matter what we can build through it, and it's not like we're scraping by, right?

我们现在的状态，是能真正去承担该承担的风险，因为有资金做后盾。

We're like building in a way where we're really being able to take the risk that we need to, because we have the capital to back them.

我们的计划是把所有基础设施都垂直整合起来，尽可能自己掌控，这样用户体验才能提升得更彻底。

Our plan is to kind of verticalize all of our infrastructure and own as much of it as we can in house, because it just makes the product experience that much better for our users.

你们怎么决定哪些东西要买，哪些要自己开发？

Do you determine what you buy versus build?

这是个非常好的问题。

It's a really good question.

凡是跟核心产品直接相关的——也就是用户能直接接触到的部分——都得自己做。

Anything that has to do with the core product, meaning the things the users face, you kind of have to build yourself.

比如有些竞争对手，直接买下交易终端，然后整合进自己的产品，但FOMO Web带来的体验完全是另一回事。

For example, some of our competitors, they just acquired trading terminals and then built that into their product, but FOMO Web is like a whole different experience.

你们拥有相同的社交关系网络。

You have the same social graph.

你可以用一个身份在平台上下单，再在手机上打开，这种基于网页的社交交易体验，以前从来没人做过。

You could place a trade there under one identity, open it on your phone, and it's this social trading experience on a web that no one has ever done before.

所以这一点很明显，我们非得自己做不可；但像数据基础设施这类东西，比如有人已经搭建了裸金属服务器，让我们不用花那么多钱在AWS或谷歌云上，或者有人帮我们做数据索引——这些后台工作，我们要自己积累相关能力得花很长时间，这种情况下，直接收购进来显然更合理。

So that's something that was very obvious that we just have to build this, but then something like data infrastructure, maybe someone who's set up bare metal servers so we don't have to pay so much money to AWS or Google Cloud, or someone who does indexing for us, or these things that are behind the scenes that would take so long for us to build up the expertise to, that's something that obviously makes sense to to kinda acquire to bring in house.

美国人总爱贬低欧洲，但看看数据就知道，现在反而是欧洲在金融科技领域碾压美国。

America likes to shit on Europe, but when we look at the numbers, right now, is kind of shitting on America when it comes to fintech.

我们的金融科技服务商Revolut大放异彩，下一轮估值将远超Robinhood。

It's our fintech provider, Revolut, which has stolen the show, and on the next round will be considerably more valuable than Robinhood.

为什么美国没人做出欧洲已经做到的东西？

Why has no one in The US built something that Europe has?

是的。

Yeah.

我觉得确实如此。

I think it's true.

我的意思是，你听我说。

I mean, listen.

每个国家的经纪法规都不一样，欧洲国家多得多，而Revolut能一下子覆盖整个欧洲，因此比Robinhood增长快得多。

Because every country has different brokerage laws, Europe is a lot larger in terms of like there's so many countries in Europe, and I think Revolut has been able to kind of saturate all of Europe at once, and has been able to grow globally a lot faster than Robinhood.

Robinhood做得很好。

Robinhood has done a great job.

他们在美国有两千多万个已充值账户，我们之前还聊到，他们还拓展了其他产品线。

I think they have a little over 20,000,000 funded accounts in The United States, and we were discussing before how they've horizontally grown to other product categories.

他们最大的举措是推出链上代币化股票，让这些产品真正走向全球，我觉得这特别重要。

Their biggest move with on chain tokenized equities is to be able to finally distribute these things globally, and I think it's really important.

看看Facebook、WhatsApp，所有那些迅速崛起的社交产品，从第一天起就是全球化的，这一点非常关键。

Look at Facebook, look at WhatsApp, like all of these social businesses that really scaled are global from day one, and I think that's really, really important.

我不确定这是否直接回答了你关于欧洲和美国的问题，但我觉得关键在于你能覆盖更多用户。

I So don't know if it necessarily answers your question of Europe versus US, but I think the reason is your ability to saturate a larger group of people.

你们到底是一家社交公司，还是一家金融公司？

Are you a social company, or are you a financial company?

我们一开始就是做交易的。

We're trading at first.

我觉得这一点特别重要。

I think that's really important.

社交功能的作用，至少在当前形态下，是让你通过看清顶尖交易者在做什么，来提升自己的交易水平。

What the social features do is they, at least right now in its current form, is they allow you to become a better trader by having transparency to what the best people are doing.

你可以立刻发现、立刻收到提醒，还能关注别人。

So you could instantly discover, you could instantly get notified, you could follow people.

