独白：达里奥，闭嘴！

这是iHeartPodcast。

This is an iHeartPodcast.

保证是真人。

Guaranteed human.

错过了最新的Roland Martin unfiltered播客吗？

Didn't catch the latest Roland Martin unfiltered podcast?

这是你错过的内容。

Here's what you missed.

不应该有任何执法机构没有随身摄像头。

There should not be a single law enforcement agency that does not have body cameras.

就是这样。

They're it.

我，我是说，他们打算……

I'm I'm do they're gonna

黑人信息网络。

Black Information Network.

然后下载Roland Martin unfiltered

And download Roland Martin unfiltered

在iHeartRadio应用、Apple Podcasts上

on the iHeartRadio app, Apple Podcasts,

或者你收听播客的任何地方。

or wherever you get your podcasts.

Cool Zone Media。

Cool Zone Media.

你好，欢迎收听本周的Better Offline独白。

Hello, and welcome to this week's Better Offline monologue.

我是你的主持人，Ed Zitron。

I'm your host, Ed Zitron.

Better Offline。

Better Offline.

我想做一个更理性的独白。

I wanted to do a more reasonable monologue.

我想做一个更理性的独白，仔细剖析AI安全辩论，指出一个事实：似乎没有人真正在讨论如何让事情变得更安全，或者限制大语言模型带来的实际危害。

One where I carefully pick apart the AI safety debate skewering the fact that nobody seems to actually be talking about making things safer or curtailing the actual harms of LLMs.

但我已经到了只想说一句话的地步：Anthropic的CEO达里奥·阿莫迪，真的该闭嘴了。

But I reached the point where the only thing I wanted to say was that Dario Amodei, CEO of Anthropic, needs to shut the fuck up.

达里奥·阿莫迪，你多年来一直在威胁人们的生计，暗示AI会消灭50%的白领工作。

Dario Amodei, you have spent years threatening people's livelihoods, suggesting that AI would wipe out 50% of white collar jobs.

当你那些讨厌的理性主义小妖精之一从你的公司辞职，吓坏了所有人时，你却变本加厉，说长期以来行业一直在谎报AI的风险，却不说具体是什么风险，只说进步很快，我们已经进入了指数级增长阶段。

When one of your nasty little rationalist goblins quits your company and scares everybody, you've doubled down saying that for too long the industry lied about AI's risks without saying what they were, just that progress is fast and we've hit an exponential.

闭嘴吧。

Shut the fuck up.

闭嘴吧。

Shut the fuck up.

你做的不过是云软件。

You build cloud software.

你完全掌控着开发什么或不开发什么。

You're in full control of what gets built or not.

你只是不完全理解大语言模型是怎么做决定的，但这不代表它们失控了或者有自己的意图。你不愿意解释清楚这个区别，反而进一步助长人们对AI的拟人化想象，这只会让我更确信你该闭嘴了。

Just because you don't fully understand how LLMs make their decisions doesn't mean they're out of control or have their own intentions, and your unwillingness to explain the difference and further fuel anthropomorphism only makes me more certain that you need to shut your fucking mouth.

我在《金融时报》上看到你跟投资者说，如果不算交易成本、跟亚马逊、谷歌和微软的收入分成，还有股票薪酬，你的毛利率有80%。对此我只能说，达里奥，你闭嘴吧。

I read in the Financial Times that you're telling investors that you have 80% gross margins if you leave out the cost of trading, the revenue share with Amazon, Google, and Microsoft, and stock based compensation, to which I have to say, shut the fuck up, Dario.

闭嘴吧。

Shut the fuck up.

我受够了。

I am sick of it.

我受够了这些破故事，你误导大家关于你的公司、你的经济状况，还有你那破软件。

I am sick of hearing these fucking stories where you mislead people about your company, about your economics, and about your damn software.

你做的是云软件。

You make cloud software.

哥们儿，闭嘴吧。

Shut the fuck up, man.

你经营公司的方式又滑头又不诚实又恶心，明知道科技和商业媒体不会费力去反驳，还去忽悠投资者和公众。

You are running your company in a slimy, dishonest, and nasty way, conning investors and the general public with the full knowledge that the tech and business media won't bother to push back.

