本集简介
双语字幕
仅展示文本字幕，不包含中文音频；想边听边看，请使用 Bayt 播客 App。
你正在观看TVPN。
You're watching TVPN.
今天是2026年3月4日，星期三。
Today is Wednesday, 03/04/2026.
新发型。
New haircut.
注意，新发型上线了。
New haircut alert.
我们正在TVPN现场直播。
We are live from the TVPN.
终极科技神殿，金融堡垒，资本之都。
Ultimate Temple of Technology, the fortress finance, the capital capital.
让我告诉你关于ramp.com的事。
Let me tell you about ramp.com.
时间就是金钱。
Time is money.
同时保存两者。
Save both.
轻松使用企业卡片、账单支付、会计以及其他诸多功能，全部集中在一个地方。
Easy use corporate cards, bill pay, accounting, and a whole lot more all in one place.
我们今天的节目会短一点。
We have a little bit of a shorter show.
让我们调出线性排期。
Let's pull up the linear lineup.
线性，当然是现代软件开发的系统。
Linear, of course, is the system for modern software development.
70%的企业工作空间在使用线性时都采用了智能代理。
70% of enterprise workspaces on linear are using agents.
我们邀请了迪恩·鲍尔来阐述鲍尔对汤普森关于Anthropic观点的另一面。
We have Dean Ball coming on to give the other side of the Ball versus, Thompson argument, about Anthropic.
接着我们有斯科特·库波尔和杰瑞德·艾萨克曼来谈论太空。
Then we have Scott Kupor and Jared Isaacman coming on to talk about space.
我们要去月球，他们会带我们去。
We're going to the moon, and they're gonna take us there.
接着我们有来自Skydio的亚当·布赖，我大约十年前曾通过安德森·霍罗威茨与他合作，然后是Eight Sleep的首席执行官马泰奥，他有一些令人兴奋的消息，还有迪伦将前来解析最新的AI代理Hermes代理。
And then we have Adam Bry from Skydio, who I actually worked with through Andreessen Horowitz, maybe ten years ago, and then Matteo, CEO of, Eight Sleep has some exciting news, and Dillon's coming on to break down the latest AI agent, Hermes agent.
好吧，不管怎样，这周有一些重大新闻。
Well, anyway, there was some big news, this week.
当然，大家都预料到苹果会推出新产品，但还是有一些惊喜，从MacBook Neo开始，有一些很酷的东西。
Of course, everyone knew that Apple was launching new products, but there were some surprises, And, there's some cool stuff starting with the MacBook Neo.
我更深远的思考是，苹果如何受到AI热潮的影响？
My the the the more downstream take is, like, how is Apple affected by the AI boom?
但让我们先从实际产品说起，因为这个东西在视频里看起来太小了。
But let's start with the actual products because, this thing, in the video, it looks so small.
它看起来像一台八英寸的笔记本电脑。
It looks like an eight inch laptop.
我想这是MacBook Neo，旨在与Chromebook竞争。
I guess this is, so this is the MacBook Neo sort of designed to compete with the Chromebook.
它的售价是5.99美元。
It comes in at $5.99.
不仅仅是Chromebook，而是所有的PC。
Well, not even just the Chromebook, just all PCs.
是的。
Yeah.
但我的意思是，如果你在专业工作环境中，肯定不会选这个，你会用带多个显示器的台式机，进行一些高级工作。
But, I mean, you're not gonna get this for, like, if you're in a professional work place, go to a desktop with multiple monitors, plug in, you're doing, like, advanced work.
这主要是给学生用的。
Like, this is for students.
这是面向低端市场的。
This is for the low end.
或者只是作为个人电脑使用。
And in Or just a personal computer.
我明白了。
I see.
我觉得它是为那些有iPhone但没有Mac的人设计的。
Think it's designed for people that Yeah.
因为价格原因。
Have an iPhone but don't have a Mac Yeah.
因为成本问题。
Because of costs.
是的。
Yeah.
对。
Yeah.
嗯。
Yeah.
现在他们可以拥有一套集成的软件。
And now they can have an integrated suite.
确实是。
It's yeah.
我的意思是，这真是个疯狂的产品。
I mean, it's a it's a crazy product.
我也超爱那个发布视频。
I I love the I love the launch video too.
一开始特别奇怪，因为只是一个空白的粗体方块，然后他们用CGI逐步添加功能。
It's very weird when they start because it's just this, like, blank, bold block, and then they're adding features to it with CGI.
它做得非常好。
It does a really good job.
有点令人失望，因为他们添加的每个功能都跟你预期的一模一样。
It's sort of anticlimactic because every feature they add is just like what you'd expect.
是的。
Yeah.
我当时还想着，哦，这会是触屏的吗？
I was kind of like, oh, is this gonna be a touch screen?
结果却说，哦，不是。
And then it's like, oh, no.
它就是一个普通的屏幕。
It's just a normal screen.
它就像，呃，猜猜看，它有个键盘。
It's like, like, guess It has a keyboard.
主页，麦克风静音。
Home, mic drop.
哦，它有两个USB-C接口。
Oh, it has two USB C ports.
他们把耳机插孔放得更低了一点，这还挺酷的。
It is kinda cool that they put the, the headphone jack a little bit further down.
我觉得这很酷。
Thought that was cool.
然后他们说，看看这个。
And then they're like, check it out.
你绝对猜不到这个。
You're never gonna guess this.
一个摄像头，我们把它装到了电脑上。
A camera, we put it on the computer.
他们说，嘿。
They're like, hey.
我们听说你喜欢用iPad，是的。
We heard you liked having an iPad Yeah.
你可以给它配个键盘。
That you could attach a keyboard to.
所以我们 basically 把一个iPad装上了键盘。
So we basically took an iPad and put a keyboard.
但他们这样扩展产品的方式真的很酷，就像在告诉你，看，好吧。
But it's very cool how how they, like, sort of expand the product, like, show you, okay.
这里有扬声器。
There's the speaker.
它相当大。
It's pretty big.
它装在侧面。
It goes on the side.
有两个，给你展示了。
There's two of them, shows you.
但在我看来，这看起来不像一台13英寸的电脑。
But this does not look like a 13 inch computer to me.
在我看来，这看起来像一台8英寸的电脑。
This looks like an eight inch computer to me.
这看起来像一个小iPad，而13英寸是非常标准的尺寸。
This looks like a small iPad, and 13 inches is a very normal size.
也许这个型号就像沙克之类的东西。
Maybe the model is just like Shaq or something.
是的。
Yeah.
是的。
Yeah.
不。
No.
我觉得他们实际上用了一些大手来制作，可能是为了让它看起来更小之类的。
I I think they actually use some of the big hands to make it, like, maybe look smaller or something.
但13英寸的显示屏，十六小时的电池续航。
But, 13 inch display, sixteen hour battery life.
价格才是最疯狂的地方。
The price is the really crazy thing.
所以是5.99美元。
So $5.99.
对学生来说，便宜100美元。
For students, it's a $100 cheaper.
所以是4.99美元，这是苹果有史以来最便宜的产品之一。
So $4.99, that is one of the cheapest Apple products they've ever made.
我敢肯定耳机的价格更高。
I'm pretty sure the headphones cost more.
苹果Pro，苹果AirPods Max的价格大约是500美元。
The Apple Pro, the Apple AirPods Max cost like $500.
非常便宜。
Remarkably cheap.
搭载M18芯片，256GB或512GB存储，四种颜色可选。
A 18 pro chip, 256 gigs or five twelve of storage, four colors.
他们推出了柑橘黄配色。
They got it in ramp yellow.
他们称之为柑橘色，但我认为每个人都会
They call it citrus, but I think everyone's gonna
而且现在它的价格和入门级iPhone一样了，是的。
be And it now costs the same amount as the entry level iPhone Yeah.
17e？
17 e?
是的。
Yeah.
它配备了Wi-Fi 6E、蓝牙6连接、8GB内存和1080p前置摄像头。
And it has Wi Fi six e, Bluetooth six connectivity, eight gigs of RAM, ten eighty p front facing camera.
人们有点觉得，哦，这性能有点弱。
Like, people are sort of like, oh, it's underpowered.
但看看这个价格。
But look at the price.
这太疯狂了。
This is this is crazy.
上一次苹果推出如此便宜的产品，真是令人惊讶。
The last time Apple had a product that was anywhere near this cheap was wow.
那是在哪儿？
Where where where was it?
让我想想。
Let's see.
我写过关于这个的内容。
I I wrote about this.
所以2014年，也就是十二年前，他们曾以8.99美元的价格出售MacBook Air。
So in 2014, so twelve years ago, long time ago, they sold the MacBook Air for $8.99.
而现在，他们把这款产品降价到4.99美元或5.99美元，这太惊人了。
And now they have this down at $4.99 or $5.99, which is remarkable.
真的非常便宜。
Really, really cheap.
20.26美元就买得到，对苹果产品来说很便宜，甚至比其他笔记本电脑还便宜。
Cheap for $20.26, cheap for an Apple product, Even cheap when compared to other laptops.
所以这就像他们的逃生通道，压力释放阀。
So this is sort of like their their escape hatch, their pressure release valve.
因此，他们可以更大幅度地提高其他产品的价格，因为如果你对价格敏感，他们还是有适合你的产品。
So they can actually increase the price on the other products more because if you're price sensitive, well, they have a product for you.
在新一代iPhone上也发生过类似的情况：iPhone Pro Max变得越来越大，摄像头凹槽非常显眼，手机更重更大，但电池续航很棒。
And there was something similar that happened with the the new suite of iPhones where the iPhone Pro Max was getting so big and the camera notch was so in your face and the phone was bigger and heavier and the battery life was great.
它性能非常强大，很多人想要这样的手机。
It was very powerful, and a lot of people want that.
但也有不少人抱怨。
But there were also a lot of people who were complaining.
我只想要一部更薄的手机。
I just want a thinner phone.
那么，你就有了iPhone Air。
Well, then then you have the iPhone Air.
而且你会面临一些疯狂的取舍。
And there's gonna be crazy trade offs you only get.
