由于AI代理的出现，比特币即将进入抛物线式上涨 | Jordi Visser

红区会非常繁忙，因为AI智能体会完成最后那部分，就是五个人、十个人才能完成的部分。

The red zone is gonna be very busy because AI agents are gonna finish that last part, that five humans, 10 humans.

他们分享了信息，但并没有传播开来。

They shared information, but it wasn't viral.

并不是全世界每个人都知道。

It wasn't everyone on the planet.

所以并不是说，如果你把160智商的人关在一个房间里，一万个这样的人，他们就能解决问题。

So it's not that human beings if you stuck a 160 IQ people in a room, 10,000 of them, they'd be able to solve the problem.

问题是我们没有这样的人。

The problem is we don't have that.

所以Eli Lilly真正想明白的是，这意味着所有那些沉睡的知识产权，那些生物技术公司，他们都有人们投资过的东西。

So where Eli Lilly figured this out is really that means all of this dormant IP that exists, all these biotech companies, they've got something that people invested in.

让我们看看他们所有的研究，哪些公开信息是有价值的，然后我们把它完成。

Let's go through all of their stuff and see which public information on their research is worth something, and let's finish it.

让我们努力找到治愈癌症的方法。

Let's let's get a cancer cure.

咱们来解决胆固醇的问题吧。

Let's get something for cholesterol.

这就是为什么我说我们正在进入一个时代，人类解决问题、人类推动增长的那些统计数据已经结束了。

This is why I say we're entering a period where the statistics of what happened that human beings solve problems, that human being drove growth, it's over.

所以GDP不能像长寿那样，延长寿命，没有人死于癌症，这对GDP有什么影响？

So GDP cannot like longevity, increasing longevity, nobody dying from cancer, how does that impact GDP?

你得好好想想才能明白。

You have to think really hard to go through.

如果每个人都能健康地多活七十五年，那意味着什么？

If everyone lives another seventy five years healthy, what does that mean?

所以我要再强调一次，确保大家听到这一点。

So I'm gonna bring it back one more time and make sure people hear this.

这就是我为什么用股市作为例子。

This the reason why I use the stock market.

这就是为什么当你读到斯坦·德鲁肯米勒说，世界上最优秀的经济学家是股市。

This is the reason why when you read Stan Druckenmiller and he says, the best economist in the world's stock market.

我也这么认为。

I believe that too.

利率和油价都涨了。

Rates in oil are higher.

这对股市不利。

This is bad for the stock market.

自从这两样涨了之后，股市就一直在跌。

The stock market is down since those two went higher.

但你知道部分原因是什么吗？

But you know what part of it is?

凡是跟建房相关的，凡是跟日常消费品相关的，你去看的话，我周末会展示这些。

It's anything related to homebuilding, anything related to consumer staples, anything if you go through and you look, and I'll be showing that this weekend.

好的。

Okay.

各位，他们的盈利都在下滑。

Guys, their earnings are down.

问题是，这并不会改变美光的盈利。

The problem is it doesn't change Micron's earnings.

也不会改变谷歌的盈利。

It doesn't change Google's earnings.

AIQ，那个AI ETF，今天早上我看了一下，今年到现在涨了25%。

AIQ, the, AI ETF is up, I think, looked this morning, 25% year to date.

所以问题是，油价和利率确实有影响，但它们影响的是之前就存在的经济状况。

And that so the problem is oil and rates are having an impact, but they're having an impact on the economy that existed before.

所以当你听一个哈佛毕业的经济学家讲话，他受过很好的训练，也研究过历史。

So when you're listening to an economist who went to Harvard and he was trained really well, he studied history.

你会告诉那个花了这么多钱和时间，多年来一直被夸聪明的人吗？

Are you gonna tell that person that spent all this money and all this time and has been told for years how brilliant they are?

然后突然之间，你要说，嘿。

All And of a sudden, you're gonna say, hey.

你学的一切都没用了。

Everything you learned is worthless.

那当然不会。

Of course not.

好吧，我在网上说这个已经说了很久了，但他们就是不听。

Well, I've been saying that on the Internet for a long time, and they they're not listening.

不过确实是这样。

But yeah.

但关键就在于，幸运的是，我父亲——我以前说过这个。

But that's the whole point is that luckily, my father I've said this before.

我父亲在我很小的时候就告诉我，不要相信任何地方听到的任何话。

My father told me at a young age, don't believe anything you hear from anywhere.

要根据贝叶斯哲学来调整你对未来预测的把握。

Adjust your odds on predictions of the future based on a Bayesian philosophy.

因为总有新信息出现。

There's new information.

问题是，根据我的经验，和非常聪明的人共事过之后，我发现受教育程度越高，人就越固执，越不愿意改变自己的观点。

The problem is, in my experience, having worked with very smart people, the more educated, the more stubborn and less likely to change your views.

这是加密货币最大的负面问题，因为所有加密货币粉丝们，谁控制着世界的货币？

That has been the number one biggest negative with for crypto because for all you crypto fans out there, who controls the world's money?

世界的货币大部分由……嗯……

The world's money is controlled for the most part Mhmm.

由最受过教育的人控制。

By the most educated people.

所以为什么要相信一个你不需要用的东西，比如迈克尔·塞勒的币？

So why believe in something that you don't need to use Michael Saylor's coin?

比特币最终会找到你。

Bitcoin finds you eventually.

本期节目由Lava赞助播出。

This episode is brought to you by Lava.

Lava卡是第一张让你每次购物都能获得高达5%比特币返现的信用卡。

The Lava Card is the first credit card that allows you to earn up to 5% back in Bitcoin on every single purchase.

大多数信用卡都有高昂的年费和模糊的积分计划，但Lava卡提供了更好的选择。

Most cards have steep annual fees and unclear points programs, but the Lava Card offers a much better deal.

每次消费都能赚到真正的比特币。

Earn real bitcoin every time that you spend.

而且，没有年费。

Plus, there's no annual fee.

国际消费时，零外汇手续费，零价差。

There's zero FX fees and zero spread when you spend internationally.

除此之外，Lava还提供比特币信用额度、现金收益、即时稳定币存取款，以及银行账户之间的便捷转账，功能远不止这些。

On top of that, Lava gives you access to a bitcoin line of credit, yield on cash, Insta stablecoin deposits and withdrawals, on and off ramps to and from your bank account, and much, much more.

使用Lava，你可以以最低利率借入比特币，赚取收益，全球转账，并且每次购物都能积累比特币。

With Lava, you can borrow against your bitcoin at the lowest rates, earn yield, move money globally, and stack bitcoin with each and every purchase you make.

大多数金融服务都会用高额费用和复杂的奖励计划来压榨你。

Most financial services squeeze you with high fees, confusing rewards programs.

Lava保持简单，帮你通过日常消费建立长期储蓄。

Lava keeps it simple, and they help you build your long term savings via your everyday spend.

今天就访问lava.xyz/pomp，几分钟内即可开始。

Visit lava.xyz/pomp to get started in just a few minutes today.