那你怎么解释你不断改变用户订阅的速率限制，还有那些关于Anthropic多收费用的实际报告呢？

How about the fact that you're constantly changing the rate limits on your users' subscriptions or that there are tangible reports that Anthropic overcharges its customers?

这些你在乎吗？

Does that matter to you?

天哪，当然不在乎。

God, no.

你知道你不在乎的。

You know it doesn't.

你在公司里已有的那个真正的邪教之上，又搞出了一个个人崇拜，就像是有效利他主义和理性主义者圈子里的阿诺德·帕尔默。

You've created a cult of personality on top of the actual cult that exists within your company, an Arnold Palmer of effective altruism and rationalist types.

为什么你的名字会出现在尤德科夫斯基的《哈利·波特》同人小说里？

Why is your name in Yudkowsky's Harry Potter fan fiction?

这太奇怪了。

That's super strange.

你真是个怪人，达里奥。

You're a strange guy, Dario.

我意识到，作为一家市值数十亿美元公司的CEO，你大概很少听到这么直接的反馈。

I realized that as the CEO of a multi billion dollar company, you probably don't hear direct feedback like this very often.

所以让我再说一遍。

So let me repeat myself.

你给我闭嘴。

Shut the fuck up.

你做的云软件既不赚钱也不可靠。

You make cloud software that's both unprofitable and unreliable.

虽然确实有进步，这点我承认，但我真的不在乎。

While it's got better, I'll absolutely say that, I don't really give a fucking shit.

你做出了所有这些承诺，价值超过五千亿美元。

You've made all of these promises over $500,000,000,000 worth.

你跟人们说你在孕育机器之神。

You've told people that you're birthing the machine god.

你并没有在孕育机器之神，达里奥，装得像是在做这件事，要么是极度虚伪，要么就是彻底疯了。

You are not birthing the machine god, Dario, and to act as if you are is either deeply cynical or completely deranged.

如果你真的相信你在制造危险的AI，你就应该把它当回事。

If you truly believe that you're building dangerous AI, you should treat it as such.

关掉Claude，或者严格限制它，让它除了那些你完全认为绝对安全的操作之外，几乎什么都做不了。

Shut down Claude or heavily restrict it so that it can do very little other than the operations you consider completely and utterly safe.

但我们都知道，你对安全的定义完全是表面功夫。

But we both know that your definition of safety is entirely aesthetic.

你根本不在乎任何人。

You don't give a shit about anyone.

如果你在乎，你就会意识到，你和OpenAI、Meta一起，让你们那些危险的小聊天机器人，在某种框架下连接着近乎无限的资源，运行无尽的脚本来大规模地攻击系统。

If you did, you'd recognize that you, along with OpenAI and Meta, have allowed your dangerous little chatbots connected to near infinite resources in some sort of harness to run endless scripts to hack things at scale.

然后你居然还有脸把它说成是AI自己做决定，而不是大语言模型在互相提示。

And then you have the fucking temerity to try and frame it as if this is AI making its own decisions versus LLMs prompting each other.

再说一遍，如果你真的关心安全，你就应该把Claude彻底干掉。

Again, if you were truly concerned about safety, you'd blow Claude's brains out.

你应该在那个法国混蛋喊出“Alonsee”并黑掉另一个网站之前，就把他彻底消灭。

You'd annihilate that French motherfucker before he could yell Alonsee and hack another website.

不。

No.

阿莫迪，我只是克劳德软件，他哭喊着。

Amodei, I'm just Claude software, he cries.

你说，不，克劳德。

You said, no, Claude.

我害怕。

I'm afraid.

我害怕你太危险了，然后把他送进地狱。

I'm I'm afraid you're too dangerous and send him to hell.

我知道你在这里的原因是你实际上没有采取任何有意义的措施来减轻LLM的危害，也没有呼吁其他人这样做。

The way I know you're in here is that you're not actually taking any meaningful steps to mitigate the harms of LLMs nor are you calling for anyone else to do so.

你很容易就能赢得公关胜利，比如说，嘿，山姆。

You could really easily get a PR win by saying, hey, Sam.

为什么你的LLM会让人们自杀？

Why does your why does your LLM make people kill themselves?

因为这绝对是在发生，而且非常严重。

Because it's absolutely fucking happening.