我觉得只有一颗摄像头，电池续航也差很多。
I think one camera, the battery life's not nearly as good.
它会卡顿。
It lags.
它会卡顿。
It lags.
但对于一些人来说，如果他们一直在抱怨Pro Max的种种特性，那我们就为你们准备了Air。
But for some people, if they're gonna be complaining about the all the features of the Pro Max, it's like, well, we got the Air for you.
别抱怨了。
Stop complaining.
我觉得这就是他们的主要策略。
And I think that's a lot of the strategy there.
所以他们推出的其他产品包括：搭载M5芯片的MacBook Air、搭载M5 Pro芯片和M5 Max芯片的MacBook Pro。
So the other products they launched, MacBook Air with an with an m five chip, MacBook Pro with the m five Pro and m five Max chip.
你最后选了多大的内存？
Did you wind up go how much memory did you wind up going with?
48。
48.
48。
48.
够用吗？
Is that enough?
我感觉有点不够用。
I feel a little I feel a little soft.
这么多内存足够在直播中加载推文并做出回应吗？
Is that enough to pull up tweets and react to them on a live stream?
萨姆·阿尔特曼。
Sam Altman.
显然，你实际上一直在用到大约32GB的内存，这让我有点惊讶。
Apparently, you you actually have been maxing out something like 32 gigs of RAM, which which took me a bit by surprise.
我从未遇到过内存不足的问题，但我确实非常喜欢选择最高内存配置，因为这是本地运行大语言模型时的主要限制因素。
I have not had I've not been memory constrained, but I do I do actually love the idea of getting the top end memory, which is the thing that constrains you on running LLMs locally.
是这样吗，泰勒？
Is that correct, Tyler?
是的。
Yeah.
我的意思是，如果你用的是旧模型，确实可以绕过内存限制，但那样会变得特别慢，因为你得不断把模型加载进来又挤出去。
I mean, you can always, like if you have a relic model, you can, like, get around the memory thing, but then it gets, like, super slow because you're, like, loading it in and out.
哦，明白了。
Oh, okay.
你每次生成一个词元时都要把模型加载进来再挤出去。
You're loading it in and out for, like, every token that you're generating.
运行模型需要大量内存。
You need a lot of memory to run.
是的。
Yeah.
但没错。
But yeah.
但如果我选择128GB的顶级配置，我应该能够相当稳定地在本地运行一个相当智能的模型。
But it so so if I go with the one twenty eight gig, the top of the line chip, I should be able to inference, like, a pretty smart model locally, pretty reliably.
我的意思是，你也可以总是把模型进行量化处理。
I mean, it's also like the you can always just quantize the models down.
是的。
Yeah.
所以，你实际上可以在像这样的开源前沿模型上运行，是的。
And and so, like, you you can actually run, like, really frontier, like, open source models on, like Yeah.
现在这已经是相当普通的硬件了。
Pretty normal hardware now.
我只是纯粹在想，比如末日求生准备。
I'm I'm just purely thinking for, you know, doomsday prepping.
你的想法是，万一末日来临，闲着无聊时还有自己的电影库。
The idea of, you know, you have your library in case you're bored during the apocalypse.
你有你的DVD收藏。
You have your DVD collection.
你可以看所有《星球大战》的电影。
You can watch all the movies from Star.
这对你来说是最重要的，因为你将能从头到尾看完整部美国电影史。
That's that's the big one for you because you're gonna be able watch all of American cinema from the beginning.
我还是觉得我实际上不会去看。
I I still still I don't think I'd actually watch.
你不会这么做的。
You wouldn't do that.
那你还会做什么？
What would you be doing?
假设在这种末日情景下互联网瘫痪了，你没法刷手机了。
Assuming that the Internet's down in this apocalyptic scenario, you can't scroll.
还有很多其他事情可以做。
There's Plenty of other things to do.
你会做什么？
What would you do?
去海滩散步。
Take a walk on the beach.
如果外面到处都是核辐射，你根本出不去怎么办？
What if you can't go outside because there's nuclear fallout everywhere?
你只能困在避难所里。
You're stuck in your bunker.
你在做什么？
What are you doing?
刻木头？
Whittling?
你开始玩刻木头了吗？
Did you get into whittling?
我小时候其实特别喜欢刻木头。
I actually used to love whittling as a kid.
我说中了吧。
Called it.
我说中了吧。
Called it.
瞧瞧，就是这样。
There you go.
他不会去看电影的。
He's not gonna be watching movies.
他会去刻木头。
He's be whittling.
我会一边刻木雕一边看电影，同时运行一个本地模型，随时向它提问并获取各种历史信息。
I will be whittling while I watch movies, while I inference a local model that I can just ask questions to and get all sorts of history.
可能会出现一些幻觉，因为它无法连接互联网，毕竟在这个假设情景中，互联网已经不存在了。
And there might be a few hallucinations because it's not creating the Internet because, course, in this hypothetical scenario, the Internet is gone.
但我仍然能访问所有预训练的信息，相当于拥有世界上最好的百科全书，这会很有趣。
But I would still have access to all the pre trained information and basically the world's best encyclopedia, which would be fun.
他们还会推出什么其他产品？
What else do they launch?
在一个电脑价格持续上涨的世界里，看到苹果推出一款5.99美元的笔记本电脑，真是有点疯狂，西奥说。
In a world a world where computers keep going up in price, it's kinda wild to see Apple drop a $5.99 laptop, says Theo.
我同意，这更像是一部手机。
A phone I agree.
再多说点。
More.
他说：天啊，他们能卖出五十亿台这种东西。
Says, oh, man, they're gonna sell 5,000,000,000 of these things.
是的
Yeah.
不知为什么，我觉得这对孩子来说比iPad更显而易见。
This is this feels this feels like more of a no brainer for kids than an iPad for some reason.
我不知道是不是我太老派了。
I don't know if I'm just like old school.
我觉得，只要它有键盘，那就是
And I'm like, if it has a keyboard, it's a
iPad给孩子们用确实看起来不太美观。
iPads with kids just have really bad aesthetics.
确实如此。
It does.
孩子拿着iPad发呆。
Kids with an iPad zoning out.
对吧？
Right?
你给他们配个键盘，他们就专注了。
You get them a keyboard, they're locked in.
但你确实把他们吸引住了。
But you get them there.
我觉得这里有些门道。
I I think there's something here.
你会想，他们在做股票交易吗？
And you're like, are they are they day trading?
是的。
Yeah.
他们在
Are they
是的。
Yeah.
他们肯定在做些什么。
They're doing something for sure.
它可能性能不足，但足以完成任务。
It might be underpowered, but it'll get the job done.
而且我觉得，打字时用键盘这种形式，确实更有利于创作而非消费。
And they I I I think there's something about, you know, typing on a keyboard that does lend itself to more creation and less consumption.
比如，iPad 是一个媒体消费设备。
Like, the iPad is a media consumption device.
大多数人用它就是干这个的。
That's where where most people use it for.
打开Netflix，玩个游戏，做点别的事。
Throw on Netflix, play a game, do something.
但真正成为重度用户的人并不多。
There there aren't that many people that are power users.
但只要你一配上键盘，就能立刻打字，立刻投入工作。
But as soon you get the keyboard, you're ready to type, you're ready to you're ready to work.
让我跟你聊聊Railway。
Let me tell you about Railway.
Railway 是一个一体化的智能云服务商。
Railway is the all in one intelligent cloud provider.
使用您最喜欢的工具部署网页应用服务器、数据库等，Railway 会自动处理扩展、监控和安全问题。
Use your favorite agent to deploy web app servers, databases, and more while Railway automatically takes care of scaling, monitoring, and security.
让我再介绍一下 Okta。
And let me also tell you about Okta.
Okta 帮助您为每个 AI 代理分配可信身份，让您在享受 AI 强大功能的同时规避风险。
Okta helps you assign every AI agent a trusted identity, you get the power of AI without the risk.
保护每个代理。
Secure every agent.
保护任何代理。
Secure any agent.
这是 Andreessen Horowitz 发布的一个搞笑图表。
So hilarious chart from Andreessen Horowitz.
苹果在资本支出上？
Apple on CapEx?
不。
No.
我们很好。
We're good.
他们查看了亚马逊、微软、Alphabet、Meta 和苹果的标准化季度资本支出数据。
They looked at the standard standardized quarterly capital expenditure for Amazon, Microsoft, Alphabet, Meta, and Apple.
所有四家从事人工智能业务的公司——亚马逊通过 AWS 和 Anthropic，微软通过 OpenAI，Alphabet 通过 Gemini，Meta 通过 MSL——在资本支出上都迅猛增长，直冲云霄；而苹果在 a16z 制定的标准化指标上，自 2015 年以来却下降了 19%。
And all of the four companies that are working on AI, Amazon, obviously, through AWS and Anthropic, Microsoft through OpenAI, Alphabet through Gemini, and Meta through MSL, they are all absolutely off to the races going to the moon on CapEx, and Apple is down 19% since, 2015 on the standardized, metric that a16z put together.
非常有趣。
Very, very funny.
有趣的是，苹果在人工智能成本上几乎毫无影响，不仅在资本支出方面如此，而其他所有超大规模云服务商都不得不承认这一点。
And, it's just interesting that Apple has been so unaffected by AI costs, not just on the CapEx side, which is where every hyperscaler sort of had to say, yep.
你知道吗？
You know what?
我们的业务现在从根本上不同了。
Our business is fundamentally different now.
我们不是高利润率、产生现金流的业务。
We are not the high margin cash flow generating business.
我们将动用这些现金。
We're gonna draw down on that cash.
我们可能需要借一些债。
We might be taking on some debt.
这可是个更大、更不同的财务模式。
Like, this is a bigger big, different financial model.
这与过去几年的财务模式截然不同。
It's a very different financial model from years past.
做好准备吧。
And buckle up.
股东们，你们将和我们同行，股价表现良好，因为许多股东相信这一目标，相信人工智能的价值，相信广泛构建计算基础设施的价值。
You're coming with us shareholders, and the stocks have done well because a lot of shareholders have believed in the goal, believed in the value of AI, believed in the value of building compute infrastructure broadly.
