帕尔默·卢基：我为何创办自己的银行

我们有帕尔默·拉基在重播等候室里。

We have Palmer Lucky in the restream waiting room.

让我们把他接入TVPN UltraDown。

Let's bring him in to the TVPN UltraDown.

帕尔默，你怎么样

Palmer, how are you

正在做？

You are doing?

直播中。

Live.

您正在直播，先生。

You are live, sir.

你能

Can you

听清楚我们吗？

hear us okay?

是的。

Yep.

不过我能听得很清楚，只是有点延迟。

I hear you loud and clear, though, with a little bit of delay.

所以我们尽量让哦，保持快速。

So we'll try to keep Oh, it zippy.

好吧，我们大概只会说五个词，然后你就疯了。

Well, we're gonna say probably five words in total and you're just gonna go crazy.

告诉我们发生了什么。

Tell us what's happening.

告诉我们发生了什么。

Tell us what's happening.

我的意思是，特朗普在抱怨关税。

So, I mean, Trump is complaining about tariffs.

对。

Yes.

伊朗正处于战争的边缘。

Iran is on the brink of war.

是的。

Yes.

而埃雷博——为科技界打造真正重要事物的全新银行——已正式获得特许并投入运营。

And Erebor, the the the new the the new bank for the tech community building things that actually matter, has officially received this charter and become operational.

所以目前发生的事情相当不错。

So it's a pretty good set of things that are going on.

大获成功。

Big hit.

恭喜。

Congratulations.

大获成功。

Big hit.

让我们回溯一下，这个项目是什么时候启动的？

Take us back in When did this project start?

触发这个项目的因素是什么？

What was the inciting element?

为什么这很重要？

Why is this important?

你怎么选中这个项目来做的？

How'd you pick this to work on?

听我说，我能列举出很多原因，但归根结底，我一直想创办一家银行，主要是为了满足我自己的需求，因为当时根本没有银行真正理解我的业务、我的那些项目。

Look, there's a lot of there's a lot of things that I could point to, but it really boils down to I've been looking at starting a bank for a while primarily for my own personal use because there weren't banks out there that really understood my business, my businesses, the things that I was doing.

硅谷银行做得还算不错，但后来他们倒闭了，卷走了所有人的钱，最后还得靠政府出手救助所有人。

Silicon Valley Bank was doing a reasonable job, but then they went out of business and took everybody's money with them and had to have the government bail everybody out.

我认为，正是这件事让我真正认真地考虑这个问题。

I think that was really kind of the thing that that got me very serious about it.

我意识到，当时并没有银行与美国的利益保持一致，也没有银行真正理解并支持深科技、硬科技、能源这些复杂难懂但却至关重要的领域，而这些领域恰恰需要银行提供切实有效的服务。

I realized that you didn't have banks that were aligned with The United States, interests that were aligned with deep tech, hard tech, energy, the things that are really complex and hard to understand, but that do really matter that could also serve them in a meaningful way.

我的意思是，有专门服务农业社区的农场银行。

I mean, you have farmers banks that serve communities that are doing agriculture.

当然有一些银行在石油和天然气领域做得不错，但说到科技领域，这方面的选择就非常少了。

You have banks certainly that are doing oil and gas quite well, but when it comes to tech, it's a pretty sparse field.

于是我开始思考，如何建立一家能够服务这些领域的银行？

And so I I I started to think, how do you build a bank that could serve those things?

你该如何以一种非常保守、风险极低的方式来做？确保自己不会倒闭，不会迫使所有人依赖政府救助——也许你无法在全面金融崩溃中幸存，但至少可以做到。

How could you do it a very conservative way, very very low risk way where you can ensure you're not gonna go out of business, you're not gonna force everyone to rely on government bail You may not be able to survive a total financial collapse.

是的。

Yeah.

银行很少能幸存，但你至少可以成为最后一个站着的人。

Banks rarely can, but you can at least be the last man standing.

因此，在思考这一点的同时，我也想到许多新技术带来的可能性，比如使用美元支持的加密货币实现3.65美元或24.07美元的即时结算，这是许多企业需要但极少能获得的服务。

And so thinking about that and then also a lot of things that new technology enables, like using US Dollar backed cryptocurrencies to have $3.65, $24.07 settlement of payments, which is something that a lot of businesses need and very few get.

把这些因素综合起来，你就清楚地看到，完全有可能出现一家真正为从事实质性业务的公司服务的银行。

You kind of stack all these things together, and it becomes clear that there's there's room for a company to to to be a real bank for real companies doing real things.

那么，这里存在网络效应吗？

So is there a network effect there?

因为如果我和Jordy都在Arabore上，我们都能实现24.07美元的结算，但其他人在一个不同的银行，无法实现24.07美元结算，那我们的体验就会好很多。

Because if I'm on Arabore and Jordy's on Arabore and we can both have $24.07 settlement, but someone else is on a different bank and they don't have $24.07 settlement, we're gonna have a good experience.

然后可能就会想脱离这个网络。

And then maybe to get out of the network.