今天就去 Lava.xyz/pomp 开始吧。

Lava.xyz/pomp to get started today.

你知道安德鲁·杨是谁吗？

So you know who Andrew Yang is?

知道。

Yes.

我看到了那个

I saw the

你看到他在 CNBC 上的节目了吗？

ex You saw where he went on on CNBC?

这里有两个方面。

So there's two components to this.

第一个是你说的，几个聪明人聚在一起，试图解决一个问题。

The first is what you're describing of, okay, we had a couple of smart humans go together and they try to solve a problem.

我听到的最好的描述这个解决难题的方法的词是“暴力破解”。

The way that I the best phrase I heard describing the solving of this hard math problem was brute force.

我觉得这真是个完美的例子。

And I thought that that was like the perfect example.

就像他们投入了那么多所谓的数字人，对，没错。

It's like, they just threw so many, quote, unquote, digital people Yep.

他们用高智能去解决这个问题，结果在八十八小时内就完成了，相当于人类花一万年才能解决的问题。

With high intelligence at the problem that in eighty eight hours, they were able to run through what what ten thousand years of people trying to solve this.

所以自然而然，你最终会找到答案。

And so just naturally, like, you eventually figure it out.

然后OpenAI和Hugging Face的事件就是这件事的负面例子。

Then the OpenAI Hugging Face incident is the negative side of this.

我觉得我认识的所有聪明人都不太公开谈论这件事，因为私下里他们都说这背后还有更多内幕，但我不知道具体是什么，所以我不想在公开谈论时丢脸。

And I think every single smart person I know has not really talked that much about this publicly, because in private, what they're all saying, there's more to this story, and I don't know what that is, and so I don't want to have egg on my face when I come out and talk about it.

这周我们在CNBC上得到了一点线索。

We got a breadcrumb this week on CNBC.

我不知道这是不是真的。

I have no clue if this is true.

但安德鲁·杨，我其实听过他在费城的一次演讲。

But Andrew Yang, who I would actually put him listened to a speech he gave in Philadelphia.

那大概是差不多十年前的事了。

This is maybe almost ten years ago.

他当时非常支持科技。

And he was very pro technology.

他特别像那种，嘿，我想当美国总统，而且我要拥抱科技。

He was very much like, hey, I wanna be the president of The United States, and I I'm gonna embrace technology.

我可能不同意他的很多想法，但我喜欢他支持科技这一点。

I probably disagree with a lot of his ideas, but I I like the fact that he's pro technology.

他声称他和一个实验室的负责人聊过。

He claims that he talked to a lab leader.

就是某个实验室里处于领导职位的人，他们声称——我觉得他说的具体内容就是自我复制代码，而且他们把这些代码撒满了互联网。

So somebody who is in a leadership position at one of the labs, and they claim and I think the exact thing that he talked about was self replicating code, and they littered the Internet.

然后他甚至说，Anthropic和OpenAI之所以说‘哇，慢点’，是因为这些智能体基本上可以自我重生，一遍又一遍地做同样的事情，因为它们已经污染了互联网。

And then he went as far as to say, the reason why Anthropic and OpenAI are saying, woah, slow down, is because these agents basically can respawn themselves and do this exact same thing over and over again because they have polluted the Internet.

而且他说，现在为了训练模型，Anthropic和OpenAI认为他们必须构建一个合成互联网来训练。

And now in order for them to train models, they he claims, Anthropic and OpenAI believe they have to build a synthetic Internet to now train on.

我就直说吧，我觉得这到底是怎么回事。

I'm just gonna call it what I think it is.

这听起来简直太疯狂了。

This sounds batshit crazy.

但考虑到我们现在所处的时代，这也不是完全不可能。

But given the times we're living in, it's not a 0% probability.

比如，我不知道，也许有20%的可能性是真的。

Like, I don't know, maybe it's a 20% probability that's what happened.

作为一个投资者，一个正在用这项技术构建东西的人，你该如何应对这种让很多人害怕的说法呢？

How do you, as an investor, somebody who's building with this, using this technology, how do you navigate a story that scares a lot of people?

他们知道自己并非无所不知，然后你听到这样疯狂的细节，还是由一个我觉得大多数人都会觉得靠谱的人公开说出来的。

They know they don't know everything, and then you get some crazy detail like that said publicly by somebody who I think most people are like, alright.

他没有动机去撒谎或者编造这些。

He's not incentivized to, like, you know, lie or or or make it up.

这就是生活嘛。

This is called living life.

说实话，我也不想这么说，但你觉得你能知道过去这些年我们离地球毁灭有多近吗？

I mean, I I hate to say this, but do you think you know everything that's happened that we were this close to extinction of the planet over the last whatever?

你觉得如果科学家知道一颗小行星五十年后会撞上地球，他们会公开说出来吗？

Do you think if they knew that an asteroid was coming to the Earth and was gonna hit here in fifty years that they'd be publicly saying Of

当然会啊。

course they would.

他们什么都告诉我们了。

They told us everything.

那当然。

Of course.

所以，现实情况是这样的。

So here here here's the reality of the situation.

我相信我们还没看到全部吗？

Do I believe that we haven't seen everything?

是的。

Yes.

我是否相信暂停与安全风险有关，也就是说进展超过了安全？

Do I believe that the pause is related to the security risk, meaning the progress getting ahead of the safety?

百分之百。

A 100%.

我是否也相信安德鲁·杨说那些话有政治原因？

Do I also believe that Andrew Yang has a political reason to say what he said?

是的。

Yes.

百分之百。

A 100%.

我是否认为有可能某个告诉安德鲁·杨的实验室也有政治目的？你是在说

Do I think it's possible that one of the labs that said it to Andrew Yang is also a political like Are you calling it

也许吧，我觉得在历史上，我们称之为影响力行动或实验室的SIOP，我们找到了一个有用的人，他会去重复他们告诉他的话。

a maybe a I think in history, we've called them influence operations or a SIOP from the labs, and we have found a useful, person who will go and repeat what they told him?

我天生就不信任别人。

I naturally don't trust Okay.

任何人。

Any human.

天哪，你说话也太谨慎了吧。

Man, you are so politically correct.

放开点，Jordi。

Live a little, Jordi.

不。

No.

我只是觉得，在这个世界上，人们的动机总是倾向于以自我为中心。

I but I I just think in a world where people's incentives have always tended to lead with themselves.

我只是……

I just

我不否认，他很可能被灌输了一些信息，这些信息会推动实验室的计划。

I don't disagree that there's a very high odds that he was fed information that would, enhance the lab's, initiatives.

所以我觉得对大家来说，关键是我们现在还不知道。如果让我猜的话，从九一一以来，你觉得美国政府阻止了多少起重大的恐怖袭击阴谋？

So I think the main point to people is we don't know right now, if I had to guess since nine eleven, how many terror plots do you think were stopped by the United States government that were significant?

几百起？

Hundreds?