在大多数情况下，他们也因此获得了回报。
And they've been rewarded for it in most cases.
但苹果只是置身事外，说：我们不会去训练一个基础模型。
But Apple's just sort of sat that out and said, hey, we're not we're not gonna train a foundation model.
我们不会去构建整个
We're not gonna build a whole
庞大的计算基础设施。
bunch of compute infrastructure.
我们只会使用 Gemini。
We're gonna serve Gemini.
我们只会使用 Gemini。
We're gonna serve Gemini.
那我们付给他们多少钱？
And we'll pay them what?
每年十亿美元左右吗？
A billion dollars a year or something like that?
是的。
Yeah.
这就像
It's like
对
Yeah.
我的意思是，今年将会非常有趣。
I mean, this this year will be really interesting.
很值得期待新的Siri到底是什么样子，它真正能做什么。
It'll be fascinating to see what the new Siri actually looks like, what it can really do.
我想我们会在第二季度开始看到这些。
We'll start to see that I think in q two.
是的。
Yes.
另外，我们预计OpenAI会推出新设备。
And then separately, we're expecting new devices from OpenAI.
对吧？
Right?
哦，是的。
Oh, yeah.
逗弄冰球，是的。
Teasing the puck Yeah.
一分钱，不管你怎么叫它。
Dime, whatever you wanna call it.
是的。
Yeah.
很有趣的是，看看他们能否打破人工智能硬件的魔咒
It'll be interesting to see if they can kind of break the the AI hardware curse
对。
Yeah.
我觉得这
That I feel like is
嗯，我想过这个问题，因为我在想，那些曾经糟糕发布的产品，它们的故事是不是已经结束了。
Well, I I wanted to think about that because I I wonder if the story's over for some of those, like, cursed bad launches product.
比如，Friend 似乎找到了一个小众市场，虽然很有争议，但 Humane 已经被 HP 收购了。
Like, saw Friend apparently has found a little niche, very controversial, but Humane has sold to HP.
看起来他们不再继续开发那个产品了，就是 AI Pin。
It seems like they're not working on that product anymore, the pin, the AI pin.
但 Rabbit R1，我觉得那个叫 Jesse 的人做的产品，还是有潜力的，经过足够多的迭代，我觉得它可能会卖得很好。
But the Rabbit R1, I think his name is Jesse, you know, that product, there's something there where with enough iterations, I feel like that might sell well.
前几天我玩过那个主板。
I was playing with the board the other day.
那个硬件就像一块巨大的触摸屏，你可以用它来玩游戏。
That hardware, it's like a huge screen, touchscreen that you can play games on.
像这样的小众应用场景还是存在的，我认为如果你有硬件团队，是有能力卖出产品的。
Like, there's still these niche applications that I think you could potentially if you have a hardware team, you have the ability to sell something.
它不够精致。
It's it wasn't polished.
AI 的表现也不够好。
The AI wasn't good enough.
嗯，AI变好了。
Well, the AI got better.
你仍然拥有供应链。
You still have the supply chain.
也许有办法突破，但可能需要更加聚焦于小众市场。
Maybe there's a way to cut through, but it probably needs to be a little bit more niche.
是的。
Yeah.
每个人都会拥有一台。
Like, everyone's gonna have one.
泰勒，你对AI时代的硬件怎么看？
Tyler, what do you think about hardware in the age of AI?
我的意思是，我觉得Mac mini是个不错的例子，
I mean, I I think the Mac mini is, like, a good example of,
比如现在的AI硬件。
like, AI hardware right now.
是的。
Yeah.
不。
No.
这是个很好的例子。
That's a great example.
是的。
Yeah.
我在想，Gemini API 在 iOS 中会开放到什么程度。
I wonder I I I wonder what like, how exposed the Gemini APIs will be in iOS.
因为如果你像调用天气 API 或地图 API 一样，完全开放 Gemini Kit，Uber 的存在就是因为当时有一个 GPS API，开发者只需在 iOS 应用里写几行代码就能调用。
Because if you if you fully expose, like, Gemini Kit in iOS, just like you can call a weather API or a maps API, Uber exists because there was a GPS API that you could just call within an iOS application with a couple lines of code.
你不需要自己去判断用户的位置，因为只要用户点击‘与 Uber 共享我的位置’，iPhone 就会自动提供。
And you didn't have to, like, figure out where the user was because the iPhone just told you that as long as you click, like, yeah, share my location with Uber.
如果你现在有一个应用，不仅能使用苹果本地那些有点平庸的 LLM，还能调用 Gemini 做各种疯狂的事情。
If you now get an app that has not just Apple's local LLMs that are sort of, you know, mediocre, but it can go and call Gemini and do a whole bunch of crazy stuff.
你会开始遇到这样的问题：会不会有人开发出类似 Open Cloth 的应用？
You you start running into, like, will someone build an Open Cloth app or something?
苹果的AI战略有多少会开放给第三方生态系统？这个生态系统长期以来为他们带来了巨大成功，但他们却逐渐忽视了它。
Like, how much of Apple's AI strategy will wind up being vended into the third party ecosystem, which has served them so well for so long, they've sort of turned their back on it.
他们对 Apple Vision Pro 的限制更多了。
They locked down Apple Vision Pro a little bit more.
那里的用户增长要少得多。
There was way less uptake there.
这其中存在巨大风险，因为如果我开发了一个免费的 iOS 应用，却产生了大量的 Gemini 令牌费用，这些费用该如何转嫁？
And there's a ton of risk because if I develop like a free iOS app that then incurs like a ton of Gemini tokens and ton of charges, How do those get passed through?
苹果是打算自己承担这笔费用吗？
Is Apple just biting the bullet on it?
但这里有个有趣的地方：我们正在让应用开发者自由地构建软件，这些软件不仅能利用本地大语言模型的能力，还能在后台调用 Gemini，并将其封装成面向消费者的完整解决方案。
But there's something interesting there where we're turning app developers loose to to build software that takes advantage not just of local LLM capabilities, but Gemini under the hood, and sort of wrap it up in a neat bow that is available to the consumer.
这可能会非常酷。
Could be very cool.
总之，我原本有个大问题，接下来我会在文章中探讨：苹果似乎不受资本支出成本的影响，而且通过这次发布，他们也避开了内存成本和内存危机的影响。
Anyway, the big question that I had, and I will get into the essay, was so Apple's unaffected by CapEx costs, but they've also, with this release, been unaffected by memory costs and the ramageddon.
所以，我会带你们看看我是如何深入研究苹果的内存定价问题的。
So I will I will take you through my how I dug into the question of memory pricing at Apple.
但首先，我要跟你们说说 Labelbox。
But first, I will tell you about Labelbox.
RL 环境、语音、机器人、评估和专家人工数据。
RL environments, voice, robotics, evals, and expert human data.
Labelbox 是全球领先 AI 团队背后的数据工厂，我还会跟你们说说 Console。
Labelbox is the data factory behind the world's leading AI teams, and I will also tell you about Console.
Console 构建 AI 代理，自动化 70% 的 IT、人力资源和财务支持工作，让员工能即时解决访问请求和密码重置问题。
Console builds AI agents that automate 70% of IT, HR, and finance support, giving employees instant resolution for access requests and password resets.
这个音效不错。
That's a good sound effect.
我喜欢这个。
I like that one.
我也有一个。
I have one too.
惊喜。
Surprise.
我真不敢相信你还没用过那个。
I can't believe you haven't used that
还没。
yet.
我，不，不行。
I I no.
不行。
No.
不行。
No.
因为上次我一拿到音效板就玩得太疯了，结果被坑惨了。
Because I was so burned by the last time I got access to the soundboard, and I went too crazy.
有一场演出，我和乔迪都拥有完整的音效板权限，使用的是完全相同的音效提示。
So there's one one show where me and Jordy both had full soundboard access, all of the same sound cues.
只有一场演出。
One single show.
那场演出真是
And it was truly
一场灾难。
A disaster.
希腊神话里那个比喻是什么？
What's the what's the metaphor from Greek mythology?
对我来说，那就是潘多拉的盒子。
It was the it was Pandora's box for me.
一旦我打开了它，就根本无法抗拒。
Like, once I opened it, I could not resist.
我上瘾了。
And I became addicted.
我当时在操作那个
And I was pushing the
东西，而且
thing and
它就是不工作。
it just didn't work.
所以我们决定，算了。
So we were like, no.
让Jordy来操作音控台吧。
Let's have Jordy play the soundboard.
我来做其他事情。
I'll I'll do other stuff.
作为音控师，是的。
The the as a as a soundboard artist is Yes.
当我没有在做演出的时候
When I'm not doing the show
是的。
Yes.
当
When
我周末闲逛的时候，是的。
I'm hanging out when I'm hanging out on the weekend Yes.
我只是在想，是的。
I'm just thinking Yes.
我只是，如果有人说了什么，是的。
I'm just I'm if somebody says something Yeah.
比如我们在吃饭，是的。
Like we're at dinner Yeah.
我在想音效。
I'm thinking about the sound effect.
是的。
Yeah.
是的。
Yeah.
我当时和你在一起，你说了些疯狂的话，我想着干脆打开手机，上YouTube找到那个音效播放出来。
I was I was hanging out with you and you said something crazy and I wanted to just be like And so I opened up my phone, I went to YouTube and I pulled up that sound and played it.
但后来我得把音量打开，花得太久，笑话已经过去了。
But then I had to turn on my volume and it took me way too long and the joke had passed.
但我们有个解决方案。
But we have a solution.
那边的泰勒·科斯格罗夫一直在开发一个手机音效板应用，你可以安装，我们还应该做个小组件，这样它就直接显示在主屏幕上。
Tyler Cosgrove over there has been working on a soundboard app that can go on your phone, and you can have we should do a widget too so that it's just there on the home screen.
哔。
Boop.
你最爱的音效板提示音。
Your favorite your favorite soundboard cues.