这怎么运作呢？

How does that work?

我认为会存在网络效应，但老实说，无论有什么网络效应，大概都会很短暂。

I think that there will be network effects, but to be honest, whatever network effects there are will probably be pretty short lived.

是的。

Mhmm.

我认为很快，每个人都会意识到他们需要支持这些功能，而且我认为，看到本届政府支持使用美元背书的稳定币。

I think that pretty quickly, everyone is gonna realize that they need to support these And I think that when like, seeing this administration's support for using specifically dollar backed stablecoins.

他们并不是说要让所有人都脱离美元。

They're not saying we're gonna move everybody off of the dollar.

而是要让美元不再是全球储备货币。

We're gonna move them off of, you know, the The United States currency as the reserve currency for world.

但我认为大多数银行都会被卷入其中。

But I think probably most banks are gonna get drug into this.

不同之处在于，我们是从一开始就试图这样做。

The difference is that we are trying to do it from the start.

那么，是否存在网络效应呢？

So so so is there a network effect?

大概有。

Probably.

但我认为，那些具有前瞻性、重视这种网络效应的公司，尤其是这一种，预计在未来几年内，所有银行都将被迫采用这种方式以保持竞争力。

But I think most of the forward looking companies that would value that network effect, that one in particular, see see see that in the next few years, probably all banks are gonna be forced to adopt this to be competitive.

那么，除了稳定性和低风险之外，您理想的银行体验是什么？

So what is what is the ideal banking experience for you besides stability and low risk?

能与人交谈，这是个问题吗？

Are are is being able to talk to somebody, is that a bug?

我们是不是应该让Airborne只是一个聊天机器人，当你要求与真人交谈时，它却说：‘很抱歉，我无法做到。’

Should we just is it is Airborne just gonna be, you know, a chatbot that you say talk to a human and it says, unfortunately, I cannot do that.

我有一个AI聊天机器人。

I have an AI chatbot.

我认为，在监管原因的限制下，我们还需要很长时间才会把事情交给聊天机器人处理。

I think it's be a long time before we hand things off to the chatbots if only for regulatory reasons.

比如，撇开风险不谈，这是一个巨大的因素。

Like, you set aside the risk, which is a huge factor.

撇开我们拥有自己极其高效的银行清算系统，并且可以随时完成结算这一事实不谈。

You set aside the fact that we have our own banking rails which are extremely efficient and that we can settle at any time.

我的意思是，这些本身就是非常重要的因素。

I mean, the those are huge factors on their own.

但即使你抛开所有这些明显的优点，你仍然能从一家与美国高度一致、与美国利益高度一致、与美国国防部以及情报机构高度一致的银行获得很多其他好处。

But even if you knock off kind of all of these all of these obvious advantages, there's a lot of other things that you get from having a bank that is strongly aligned with The United States, strongly aligned with US interests, strongly aligned with US Department of War and also the intelligence community.

而这些优势在很多方面都有所体现。

And then there there's there's a lot of ways that that manifests.

例如，我们正在做的一件事是，不是只在法院下令后才与情报机构合作，去应对平台上可能发生的犯罪行为，而是主动出击，提前介入，说：不。

For example, one of the things that we're doing is rather than working with the intelligence community only under court order to kinda go and, you know, respond when they think there's crime going on in the platform, we're preemptively going out there and saying, no.

我们会从一开始就与他们合作，帮助让欺诈行为几乎不可能发生在我们的平台上，这对我来说也很棒，因为这意味着想从事欺诈的人会去其他银行。

We're gonna work with them from the very, very beginning to help make fraud almost impossible on our platform, which is great for me too because it means that people who want to engage in fraud are gonna go to other banks.

这意味着不想从事欺诈的人会非常高兴能在一个没有此类活动的平台上交易。

It means that people who don't want to engage in fraud are going to be thrilled about being on a platform that doesn't have that type of activity on it.

即使在他们自己的平台上，欺诈行为在他们不知情的情况下存在，他们也知道有一个愿意与政府合作的伙伴，而不是隐瞒这些事情。

To the extent that it exists inside of their own platform against their will, they know that they have a willing partner that is willing to work with the government rather than hide these, you know, hide these things.

你可以把它想象成汇丰银行的反面，汇丰曾为贩毒集团提供银行服务，并拼命避免任何可能让公众市场知晓此事的政府干预。

I'd say maybe think of it as like the opposite of HSBC where they were banking the cartels and desperately trying to avoid any sort of government intervention that would make that clear to the public markets.

把Erobor放在这个光谱的另一端。

Put Erobor way at the other end of that spectrum.

我认为还有很多其他银行，不管他们是好人还是坏人，这都无关紧要。

I think there's also a lot of other banks that, you know, whether whether they're good or bad people is irrelevant.

你可能会遇到一些好人，但他们却极度依赖欧洲市场、中国市场或其他外国市场，他们必须让这些市场满意，要么是为了保持盈利，要么是不想让他们的高管在这些司法管辖区被逮捕。

You might have good people who nonetheless are very beholden to European markets, Chinese markets, other foreign markets that they need to keep happy either because they wanna keep them happy and make money or because they don't want their executives in those jurisdictions to get arrested.