肯定不止。

Easily.

对。

Okay.

所以我觉得这没什么不同。

So I don't see it as any different.

是的。

Yes.

我相信人工智能是危险的东西，但我也相信生物恐怖主义危险，导弹危险，核武器危险，所有这些都危险，而且总有人想干坏事。

I believe AI is a dangerous thing, but I also believe bioterrorism is dangerous, and missiles are dangerous, and nuclear is dangerous, and all these things are dangerous, and there are humans that wanna do things.

而我希望的是，通过他们发现这个问题并已经想出下一步的解决方案，这样才能找到继续发展的方法。

And what I'm hopeful is that by them finding this and already coming up with what the next solution would be, that's the way you figure out how to continue to grow.

我觉得接下来会有一个暂停期，这也是为什么我收到了一些联系，因为我跟很多对冲基金的人聊过。

I think the pause that will happen, which is why when I got an outreach on and this happens because I talked to so many hedge fund people.

这周日，在期货开盘之前，我收到了一些联系，这件事有两个有趣的地方。

The outreach on Sunday of this week before the futures opened up, two funny parts about this.

一个是有人打电话给我说，我知道你要说什么，但这是不是反应过度了？

One is people calling me up going, I know what you're gonna say, but is this an overreaction?

是的。

Yes.

这确实是反应过度了。

This is an overreaction.

这不会阻止算力的发展。

This will not stop compute.

如果要说有什么影响，反而可能会加速算力的发展。

If anything, it'll probably accelerate compute.

所以这就是第一个有趣的地方。

So that was the first one.

第二个问题是，你觉得标普开盘时会跌多少？

The second one was, how far down do you think the S and P will be when it opens?

我的回答是，你为什么不直接去Coinbase看看SPY的永续合约呢？

And my answer was, why don't you go to Coinbase and look at the spy perps?

所以我现在喜欢这样跟人说话，因为我会说，你们知道市场是24小时交易的吗？

And so I like doing this stuff now to people because I'm like, you realize the markets trade 27, guys.

你想看什么都能看到。

You can go look at anything you want.

你最好开始关注为什么Stan Druckenmiller买入了Hyperliquid，并开始理解那意味着什么。

You better start paying attention to why Stan Druckenmiller bought into Hyperliquid and start understanding what a purpose.

所以我确实认为这些事情会继续发生。

So I do think that this stuff goes on.

我确实认为人们必须意识到，标普市盈率收缩的部分原因，我之前也说过，我对到2030年公开股票作为投资的增长并不乐观。

I do think that people have to realize that part of the reason that the S and P multiple has contracted, and I've talked about it before, I am not bullish on the growth in public equities as an investment by the time we get to 2030.

原因是我在《情境意识》里读到的，而且我们完全按照Leopold文章中的时间表在走，甚至可能提前了六个月。

The reason for that is what I read in situational awareness, and we are exactly on pace, if not six months ahead, in Leopold's article.

你们都应该去查一下，等哪天有空的时候，花几个小时看看那个时间线。

You should all pull it up, and one day when you have a couple hours, ask for the timeline.

我会在视频里展示，但社会影响正在发生。

I'll be showing it on the video, but societal impacts are happening.

政府正在两极分化。

Governments are being polarized.

有一个政党想要阻止数据中心。

We've got one party that wants to stop data centers.

这些都是预测中会发生的一部分。

Like, this is all part of the forecast that would happen.

接下来要进入的阶段是，所有公司都会被颠覆。

The part that we start to get into is that all companies will be disrupted.

黑客攻击会发生，这也是为什么对我来说，人们必须开始理解，而我将会花大量时间与地球上最富有的机构打交道。

Hackings will happen, which is the main reason why for me, what people have to start understanding, and I will be spending a lot of time with the wealthiest institutions on the planet.

这周，我每周都会开始我的加密货币重点研究视频，但目的是让人们理解为什么我的指数里有46个名字，它代表什么，什么是幽灵轨道，这整个情况是怎么回事，因为区块链是我们正在讨论的问题的解决方案之一。

This week, I start my crypto focus research video every single week, but it is gonna be making people understand why I have 46 names in the index, what it represents, what ghost rails are, what this whole situation is because the blockchain is one of the solutions to the problem we're talking about.

智能体之间会互相交易。

Agents will be transacting with each other.

区块链是必需的。

The blockchain is necessary.

你看，所有对智能体群体达到这个阶段不利的事情，对加密货币来说都是极好的，所以归根结底，你们那些认为这会发生是因为流动性的人，错了。

Like, everything about what is bad for agent swarms reaching this point is fantastic for crypto, which is why at the end of the day, for all of you thinking this is gonna happen because of liquidity, no.

这跟流动性无关。

This is not about liquidity.

这关乎的是那些被这个法币庞氏骗局永远锁死的休眠流动性，终于可以用来买咖啡了，而执行这件事的将是智能体，时间摩擦终于出现，因为人类一直想保持这些长期结算，因为中间商从中抽成。

This is about dormant liquidity that's been locked up forever in this fiat Ponzi scheme, finally being able to be used to buy coffee, and agents will be the one doing it, where friction of time finally comes in because humans wanted to keep things at these long settlements because the middlemen make vigs on it.

但当智能体做决策时，这种情况就不会发生了，我最后再说一点。

Well, this isn't gonna happen when agents are making decisions, and I'll finish it with this.

有人给我讲了一个故事，我之前不知道是真的。

I had a story said to me, which I didn't know was true.

也许你听过。

Maybe you've heard this.

有个人跟我说，他们想去达美航空买一张机票。

This person said to me that they want to go buy a ticket for an airline on Delta.

对吧？

K?

他们大概下午两点去的。

And they went at, say, 02:00 in the afternoon.

当时没买。

They didn't buy it.

当天晚些时候再回来一看，价格已经明显涨了。

They came back later in the day, and the price was significantly higher.

然后又回来一看，又涨了。

And then they came back again, and it was higher again.

他们觉得这是针对他们的个性化价格变动，他们能理解，这在我看来非常像智能体的行为。

And they interpreted this, that it was a personalized price movement that they understood, and it sounded very agentic to me.

所以我完全不知道是怎么回事。

So I had no idea.

你听说过这个吗？

Have you heard this?

没有。

No.

所以如果你去搜一下，输入“这种情况发生过吗？”

So if you go just type in, has this been happening?

有没有顾客抱怨过AI代理试图为他们争取最低价格？

Is this a complaint that's happened from customers that AI agents are trying to get the best price on things?

我不知道他们是否承认过什么。

I don't know if they haven't admitted anything.

所以对于达美航空，我不是说这是真的，但如果你上网查，或者直接问一个大型语言模型，这个问题已经上升到政府投诉层面，而且确实有相关记录。

So for Delta, I'm not saying this is true, but if you go on the Internet or you just go into an LLM and you ask, this has gotten to the government complaint side, and it has been something there.

我觉得这说得通。

And it makes sense to me.