所以，很快就能看到了。
So, expect that soon.
展开剩余字幕（还有 480 条）
总之，让我跟你聊聊AppLovin。
Anyway, let me tell you about AppLovin.
通过axon.ai，让盈利广告变得简单。
Profitable advertising made easy with axon.ai.
获取超过十亿日活跃用户的访问权限，今天就开始扩大你的业务。
Get access to over 1,000,000,000 daily active users and grow your business today.
所以点击
So Hit
我们。
us.
这个那个，问题在于，就像我明白苹果的策略一样，看。
The the the quest the question, like like, I get I get the strategy of Apple saying, look.
我们让谷歌来处理AI推理部分没问题。
We're good with Google handling the AI inference stuff.
他们有GCP。
They have GCP.
他们拥有TPU。
They have the TPU.
他们有丹尼斯，塞巴斯蒂安·马拉巴伊还为他写过一本书。
They got Dennis, who's got a book written about him by Sebastian Malabai.
他觉得DeepMind很棒。
He like like, DeepMind is great.
我们正在与他们合作。
We're working with them.
所以资本支出不受影响。
So CapEx unaffected.
我明白这一点。
I understand that.
他们怎么不受内存价格影响？
How are they not affected by memory prices?
内存危机已经降临到我们所有人头上。
Ramageddon has come for us all.
本·汤普森来自Stratechery，约翰·格鲁伯来自Daring Fireball，他们为此打过赌。
And and, Ben Thompson of Strathecari and John Gruber of Daring Fireball actually had a bet about this.
我希望我没有把约翰·格鲁伯的名字拼错。
And I hope I didn't misspell John Gruber's name.
他们确实为此打过赌。
They had about they had a bet about this.
他们说，赢的人可以吃一顿牛排大餐。
They said whoever wins this bet gets a steak dinner.
牛排大餐是在牛排店吃的，而本·汤普森输了。
A steak dinner was at steak, and Ben Thompson lost.
约翰·格鲁伯赢了，因为他认为苹果公司总是对内存收取过高的费用。
John Gruber won because John Gruber said, Apple always overcharges for RAM.
如果你自己组装一台电脑、手机、笔记本电脑，或者像iPad这样的设备，去零售店多买128GB内存，或者做个小小的内存升级，可能要花30美元。
Getting a if you went to build a computer or a phone or laptop or, you know, iPad equivalent, you just went to retail, getting an extra 128, gigs of memory or or a slight memory upgrade, might cost you $30.
但在苹果生态系统里，这差不多要200美元。
In the Apple ecosystem, that's like $200.
因此，苹果公司长期以来一直内置了这种定价缓冲机制，而今年正是这个缓冲帮了他们大忙。
So Apple has built in this, this pricing buffer for a very, very long time, and and that buffer really saved them this year.
但事情远不止如此。
But there's a there's a lot more to it.
今年，许多公司都在应对内存危机，一些数据非常有趣。
So, many companies are dealing with Ramageddon this year, and some stats that are interesting.
2025年全年，DRAM价格飙升了172%。
DRAM prices rose 172% throughout 2025.
自2025年9月以来，DDR5现货价格已经翻了四倍。
DDR5 spot prices have quadrupled since September 2025.
TrendForce预计，2026年PC用DRAM价格将大致翻倍，而LPDDR价格则将创下历史最大涨幅。
TrendForce expects PC DRAM to roughly double in price just in 2026 with LPDDR pricing seeing the steepest increases in history.
每一片用于NVIDIA GPU高带宽内存（HBM）堆栈的晶圆，都意味着有一片晶圆无法用于中端智能手机或消费级笔记本电脑的LPDDR5X模块。
So every wafer that's allocated to an HBM stack, a high bandwidth memory stack for an NVIDIA GPU, is literally a wafer denied to the LPDDR5X module of a mid range smartphone or consumer laptop SSD.
因此，它们确实是真正的大宗商品。
So they are truly commodities.
它们的数量有限，只能流向其中一个地方。
There's only so many of them, and they can only go to one place or the other.
据称，仅OpenAI的Stargate项目就可能消耗当今全球DRAM产量的40%。
Apparently, OpenAI's Stargate initiative alone could consume up to 40% of global DRAM output as it stands today.
这就是问题所在，因为内存领域只有三家主要公司。
And this is the problem because there's only three major companies in the memory category.
它们是三星、SK海力士和美光。
It's Samsung, SK Hynix, and Micron.
这三家公司共同控制了全球95%的DRAM产能。
Together, they control 95% of the global DRAM production.
别忘了，美光一直在努力建设一家新工厂，是的。
Remember, Micron has been trying to build a new facility Yes.
在纽约。
In New York.
是的。
Yes.
有六个人基本上阻挠了它。
And six people basically blocked it.
哦，我们之前聊过这个，
Oh, I I we we talked about that,
但传言说这背后是外国资金在操作。
but I And the rumor the rumor is that it's, like, foreign funding.
哦，有意思。
Oh, interesting.
实际上是在故意拖慢美光的进度。
Actually trying to hold Micron up.
是的。
Yeah.
对。
Yeah.
我的意思是，美光一直是人工智能利润和高价的巨大受益者。
I mean, Micron has been very they've been a huge beneficiary of AI margins and higher prices.
因此，他们甚至将所有产能都转向了AI领域，并退出了名为Crucial的消费品牌。
And so much so that they've actually reallocated all of their capacity to AI, and they pulled out of their consumer brand, which is called Crucial.
所以如果你曾经组装过游戏电脑，很可能用过Crucial Ballistics，那是款极具街头电竞风格、专门面向PC玩家的DDR内存，
So if you ever tried to build a gaming PC, you probably found Crucial Ballistics, which was, like, this very aggressively X Games branded, like, PC gamer DDR memory, which was,
你知道的，就像是RAM界的Monster Energy能量饮料。
you know Like the monster energy of RAM.
没错。
Exactly.
说得太到位了。
Couldn't have said it better myself.
不是的。
No.
确实如此。
It really is.
Crucial Ballistics的包装设计简直疯狂到离谱。
Like, the the the the branding on Crucial Ballistics is, like, insane.
总之，玩家们很快就会手握‘怪物能量’，不惜一切代价夺回内存条。
Anyway, gamers will be rising up soon with monster energies in hand to to take the the the Ram by by force if they need to.
我经常看到关于这件事的梗图。
I already see memes about this all the time.
我看到那些最糟糕的AI生成视频，简直一塌糊涂。
I see these, like, the worst AI slop video you've ever seen, just absolute junk.
然后配文会说：这就是我买不起内存条的原因？
And then the caption will be like, so this is why I can't afford RAM?
你知道的。
You know?
或者说是为什么内存条要卖800美元一根。
Or this is why RAM is $800 a stick or whatever.
而另一方面，偶尔也会出现一些不错的AI生成视频，人们就会说：好吧。
And then on the flip side, people will say, you know, occasionally, there'll be a good AI slot video and people will like, okay.
为了这样的视频，花800美元买内存条是值得的。
It's worth paying $800 for RAM because this is so good.
对吧？
Right?
双方都有。
It goes both ways.
双方都有。
It goes both ways.
但显然，人们正感受到压力。
But clearly, people are feeling the the the pinch.
但苹果粉丝会置身事外，因为苹果今年采取了几项关键举措，帮助他们避免了灾难。
But Apple fans will be sitting out the fight, thanks to a few key moves from Apple that help them avoid disaster this year.
首先，苹果有长期的供应协议，可以提前十二到二十四个月锁定内存。
So one, Apple has long term supply agreements that allow them to secure memory twelve to twenty four months in advance.
他们还受益于垂直整合和自研芯片。
They also benefit from vertical integration and custom silicon.
因此，他们不会购买你在游戏PC或NVIDIA显卡等设备中看到的那种消费级通用DRAM模块。
So instead of buying that consumer grade commodity DRAM modules that you see go into either gaming PCs or NVIDIA GPUs, for example.
他们与三星和SK海力士直接谈判多年期合同，定制符合苹果精确规格的LPDDR封装。
They negotiate multiyear contracts directly with Samsung and SK Hynix for custom LPDDR packages that meet Apple's exact specifications.
所以我认为是的。
And so I think Yeah.
所以这就像内存一直都有周期性。
So this is like Ram's always been cyclical.
对。
Yep.
因此，苹果说锁定价格是非常合理的。
And so it makes a lot of sense for Apple to say basically like lock up pricing.
嗯。
Mhmm.
三星在锁定需求。
Samsung's locking up the demand.
他们，我都不认为三星会预见到这种程度的行情。
They they I I don't even think Samsung would have predicted this level Yeah.
甚至在三四年前的需求。
Of demand even three, four years ago.
是的。
Yeah.
否则，他们可能不会以那样的方式定价。
Otherwise, they may not have priced it in the way that they did.
当然。
Of course.
另一个你可能想到的是
The other thing that maybe you get to is
而且似乎仍在保持他们的常规利润率。
And seems like still making their default margin.
他们只是觉得机会成本太高了。
They're just like the opportunity cost is so high.
我们可以重新分配。
We could reallocate.
另一点是，他们在追加销售上可能有非常大的利润空间，是的。
The other thing is is they they probably have so much margin in the upsell Yeah.
所以即使利润空间被压缩一些，也没关系，没错。
That they could have some margin compression Exactly.
仍然可以保持盈利。
Still be
苹果在不同时期曾拥有超过一千亿美元的现金。
So Apple at various times has had over a $100,000,000,000 in cash.
他们目前的现金超过六千六百亿美元。
They currently have over 66,000,000,000 in cash.
他们完全可以承受这样的损失。
Like, they can take the hit there.
他们也可以直接承受利润率的下降。
They can also just take a hit to their margins.
每个AI实验室看到这笔现金都在想：蒂姆，我真能好好利用这笔钱。
Every AI lab looking at that cash being like, Tim, I could really put that to you
现在。
right now.
就像，是的。
Like, yeah.