Arabore采取了截然不同的做法。

Arabore is taking a very different approach.

我们说的是，你知道吗？

We're saying, you know what?

我们是一家美国公司。

We're an American company.

我们是一家支持美国企业的美国银行，我们会遵守美国法律，但不会服从那些对我们根本没有管辖权或权威的人发布的无理裁决。

We're an American bank that supports American companies, and we will comply with US law, but we will not comply with spurious rulings from people who have no real jurisdiction or authority over us.

你很难找到一家持这种立场的银行。

And you'll find a very hard time finding a bank that takes that position.

我不了解除我们之外还有哪家。

I'm not aware of any other than us.

你如何看待产品的推出？

How do you think about rolling out products?

银行可以提供从抵押贷款到设备融资，再到风险债务和资本募集，甚至帮助企业上市的所有服务。

A bank can do everything from mortgages to equipment financing to venture debt to capital raises, take someone public.

这就是你去大银行时，他们会帮你上市的原因。

That's kind of what you know, you go to a big bank, they're gonna help you IPO.

你最先想做什么？

What do you wanna do first?

目前，你知道，我们刚刚开业，所以请给我们一点时间。

That right now, you know, we we look, we've just barely opened, so give us a little bit of time.

我们优先推出的產品完全基于客户反馈，即他们告诉我们需要什么，以及我们最能帮助他们解决什么问题。

The products that we are prioritizing are based entirely on the customer feedback that we're getting, the things that they tell us they need, the things that they tell us that we can most help with.

在我看来，银行的存在几乎是一种过时的东西。

In a way, the existence of a bank to me is almost this antiquated thing.

你本希望公司能自己完成所有这些事情，或者至少大部分能自己完成，但监管环境和银行体系的运作方式要求必须有一个受监管的银行作为中介来促成这些事情。

Like, you you would hope that companies could really just do all of this stuff themselves or at least most of it themselves, But the the regulatory climate, the way that the banking system works requires that you have this intermediary in the form of a regulated bank to make these things happen.

因此，我基本上把我们的业务视为那些需要我们的公司的组成部分，我们会与他们紧密合作，专注于他们关心的事情。

And so I I kind of just think of our business as part and parcel of these companies that need us, and we're gonna work very, very close with them closely with them and focus on the things that they care about.

这可能不会包括，例如，住宅抵押贷款，而是像设备融资这类非常保守、不会演变为风险管理问题的业务，帮助他们以最不痛苦的方式与传统银行系统对接。

That's probably not going to be, for example, you know, residential mortgages, but things like equipment financing, very conservative, you know, very conservatively played where it's not gonna turn into a risk management problem, allowing them to work with the legacy banking system and plugging into that in the least painful way.

这些正是我们许多客户非常感兴趣的方向。

These are things that we have a lot of customers who are very interested.

哦，我们不能丢掉他们的钱。

Oh, and we need to not lose their money.

是的。

Yeah.

如果我们能做到这一点，我们就比许多其他银行做得更好了。

If we can do that, we'll be doing better than many other banks.

是的。

Yes.

是的。

Yes.

谈谈走特许经营路线的利弊。

Talk about the trade offs of going the charter route.

你知道，过去十年，创始人使用的大多数银行产品都是建立在现有特许经营机构之上的，或者说是的。

You know, the last ten years, most of the banking products that founders have used were built on top of existing charters or or Sure.

将存款分散到多个

Spreading deposits across multiple

避免获取特许经营权。

avoid getting a charter at all.

你知道，他们想成为一家金融科技公司，实际上他们回避了一个事实：他们本质上就是在做银行的事情。

You know, they wanna be a fintech company and they basically will dance around the fact that realistically what they're doing is basically being a bank.

听好了，没有自己的特许经营权，你依然可以做很多事情。

Look, there's a lot of things that you can do without having your own charter.

你可以通过与其他公司合作来完成一些事情。

There's things you can do partner with other companies.

没有自己的特许经营权，你无法做很多事情。

There's a lot of things you cannot do without having your own

反对过度快速增长。

anti anti hyper growth.

比如，即使市场需求存在，你也不可能明天就让存款达到500亿美元。

Like, you can't just to 50,000,000,000 of deposits tomorrow even if there's the demand.

对吧？

Right?

所以也许

So maybe

我们决定要申请自己的银行牌照，因为我们打算运营的银行类型，以及希望由我们自己承担最终责任。

we we decided that we needed to get our own charter because of the type of bank we were gonna try to run and the ability to have the buck stop with us.

一旦你依赖他人提供关键基础设施、关键风险管理、关键许可事宜，他们就可以把你踢出平台。

The moment that you're dependent on somebody else for the key infrastructure, the key risk management, the key licensure questions, they can kick you off the platform.

他们可能被迫把你踢出平台。

They can be forced to kick you off the platform.

他们可能会这么做，也许是因为正当的理由。

They might and maybe they're doing it for good reasons.