我觉得随着时间推移，你得想办法让人愿意持续使用这个应用，哪怕他们并不天天交易。

I do think over time, you want to create momentum for people to use the app, maybe even if they're not trading.

你不能把应用的核心优势藏起来。

You don't want to obfuscate what the app is good at.

我们得始终做最好的交易应用，不然顶尖交易员根本不会用我们。

We have to always be the best trading app in the because the top traders won't use us otherwise.

不过长远来看，如果你做出了这些社交功能，那些不天天交易的人也可能更愿意在平台上互动。

However, over time, if you build these other social products, maybe the people who aren't necessarily trading every day can interact on the platform more.

所以我觉得，我们现在还是更侧重交易功能，但正在朝着更社交化的方向发展。

So I'd say we're much heavier on the trading side today, but heading in the direction of becoming more social.

老兄，我想跟你来个快速问答。

Dude, I want to do a quick fire round with you.

我跟你聊一整天都没问题，但我先说一句，你马上告诉我你的第一反应。

I could talk to you all day, but I say a short statement, you give me your immediate thoughts.

这样可以吗？

Does that sound okay?

好，来吧。

Okay, let's do it.

过去十二个月里，你对什么事情改变了看法？

So what have you changed your mind on in the last twelve months?

你觉得社交功能在跟风心理中的重要性如何？

How important social is on FOMO?

我们之前一直专注于优化交易产品，觉得只要交易体验够好，用户自然会来，社交关系也会随之建立。

I think we were doubling down the trading product and assume that people will come to trade, and then the social graph will grow from there.

但你得非常有意识地去经营这个社交网络。

But you need to be very intentional about the social graph.

我觉得这事儿特别依赖势头，一旦势头一掉，用户就不用了，整个东西就

And I think that's something that is very momentum based, and as soon as you start losing momentum there, people stop using the app, and then the whole thing kind

开始崩盘。

of unravel.

Revolut 和 Robinhood 对比。

Revolut versus Robinhood.

我是 Robinhood 的用户，因为我在美国，所以我选 Robinhood。

I'm a Robinhood user because I'm in The US, so I'm gonna say Robinhood.

但我认为 Revolut 的定位其实非常好。

But I do think Revolut is very well positioned.

也许是 Revolut 更有优势。

Maybe Revolut.

对现在大学里学计算机科学的学生，你最大的建议是什么？

Biggest advice to someone studying computer science at universities today?

少用点人工智能。

Use less AI.

少用点人工智能。

Use less AI.

我觉得应该少用AI，因为将来AI触手可及，但今天最顶尖的工程师都是在没有AI的情况下练出来的，才变得特别厉害。

I think use less AI, because you're gonna have AI at your disposal, and all of the best engineers today had to learn not using AI to become really, really good.

工作中你肯定会用AI，但我觉得这个观点可能有点争议。

And I think that when you're in practice in your job, you're gonna use AI, but I think this is a controversial take.

我尽量少让AI帮我写东西，因为如果全靠AI，我迟早会失去写作能力，也记不清什么才是好文字了。

I try to use AI for as little of my writing as possible because I think if I use AI for all my writing, I'm not gonna be able to write anymore, and I'm not gonna be able to remember what is good writing.

说实话，我对这点挺担心的。

I'm pretty scared of that, to be honest.

我同意你的看法，

I agree with

尤其是发社交媒体内容的时候。

you, especially on social posts.

我完全同意你的看法。

I completely agree with you.

我有些团队成员用那个叫ChattyPuji的东西发社交媒体帖子，我觉得简直糟透了。

Some of my team were using, you know, chattypuji for social posts, and I was just like, it's shit.

我一眼就能看出来。

And I can tell.

一点幽默感都没有。

There's no humor.

完全没有个性。

There's no personality.

干巴巴的，毫无质感。

There's no texture to it.

但说到工程师写代码，那又是另一回事？

But I'm like, with engineers and with coding?

去他的。

Fuck.

如果我给计算机专业的学生建议，我会说尽量多用，越频繁越好，因为你真的会发现

If I'm advising CS students, use it as much as freaking possible because you're like

我不是工程师，所以别听我的意见。

I'm not an engineer, so don't take my advice on that one.

说来听听。

Tell me.

那真的特别有意思。

That that that's very funny.

你最想拿到但还没拿到的投资方是哪家？

What investor do you not have that you would most like to have?

Ribbit Capital。

Ribbit Capital.

你们找他们谈过吗？

Have you pitched them?

我们跟他们聊过。

We spoke with them.

是的

Yeah.

我真的很

I really

喜欢米奇和整个团队。

like Mickey and the team.

他们非常出色。

They're they're great.