他们是唯一一家还没有投入高达100亿、200亿、300亿美元到前沿AI实验室的超大规模云服务商。
They're the only hyperscaler that hasn't really shelled out to the tune of $10.20, 30,000,000,000 into one of the frontier frontier labs.
也许这种情况会改变。
Maybe that changes.
我不确定。
I don't know.
有人提出，苹果应该收购Anthropic，或者亚马逊应该收购Anthropic，但我认为目前价格太高了。
Somebody was throwing out, like, Apple should buy Anthropic or or or Amazon should buy Anthropic, but I think the price is too high at this point.
实在太贵了，价格正在飞涨。
It's just too it's just getting away.
是的。
Yeah.
但我对内存制造供应链的具体运作并不完全确定，不过我认为，由于苹果的特定需求，三星很难将已经生产出来的内存产品迅速重新分配——比如，如果这些晶圆原本是用于游戏PC的，那就更容易重新调配。
But so I I'm not I'm not a 100% sure about the how the manufacturing supply chain for memory works, but I believe that because Apple's specific requirements, it's harder for Samsung to just instantly reallocate product that's been already manufactured in a way where crucial ballistics, if it's going into gaming PC, it's easier to reallocate those wafers.
所以，即使没有那些合同，我认为由于苹果是垂直整合的，供应链在与苹果合作时会更僵化一些。
So even if they didn't have those contracts, I think I think the the supply chain's a little bit stickier with Apple because they're vertically integrated.
然后，当然，他们也有能力支付更高的价格。
And then, yeah, of course, they have the money to overpay.
我们之前讨论过，MacBook Neo就像一个压力释放阀。
We talked about how the the the MacBook Neo is this pressure release valve.
所以，即使内存价格上涨，他们也有充足的现金缓冲。
So even if the memory does increase, they have this they have the cash buffer.
他们还对成本仅30美元的产品收取200美元的费用。
They also have the they're charging $200 for something that costs 30.
如果成本涨到60美元，他们完全可以承受利润率的下降来缓冲影响。
If it goes to 60, they can just take a hit on margins buffer.
此外，他们还有合同缓冲。
Then they have the contract buffer.
然后他们还有这款Neo，如果MacBook Pro涨价了，MacBook Pro的买家对价格可能没那么敏感。
Then they also have this Neo where if the MacBook Pro goes up in price, the MacBook Pro buyer is probably less price sensitive.
是的，没错。
They're like, yeah.
你知道，今年它从3000涨到了4000。
You know, this year, it was $4,000 instead of 3,000.
我是为了商务用途才买的。
I'm buying it for business purposes.
这并没有造成太大的压力。
It's not that big of a squeeze.
而价格敏感的买家会说，是的。
And then the price sensitive buyer says, yeah.
你知道吗？
You know what?
我不会去买那款2000美元的笔记本了。
I'm not gonna go for the $2,000 laptop.
今年我会买一台600美元的笔记本电脑，但仍然留在苹果生态系统中。
I'm gonna go for the $600 laptop this year, and they're still in the Apple ecosystem.
但我的体会是，苹果能在如此规模下做到这一点，真是对其运营能力的绝佳证明。
But my takeaway is, like, this is an incredible testament to the operational prowess of Apple to pull this off at this scale.
据称，索尼PlayStation 6会因内存成本上升而推迟到2028年或2029年发布。
The Sony PlayStation six is allegedly gonna be delayed until 2028 or 2029 because of rising memory costs.
他们会等一等。
They're gonna wait it out.
因为这里有三家厂商，不只是台积电这一家。
Because there are three players, it's not just TSMC in this case.
感觉价格竞争会加剧，同时也会更激烈地争夺产能。
It does feel like there'll be more pricing competition, and and there's more of a race to build capacity.
而台积电如果不去建新的晶圆厂，就可以不断提价，并在下一波合同续约时提高利润率。
Whereas the TSMC, if they don't build the next fab, they can just continually raise prices and and increase margins once the next contracts go up for renewal.
而如果三星有大量产能而美光没有，那么三星就会胜出，因此这是一种博弈论式的局面——他们必须扩建产能。
Whereas if if Samsung has a bunch of capacity and Micron doesn't, like Samsung wins, And so there's more of this game theoretic, like they got to they have to build.
所以我认为很多人，包括泰勒之前跟我聊过，我们都觉得这场内存短缺只会持续几年，不会长达五到十年。
And so I think a lot of people and Tyler was talking about this with me earlier, like, we we all expect this memory shortage to be few years, not more not like five or ten years.
所以如果你把索尼PS6推迟到
So if you push out the Sony p s six to
泰勒，你看好吗？
You're bullish, Tyler?
28年。
28.
他看好的。
He's bullish.
不。
No.
我非常看好，但我觉得是的。
I'm very bullish, but I I think yeah.
因为有更多玩家参与，我觉得这场短缺会
Because because there's more players, like, I I think the the shortage will be
你就说你不重要好了。
Just say you're not important though.
AGI药丸。
AGI pill.
你什么意思？
What do you mean?
我的意思是，如果你真的一直呈指数增长，那就真的会变成这样。
I mean, there is a question about, like, if you truly go exponential forever and it just like, okay.
是的。
Yeah.
有成千上万亿美元正在投入建设数据中心。
There's, like, you know, trillions and trillions of dollars of of data centers being built.
就连三星、美光、SK海力士能永远跟得上吗？
It's like, can even Samsung, Micron, SK Hynix keep up forever?
就像，它真的能吗？
Like, can it yeah.
不。
No.
所以，我们可能正生活在一个永久性的更高价格环境中，因为这是一种更有价值的产品。
So it might we might be living in, like, a permanently higher price just because, like, it's a more valuable product.
比如，我们获得了更多的经济产出。
Like, we get more economic output.
会不一样。
Be different.
可能会不一样。
Might be different.
此外，任天堂Switch 2明年的价格可能会上涨。
Also, the Nintendo Switch two might go up in price next year.
这对游戏玩家来说太糟糕了。
It's really bad for for gamers.
游戏玩家的显卡价格正在上涨。
Gamer gaming graphics cards are going up in price.
英伟达正在退出游戏显卡市场，或者跳过一年。
NVIDIA is pulling out of gaming graphics cards or skipping a year.
这对游戏玩家来说简直是混乱。
So it's mayhem for gamers.
但从长远来看，内存制造商之间的竞争应该会稳定价格。
But long term, the competition between the memory manufacturers should stabilize prices.
但你知道，看到所有那些非苹果的大公司都措手不及，真是令人难以置信，因为任天堂或索尼的团队可不是那种会说‘哦，是的’的人。
And but, you know, yeah, just watching all the big companies that are not Apple get caught flat footed is just crazy to watch because it's not like it's not like the folks at Nintendo or the folks at Sony are like, oh, yeah.
这是他们第一次制造产品。
First time manufacturing something.
如果你告诉我，Rabbit R1背后的团队正遭受内存成本上涨的困扰，我会说，是的。
Like, if you told me that, like, the team behind the Rabbit r one was, like, suffering from higher memory costs, I'd be like, yeah.
这是一家新公司，推出新产品。
Like, it's a new company, new product.
他们以前没做过这个。
They haven't done this.
索尼已经生产PlayStation几十年了，却还是在这上面措手不及。
Like, Sony has been making PlayStations for decades, and they still got caught flat footed on this.
这其中很大一部分证明了苹果的品牌影响力、网络效应以及他们所建立的规模——他们处于这样的位置，即使其他大型云服务商在各个方面都被挤压，他们依然没有被压垮。
And a lot of that is a testament to the brand power that Apple has, the network effects, the the the the scale they built that they're in this position and still not getting squeezed, even when other hyperscalers getting squeezed on all sorts of stuff.
蒂姆·库克依然被低估。
Tim Cook, still underrated.
他确实干得漂亮。
He cooked.
他确实做得很好，而且没错。
He has cooked and and yeah.
苹果的运营方面总是被低估，即使面对功能、产品、营销上的小失误，或者‘这个广告没打中’之类的批评。
The operational side of Apple is always it's it's just habitually underrated, even in the face of, like, slight misses on features and products and marketing and, oh, this ad didn't land.
但业务依然蓬勃发展，不断应对一个又一个危机。
Well, the business is still thriving and dealing with just crisis after crisis.
关税本应是一场重大危机，但他们挺过去了。
Like the tariff should have been a major crisis, They got through it.
这场内存危机本应是一场危机。
This memory Ramageddon should have been a crisis.
但他们挺过去了。
They got through it.
这对苹果来说全是非常好的消息。
It's all been very, very good news for Apple.
让我跟你聊聊 Eleven Labs。
Let me tell you about Eleven Labs.
构建智能的实时对话代理。
Build intelligent real time conversational agents.
用 Eleven Labs 重新想象人机交互。
Re imagine human technology interaction with Eleven Labs.
准备好听钱进账的声音了，因为我马上要跟你聊聊 Phantom Cash。
And get that cha ching ready because I'm gonna tell you about Phantom Cash.
无需交易所或中间方，为你的钱包充值，并使用 Phantom 卡消费。
Fund your wallet without exchanges or middle end and spend with the Phantom card.
苹果正在践行协同效应假说。
Apple is exercising the cose conjecture.
他们正在等待开源模型变得更好，对前沿实验室的定价施加压力，例如。
They're waiting for open source models to get better, applying pressure to frontier lab pricing, e.
G。
G.
拒绝了Anthropic的提议。
Rejecting Anthropic overtures.
Anthropic希望从苹果那里获得更多的资金。
Anthropic wanted more money from Apple.
他们选择了谷歌和Gemini。
They went with Google and Gemini.
这是来自Larsen的观点，他是一位我非常尊重的经济学家。
This is from Larsen, who is an economist who I really respect.
更有可能的是，缺乏耐心的买家存在逆向选择，暴露了其实不佳的商业定位。
It may rather be that impatient buyers have adverse selection, revealing actually poor business positioning.
所以他是在识别苹果在系统中的杠杆作用。
So he's identifying Apple's leverage in the system.
他们是一个消费者聚合平台，一直表现得非常好。
They're this consumer aggregator, and they've been doing very well.