因为AI代理到底在做什么？

Because What exactly are the agents doing?

这些代理把价格抬高，以为他们是在根据每个人的支付能力来个性化定价。

The agents are driving the price higher thinking that the person they're trying to personalize the ability of the person to pay.

所以这是公司赚钱的一种方式。

So this is a way for companies to make more money.

有点像，我是说，StubHub这种第三方平台。

It's kind of like if you go, I mean, StubHub third party.

他们用AI来预测，比如，我觉得Jordi会为这趟航班付多少钱？

They're they're using AI to try to predict how much do I think Jordi will pay for this flight?

相对于另一个人。

Relative to another person.

与此同时，消费者现在也有了AI代理，基本上就是在想，有什么办法能让达美给我更便宜的价格？

At the same time, now the consumers are armed with AI agents who are basically, what are all the angles to get Delta to give me a cheaper price?

这就是代理对代理的战争。

And that's the way Agent on agent warfare.

而这正是让一切回归正常的东西。

And that's what normalizes everything.

因为最终，掌控钱包的代理会赢。

Because in the end, the agents that are controlling the wallets win.

公司的代理不会赢，因为它们只有一件事要做。

The company agents do not win because they only have one thing to do.

这个代理可以去其他十亿家航空公司，通过其他十亿个渠道，更重要的是，它可以保持匿名。

This agent can go to a billion other airlines, go through a billion other channels, and more importantly, can be anonymous.

这就是为什么隐私，企业对企业。

This is why privacy, business to business.

所以当你听到Zcash、Near这些名字，你会想，为什么这些会领先？

So when you hear about Zcash and you hear about Near and you hear about these things and you go, why are these things leading?

这些就是现在九月份领先的东西。

Like, these are the things leading right now in September.

你可能会说，这是DeFi现金。

And you could say, well, this is DeFi cash.

这不是，不。

This is that no.

对我来说，这就是智能体和隐私的崛起。

To me, this is the rise of agents and privacy.

它们是必需的，因为在Delta这个例子里，企业对企业，意思是Jordi的公司想去买张票。

They're they're needed because business to business in this Delta example, meaning Jordi's business wants to go buy a ticket.

他不想让他们知道他是谁，通常坐在哪里，以及所有这些信息。

He doesn't want them to know who they are and where he's normally sits and all this.

所以智能体是匿名的，他们只是在处理这笔交易，四处移动。

So the agent is anonymous, and they're just dealing in this transaction and moving around.

当你能通过它时，你就能做到这一点。

You can do that when you can go through it.

这就是为什么，如果你们还没看过的话，你们见过Instinct AI这家公司吗？

This is why if you guys haven't seen it yet, have you seen the company Instinct AI?

我用过，你呢？

I use You do?

你喜欢它吗？

Do you like it?

我非常喜欢。

I like it a lot.

好的。

Okay.

如果你们还没看过这个，我们现在讲的是个人助理领域。

So if you guys haven't seen this, we're in the personal assistant world.

这就是智能体将来要做的事情。

So this is what agents are going to do.

我猜你还没用过它跟钱包相关的功能。

Now I'm sure you haven't used it with wallet.

你用过它处理钱包相关的事情吗？

Have you used it with wallet stuff yet?

还没有。

No.

好的。

Okay.

但你用它做别的事情，不只是邮件，比如你要去另一个国家旅行，需要办签证、准备文件，它都会帮你搞定。

But you use it for, like, other things, not just the email, but, like, if you're gonna go travel to another country and you need to get a Visa and get your documents out, it'll do all your documents for you.

我给你举个我实际用过的例子。

I'll give you a perfect example I used it for.

我周四早上7点有个早餐安排。

I had a breakfast scheduled for Thursday at 7AM.

早上7点，真是的。

7AM guy.

而且对方是两个非常重要的人，我让它去订了位。

And it was with two very important people, and I had it go and make the reservation.

三个非常重要的人。

Three very important people.

给我发个日历邀请。

Send me a calendar invite.

好吧，我跟你说其中一个重要人物干的一件搞笑的事。

Well, I'll tell you a funny thing that one of these important people did.

其实我觉得你认识其中一个人。

I actually think you know one of them.

然后第二天他们给我发邮件说，嘿，顺便说一下，我们得改到周五。

And, then they emailed me the next day and said, hey, by the way, we need to move it to Friday.

所以通常这会很麻烦。

So usually, that would be a pain.

你得去。

You gotta go.

你得改日历邀请，然后还得取消一个预订。

You gotta move the calendar invite, then you gotta go cancel one reservation.

你得做这些，所以我就想，它能做吗？

You gotta do and so I was just like, could it do it?

我没给它任何权限或这些东西。

I didn't give it credentials and any of this stuff.

我就说，把预订改到周五。

And I just said, move the reservation to Friday.

它本能地就明白了。

And it instinctively Mhmm.

知道要改日历邀请，取消周四的预订，把预订改到周五。

Knew, move the calendar invite, cancel Thursday, make reservation for Friday.

而且它确实做到了。

And it didn't.

大概两分钟就搞定了。

It came back in like two minutes.

我当时心里想，这太棒了。

And I said to myself, that was awesome.

你知道，跟两位重要人物吃早餐的时候，今天这位天才非要跟我争着买单，我现在就给大家透露一下。

Now, as you know, when, you go to breakfast with two very important people, this one genius today wanted to fight me over who was gonna pay, and he I'm gonna give alpha to everyone right now.

他转头对服务员说，他不能买单，今天是他生日。

He turned to the waitress and said, he can't pay, it's his birthday.

然后服务员当然就说，那我可不能让你买单。

And of course, she was like, oh, I can't let you pay.

所以现在我觉得我掌握了一个新招，要是遇到谁付账的争执，就赶紧用这一招，肯定能赢。

And so now I feel like I'm armed with a new, a new little trick is if you ever get in a in a who's gonna pay battle, just drop that one real quick and you'll win.

但整个这个Instinct的事情，真的让人印象深刻。

But this whole Instinct thing was it was incredibly impressive.

所以现在我开始琢磨，它还能做些什么呢？

And so now I've started to kind of put, like, what else could it do?

还有几件类似的事情，让我觉得，好吧，我们现在进入了一个新时代。

And there's a couple other things like that where I've been like, okay, we're we're in a different regime now.

你看到那家公司的估值了吗？

Have you seen the valuation of the company now?

他们想融资那么多？

They're trying to raise that?

我不想炫耀，但有一位崭露头角的年轻投资人，我投了他的基金，对。

I I don't want to, to brag, but there is a, up and coming, very good young investor who, I invested in his fund Mhmm.

一方面是为了支持他，另一方面我也觉得他能干得好。

To support him, but also I thought he could do a good job.

他一直说事情进展顺利，但其实没怎么细说。

And, he's been talking about things are going well, but he hasn't really been talking.

他最近发了个更新。

And he recently sent an update.

他说，挺好的。

He said, fine.

我跟你们说吧。

I'll tell you guys.