也许是因为不好的理由。

Maybe they're doing it for bad reasons.

我的意思是，你已经看到这种情况发生过，比如Steam和他们的支付处理商、他们的银行以及支付处理商的银行，他们基本上是在审查游戏，因为有一群人互相推诿，都说不是自己在施加影响。

I mean, you've seen this play out, for example, with Steam and their payment processors and their banks and their payment processors' banks where they're basically censoring games because there's some circle of people who all claim it's not them pointing down the chain.

是那个人，是那个人，是那个人，但他们都一致认为，即使无法达成一致谁在导致审查，只要你不审查，他们就会抛弃你、关闭你的账户、把你从银行系统中剔除、切断你的支付通道。

It's that guy, it's that guy, it's that guy, But they all agree, even if they can't agree, who's causing the censorship, that if you don't censor, they're going to drop you, they're going to drop your account, they're going to debank you, they're going to depayment you.

因此，当你让自己依赖另一家银行或另一家支付公司时，你必须问自己：我是不是在制造一种局面，让他们替我做决定？

And so one of the things that you have to ask yourself when you're making yourself dependent on another bank or another payment company is, am I setting this up where they get to make my decisions for me?

他们是替我做决定，还是中国共产党在做这些决定？

And are they making their decisions for me or is the Chinese Communist Party making those decisions?

是不是某个对我们的网络上的某些业务极度敌视的政治团体在真正做出这些决定？

Is some political party that is extremely hostile to some business on our network the one actually making that decision.

所以，如果你不掌控自己的特许经营权，不掌控自己的命运，责任就不会落在你身上，你总是得让其他方满意。

So if you don't control your own charter, if you don't control your own destiny, the buck doesn't stop with you and you're always gonna have to keep other parties happy.

我认为这并不是我们感兴趣的事情。

And I I think that was not something that we're interested in.

我们希望按照自己认为应该的方式运营这家银行。

We wanted to be able to run the bank the way that we think it needs to be run.

如果我们决定终止某人的账户，我们必须为此承担责任。

And if we are debanking somebody, we are gonna have to take responsibility for that.

我认为这实际上会带来更好的行为。

And I think that that actually leads to much better behavior.

我非常希望看到这种行为在整个行业推广开来。

I would love to see that type of behavior across the industry.

实体网点。

Physical branches.

你对此有什么想法？

How are you thinking about that?

我认为我们会从诺克斯堡开始，然后逐步扩展。

I think we're gonna start with Fort Knox, and we'll expand from there.

我喜欢这个想法。

I like it.

是的。

Yeah.

也许我们可以更深入地探讨一下。

Maybe actually dive deeper there.

比如，如今的银行到底是什么？

Like, what is a bank these days?

我认为大多数人能想象到，设置一个存放加密货币的钱包。

I think most people can conceptualize, like, setting up a crypto wallet that holds coins.

但当你想持有真正的美元时，我猜你并不会真的拥有一座实体金库。

But when you wanna hold real dollars, you don't actually have a physical bank vault, I imagine.

所以一旦你获得了特许状，这是否只是赋予了你维护数据库和账本的能力？

So once you get the charter, does that just give you the ability to, like, maintain a database and a ledger?

那么，什么是银行？

Like, what is a bank?

什么是银行？

What is a bank?

我的意思是，银行可以有多种形式。

I mean, there's lots of forms it can be.

如果你想要所有细节，实际上可以阅读我们大量的申请材料。

I think if you want all the details, you can actually read through a lot of our applications.

其中一些内容实际上是公开记录，你可以查看到。

Some of it's actually a matter of public record and you can see that.

Andoril 会抱歉。

Andoril will sorry.

不是 Andoril。

Not Andoril.

Erebor 将拥有一个实体金库。

Erebor will have a physical vault.

哦，而且我们会在里面存放大量财宝。

Oh, And we will have a lot of treasure in it.

所以，好吧。

So Okay.

我们并不是想成为一家纯粹数字化的、纯粹有趣的公司。

We are not trying to be a pure digital, pure Interesting.

你知道，一家纯粹的以太坊公司。

You know, pure ether company.

事实上，许多对与我们合作感兴趣的公司都希望我们拥有安全的实体金库存储。

In fact, a lot of the companies that are very interested in working with us want us to have secure actual vaulted storage.

我不是说这会是我们首先关注的重点，或者是我们主要的差异化优势，但我们并不是一家以太坊银行。

I'm not saying that that's going to be the first thing that we focus on or that's our key differentiator, but we are not a we are not a bank of the ether.

我们是TerraFirma公司。

We are we are we are we are we are of TerraFirma.

是的。

Yeah.

上次你来的时候，我们聊了很多关于ModRetro的内容。

Last time you came on, we talked a lot about ModRetro.

我很想了解一下M64的最新进展。

I'd love an update on the m 64.

这个产品有什么特别之处？

What makes that product special?

相比Chromatic项目，你们在M64上走得更远了哪些方面？

Where did you go further with this project than with the chromatic?