凯尔·哈里森说，我们算过账了。
Kyle Harrison says that, We ran the math.
OpenAI 购买的内存是其短期内实际所需量的三到四倍。
OpenAI is buying three to four times more memory than it could possibly need in the short term.
最善意的解释是他们采取了积极的前瞻性布局。
The most charitable explanation is aggressive forward positioning.
不太善意的解释是，他们正在垄断供应，以在设备端AI兴起之前将其扼杀。
The less charitable one is that they're cornering the supply to kill on device AI before it starts.
如果设备端AI发生在苹果产品上，这似乎并不是一个现实的场景。
Well, if the on device AI happens on Apple products, that doesn't seem like that real of a scenario.
但这确实是一个有趣的论点。
But it is an interesting thesis.
我认为，在前沿实验室中，所有人都被AGI思维深深影响了。
I think that across the frontier labs everyone is just very AGI pilled.
他们看到了能力的提升，但也看到了那种叠加的指数级增长。
They're seeing the increase in capabilities, but also the sort of like stacked exponentials.
本·汤普森今天解释说，当你从大语言模型转向推理时，为了得到一个答案所需的令牌数和计算量突然增加了十倍。
Ben Thompson was explaining this today, that when you went from LLMs to reasoning, the number of tokens, the number of compute required just to get a single answer went like, you know, 10x all of a sudden.
然后我们在编码代理上也看到了同样的情况，这些长时间运行的任务使计算量呈平方级增长。
And then we saw the same thing with coding agents, where these long running tasks are sort of like squaring the amount of compute.
因此，你不断叠加这些技术，计算需求持续上升，但经济产出也在增加。
And so you just keep stacking these things on and you just get more and more compute demand, but you get more economic output.
所以最终这一切都是值得的。
So it winds up being all worth it.
苹果什么时候会发布iPhone？
When will the when will Apple release the iPhone
关于OpenAI内存方面的另一件事是，他们计划销售大量消费类设备。
The only other thing only other thing on the OpenAI memory side is they plan to sell a lot of consumer devices.
他们哦，所以他们可能就是为了这个而购买。
They Oh, so they might be buying it for that.
对吧？
Right?
是的。
Yeah.
所以我们其实并不清楚。
So so we we don't really know.
但没错，我认为OpenAI确实已经回调了那个11.2万亿美元的数字。
But but yeah, I think that obviously OpenAI has kind of rolled back some of like that that $11,200,000,000,000.0 number.
没错。
Yep.
但看到今年头几个月需求的增长，似乎开始让人觉得，嘿，也许他也没那么疯狂了。
But seeing seeing the growth in demand just in this first couple months of the year starts to look like, hey, maybe maybe maybe he wasn't so crazy Yeah.
之后
After
是的。
Yeah.
对。
Yeah.
完全正确。
Totally.
这些模型正在证明计算资源的合理性。
Like, the the models are justifying the compute.
我认为我们昨天也聊过这一点，关于编排器，就是下一个10倍增长会从哪里来？
And I think that and we talked about this yesterday a little bit with the orchestrators, just this idea that, you know, where will the next 10x come from?
很可能来自运行10个Cloud Code或Codecs实例，然后将推理量提升10倍。
Well, it's probably from just running 10 instances of Cloud Code or Codecs and and just firing off 10 times as much inference.
而且人们已经准备好这么做了。
And and people are ready to do that.
你经常在Twitter上看到，人们讨论他们进行了多少次推理。
They're you you see it all the time on Twitter, people talking about how much inference they're they're doing.
苹果什么时候发布iPhone 18？
When will Apple release the iPhone 18?
她认为在10月前发布的概率是35%。
Call she has before October at 35%.
在2027年之前发布的概率是34%。
Before 2027 at at 34%.
他们今年不发布这款手机的概率有多大？
What is the chance that they don't release it this year?
那可能是零。
That would be probably zero.
我认为就是零。
And I think it is it is zero.
所以请持续关注。
So stay tuned.
我肯定马克·德国会对下一次发布进行详细报道。
I'm sure Mark German will have good coverage on on the next
让我们把格米纳托尔请回来。
iPhone Let's get the Germinator back on.
是的。
Yeah.
嗯，我本来想邀请约翰·格鲁伯来讨论这些话题，但马克·古尔曼当然随时都欢迎。
Well, I I I I wanted to invite John Gruber to talk about these, but Mark Gurman, of course, is welcome at any time.
但我们至今还没和约翰·格鲁伯聊过，他整个职业生涯都在出色地报道苹果。
But we've but we have yet to talk to John Gruber, he has been covering Apple fantastically throughout his entire career.
所以他会是一个很有趣的对话人选。
So he would be an interesting gentleman to talk to.
让我你
Let me you
先说说Gusto。
about Gusto first.
Gusto是一个集薪资、福利和人力资源于一体的平台，专为适应现代中小企业的演变而设计。
Gusto is the unified platform for payroll benefits and HR built to evolve with modern small and medium sized businesses.
让我跟你聊聊 Restream。
And let me tell you about Restream.
一个直播流，覆盖三十多个平台。
One live stream 30 plus destinations.
如果你想多平台直播，请访问 restream.com。
If you want to multi stream, go to restream.com.
上周提交了在南部科学园区新建半导体晶圆厂的计划。
Submitted a new semiconductor fab construction plan for the Southern Taiwan Science Park last week.
嗯。
Mhmm.
该地块位于台南市安南区，紧邻 F18 晶圆厂，嗯。
The site is located on Tainan's Andeng District adjacent to the F 18 fab Mhmm.
占地面积为 15.46 公顷。
And covers 15.46 hectares.
他们正在推进更先进的 2 纳米制程生产，其中八公顷用于生产区，一公顷用于行政设施。
They're doing more two nanometer production, the leading edge, with eight hectares allocated for the production area and one hectare for administrative facilities.
据半导体分析公司称，该项目预计将于2028年完工并开始生产。
The project is expected for completion and production start in 2028, says semi analysis.
所以我们得看看台积电实际发布的资本支出指引会怎样。
So we'll see what happens in the actual CapEx guide from TSMC.
本·汤普森对台积电此前的资本支出指引一直很失望，但也许这会是一个转变。
Ben Thompson has been very disappointed with where TSMC has been guiding on CapEx, but maybe this is a shift.
我想知道在那里再建一座台积电晶圆厂会快多少。
I wonder how much faster it will be to build another TSMC fab over there.
如果他们在亚利桑那州之前就实现这个新工厂的量产，并且赶超亚利桑那州，让这个项目率先达到规模，那将暴露出美国在建设和许可体系上的严重问题，我想。
If they ramp production of this new facility before Arizona, if they eclipse and there's a flipping and this gets to scale before Arizona, it will reveal some very serious problems with America's building and permitting system, I would imagine.
或者也许是人才储备的问题。
Or maybe maybe talent pool.
我当然希望
I would I would certainly hope that
两者都是。
Both.
台湾，我的意思是，这是他们的瑰宝。
That that Taiwan, I mean, this is their crown jewel.
他们应该能比较快地让这个项目投入运行，不会有很多延误，因为他们以前做过，而且他们对台积电有重大投资。
They should they should get this up and running fairly quickly without many delays since they've done it before and they are heavily invested in TSMC, of course.
更重要的是，温迪的美国总裁在与麦当劳和汉堡王的汉堡大战中发布了新视频。
More importantly, Wendy's US president posts new video amid burger battle with McDonald's and Burger King.
我们看看。
Let's see.
我们来评分吧
Let's Let's rate
它。
it.
他有没有说好的。
Does he does he put Okay.
我喜欢他拿的方式，你知道的。
I like the way that he's holding it, you know.
好的。
Okay.
看起来正常。
Seems normal.
是的。
Yeah.
太棒了。
Wonderful.
有点儿。
Kinda
那是一个汉堡。
That's a burger.
这不错。
This is good.
我喜欢聊这个，正是关于这个的。
I like talking This is exactly about.
什么，好的。
What Okay.
是的。
Yeah.
我们倒回去再说一下，
Going going going back for a second and
你得给它加个结尾。
You gotta top it off.
他在说汉堡。
He's saying burger.
他不是在说产品吗？
He's not saying product?
产品。
Product.
好了，就这样。
There we go.
是的。
Yeah.
毫不迟疑。
No hesitation.
毫不迟疑。
No hesitation.
等等。
Wait.
等等。
Wait.
他蘸的是霜冻酱吗？
What did he dip in frosty?
那是枣泥吗？
Was that date?
他最后做了什么？
Did he what did he do at the end there?
他太厉害了。
He Amazing.
他把薯条蘸在霜冻饮料里？
He dips a french fry in the frosty?
哇哦。
Woah.
好吧。
Okay.
好吧。
Okay.
他不称之为产品。
Doesn't call it a product.
毫不迟疑。
No hesitation.
嘴里含着食物说话。
Talking with the food in his mouth.
是的
Yeah.
是的
Yeah.
这很大胆。
That's bold.
这很大胆。
That's bold.
大胆。
Bold.
我给这个打八分。
I'm giving this an eight out of 10.
如果你刷下去看了，我看完这个后反而喜欢上桑德斯上校了，只是因为知道
If you scrolled out, I like Colonel Sanders after seeing this, just knowing
是的
Yeah.
还有，他那种放薯条的方式，是的。
Also also the way he he kinda puts the fry in Yeah.
然后迅速地，你知道的，把手指收回来。
And quickly, you know, kinda pulls pulls the hand back.
是的。
Yeah.
这很有风格分。
I I I it's got style points.
伙计，如果你在找一份能赚大钱的工作，那就去申请麦当劳的CEO职位吧。
Dude, here's a good job if you're looking for a gig that will bring home the bacon, apply for the job of CEO at McDonald's.
据说这家伙年薪一千八百万。
Apparently, the guy makes 18,000,000 a year.
等等。
Wait.
他们有公开的职位申请吗？
Do they have an open job application?
不。
No.
但在这之后，完全有可能。
But after this, it's entirely possible.
哦。
Oh.