用错失恐惧感去赞助任何体育队伍。

Sponsor any sports team with FOMO.

你最想赞助哪支体育队伍？

What sports team do you want?

纽约尼克斯队。

The New York Knicks.

这根本不用想。

Not even a question.

如果你重新开始做FOMO，给自己一条建议吧。

You can give one piece of advice to yourself starting FOMO again.

如果现在的你了解一切，你会对过去的自己说什么？

What would you tell yourself if you knew everything you know now?

早点谈那些难开口的事。

Have the hard conversations sooner.

我觉得当CEO最难的，就是得谈这些棘手的对话——不管是跟员工、早期投资人，还是帮你的朋友，都一样。

I think the hardest thing about being a CEO is having hard conversations, whether that's with employees, whether that's with early investors, whether that's with friends who are helping you, all of these things.

很多人躲着不谈，怕冲突，因为确实不舒服，但越早谈越好。

And I think that people try to avoid confrontation, avoid having the hard conversations, because it sucks, but have those sooner.

直接坦诚相待，真心实意地沟通，对方是会理解的。

Just be completely transparent and honest and come from the best place you can, and they'll understand.

很多时候，你觉得难开口的对话，其实没你想的那么糟。

Often when you have a hard conversation, it's not as bad as you think.

谈完反而有种成就感，之后你还会主动去谈更多难事，因为你发现：也没那么难。

You feel a sense of accomplishment, and then you actually take on more hard things because you're like, it wasn't as hard as I thought.

这会引发一连串主动迎接挑战的连锁反应。

It leads to, like, a domino effect of taking on hard things.

你明白我的意思吧？

Do you know what I mean?

我也是后来才意识到这一点的。

That's exactly what I realized too.

九九六。

Nine nine six.

你怎么看这种拼命工作、没日没夜的文化？

How do you feel about this grind, slot, hustle culture?

在FOMO，我们从不设什么硬性工时标准。

I think at FOMO, there's no we don't put numbers to it.

我们不搞九九六。

There's no 996.

我觉得这可能跟我们公司的特点有关——如果你给团队充分的自主权，而他们又真的愿意承担责任，这些人就会像对待自己的事业一样全心投入，每个人都觉得自己是FOMO的主人。

I think if you give and this might be unique to our business, but if you give a lot of ownership to a team that takes a lot of ownership, like these people are fully autonomous and they really care about what they're doing, they all feel like they're owners of FOMO.

他们觉得自己是企业的主人，所以我相信他们会尽全力做好工作。

They feel like they're owners of the business, so I trust that they're going to do their best work.

有时候大家周末不用上班，但有时候像这周我们在开发新产品，整个周末都在忙，可大家还是乐在其中。

And there are times where people don't have to work on the weekends, and then there are times where we're building a new product like this week and we're working all weekend, but people love to do it.

我觉得没有什么比一支真正享受工作的团队更厉害了，我们就是特别热爱自己做的事。

I think you can't beat a team that's having fun, and we're just loving what we're doing.

我总是处于待命状态，有什么事发生就得立刻处理，但我们真的打心底喜欢这份工作。

So I'm always on in the sense that if something comes up, I have to work on it, but we love what we do.

你最大的优势是什么？同时，最大的弱点又是什么？

What's your greatest strength, but also your greatest weakness?

我有一些坚定但不僵化的信念，有时候这也有副作用——该做决定的时候，你得果断拍板，但我始终愿意倾听。

I have loosely held strong beliefs in the sense that sometimes the downside of it is sometimes you just need to make the decision, but I always listen.

每回FOMO要做决定，我们都会问每一个人，尤其是核心团队，听听他们的想法，大家一起讨论，直到找到最对的方案。

Like, whenever there's a decision at FOMO, we're asking every single person, especially the core team, what they think about it, and we're talking it out and coming to the right conclusion.

我觉得这绝对是优势多于劣势，因为它逼着我们始终去认真对待另一方的观点。

I think it's definitely more of a strength than a weakness, because it's a forcing function for us to strongman the other side always.

我们总是会想，要是换种方式做会怎么样？

We're always like, What if we did it this way?

我有时候的想法跟你正好相反。

Or I have the opinion that's inverse to yours.

咱们好好聊聊吧。

Let's talk it out.

但我确实觉得，有时候你得干脆做个决定，别老纠结细节。

But I do think that sometimes you need to just make a decision and we get hung up on things.

特朗普让美国的商业环境变好了吗？

Has Trump made business better in The US?