还有就是，总的来说，你对M64有什么想法？

And just general, like, you're thinking about the m 64?

对于不熟悉的人，M64 是我对任天堂64的致敬之作，而任天堂64是有史以来最伟大的多人游戏主机之一。

Well, so for people who aren't familiar with it, the m 64 is a tribute to the Nintendo 64 that I'm building, one of the greatest multiplayer consoles of all time.

我们正在打造一款兼容所有经典N64游戏的主机。

We're building a console that is compatible with all classic n 64 games.

我们正在重新发布许多经典N64游戏，以便你如果还没有的话可以购买。

We're re releasing a lot of classic n 64 games so that you can buy them if you don't already have them.

我们还在推出一些被取消的N64游戏新版本，这些游戏从未上市，要么是因为被取消并移植到了GameCube，要么是因为资金耗尽，要么是因为工作室因人际问题而崩溃。

We're actually doing new releases of canceled n 64 games that never even made it to market either because they were canceled and ported to GameCube or because they ran out of money or because the studios imploded for interpersonality reasons.

有很多精彩的内容即将推出。

It's there's a lot of cool stuff coming out.

M64的主要更新是，它现在已进入大规模生产阶段。

The m 64, guess, main update is that it's in mass production right now.

它不是‘迟早会来’，

It is it is not eventually coming.

它现在就已经在大规模生产了。

It is in mass production right now.

我们决定，直到我们备足足够的库存以满足需求后，才会正式发布。

We've decided we're not gonna launch it until we build up sufficient stock to meet the demand.

我认为我们在推出Mod Retro Chromatic（致敬Game Boy和Game Boy Color）时犯了一个错误：我们一开始就开始销售，结果突然就卖断货了。

One of the mistakes I think we made with the mod retro chromatic, which was a tribute to the Game Boy and Game Boy Color, was that we started selling them and then all of sudden we sold out.

然后，当人们听说了它、读到相关报道、朋友都在讨论时，他们想买一台，却发现已经售罄了。

And then you have people who hear about it, they read about it, their friends are all talking about it, they go to try to buy one and they can't buy one because it's sold out.

那么问题来了：他们还会记得回头再来购买吗？

And there's a question, will they remember to come back and buy it some other time?

他们还会保持立即购买的冲动吗？

Will they still feel like going and immediately buying it?

是的。

Yeah.

没错。

Exactly.

因此，我们最终陷入了一种毫无后续的行动陷阱。

And so we we you kind of end up with this with this dead end call to action.

因此，这次我们决定，顺便说一下，这对与我们合作的游戏发行商来说尤其糟糕。

And so we've decided this time and and by the way, this is especially bad for the game publishers that we're working with.

假设有人在重新发布一款游戏，或者有人在开发一款全新游戏，我们正在与一些独立开发者合作推出Chromatic系列。

Let's say that it's someone who's re releasing a game or somebody who's doing a brand new game, some of the indie developers we're we're working with on the Chromatic.

他们正在积极推广自己的游戏。

They're out there promoting their game.

他们正在努力营销自己的游戏。

They're trying to market their game.

他们正在进行一整套营销活动，试图说服人们：快来买我的新Game Boy游戏吧。

They're doing a whole marketing activation, trying to convince people, hey, come and buy my new Game Boy game.

快来买我的新Nintendo 64游戏吧。

Come and buy my new Nintendo 64 game.

但如果买不到主机，会发生什么？

Well, what happens if you can't buy the console?

很多人会说：哦，我改天再来买，但后来就再也不买了。

A lot of people are like, oh, I'll come back later and buy it, and then they don't.

因此，这次我们不想让我们的开发合作伙伴失望，一定要确保库存充足。

And so what we wanna do is not let down our developer partners this time and make sure that we maintain stock.

所以我们的目标是，无论有多少人访问我们的网站，N64都不会售罄。

So the goal is that the m 64, no matter how many people hit our website, is not going to sell out.

因此，我们正在大规模生产。

And so we are mass producing.

到目前为止，我们已经囤积了数万台。

We're stockpiling, I think, tens of thousands of them so far.

还会继续增加更多。

More and more to come.

那么，接下来的供货情况如何？

What is the what is the supply soon.

目前供应链的实际状况是怎样的？

What's the supply chain actually like right now?

我们已经讨论过玩家对医疗问题的反抗。

We've covered the the gamer rebellion against Medical.

反对AI，是的。

Against AI Yeah.

内存。

Memory.

玩家们正在崛起。

Gamers are rising up.

你知道，因为我们所使用的硬件类型，供应链还行。

You know, the supply chain is okay because of the type of hardware that we're building.

当然。

Sure.

现在有很多消费类硬件，人们都在抢购内存，这成了大问题。

There's a lot of consumer hardware right now where people are trying to buy memory and it's a huge problem.

但别忘了，Nintendo 64的内存本身并不多。

But remember that the Nintendo 64 doesn't have exactly a lot of memory.

它默认是16兆字节，加上扩展包的话是32兆字节。

You know, it had the what was 16 megabytes by default, 32 megabytes with the expansion pack.