那那那
That that that
如果你能精彩地大口吃掉，就有可能成功。
if you if you put on a masterful performance chomping down, you can make you can make a run.
我不知道。
I don't know.
有趣的是，CEO一个接一个地出现。
It is funny that there's just like one after another of CEOs.
这对他们所有人来说可能都是好事。
This is probably good for all of them.
我不知道。
I don't know.
这挺有趣的。
This is this is fun.
人们都，我不想有跳跃剪辑。
People are I want no jump cuts.
应该有人看到有人在吃东西。
Somebody should have seen somebody should be seen swallowing the food.
哦。
Oh.
好的。
Okay.
人们都在分析，这有点公平。
People are people are analyzing That's somewhat somewhat fair.
分析细节。
Analyzing the details.
让我告诉你关于 Gemini 3.1 Pro 的情况。
Let me tell you about Gemini 3.1 Pro.
Gemini 3.1 Pro 现已推出，基础能力更强大。
Gemini 3.1 Pro is here with a more capable baseline.
它非常适合处理超复杂的任务，比如可视化困难的概念、将数据整合为单一视图，或让创意项目成真。
It's great for super complex tasks like visualizing difficult concepts, synthesizing data into a single view, or bringing creative projects to life.
伯伯·金发表了文章。
Berber Jin came journal.
昨天在全员会上，阿尔特曼为 OpenAI 与美国国防部的合同辩护，称外界的反对‘非常痛苦’。
In and shared Altman defended OpenAI's Defense Department contract in an all hands yesterday, calling the backlash, quote, really painful.
这是一个复杂但正确的决定，尽管带来了极其严峻的品牌影响和负面公关。
This was an example of a complex but the right decision with extremely difficult brand consequences and very negative PR for us.
他告诉员工，在短期内。
In the short term, he told staff.
是的，我仍然坚持这一立场，考虑到世界上发生的这些事件。
Yeah, I I still I still stand by that given given the events that have transpired in the world Yeah.
自从上周以来，就像去那里会见政府官员一样。
Since last week, like being there to like meet meet the government Yeah.
感觉像是走在正确的道路上。
Feels like being on on the right path.
就是这样。
It was just Yeah.
奇怪的是，谷歌的策略居然是什么都不说。
It was just odd that the that like Google strategy was to not say anything.
基本上，我认为什么都不做就是赢。
Do nothing win, basically, I I would think.
或者什么都不做，不进不退。
Or or do nothing Don't go up or down.
比如，Gemini只是……
Like, Gemini just
但好的是，我认为Gemini并没有参与这个合同。
But good but but I don't think Gemini is involved in this contract.
是的。
It's Yeah.
是Grock。
It's Grock.
是Grock？
It's Grock?
Grock是另一个那个。
Grock was the other one that that
我会。
I would
用的是同一个术语。
came in with the same term
所以Gemini和国防部没有任何关系吗？
So Gemini doesn't have any relationship with the Department of War?
我觉得有。
I think they do.
我觉得我原本以为那里有一个，呃，我
I I I think I I thought that there was a there was an I
可能他们可能会。
might be they they might.
我只是觉得他们没有参与这个合同谈判。
Just don't think they were involved in this contract negotiation.
因为我看到Palantir展示过，他们有三四个模型，会混合使用来生成回应。
Because I've seen I've seen Palantir, like, show that they have, like, three or four models that they sort of, like, mix together to get responses.
我不确定。
I don't know.
OpenAI在周五宣布了与国防部的交易，就在国防部长佩特·海格塞斯将竞争对手Anthropic和供应链风险指定为对手后数小时。
OpenAI announced its defense department deal on Friday, hours after the defense secretary Pete Hegseth designated rival Anthropic and Supply Chain Risk.
目前仍不清楚这笔交易是否会最终达成。
There's still a question as to whether or not that's going to go through.
所以日程表上显示有42.8%的可能性，但这个数字几乎没有变化。
So call sheet has it at a 42.8% chance, and it hasn't really moved.
不清楚这是否只是停留在推文层面的虚张声势，还是会有什么实质动作？
It's unclear if that's just, like, staying in tweet land and is, like, bluster, or will there be something there?
我的意思是，我不得不认为，此时此刻，国防部和Anthropic一定在互相沟通。
Will will there I mean, I have to imagine that at this point, both the Department of War and Anthropic are talking to each other.
你有更多信息吗？
Do you have more information?
是的。
Yeah.
Gemini确实与他们达成了协议。
Gemini does have deals with They do.
国防部。
The DD.
是的。
Yeah.
是的。
Yeah.
所以还有七月那笔两亿美元的交易。
So there's the 200,000,000 one that was in July.
是的。
Yeah.
Anthropic 也参与了那笔交易。
That was Anthropic was also in that one.
是的。
Yeah.
然后他们最近还做了那个生成式AI相关的。
And then they did there was the recent, like, genai.
邮件。
Mail.
你还记得吗？
Do you remember
不记得。
No.
塞斯在推特上谈论过这个吗？
Seth was tweeting about this?
哦，有意思。
Oh, interesting.
好的。
Okay.
但他们也参与了这件事。
But they're, like, involved in that too.
是的。
Yeah.
但我感到震惊的是，AWS在2026年如何在FedRAMP和联邦机密网络方面占据优势？
But I I am what I'm so shocked by is how does AWS have an advantage in 2026 on FedRAMP and, like, federal classified networks?
我知道这很难。
Like, I get that it's, like, hard.
你需要做大量的网络安全工作。
You have to do a lot of cybersecurity.
你必须真正加强服务器的安全性。
You have to really harden the servers.
但你可能会觉得，微软和谷歌早就应该掌握了如何全面支持美国政府最深层需求的方法，并在亚马逊之前就获得全部认证，但我原以为微软和谷歌至少会比亚马逊晚一两年才做到。
But you would think that Microsoft and Google would have figured out how to fully support the deepest depths of the US government and get full approval before like like, May maybe Amazon would have gone first, but I would have thought that Microsoft and Google would be, like, a year or two behind.
因此，你会看到Gemini和OpenAI迅速完成这些合规要求。
And so you would see you would see Gemini and OpenAI, like, checking those boxes really fast.
但我也不确定。
But but I I don't know.
我想，AWS在这方面可能确实仍有明显优势。
I guess I guess AWS does have, like, a a material advantage there still.
阿尔特曼表示，虽然他不后悔与国防部达成协议，但他后悔过早宣布这一决定，告诉员工这看起来像是投机行为，且与行业整体立场不一致。
Altman said that while he didn't regret signing a deal with the defense department, he wished he hadn't announced the decision so quickly, telling staff that it looked opportunistic and not united with the field.
我认为这是一个很好的观点。
And I think that's I I I think that's a good point.
我同意他的看法。
I agree with him on that.
他的言论呼应了他周一在X平台上与员工分享的一份备忘录，其中他表示这笔交易显得机会主义且草率。
His remarks echoed a memo he shared with the staff and on x Monday in which he said the deal looked opportunistic and sloppy.
公司内部及整个硅谷的一群人工智能研究人员在过去几天里批评阿尔特曼和该公司，认为他们本质上同意允许人工智能用于所有合法用途的协议，是对五角大楼的妥协。
A groundswell of AI researchers in the at the company and across Silicon Valley spent the past few days criticizing Altman and the company for what they saw as a capitulation to the Pentagon by essentially agreeing to a deal that allowed AI to be used in all lawful cases.
我们会看到法律将走向何方。
And we will see where where where the laws go.
我最近又深入研究了一下关于监控的相关规定。
I've I've been digging a little bit more into, like, the rules around surveillance.
实际上，相关规定非常多，而且非常复杂。
There's there's actually, like, a ton of them, and it's a very nuanced thing.
因此，感觉很难将这些规定浓缩成一句话。
And so it feels like it's hard to to to bake it down into a single line.
但无论好坏，社交媒体上能引发关注的往往只是一句话。
But a a single line is what hits on social media for better or for worse.
是的。
Yeah.
所以你可以看到一句话，然后觉得：太棒了。
And so you you you can see a single line and be like, that's amazing.
这正是我所理解的意思，而这可能意味着好事，也可能意味着坏事。
It it's exactly what I what I think it means, and that and that could either mean good or bad.
所以
So
非常非常棘手。
very, very tricky.
斯科特·库波尔今天早上在CNBC上表示，没有任何私营公司能够决定我国国家安全的条款。
Scott Kupor went on CNBC this morning and said no private company will ever dictate the terms of our national security.
Anthropic试图在其与美国政府的合同中加入使用条款是不可接受的，其产品将不再被美国财政部或任何其他政府机构使用。
Anthropic's attempts to push to push use clauses into their contracts with the United States government are unacceptable, and their products will no longer be utilized by the US Treasury or any other government agencies.
他们在此问题上进一步强化了立场。
Doubling down there.
Semaphore 上还有一篇报道。
There was also a story in in in Semaphore Yeah.
今天早上来自里德的消息。
This morning from Reed.
他写道，Anthropic的投资者在它与五角大楼的对抗中并未支持它。
He writes, Anthropic's investors don't have its back in its fight with the Pentagon.
非常有趣。
Is very interesting.
Anthropic可能正在与五角大楼对峙，但这家人工智能巨头却在没有其知名支持者明显声援的情况下独自坚持。
Anthropic may be standing its ground against the Pentagon, but the AI powerhouse is doing so with a noticeably quiet quarter, its own high profile backers.
尽管该公司态度强硬，但硅谷最大的企业们却保持沉默。
Despite the company's defiance, Silicon Valley's biggest players have remained silent.
据两位了解此次会面情况的人士称，在亚马逊首席执行官安迪·贾西与皮特·赫格塞斯最近的一次会晤中，提到了Anthropic的问题。
In a recent meeting between Amazon CEO Andy Jassy and Pete Hegseth, the issue of Anthropic came up according to two people who were briefed on the encounter.
亚马逊已向这家初创公司投资了数十亿美元，这是亚马逊定制芯片战略的关键部分，因为亚马逊是该公司TraniumAI芯片的最大买家。
Amazon has invested billions in the startup, a crucial part of Amazon's custom chip strategy as the largest consumer of the company's Tranium AI chips.