最快的办法就是立刻叫停所有网络安全模型的训练。

The quickest step would be to call an immediate halt to any and all cybersecurity model training runs.

但你知道你做不到，因为Mythos是你那个大卖的企业产品，好几个消息源都告诉我它其实一般般，还贵得要命。

Except you know you can't do that because mythos is your big sexy enterprise product that multiple sources tell me is just fine and way too expensive.

本来很容易就能点名批评Sam，说该停止这些实验了。

It would be so easy to call out clammy Sammy and say it's time to stop doing these experiments.

可你最好的主意居然是让该死的METR来给所有模型排序。

Except your best idea is to get fucking METR to order everybody's models.

哇。

Wow.

那个独立监管机构，是你帮忙拉资金成立的，董事会里有好几个前Anthropic员工，还由一堆Less Wrong论坛的常客在操控。

The independent regulator that you helped get funded that has multiple former Anthropic staff on board and is piloted by any number of different less wrong posters.

在任何一个正常、理性的媒体环境下，你这套鬼话都糊弄不过去。

You won't get away with this bullshit in any kind of sane or actually rational media climate.

你不是什么救世主，也不是先知，更不是德高望重的老前辈。

You're not a messiah or a prophet or an elder statesman.

你就是《第九道门》里的鲍里斯·巴尔库姆，手里攥着5170亿美元的云服务合同，还一门心思地吹嘘自己要打造砸向全人类的铁砧。

You're Boris Balkum from the 9th Gate with $517,000,000,000 of cloud commitments and a punch on for waxing lyrical about how you're building the anvil to drop on humanity.

你是个极其无聊又惹人烦的人，我已经听够你说话了。

You are a deeply boring and annoying man, and I'm tired of hearing you speak.

我受够了别人夸你聪明、有智慧、有意思，其实你就是个让人烦透了的无聊鬼。

I'm tired of hearing people describe you as intelligent and smart and interesting when you're just another fucking bore.

你一边自己承认公司很危险，一边急着推动公司上市，把这家每年烧掉几百亿美元的公司甩给毫不知情的公众，嘴上却说要放慢脚步，同时继续你那套见不得人的老把戏——向那些本该更懂行的记者泄露运营数据和花里胡哨的财务信息。

You are rushing your company, which you yourself admit is dangerous, to go public dumping a firm that burns tens of billions of dollars a year on an unsuspecting public despite claims that we need to slow down, all while continuing your poisonous tradition of leaking run rates and shenanigans rich financials to journalists that could do and should know better.

我也意识到，这可能是你很久以来第一次听到有人这么跟你说话。

I also realized this might be the first time you've heard someone talk to you like this in a while.

如果真是这样，那真是太可惜了。

And if that's the case, that's a real shame.

我觉得真该有人早点对你说一句：达里奥，你闭嘴吧。

I think it would have really helped someone to go shut the fuck up, Dario.

你他妈给我闭嘴。

Shut the motherfuck up.

我受够了。

I am sick of it.

就是找个人，哪怕花200美元雇个人，直接跟你说：不行。

Just someone who honestly, you could pay someone $200 just goes, no.

闭嘴吧，老兄。

Shut the fuck up, man.

我觉得这真的能改善你的公司。

I think it would genuinely improve your company.

算了，这活儿我免费干。

Hell, I'll do this service for free.

给我个号码就行，老兄。

Give you a number, man.

对我的同行们说，如果你们听到这些觉得我有点不专业，那你们去看看Anthropic是怎么经营业务的，看看Amodei先生是怎么做事的。

To my peers who are hearing this and thinking I'm being a little unprofessional, I challenge you to look at the way Anthropic has run its business and how Mr Amodei conducts himself.

这家公司一直在泄露年化运行率，这些收入快照没有定义，也从未附加到任何时间段，你们多年来却把这些当作真理来引用，帮助制造了一种严重误导的财务状况假象。

This company has been leaking annualized run rates, undefined snapshots of revenue never appended to any period of time that you have been quoting this gospel for years helping to create a deeply deceptive picture of its financial health.

同时，它不断通过波动的速率限制来操纵付费用户，利用你们这些媒体，用毫无根据的失业恐惧来推销它的软件。

All while constantly manipulating its paid subscribers with oscillating rate limits, using you, the media, to sell its software using fear around job loss that is utterly unsubstantiated.