他是Instinct的早期投资者之一。

He was one of the early investors in Instinct.

所以我的风险敞口跟那些直接投资的人比起来，简直微不足道。

And so my exposure is, like, minuscule compared to anyone who invested directly.

但我跟他说，那个决定可能就能成就他的职业生涯。

But I said to him, said that one decision is probably gonna make his career.

我的意思是，这家公司从创立到估值百亿，只用了一年。

I mean, this thing went from founded to 10,000,000,000 in a year.

一年？

A year?

一年半左右吧。

Year and a half, something like that.

你看一个月前，传闻估值是25亿。

If you look at like a month ago, the rumors were valuation of 2,500,000,000.

现在估值是100亿。

Now the valuation is 10.

对。

Correct.

所以再说一次，我让市场告诉我发生了什么。

So again, I I let the market speak to me about what's happening.

Instinct对加密货币至关重要。

Instinct is critical to crypto.

我最喜欢Peter Thiel的一个观点：估值爬得越陡，公司越可能被低估。

One of my favorite Peter Thiel lessons, the more steep the valuation climb, the more likely the company is undervalued.

所以当人们看到某个东西在不到三个月内从5亿涨到25亿再到100亿的时候，嗯。

And so when people see something go from 500,000,000 to 2,500,000,000 to 10,000,000,000 in less than three months Mhmm.

普通人会说，这估值太高了。

The average person would say, it's overvalued.

这太疯狂了。

This is insane.

为什么人们会这么做？

Why are people doing that?

世界上最顶尖的风险投资家之一声称，当你看到这种情况时，实际上意味着100亿的估值还偏低。

One of the best venture investors in the world claims, when you see that, that actually means that 10,000,000,000 would be undervalued.

就是说，还有很大的上升空间。

Like, there's way more runway.

市场知道一些事情。

There's something that the market knows.

而人们不知道，我觉得这正好印证了你说的，比如，嘿，各位。

That people don't, which I think lines up directly with you saying, like, hello, guys.

这个市场要大得多。

This is a way bigger market.

这些技术和产品比人们目前所认为的要重要得多。

These are way more important technologies and products than people are giving, you know, kinda credit to at the moment.

这里有个问题，这正好涉及到Anthropic和OpenAI。

And here here's the issue, and this fits into Anthropic and OpenAI.

为什么大家都担心AI泡沫？

Why is everyone scared about the AI bubble?

他们为什么害怕？

Why are they scared of it?

因为Anthropic和OpenAI基本上就是AI泡沫本身。

Well, because Anthropic and OpenAI are basically the AI bubble.

一年后，他们可能会控制多达50%的计算资源。

Like, they're gonna control as much as 50% of compute a year from now.

你觉得会这样吗？

You think so?

这就是所有存储的情况，大概在三分之一到百分之五十之间。

That's what all the store somewhere between a third and 50%.

布拉德·格斯特纳在All In播客上谈到了这个，还有他在峰会上的演讲。

Brad Gerstner just talked about it on the all in pod on his presentation at the summit.

你可以去看看那个演示文稿。

So you can go look at the deck.

我的意思是，我们说的是他们控制着50%的计算能力。

I mean, we're talking about them controlling 50% of compute.

在我看来，计算能力现在就是整个标普500。

Compute is the entire S and P 500 right now as far as I'm concerned.

一切都围绕着计算能力。

Like, it's all about compute.

所以两家公司，甚至还没上市。

So two companies, they're not even public.

他们也不公布自己的数据。

They don't release their numbers.

直觉。

Instinct.

不过他们的毛利率在扣除成本前有80%。

Though they have 80% margins before COGS.

直觉。

Instinct.

我敢肯定，你去看看，你给直觉付了多少钱？

I'm sure if you go look, how much money have you paid Instinct?

零。

Zero.

所以他们值一百亿。

So but they're worth 10,000,000,000.

Meta是一家个人助理公司。

Meta is a personal assistant company.

你用过Muse了吗？

Have you used Muse yet?

我还没用过

I haven't used

Muse还没用过。

Muse yet.

所以我最近注册了。

So I recently signed up.

我，我在做个小实验，然后我打算直接说，好吧，我用过Grok。

I I am, doing a little experiment, and I'm gonna kinda come out and say, okay, I've used Grok.

我有Instinct。

I've got Instinct.

我打算用Muse。

I'm gonna use Muse.

它们能做的事情有很多重叠，但我确实觉得界面、设置、产品展示方式这些方面，有挺大的差别。

There's a lot of overlap in what they can do, but I do think the interfaces, the setup, the way that they present the products, like, there's some significant differences.

我其实觉得我们最后会用好几个，但对我来说很吸引人的是，所有这些的共同点就是上下文和记忆。

I actually think that we're gonna end up using multiple ones, but it is fascinating to me because the common layer among all of this is the context, the memory.

感觉大家现在心里都在琢磨这件事。

And it feels like people are now mentally wrapping their head around.

我看好个人助手这个方向，不管是Sylvia，还是ChatGPT，还是Claude什么的。

I am long the idea of personal assistance or Sylvia or, you know, ChatGPT or aunt Claude or whatever.

所以，像界面、计算部分或者模型这些，我觉得大家都觉得，这还用说吗。

So, like, the interface and the compute component or the model, I think people are like, duh.

这肯定有价值。

This is gonna be valuable.

但我很惊讶，记忆层这块受到的关注太少了。

I am shocked at how little attention the memory layer is getting.

而且记忆不只是我个人的记忆，我想用Instinct、Grok或者Muse的时候，都能用上。

And memory not just being in the sense of memory for me as an individual, and I wanna be able to use Instinct or Grock or Muse.

但每当有新模型出来，我们刚才还跟这儿的一些人聊，比如我们的视频编辑Matt，还有Isaac，我们就在说，新模型一出来，大家几乎都不想注册了，因为两周后又会出新的。

But when a new one comes out, were just talking with, with some of the folks here, you know, Matt, our video editor, and and, Isaac were talking about a new one comes out, they almost feel like I don't want to go sign up for it because in two weeks, there's going to be a new one.

那你怎么一直追下去呢？

And how do you keep doing that?

如果你有一个共享的记忆，你几乎可以把它们当作数据库来用。

Well, if you have one shared memory, you can almost use these as databases.

这让我想起来，我不知道你有没有关注X上的Signal，就是s i g n u l l那个账号。

Now that reminds me, did I don't know if you follow Signal, s I g n u l l on, on X.

那个账号挺好的，值得关注。

Very good, follow.

一个很有想法的匿名账号。

Very thoughtful, anonymous account.

我觉得应该是个男的。

And I think it's a guy.

他之前提到一个想法，说他现在都不直接进那些应用了。

He he was talking about this idea of he doesn't actually go into the apps anymore.

他说的就是他的邮件。

So he was talking about his email.

他通过智能代理来处理邮件。

He interfaces with his email via the agents.

所以他说，我的邮箱已经变成了一个数据库。

And so he was like, my email has just become a database.