并且还曾威胁要将Anthropic列为供应链风险企业，这将使许多军方供应商无法与该公司开展业务。
And has been threatening to designate Anthropic a supply chain risk, which would make it impossible for many of the military suppliers to do business with the company.
据知情人士称，贾西回避了，拒绝站在Anthropic一边。
Jassy demurred declining to take Anthropic side, the people said.
他并不是唯一一个。
He wasn't alone.
尽管Anthropic首席执行官达里·阿马德对五角大楼的要求以及竞争对手OpenAI的萨姆·阿尔特曼的持续评论表达了强烈反对，
Despite Anthropic CEO, Dari Amadeh's very vocal opposition to the Pentagon's demands and rival OpenAI, Sam Altman's running commentary on the matter.
一些Anthropic投资者却保持了沉默。
A number of Anthropic investors have remained silent.
一位投资者告诉Semaphore，公开发声可能会进一步激化与政府的关系，他们仍希望这个问题能够得到解决。
One investor told Semaphore that speaking up might further inflame things with the admin and that they were still holding out hope the issue could be resolved.
另一位表示，Anthropic要求他们不要发声。
Another said that Anthropic requested they say nothing.
Anthropic和亚马逊均拒绝置评。
Anthropic and Amazon declined to comment.
五角大楼也未立即回应置评请求。
The Pentagon didn't immediately respond to a request for comment as well.
这里最重要的是我们要让皮特·赫格塞斯上Twitch。
Most important thing here is that we get Pete Hegseth on Twitch.
亚马逊的安迪·贾西正在与皮特·赫格塞斯会面。
Amazon, Andy Jassy is sitting down with Pete Hegseth.
赫格塞斯喜欢推特。
Hegseth likes Twitter.
他喜欢X。
He likes X.
让他现场直播那些记者会。
Get him live streaming those press conferences.
把他放到Twitch上。
Put him on Twitch.
这才是我们的目标。
That's the goal here.
直接去做。
Go direct.
直接进行。
Go directly.
我们会我们会我们会谈谈
We'll we'll we'll talk about
关于来回往复的讨论。
ball about the back and forth.
诺亚·史密斯总结了这一点。
Noah Smith summed it up.
老实说，在本·汤普森与迪恩·鲍尔的辩论中，我们中午请来了迪恩·鲍尔做客节目。
Honestly, in the Ben Thompson versus Dean Ball debate, we have Dean Ball on the show at noon.
我觉得本是对的。
I think Ben is right.
诺亚同意本的观点。
Noah agrees with Ben.
在任何国家，都不可能让私营公司完全掌控人类历史上最强大的武器。
There was just no way or in any nation state there was no way America or any nation state was ever gonna let private companies remain in total control of the most powerful weapon ever invented.
这是一个关于‘我们身处何地’的大问题。
This is a big question about, like, where are we?
我们是不是已经到了那里？
Like, we like like, are are are are we there yet?
这是否是人类有史以来最强大的武器？感觉这更像是为未来即将来临的事物进行的一次演练，这确实是一个重要的对话，但我们还远未到达那个阶段。
Like, is this the most powerful weapon ever invented that feels like this feels very much like like a training run for something that's coming in the future, and it's an important conversation, but we're not necessarily there yet.
所以我们会看到迪恩·鲍尔的回应。
So we'll see Dean Ball replied.
他说，迪恩对本的争论真是令人惊讶，而作为从Android时代就开始阅读本的人，我视本为我当前项目的灵感源泉，尽管我在这里并不认同本的观点。
He said, also Dean versus Ben is wild and an honor for me as a guy who's been reading Ben since the days of Android will not commodify iOS and views Ben as a core inspiration for my own project today, even though I disagree with Ben here.
不久后，我将在某个播客中更深入地谈论这个话题。
You will hear much more from me on this soon on a certain podcast.
我希望他指的是TBPN。
I hope he's referring to TBPN.
但问题是，本反对监管，也不承担国家没收AI的后果，你同样不承担。
But the thing is Ben is anti regulation and does not own the consequences of state seizure of AI, neither do you.
迪恩·鲍尔又发了一篇帖子。
There was another post by Dean Ball.
我想让他再深入解释一下。
I wanna get him to to unpack a little bit more.
他说，人工智能政治中最重要的分歧现在并不是自由派与保守派、民主党与共和党、EAC与EA，或安全与反安全，而是是否认真对待先进人工智能这一概念。
He said, it is so clear that the important fissure in AI politics is not now right now is not liberal versus conservative, Democrat versus Republican, EAC versus EA, or safety versus anti safety, but instead takes advanced AI seriously as a concept versus does not take advanced AI seriously.
这让我觉得特别有趣，因为迪恩和本似乎都属于‘认真对待先进人工智能’这一阵营。
And why this is so interesting to me is because it feels like both Dean and Ben are in the takes advanced AI seriously as a concept camp.
所以我们会看看他是否同意我对他们两人立场的理解，因为看起来他们都在‘认真对待先进人工智能’这一阵营，但他们对政府应如何施压私营企业却存在分歧。
So we'll see if he agrees with my understanding of both of their positions because it seems like they both are in the same camp of takes advanced AI seriously, but then they disagree about the level of pressure that the government should place on a private corporation.
因此，我非常感兴趣，想深入探讨这一点。
And so very, very interested into in in digging into that.
让我跟你聊聊Figma。
Let me tell you about Figma.
无论你的创意从哪里开始——Figma、Make、VOD、Code、Codex还是Sketch——Figma画布都是创意交汇、产品成型的地方。
No matter where your idea starts, Figma Make, VOD Code, Codex, or Sketch, the Figma Canvas is where ideas connect and products take shape.
用 Figma 沿着正确的方向构建。
Build in the right direction with Figma.
我们继续来看威尔·米尼蒂斯。
Let's move on to Will Minitis.
他也对达里乌斯的言论发表了看法。
He has also been chiming in on Darius' statements.
科技界选择相信一系列奇怪的小虚构故事，作为其工作道德观的基础。
There's a There's a set of strange little fictions that tech has chosen to believe as some sort of basis set for the perceived morality of their work.
其中最主要的是认为法律体系不存在，即使存在，也是一个精确的机械过程。
Chief amongst these is the idea that the legal system doesn't exist and when it does, it it is it's a precise mechanistic process.
科技被视为一种正和博弈，这意味着你假装现代企业所使用的标准武器——胁迫性诉讼、知识产权封锁和恶意诉讼、标准的白鞋律师事务所手段——都是负和或某种坏事，我们这里不这么做。
Tech is a positive sum in a community which means you pretend that the standard weapons that exist to the modern corporation, coercive litigation, IP blocking and trolling, standard white shoe law firm stuff
你这是在狠狠批评白鞋律师啊。
Really taking shots at white shoe lawyers there.
是负和的，或者是什么坏事，我们这里不这么做。
Is negative sum or something bad we don't do here.
是的
Yeah.
这导致了一种对法律程序的奇怪形式主义观点，我们必须不惜一切代价避免它。
This leads to a weirdly formalist view of the legal process that we must avoid it at all cost.
但如果我们必须参与其中，那么它就是一个公正且可重复的过程，必然会产生结果，即使不是真相，至少也会反复且可证明地得出相同的结果。
But if we must engage in it, then it's a fair and repeatable process that must necessarily yield, if not the truth, then at least the same outcome repeatedly and provably.
正是这种本能造就了Anthropic那奇特的合同形式主义。
It's this instinct that gives you Anthropic's odd contract formalism.
这就是为什么你会认为最好的策略是在一开始就亮出底牌，并以为自己已经赢了。
This is how you end up thinking your best move is to show your cards at the forefront and thinking you've won the hand.
只要你背叛了科技行业用来让自己感觉良好的这些荒谬规范，你就拥有巨大的超越机会。
There's incredible potential for outperformance available to you for betraying any of these silly norms that tech uses to feel good about itself.
现代企业必须利用一切可能的优势，而圣战需要神圣的武器，所以去获取
The modern corporation must avail itself of all possible edge and holy wars require holy weapons, so get The
我在这里印象深刻的观点是，如果我们参与其中，那么它就是一个公正且可重复的过程。
idea that sticks with me here is that if we engage in it, then it's fair in a repeatable process.
我跟一个正在处理法律事务的人聊过。
I talked to someone who is who is who is doing something with the legal system.
他说，就这个特定问题而言，这确实是一个典型的左右派分歧问题。
And he said that, like, on this particular issue, it was a true, like, left right issue.
自由派法官总是支持案件的某一方，而保守派法官则总是支持另一方。
Like, a liberal judge would always cite on one side of the case, and a conservative judge would always cite on the other side of the case.
所以，无论他论证得多好，最终都取决于他的案件遇到了自由派法官还是保守派法官。
So no matter how well he argued, it came down to did his case get a liberal judge or a conservative judge.
我当时想，哦，这看起来挺简单的。
And I was like, oh, well, like, that seems simple.
就是找个跟你公司和立场一致的法官就行了。
Like like, get a judge that's aligned with your company and your position.
他回答说，是的。
And he was like, yeah.
坏消息。
Bad news.
他们决定我该得到哪位法官的方式，是真正地转一个物理轮盘。
The way they decide which judge I get is they literally spin a physical wheel.
我当时就说，所以你什么都做不了？
And I was like, and so there's nothing you can do?
他说，是的。
He's like, yeah.
不行。
No.
它确实是安全的。
Like, it it is secure.
这是一个真正的流程。
It's like a real process.
而且，这是有意为之的。
And, like, that's by design.
有些事情最好交给民主程序来处理，加入一点随机性，这样你就无法操控轮盘，选到你想要的法官，得到你想要的结果。
Like, some things are better left in the democratic process to inject a little random randomness so that you can't put your finger on the wheel and pick the right judge and get exactly what you want.
关于 Bayt 播客
Bayt 提供中文+原文双语音频和字幕，帮助你打破语言障碍，轻松听懂全球优质播客。