结果发现，你并不需要那么多内存

And so it turns out that you don't need that much memory

是的

Yeah.

就能完美复制任天堂64的原始硬件

To perfectly replicate the Nintendo 64 original hardware.

你能

Can you

也许这实际上是今年受内存影响最小的消费电子产品。

Maybe it might be it might it might it might be actually the least impact less the least memory impacted consumer electronics product of the year.

所以，这从一开始就是你的计划。

And so this was this was your plan then all along.

对

Yeah.

你预见到了AI热潮的到来，于是决定打造最后一款主机。

You saw this AI boom was coming and you said, I'm gonna make the the final console.

是的。

Yeah.

现在我们得回到猴子的话题。

Now we gotta return to monkey.

我们必须回到起点。

We got we gotta go to go back to the beginning.

我的意思是，我在这里用32兆内存到底在干什么？我的意思是，你还记得比尔·盖茨说过什么吗？

I mean, really, what am I doing here with some with 32 megabytes of I mean, you you remember what you remember what Bill Gates said.

对吧？

Right?

那是什么来着？

What what was it?

64KB的内存对任何人来说都够用了？

64 kilobytes of RAM ought to be enough for anyone?

没错。

Yep.

让我们来想想怎么用AI来，嗯。

Let let let's let's let's figure out how to use AI to yeah.

让我们想想如何用AI构建内存占用合理的应用程序。

Let's figure out how to use AI to build applications that fit within reasonable memory footprints.

哦，是的。

Oh, yeah.

我只说了一半是认真的。

I'm only half kidding.

我想你可能看过我和Shamath在Twitter上争论，是的。

I I think you might have seen me going back and forth with Shamath on Twitter Yeah.

几周前，我们讨论过，这个用AI构建软件的AI软件平台，应该明确支持导出兼容Windows XP Service Pack 3的应用程序。

Couple weeks ago talking about how the this AI software platform for building building software using AI should definitely support export to Windows XP Service Pack three compatible applications.

我真的只说了一半是认真的。

And I really am only half kidding.

实际上，还有很多政府硬件和金融硬件仍在运行这些系统，主要是离线设备。

There there is actually a lot of government hardware that is still running those, financial hardware that's still running these, mostly offline stuff.

是的。

Yeah.

因此，让应用程序能在这些系统上运行实际上是有用的。

And so having apps that can run on those is actually useful.

但天啊，要几GB内存才能有个开始菜单，这感觉太疯狂了。

But, man, it feels crazy to need a few gigabytes of RAM to have a start menu.

需要16GB内存花半小时去搜索一堆文本文件，这简直太离谱了。

It feels pretty crazy that you need 16 gigabytes of RAM to take thirty minutes to search through a bunch of text files.

我们肯定能做得更好。

Surely, we can do better than this.

你能谈谈你如何看待游戏在培养下一代孩子中的作用吗？

Can you talk a little bit about how you think about gaming in the concept of just raising the next generation kids?

我有三个孩子，没错。

I have three three kids, and Right.

我想我完全可以弄个模拟器，跑在iPad上，直接塞到我孩子面前。

I think I could probably get an emulator and run it on an iPad and slam that over to my kid's face.

他可能根本分不出区别，但从另一个角度看，M64 有些你不能做的事反而是好事，确实如此。

And he might not know the difference, but is there another angle to the m 64 that is about, like, some of the some of the things that you can't do with it are actually good Sure.

对这个世界来说？

For the world, maybe?

我的意思是，你得记住，游戏行业和许多其他行业一样，早就找到了真正有效的方法。

I mean, you have to remember that the gaming industry, like many industries, figured out what actually worked a long time ago.

他们弄清楚了如何制作有趣的游戏。

They figured out how to make fun games.

他们找到了让人持续回来的方法，但如今已从创新转向了榨取。

They figured out to make things that make people keep coming But it's moved from innovation to extraction.

我要说明一下，这句话是我从 Framework 的创始人 Nirav Patel 那里借来的，因为他表达得比我好得多。

And I will note, I stole that line from Nirav Patel, founder of Framework, because he put it so much better than I did.

但确实有一些行业已经从创新转向了榨取。

But there's these industries that have moved from innovation to extraction.

现在它们进入了榨取阶段，拼命赚钱，而对大多数行业资金而言，突破极限已不再是首要任务。

And they're now in extraction phase where they try to make tons and tons and tons of money, and actually pushing the limits is not the top priority for some huge majority of the dollars that are in industry.

因此，有很多东西你可以学习，尤其是当你从零开始，回到人们刚刚掌握基础知识、仍在创新、打造真正吸引人的事物的时候。

And so there's a lot of stuff that you can learn, especially when you're learning it from the ground up by going back to when people were just getting the basics right, when they were still innovating, building things that were really compelling to people.

我最近当爸爸了。

I I recently had a kid.

我可能很快就要迎来第二个孩子。

I I might have a second on the way.