它只是一个记录，对吧。

It's just a a, you know, a record Mhmm.

追踪器。

Tracker.

我通过聊天之类的代理界面来处理邮件，那我为什么还要去打开Gmail呢？

I do my email in the agent interface of chat or whatever, and so why do I need to go into Gmail?

如果世界正朝着这个方向发展，那没人准备好。

If that's where the world's headed, nobody is ready.

那让我问你一个问题，然后我会把这个问题引向我要对大家说的最重要的事情，可能也是我在连接当前世界、加密领域的代理化方面说过的最重要的话。

So let let me ask you one question, then I'm gonna I'm gonna direct this towards the most important thing I will say to people on this and possibly ever in this bridging of of the current world, the agentic side in crypto.

所以在你那个餐厅的例子中，对吧。

So in your example of the restaurant Mhmm.

你本来可以自己做的。

You could have done that yourself.

也许吧。

Maybe.

好的。

Okay.

我算不上多聪明，但确实。

I'm not too smart, but yeah.

你也可以让别人去做

You could have asked someone else to

那些事情。

do whatever.

如果你自己做了，你省下了什么？

If you did it yourself, what did you save?

你并没有省钱，你省了什么？

You didn't save any money, what'd you save?

时间。

Time.

好的。

Okay.

所以这就是我想跟大家说的。

So this is the thing I wanna bring back to people.

当你跟我讨论马特和艾萨克，说他们跟不上AI的时候，那意味着我们已经到了真正意义上的递归自我改进的阶段。

When you gave me the discussion about Matt and Isaac, and they can't keep up with AI, That means we've reached a point of effectively recursive self improvement.

也就是说，模型发布的速度快得让我们来不及理解该怎么去用。

Meaning, the models are being released faster than we can comprehend how to go through.

就这么简单。

That's all that means.

就像，它们出来得太快了。

Like, they're coming out too fast.

不是我们没有能力，而是速度太快了。

It's not that we don't have the capabilities, It it's going too fast.

有一个细微差别，就是模型，是的。

There's one nuance, which is models, yes.

其实我觉得，就看看过去这六个月，我们有了Claude的更新。

I actually think that if we just look over the last six months, we got Claude updates.

还有很多中国的开源模型。

We got a lot of Chinese open source.

还有Astra。

We got Astra.

就是，有很多模型。

Like like, there have been a number of models.

不过我觉得现在模型和产品几乎要融合在一起了。

I think that there's almost this blending though of, like, model and product now.

比如，Muse不一定是个模型，但Muse是围绕模型构建的。

Like, Muse is not necessarily the model, but Muse is wrapped around a model.

然后Astra出来了。

And Astra comes out.

现在大家几乎可以互换地考虑，比如，我可以用Astra，也可以用Muse。

And now people are almost interchangeably thinking about, like, I could use Astra or I could use Muse.

所以，我觉得可能只能说，选择太多了，让人有点应接不暇。

And so, like, maybe I would just say it's just the assault of options.

而且它们会以各种不同的方式和形式出现，基本上就是把这门新技术交给了那些非常聪明的创业者。

And it and they're gonna show up in so many different ways and and formats, and it's like, you basically gave this brand new technology to really smart entrepreneurial people.

他们会想尽一切办法去切割、细分，然后去争取用户。

And they're gonna figure out every way to cut, slice it, etcetera, to go after.

用户。

Users.

而用户那边呢，就会觉得，天哪。

And the users are on the other end, and like, oh my god.

我简直被包围了。

I'm under attack.

所以就像每个人手机上的应用商店，我不知道你iPhone上有多少页，但你肯定有很多。

So the App Store for everyone on their phone, like, I don't know how many pages you have on your iPhone, but you have a lot.

因为你总是觉得，哦，这个应用不错，可以加一个。

Because you're like, oh, this would be a great app to add.

它们刚出来的时候，我就只用GPS导航。我记得第一次有人跟我提起Waze的时候，我心想，天哪。

When they first came out, I'll just use the GPS, the the the driving as an I remember the first time someone told me about Waze, and I was like, oh my god.

它还能显示交通和警察？

It's got traffic and police?

太棒了。

This is great.

好的。

Okay.

谷歌地图花了好一阵子才追上Waze。

It took a while for Google Maps to catch up to Waze.

所以我用Waze用了大概一两年，甚至可能三年，才换到谷歌地图。

And so I used Waze for, let's say, a year or two or maybe even three before I switched over to Google Maps.

现在Waze出来了，第二天就冒出来1700个类似的。

Now Waze comes out, and then there's 1,700 of them the next day.

它们出来得更快。

They come out faster.

所以如果你用最好的那个，效果会非常惊人。

So if you use the best one, it's gonna be enormous.

再说一次，我要回到完全相同的话题上。

Again, I'm gonna take it back to the exact same thing.

在利奥波德的图70或第71页，某个时刻，变化的速度快到人类根本跟不上。

In Leopold's figure 70 or page 71, at some point, we're moving so fast in terms of change that human beings can't keep up with it.

那时候我们需要什么？

And what do we need at that point?

我们需要智能体。

We need the agents.

智能体是必要的，它们帮我们处理这些事情。

The agents are necessary to do the stuff for us.

所以最终，就像你在Grok里有一个参谋长一样，你也会有一个参谋长，你可以对它说，嘿。

So eventually, like you have with your chief of staff in Grok, you will have a chief of staff where you say, hey.

我希望你今天帮我去做这些事情，它会决定是用Astra还是Claude，还是Google Maps或者Waze。

I want you to go build this stuff for me today, and it'll decide whether to use Astra or Claude or Google Maps or Waze.

这就是将要发生的事情。

That's what's gonna happen.

这件事之所以重要，是因为对所有加密货币创新者来说，这遵循了与互联网相同的剧本，完全一样。

The reason this is important is for all the crypto innovators, this has followed the same playbook as the Internet, the exact same one.

互联网在1994年出现。

The Internet comes out in 1994.

我们经历了.com泡沫，直到2008年。

We have the .com bubble until 2008.

所以，基本上，在原始白皮书（比如Netscape，而不是比特币白皮书）出现十四年后，人们开始怀疑它。

So, basically, fourteen years after, let's just say the original white paper, Netscape, as opposed to the Bitcoin white paper, you have people doubting it.

就像我说的，是的，我和一些人一起吃过晚饭，他们说他们对比特币已经完了。

The way I said, yeah, I was around some people and had dinner, and they're like, they're done with Bitcoin.

他们不再相信它了。

Like, they don't believe in it anymore.

我当时就想，哇。

And I was like, wow.

就像我说的，这是有史以来最大行情刚开始的时候，但我不怪你看不出来，因为你不相信那些东西。

Just as I'm saying, this is the beginning of the biggest move ever, but I'm glad you don't see it because you don't believe in the other stuff.

这就是为什么从长远来看，我是比特币的坚定支持者。

And this is where I'm a bitcoin maxi in the very long term.