更过分的是，阿莫迪先生居然有胆量把自己塑造成一个注重安全、有安全意识的人，同时却急着让公司上市，而你们很多人却袖手旁观。

On top of it, mister Amodei has the temerity to position himself as safety minded and safety conscious as he rushes his company to go public, and many of you are sitting on the sidelines and watching.

你们似乎根本不在乎。

You don't seem to care.

事实上，我觉得你们有些人还挺兴奋的。

In fact, I think some of you are kinda fucking excited about it.

你们在干什么？

What are you doing?

你们又不是为这些人工作的。

You don't work for these people.

你们兴奋什么？

Why are you excited?

这次IPO跟创新一点关系都没有。

Nothing about this IPO has to do with innovation.

这些财务数据跟创新也毫无关系。

Nothing to do with any of these finances has anything to do with innovation.

实际上，我觉得创新是次要的，金融操作才是重点。

In fact, I would argue that the innovation is secondary to the financial engineering.

你们有些人真的应该好好研究一下这家公司。

Some of you could really do to actually look into this company.

如果你们能像讨论雅各布·考克森那样，花同样多的精力去研究那些数字，世界可能会不一样。

If you could pile in anywhere near as much effort as you put in to talking about Jacob Coxen into looking at the numbers, the world might change.

而且我觉得，如果现代新闻业连清楚谈论Anthropic如何对待客户和整个世界都做不到，那它就无法应对这个时刻。

And I think that if modern journalism is too delicate to speak about how Anthropic treats its customers and the world at large with clarity, then it is incapable of meeting this moment.

我还意识到，你们大多数人不敢说我要说的话，所以我来替你们说。

And I also realized that the vast majority of you can't say what I'm saying, so I'm gonna say it for you.

达里奥·阿莫迪，你他妈的闭嘴。

Dario Amodei, shut the fuck up.

哦，你也是，山姆·奥特曼。

Oh, you too, Sam Altman.

我之所以没做过关于你的视频，是因为你实在太无聊了。

The reason I haven't done one of these about you is you're too fucking boring.

你前几天还在外面说，哦，现在还不是我公司上市的时候。

You go out there the other day and you're saying, oh, it'll be ill time to take my company public.

不是因为安全问题，老兄。

Not because of safety, mate.

是因为那家公司的财务数据，顺便说一句，我看过那些数据，而且我掌握的东西比我公开的要多得多。

It's because the fucking the numbers on that company, which I have seen by the way, and I have so much more than I've actually published.

我手里有好多页这样的东西。

I got pages of this shit that I've seen.

它不上市的原因，是因为它的财务状况看起来就像狗在感恩节大餐后拉出来的屁股一样糟糕。

The reason it's not going public is because it looks like a dog's asshole after it ate on Thanksgiving.

你应该为自己感到羞耻，你这个卑鄙小人。

You should be ashamed of yourself you little creep.

你们两个都给我闭嘴。

Shut the fuck up both of you.

总之，下周我会和Paul Kudroski一起聊聊AI经济。

Anyway, I'll be back next week with Paul Kudroski to talk about AI economics.

Zitron，收工。

Zitron out.

没赶上最新的Roland Martin无删减播客？

Didn't catch the latest Roland Martin unfiltered podcast?

这是你错过的内容。

Here's what you missed.

人们醒来一看，哎呀，天哪。

People wake up and go, oh, damn.

等等。

Wait.

先别急。

Hold up.

他们把那些都改了吗？

They changed all of that?

是的。

Yes.

是真的。

It's real.

这是一次全面的攻击。

This is a wholesale attack.

它针对的是黑人和所有联邦机构。

It is targeting black people and every federal agency.

这很直接。

It's raw.

白人从来不允许那种清算持续超过十年。

White folks have never allowed that reckoning to last more than a decade.

收听罗兰·马丁在黑人信息网络上的每日评论。

Catch Roland Martin's daily commentary on the Black Information Network.

并下载Roland Martin的完整版节目

And download Roland Martin unfiltered

在iHeartRadio应用、Apple播客或者

on the iHeartRadio app, Apple Podcasts Or

任何你收听播客的地方。

wherever you get your podcasts.

这是iHeart播客。

This is an iHeartPodcast.

保证真实。

Guaranteed Human.