我计划从经典游戏开始教他们，即使我都不好意思说出口。

I'm planning on starting them with I would don't even if I'd say the classics.

这并不是什么复古玩法。

This isn't like a vintage play.

这是给他们打下良好的基础。

It's it's starting them with a good foundation.

这个基础恰好主要是由老游戏构成的。

That foundation happens to be largely old games.

是的。

Yeah.

但我不会让他们陷入如今主导儿童游戏的老虎机和微交易赌博陷阱中。

But I am not gonna be throwing them into the slot machine micro transaction gamble rama that is dominating kids games today.

这看起来简直是一种疯狂的做法。

It just seems like a like a like a crazy thing to do.

这是一个斯金纳箱。

It's a Skinner box.

是的。

Yeah.

别进斯金纳箱。

Don't go in the Skinner box.

谈谈氛围编程。

Talk about vibe coding.

如果你20岁时就有氛围编程，你的生活会是什么样子？

What would your life be like if vibe coding existed when you were 20?

对……的建议

Advice for

或者12岁。

Or 12.

现在20岁、生活在 vibe coding 世界中的人们。

People who are 20 now in the world of vibe coding.

你很早就进入了硬件领域。

You went into hardware very early.

这感觉是个很棒的决定。

That feels like a great move.

这现在还是个很棒的决定吗？

Is that still a great move?

给我们详细讲讲所有

Walk us through all

这些。

that.

我的意思是，vibe coding 最大的受益者将是我这样的硬件极客。

I mean, the the the biggest beneficiaries of vibe coding are going to be the hardware nerds like me.

这将是那些擅长形状旋转的人，而不是那些处理文字单元的人。

It's going to be it's going to be the the shape rotators, not the word cells.

每个人都关注文字单元如何被AI取代，但对于形状旋转者来说，这将变得不可思议。

Everyone's focused on on on on how the words on how the word cells are going to be wiped out by AI, but but but for the shape rotators, it's gonna be incredible.

你知道，我一直是个很糟糕的软件工程师。

You know, I was always a pretty terrible software engineer.

我不是科班出身的程序员。

I'm I'm I'm not a programmer by trade.

我自学了一些技能，足以把各种东西拼凑起来并让它们运行。

I've taught myself enough to glue things together and make them work.

但当我创办第一家公司的时侯，Oculus，或者更早的时候，我14、15岁创办Mod Retro时，我对软件的了解 barely 够用。

But when I started my first company, Oculus, or actually, when I started Mod Retro when I was 14 or 15, you know, I I knew just barely enough about software.

如果当时我能把烂活儿外包给电脑去做，我其实能比自己动手快得多地完成事情。

If I would have been able to build crappy stuff by outsourcing it to a computer, I actually would have been able to accomplish things a lot faster than trying to do it myself.

有人可能会说，帕尔默，你本该去学编程的。

And people might say, Palmer, you should have just learned to code.

你知道，你本该自己学会去做这件事的。

You know, you should have should have should have just learned to do it yourself.

但我觉得，看看我的经历就很容易明白，我19岁就创办了Oculus，而从15岁起我就一直在做原型。

But I think it's pretty easy to look at my track record and realize I started Oculus when I was 19 after building prototypes since I was 15.

我之所以能取得这些成就，是因为我专注于自己擅长的领域——光机设计、一点电子工程，以及将所有这些不同组件整合成产品。

I was only able to accomplish what I accomplished because I focused on what I was good at, which was optomechanics, a little bit of electrical, and then the product integration of all of these different components.

我没有时间去学习编程。

I didn't have time to learn to program.

如果我再花一两年时间去学编程，哪怕只是达到普通水平，我也会在其他所有事情上落后两年。

If I'd spent another year or two learning to program at even a reasonable level, I would have been two years behind on everything else.

所以我非常支持‘氛围编程’，即使产出的东西全是垃圾，即使全是烂代码。

And so I am a big fan of vibe coding even if everything that comes out of it is slop, even if it is all shit.

但那也比我亲手做的要好。

It's better than I was able to make.

嗯。

Mhmm.

你最近怎么看待VR的？

How are you thinking about VR these days?

我最近在Apple Vision Pro上从头到尾看了一遍《黑客帝国》。

I recently watched The Matrix from start to finish in an Apple Vision Pro.

我看得很享受。

I enjoyed it.

如果你没有家庭影院，现在唯一真正能做的事就是搞个家庭影院。

It feels like the only real thing that you can do these days is just the home theater if you don't have a home theater.

但感觉整个行业根本没拥抱这一点。

But it doesn't feel like the industry's embraced that at all.

在显示技术似乎已经足够好，能够满足这种狭窄用途（只要减轻重量并重新布局）的时候，每个人却都在削减投资、撤回资金。

And everyone's sort of writing down their investments and pulling back at a time when the display technology does seem to be good enough for that narrow use case if you cut out the weight and move things around.

但你希望消费级VR未来走向何方？

But where do you see this where do you where do you want consumer VR to go?

而你认为它实际上会走向哪里？

And then where do you think it actually is going?