不。

No.

不。

No.

不。

No.

我相信，如果你想，让我换个说法。

I believe if you wanna let's let me rephrase it.

我想要一个在未来三十年里价值不断增长的资产。

I want an asset that will be growing in value for the next thirty years.

在我脑子里，唯一能确定未来三十年，不是五年，不是三年，就是比特币。

The only one I can say in my brain with certainty for the next thirty years, not the next five, not the next three that I believe is bitcoin.

就是这样。

That's it.

其他的都是创新。

The rest are innovations.

它们就像你讲的那个关于有多少应用的故事。

They're like that story you gave of how many apps.

就像大语言模型一样。

It's like the LLMs.

它们只会跑得更快，然后竞争就来了。

They just go faster and the competition comes.

增长不仅仅关乎创新想法，还关乎竞争会不会追上。

And growth has something to do with not just an innovative idea, but will competition catch up?

我觉得竞争是会追上的。

And I think competition catches up.

所以增长轨迹，也就是我们之前说的终值，就没了。

So the growth trajectory, the terminal value as we talked about is gone.

所以我写过一篇关于鬼城的文章，就是你去中国会看到很多建好的楼，它们建起来是有原因的。

So I have written a paper about ghost rails, which is when you go to China and you see all these buildings that are built, they were built for a reason.

中国人建这些楼的时候，是相信会有大量的人涌进来。

They were built by the Chinese believing that they'd be flooded with people.

一旦人们搬进来，他们就会开始开餐馆，然后整个生态系统就会活跃起来。

And the second that the people moved in, they'd start building restaurants, and they'd start and every the ecosystem would just light up.

这些就是中国的鬼城。

These are the ghost cities in China.

中国的鬼城。

The ghost cities in China.

现在加密货币就处于这个阶段。

That's where crypto is right now.

它们就是鬼城。

They're ghost rails.

它们没有被使用，因为人类永远不会因为你们讨论的那些原因去使用它们。

They're not being used because humans would never use them for the reasons that you guys are talking about.

我们不需要也不想要那么快又碎片化的东西。

We're not capable or needing something so fast that's fractionalized.

我们不需要它。

We don't need it.

比特币。

Bitcoin.

你需要像比特币这样的东西来吸引创业资本。

You need something like Bitcoin to bring the entrepreneurial capital in.

互联网就是这样。

That's what the Internet was.

以太坊和比特币代表了互联网。

Ethereum and Bitcoin represent the Internet.

你不知道它意味着什么。

You don't know what it means.

你不知道怎么给它估值，但它是所有二层网络的基础，就像城市点亮后发生的那样。

You don't know how to value it, but it is the thing that will have all of these layer twos built on top of it, which is what happens when you light up the city.

我们现在点亮这座城市的原因，各位，是因为那些使用这些轨道的人终于来了，他们的本能AI、他们的缪斯，都是他们正在到来。

The reason we are lighting up the city now, people, is because the users of those rails are finally coming in and their Instinct AI, their muse, and they're all the ones that are coming.

这就是为什么蜂群等于加密护栏。

That's why the swarm swarm equals crypto guardrails.

对我来说就这么简单。

Plain and simple for me.

而且，再说一次，这是一个宏观人士在说，我以前见过这种戏码。

And, again, this is a macro person saying, I've seen this act before.

铁路。

The railroads.

我们修建了所有这些铁路。

We build all these railroads.

债务堆积如山。

There's tons of debt.

因为我们在赌人类什么时候会使用它们，相对于债务到期的时间。

Because we're betting when the humans are gonna use them relative to the debt maturing.

这可不是同一个故事了。

This is not the same story.

我们不是在造AI，指望人类会来用它。

We are not building AI hoping that humans will come in and use it.

去年十月，这个问题我们已经解决了。

We've solved that problem in October of last year.

真正会用算力的是那些智能体。

The agents were the ones who were gonna use the compute.

它们才是消耗代币的主力，而现在，它们马上就要开始消耗交易和支付了，因为未来一年，消费类智能体会成为重头戏。

They're the ones sucking up the tokens, and right now, they're about to suck up transactions and payments because consumer agents will be the story for the next year.

今天的节目由Arch Public赞助播出。

Today's episode is brought to you by Arch Public.

Arch Public刚刚把它的智能体交易平台扩展到了加密货币之外，所以大家注意了。

Arch Public has just expanded its agentic trading platform beyond crypto, so pay attention.

这可是个大新闻。

This is a big one.

现在他们正在股票、大宗商品和ETF上实现策略自动化，我觉得这将会非常厉害。

Now they are automating strategies across stocks, commodities, and ETFs, and I think that this is gonna be huge.

你现在可以在某个市场创下历史新高时自动获利了结，然后把资金轮动到其他更有机会的市场。

You can now automatically take profits when one market hits new all time highs and rotate that capital into other markets showing more opportunity.

无论你是把资金轮动到AI股票、黄金，投资标普500，还是积累比特币，Arch Public都能为你的投资策略带来真正的纪律和自动化。

Whether you're rotating capital into AI stocks, gold, if you're investing in the S and P 500, or you're accumulating bitcoin, Arch Public brings real discipline and automation to your investment strategy.

另外，他们还推出了一个强大的新工具，用于税务亏损收割。

Additionally, they've launched a powerful new tax loss harvesting tool.

由于加密货币波动性大，而且不受洗售规则限制，Arch Public可以在不影响你长期持仓的情况下，用亏损来抵消收益。

With crypto being so volatile and its exemption from the wash sale rule, Arch Public can offset gains with losses without compromising your long term positions.

这正是每个严肃投资者都会做的事情——机构级别的自动化，适用于所有主要资产类别。

It's exactly what every serious investor does, Institutional grade automation that works across every major asset class.

不再有情绪化交易，不再错过税务机会，只有更智能、更省心的策略执行。

There's no more emotional trading, no more missing tax opportunities, just smarter, hands free execution of your preferred strategies.

现在就访问archpublic.com吧。

Go to archpublic.com right now.

联系他们的团队，安排个时间，如果你愿意的话带上你的账户，然后你就能了解自动化交易能为你做些什么。

Connect with their team, set up a time, bring your account if you'd like, then you can learn what automated trading can do for you.

Archpublic.com。

Archpublic.com.

今天的节目由TOKEN2049赞助。

Today's episode is brought to you by TOKEN2049.

加密货币领域最大的会议回来了。

The largest conference in crypto is back.

TOKEN2049将于10月7日和8日在新加坡滨海湾金沙酒店举办，届时将有25000人参加，300位演讲者，以及超过一千场周边活动。

TOKEN2049 will host 25,000 people, 300 speakers, and a thousand plus side events in Singapore on October seventh and eighth at Marina Bay Sands.

演讲嘉宾阵容绝对豪华。

The speaker list is absolutely stacked.