我觉得监管方面确实有不少积极进展，但与此同时，普通大众对加密货币的信心也明显下滑了。

I think there's been a lot of positive movement on regulatory clarity, but at the same time, I think that sentiment for crypto has also gone down significantly from the general public.

是什么导致了这种对加密货币信心的下降？

What has driven that crypto sentiment downgrade?

FOMO的目标，虽然听起来有点老套，就是要成为一盏明灯。

The goal for FOMO, as cheesy as it sounds, is to be a beacon of light.

看起来加密货币圈里的其他人总是只盯着短期利益，我觉得任何新兴行业刚开始都是这样——满是短期套利的人，趁机捞一把。总得有人拿扫帚把这摊子乱局收拾干净。我们的目标是做出真正为交易者着想的产品，让他们喜欢，而且能长期用下去。现在大家对这类产品印象这么差，主要是因为缺乏消费者保护，很多人根本不知道自己买的是什么就冲进去了。

It seems like everyone else in crypto has always been so short term focused, and I think in any early industry you go to, the industry at first is kind of like riddled with like the short term gains, and the people who are like taking advantage for themselves, and then someone's got to come with the broom and sweep up the mess, and our goal is to create a product that's for our traders, and something that they'll love, and something that's with them for the long term, and I think what's given such a bad rep is there's no consumer protection here on a lot of these products, so people will buy it.

本来做这件事就很难，等他们终于跨过第一步，结果钱全亏光了。

It's so hard to do in the first place, but then they get over the leap of actually doing the thing, and then they just lose all their money.

不是因为他们不知道这币可能跌，而是根本就是个骗局币，他们本来想买的是正经项目，结果买错了。

And it's not because they didn't know that this coin might go down, it's because it was a scam coin, and there was actually a real one they were trying to buy, but they bought the wrong one.

连个预警都没有。

There was no warning signal.

人们接连被坑了好多次，这才积累出这么多负面偏见。

And I think people just got burned so many times in a row that it led to all this negative dogma.

这叫三层安提克SPV。

That's called a triple layered anthropic SPV.

没错，就是这样。

Yeah, there you go.

Harry，这就是为什么永续合约这么重要——有了永续合约，你根本不需要直接交易底层资产。

And this, Harry, is why perps are so important, because with the perp, you don't actually need to trade the underlying thing.

所以我们能做一个关于Anthropic股价涨跌的永续合约，但根本不需要实际转移Anthropic的底层资产。

So like we could have a perp on whether anthropic will go up or down, but we don't need to actually transfer the underlying anthropic asset.

只要Anthropic股价变动，所有其他资产的转移都无关紧要，因为我们只是在赌它的价格走势。

So if anthropic goes, all secondary asset transfers are null, it doesn't matter because we're just betting on the price of the thing.

你根本不用跟我实际交换任何底层资产。

You're never actually exchanging the underlying thing with me.

这样是不是就不用SPV了？

Does that remove SPVs?

是的。

Yeah.

做永续合约根本不需要SPV。

You don't need an SPV for a perp.

对。

Yeah.

但你想想，接下来你得面对Anthropic、OpenAI和SpaceX这些公司了，对吧。

How do you expect and the because you're gonna now have Anthropic, OpenAI, and SpaceX Yeah.

我想这些衍生品会出现在平台上吧。

Where you'll have perps on the platform, I guess.

真想知道接下来会发生什么。

It's gonna be really interesting to see what happens.

我觉得零售市场的需求终究是有限的。

I think there's only so much demand in the retail markets.

我觉得SpaceX率先上市，肯定会非常有意思。

I think SpaceX being first to market, it's going be very interesting.

现在大概有三成资金是来自散户的。

I think like 30% is devoted to retail now.

如果散户在SpaceX、Anthropic或者OpenAI上亏了钱，还会有那么多人感兴趣吗？

Will there still be a ton of retail interests if people get burned on SpaceX in Anthropic or OpenAI?

如果SpaceX表现好，OpenAI和Anthropic的IPO也会跟着顺风顺水。

Think if SpaceX performs well, the OpenAI anthropic IPOs will also go very well.

如果SpaceX表现差，那另外两家恐怕就很难了。

I think if SpaceX performs poorly, then those are gonna have a hard time.

最后一个问题给你。

Final one for you.

别人对你做过最温暖的事是什么？

What's the kindest thing that anyone's ever done for you?

说实话，真正让我感激的是父母给了我一切。

I mean, the true answer is like my parents giving me everything I have.

我觉得很多人都有类似的经历，我真的非常感激他们，所有的一切都亏欠他们。

That's like, feel like maybe everyone kind of has a similar thing there, but yeah, do really owe them everything.