所以我要挑战你这个前提，即人们正在撤资。

So I will challenge your premise there, which is that people are pulling back.

我认为这主要是由一些人炒作Meta公司裁员造成的。

I think that's driven mostly by people sensationalizing meta, firing a bunch of people.

但你必须记住，他们在一次裁员中就裁掉了Reality Labs团队的10%。

But you have to remember, they got rid of 10% of the Reality Labs team in one layoff.

别忘了，他们原本每年就有大约8%到9%的自然流动率，这还不包括被解雇的人。

Remember that they have something like eight or 9% annual churn naturally and organically, not including firing.

是的，没错。

So Sure.

你真正谈论的，其实就是把六个月的自然流失提前压缩到一个月内。

I mean, what you're really talking about is basically pulling six months of churn forward into a single month.

是的。

Sure.

而且他们在VR上的投入仍然比任何其他公司高出一个数量级。

And they're still spending more on VR than anybody by an order of magnitude.

他们表示并没有减少在内容上的投入。

They've said they are not spending less on content.

他们只是不再把资源投入到像Facebook或Meta Horizon这样的自有产品上。

They're just not putting it into first party stuff like Facebook or Meta Horizon.

所以你已经看到，截至昨天，Meta Horizon已不再被定位为一款VR应用。

So you've seen they've announced Meta Horizon is no longer a VR app as of yesterday.

它现在仅是一款以移动设备为重点的应用。

It is now a mobile focused app only.

因此，他们实际上终止了一些没有意义的失败项目，并将这些资源转向了正在取得成效的领域。

So what they've done is they basically killed a few of these failing efforts that didn't make sense, and they're now putting those resources into things that are working.

我曾在Twitter上写过一些关于这一点的内容，人们需要明白，这并不意味着VR的终结。

And so I I I've I I wrote a I wrote a I wrote a bit about this on on Twitter where people need to understand this is not the end of VR.

它并没有崩溃。

It's not it's not collapsing.

他们仍在投入巨额资金。

They are still spending an enormous amount of money.

它们的规模比其他任何公司都大一个数量级。

They're bigger than anybody else by an order of magnitude.

数量级。

Magnitude.

我认为，如果你关注未来十二个月内Meta和其他公司即将推出的产品，你会看到大量进展。

I I and I think if you pay attention to what's gonna be coming out from Meta and others over the next twelve months, you're gonna see a lot of progress.

比如，Apple Vision Pro，没错，它在视觉质量和显示质量上确实领先于所有人。

Like, Apple Vision Pro, yes, they've so, yes, Apple Vision Pro was ahead of everybody on visual quality, on display quality.

但让我告诉你一个简短的嗯。

But let me tell you a short Mhmm.

想象一家美国公司去找一家日本显示屏供应商，说：我们想买你们最新的4K微OLED显示屏。

Imagine that an American company went to a Japanese display vendor and they said, we want you to sell us your new four k micro OLED displays.

这家公司向他们提供了这些微OLED显示屏的样品。

The company gives them samples of those micro OLED displays.

这些样品的成本大约是每块一千美元，因为这条生产线的良率只有大约10%。

The samples cost about a thousand dollars each because the yield on that production line is only about 10%.

换句话说，生产线上90%的显示屏都不达标，无法正常工作，因为这些只是工程样品。

In other words, 90% of the displays coming off the production line are not up to par, they're not working, and that's because these are engineering samples.

它们本就不是设计用来作为实际产品的。

They're not meant to actually be a working product.

他们说：听好了，这是我们的工程样品。

And they said, listen, here's our engineering samples.

再等两年，良率就会提升到90%。

Just wait two years and the yield is gonna be up to 90%.

到时候，我们可以以几百美元的价格把这些卖给你。

We'll be able to sell these to you for a couple $100.

然后想象一下，这家美国科技公司接到一个叫蒂姆·苹果的人打来的电话，他说：不行。

And then imagine that that American tech company had a a a guy named Tim Apple call them up and say, no.

我们现在就要用这些工程样品来打造一款产品。

We're gonna build a product with this right now today using our engineering samples.

他们说：但蒂姆·苹果，这太疯狂了。

They said, but Tim Apple, that's crazy.

这意味着你的头显成本将高达3000美元。

That means that your headset was gonna cost like $3,000.

他说，我不在乎。

He said, I don't care.

我要推出一款头显，计划在2027年发布，是的。

I'm going to sell a headset that should launch in '27 Yeah.

在2024年、2025年，我要以3500美元的价格推出它。

In 2024, 2025, and I'm gonna do it for $3,500.

这就是Apple Vision Pro的定位。

That's what Apple Vision Pro is.

它从来就不是为当下而设计的产品。

It was never intended to be a product of the times.

它是通过投入巨额资金，将未来的产品拉到了现在。

It's a product of the future hauled into the present by spending enormous amounts of money.

因此，Meta、索尼、苹果、谷歌以及其他所有公司都将达到这种视觉保真度。

And so Meta and Sony and Apple and Google and all of these other companies, they are going to be hitting that level of visual fidelity.