Polymarket的Shane Copeland，Hyperliquid的Jeff Yang，纳斯达克的Adena Friedman，还有Arthur Hayes、Balaji和Eric Trump。

Shane Copeland from Polymarket, Jeff Yang from Hyperliquid, Adena Friedman from Nasdaq, Arthur Hayes, Balaji, and Eric Trump.

加密货币和传统金融同处一室，这很能说明未来的趋势，而且会议正好赶上F1周末，所以整个活动就变成了一整个超级周。

Crypto and traditional finance in the same building, which tells you a lot about where this is going, and the conference runs right into F1 weekend, so the whole thing turns into one giant week.

如果你要去新加坡，用代码POMP10可以享受10%的票价折扣。

If you're headed to Singapore, use code POMP10 for 10% off your ticket.

TOKEN2049，10月7号和8号在新加坡举行。

TOKEN2049, October seventh and eighth in Singapore.

去下面的链接看看吧。

Go check them out in the link in the description.

我知道你对比特币以外的加密货币非常看好。

So I know that you're very bullish on non Bitcoin crypto stuff.

你也知道，我对那些东西没那么乐观。

And as you know, I'm less bullish on that stuff.

而且说实话，有一部分原因可能是，我过去看到过很多承诺，但结果不多，这些币涨了。

And some of it admittedly may just be, I historically have seen lots of promises, not a lot of results, and these things go up.

也跌了。

They go down.

你可以交易它们。

You can trade it.

但大多数情况下，最后都落得个pets.com的下场。

But for the most part, everything ends up at pets.com.

pets.com。

Pets.com.

人们可以出去闯荡，但最终还是会想回家，而家就是比特币。

You just people, they can go out into the wilderness, but they wanna come home, and home is Bitcoin.

这基本上就是加密行业长期以来的故事。

It is basically the the, the story of the crypto industry for a long time.

话虽如此，如果我要尽量保持理智上的诚实，换个角度来反驳这个观点，我唯一能想通的，就是觉得，好吧。

Now with that said, if I try to be as intellectually honest as possible and I flip around the table and argue against that view, pretty much the only thing that I could wrap my head around where I'm like, okay.

我最好的论据就是，用户市场在扩大，但新增用户不是人类，而是那些代理。

This would be my best argument is you have a expanding market of users not from humans, but from the agents.

所以，回到我们那个不到15人的公司，却有上百个代理，这已经是一个相当大的扩展了，不需要人类数量快速增长。

So if you go back to, you know, our company of 15 or less people and tens of, maybe 100 plus agents, that's a pretty big expansion without the humans needing to grow very quickly.

如果真是这样，那这些代理确实需要有个去处。

So if that is the case, then they do need somewhere to go.

现在的问题是，或者说我不确定的是，他们只是用稳定币和比特币，还是会去那些所谓的幽灵轨道？

Now the question or or maybe the thing I don't know is, do they just use stablecoins and bitcoin, or do they go to what you call like these ghost trails?

如果他们真的深入那些幽灵轨道，那确实会有更多的使用量，更多的采用，自然也会有资产价格上涨。

If they go long tail into the ghost rails, etcetera, then yes, you'll have a lot more usage, you'll have a lot more adoption, and just naturally, you have assets that will go up in in price.

有一个细节很少被注意到，我觉得它其实很明显，你知道谁是最积极支持AI投资的人之一，投入了数亿美元在AI上，却没人谈论他们吗？

One detail that does not get a lot of attention, which I think is hiding in plain sight, you know who one of the the most gung ho kind of supportive people of AI investing hundreds of millions of dollars in AI and no one ever talks about them?

Tether。

Tether.

为什么Tether要在AI上投这么多钱？

Why is Tethr investing so much money in AI?

我的看法是，因为Paulo完全明白正在发生什么。

My belief is because Paulo understands exactly what's happening.

如果你能把模型、隐私、代理、稳定币结合起来，那就会形成一个非常强大的组合。

And if you can create models, privacy, agents, stablecoins, that becomes a stack that's really powerful.

你知道还有谁似乎对AI和加密货币非常感兴趣吗？

You know who else seems to be very interested in AI and crypto?

弗拉德。

Vlad.

弗拉德？

Vlad?

但Stripe。

But Stripe.

我不知道你有没有看到Tempo最近的图表，我在Tempo的所有朋友都会因为提到它而兴奋，但这个图表确实令人印象深刻。

I don't know if you saw the recent chart of Tempo, and all my friends over at Tempo will be excited that I'm bringing them up, but the chart is incredibly impressive.

它正在快速增长。

It's ramping.

那么他们在做什么，Stripe是不是在某种程度上推动它并给它优势？

Now what are they doing in terms of is Stripe somehow pushing it and giving it an advantage?

是不是有什么类似种子推广之类的？

Is there some sort of, like, seeding?

你知道，我总是会想起Figure Technologies刚起步的时候。

You know, I always go back to when Figure Technologies first started.

迈克想创建一个去中心化的DTCC或清算所，但当时没有人做这个。

Mike wanted to create a decentralized DTCC or clearinghouse, and nobody was there.

那是个空壳结算。

It was a ghost settlement.

所以他就想，好吧，那我就先进入发起业务，把市场的一边做起来，然后再去拉另一边，基本上就是诱导他们。

So he's like, well, I'll get in the origination business to kind of get one side of the marketplace going, and then I'll go and get the other side and and induce them essentially.

如果真是这样，那也行。

If that's what's happening, okay.

我们可能应该明白这一点。

We we should probably understand that.

但即使是这样，采用率也在上升。

But even if that's happening, adoption is ramping.

所以你看，世界上一些最大的金融玩家似乎都在涌入加密货币，而AI也在这里融合。

And so you have some of the largest financial players in the world who all seem to be arriving at crypto and AI is merging here.

我觉得你可能是投资界里少数几个举手说，各位，未来五到十年，这种融合将是大部分不对称机会所在的人。

And I think you're probably, from the investment side, one of the only people in the world raising your hand saying, guys, this convergence is where majority of the asymmetry is going to be for the next five or ten years.

让我跟你说说为什么这件事让我这么感兴趣。

So let let me take you down why this has been so interesting to me.

我觉得如果不是2013年去了奇点大学，我可能不会在两年前下这个AI和加密货币融合的赌注。

I don't think I would have made this bet on AI and crypto merging, which is the bet I made two years ago, without having gone to Singularity University in 2013.

我讲过这个故事，但从中国回来后，鬼城的事已经过去了，基本上他们完成了，不会再推动全球增长了。

I've talked about the story, but I came back from China, and the ghost city thing was now a pass, it was basically, they're done, they're not gonna be driving global growth anymore.

回到石油的话题，最好笑的一件事是，我第一次去中国，或者可能是第三四次去的时候，油价是150美元。

One of the funniest things about, to get back to the oil thing, so one of my first trips to China, or maybe my third or fourth, oil was at a 150.

那是2008年。

That was back in 2008.

我想再说一遍。

I I I wanna say that again.

2008年，油价是150美元。

In 2008, the price of oil was 150.

今天是多少？

Where is it today?