我成长过程中有些阶段挺艰难的，但我爸那时候积蓄很少，却拼命工作，供我上大学。

I think that there were parts of my upbringing that were tough, but like my dad, for example, didn't have much savings and, like, worked his way to help pay for my college.

那是我这辈子见过最了不起的事之一，我的目标就是好好回报他们。

And, like, that was, like, one of the most incredible things anyone's ever done for me, and my goal is to just continue to give back to them.

不过说到底，还是父母给了我今天拥有的一切。

But, yeah, I think, like, my probably my parents giving me everything I've had today.

所以

So

我妈妈也完全一样。

My mother is absolutely the same.

我妈妈教我，重要的不是你说什么。

My mother taught me it's not what you say.

也不是你做什么。

It's not what you do.

而是你让人感觉如何。

It's how you make people feel that matters.

没错。

Absolutely.

我总是说，赶紧去告诉父母你有多爱他们，因为总有一天他们会离开，到那时你会后悔没早点表达。

I always say, like, cool your parents up and tell them how much you love them because there's a time when they won't be there, and you'll regret not making that cool.

正是如此。

Exactly.

老兄，这真是一次特别愉快的交谈。

Dude, this has been such a pleasure.

非常感谢你让我参与这段旅程。

Thank you so much for letting me be a part of the journey.

非常感谢你的到来。

Thank you so much for coming.

面对面交流真的特别好，你表现得非常棒。

It is so good to do in person, and you've been fantastic.

谢谢你邀请我参加。

Thanks for having me on.

这是一次很愉快的经历。

It was a pleasure.

但在结束之前，你有没有在工作中觉得卡住了的时候？

But before we leave you today, have you ever felt a little stuck at work?

比如连续不断的会议、缺乏明确反馈，或者正在学一个新的AI工作流程？

Maybe you're stuck in back to back meetings, stuck without clear feedback, or trying to learn a new AI workflow.

现在，你可以用Loom来推进工作了，这是Atlassian推出的AI优先视频平台。

Well, now you can move work forward with Loom, the AI first video platform from Atlassian.

用 Loom 同时录制屏幕和你自己，快速更新进度、提供反馈、培训同事，各种场景都适用。

Use Loom to simultaneously record both your screen and yourself to share a quick update, provide feedback, train a colleague, and everything in between.

AI 会自动生成标题、摘要，并帮你剪辑视频，让信息传递更清晰。

AI instantly adds a title, a summary, and edits to your video so clarity travels with the work.

这样一来，你就能腾出更多专注时间，减少没必要的会议，让剩下的会议更高效。

Finally, you'll have more focus time, fewer unnecessary meetings, and more effective ones.

这真正改变了团队协作方式，也极大提升了个人工作效率。

It's a real team changer and game changer for personal productivity.

现在就去 loom.com 试试 Loom 吧。

Try Loom today at loom.com.

Loom 把信息传达清楚后，Finn 就接着跟客户继续对话。

Once Loom gets the message across, Finn keeps the conversation going with customers.

随着 AI 代理在客户服务中越来越普遍，团队常常得用一堆孤立的工具来应付各种任务。

As AI agents become more common in customer experience, teams often end up juggling multiple silo tools for every job.

Finn 就是为解决这个问题而生的。

Well, Finn was built to change that.

它是一个统一的智能代理，贯穿客户服务、销售、成功体验等整个客户旅程。

It's a single unified agent that works across your entire customer experience from service to sales to success and beyond.

Finn 这个代理正在为成千上万的客户打造完美的体验。

Finn is the agent making perfect customer experiences possible for thousands of customers.

它基于定制模型训练而成，整合了多年真实的客户互动数据，因此比其他任何代理都更懂客户服务的细节与复杂性。

It's powered by custom models, trained on years of real customer interactions, so it understands the nuance and complexity of customer service better than any other agent.

这意味着问题解决更快、服务更稳定，每位客户都能获得更好的体验。

That means faster resolutions, more consistent support, and just better experiences for every customer.

它还设计为完全自主管理，随着业务发展，你可以轻松优化和调整它。

It's also designed to be fully self manageable so you can easily improve and adapt it as your business evolves.

不需要任何第三方工具。

No third party's required.

Gamma、Asana、DoorDash 和 Crypto.com 等领先企业已经在使用 Finn，并且非常满意它的表现。

Leading companies like Gamma, Asana, DoorDash, and crypto.com already use and love Fin to deliver better customer experiences.

所以，限时优惠：前三个月，每月赠送 500 美元的 Finn 积分。

So for a limited time, you can get $500 a month in Finn credits for your first three months.