人工智能商业秘密战争 + 经济学家称“我们必须立即行动” + HatGPT

据我所知，这份协议一直没被撤销。

And that has never been rescinded to my knowledge.

对。

Yeah.

所以，从法律上讲，他们手里还有一份文件，规定不能在公开场合互相抹黑。

And so, technically, they still have a document on the books that says we're not supposed to trash each other in public.

从外面看，听他们彼此怎么评价对方，我本来以为。

From the outside, reading sort of how they talk about each other, I would assume that

他们签的是个诋毁协议。

they had a disparagement agreement.

是的。

Yes.

而且是唯一一份诋毁协议。

Which An only disparagement agreement.

咱们一离开这儿，马上就得干起来。

The minute we leave here, it's on, bro.

这下可真要开打了。

It's on like Donkey Kong.

没错。

Yes.

这些公司彼此之间恨透了对方。

So these companies absolutely hate each other.

他们正在为模型对订阅用户的使用条款争得不可开交。

They are bickering over the the sort of terms of these models being accessible to subscribers.

我觉得他们很快就要打起价格战了。

I think they are going to start sort of doing some kind of price war with each other.

我的预感是，Anthropic先降低token价格，然后OpenAI再压得更低。

That's my expectation is like, you know, Anthropic will kind of reduce the price of tokens, and then OpenAI will reduce it below that.

我觉得这对消费者来说是好事，至少从狭义的消费者利益角度看，前沿AI实验室之间的竞争是有益的。

And it'll sort of be, I think, good for consumers in a very sort of narrow, you know, consumer interest perspective because competition among Frontier AI labs is good, I would argue.

对。

Yeah.

这对消费者来说确实是个好事。

It's it's definitely good consumers.

我觉得人工智能竞赛本身有些令人担忧的地方，不过那可能是另一个话题了。

I think that there are elements of the, you know, AI race itself that are scary, but perhaps that's a different conversation.

不对。

No.

我想把讨论重点放在这上面。

I think that's that's where I wanna go with this.

是的。

Yeah.

因为我觉得，OpenAI和Anthropic之间的这场争斗不仅没有缓和，反而愈演愈烈，这非常危险。

Because I think, like, this feud between OpenAI and Anthropic, which does not seem to be subsiding and, in fact, seems to be getting worse, I think it is very dangerous.

我不喜欢这些实验室彼此相处的方式。

I do not like this way that these labs are conducting their relations with each other.

我不喜欢OpenAI的人对Anthropic充满敌意，甚至一心想要压垮对方。

I do not like how much the people at OpenAI hate Anthropic and are motivated by wanting to crush them.

我不喜欢Anthropic的人对OpenAI的敌意，以及他们想彻底压垮对方的态度。

I do not like how much the people at Anthropic hate OpenAI and wanna crush them.

我觉得我们正步入一个时期，这些强大系统可能带来真实而危险的风险，这些公司应该愿意合作，共同解决这个问题，实验室之间的协调会变得越来越重要。

I think that we are heading into a period where there may be real and dangerous risks associated with these very powerful systems, and these companies should be open to collaboration and solving this problem together, and coordination between labs, I think, is going to become increasingly important.

如果你们一直互相争斗，就根本做不到这一点。

And you can't do that if you're constantly fighting each other.

我完全同意。

I completely agree.

你知道吗，这周DeepMind的负责人发表了一篇文章，提出一个想法：设立一个政府监管机构，来制定流程，控制前沿实验室何时可以发布新模型，或许能解决最近几个月出现的那些问题。

You know, who runs DeepMind put out a piece this week where he talked about an idea for a kind of government regulator that would, you know, create a process around when a Frontier Lab could release a new model, maybe solve some of the issues that we've seen in recent months.

他提到，如果各实验室未来能成功合作，共同放缓AI的发展步伐，那会是件好事。

And one of the things that he talks about is it would be great if the labs at some point could successfully collaborate around a slowdown in the development of AI.

也许到时候政府会直接下命令强制这么做。

And now maybe that's something that the government will just, you know, decree.

但在没有这种强制措施的情况下，你所提到的这种情况——也就是这些公司领导人能随时打电话沟通，考虑对整个世界有利的事，而不是只想着怎么获得更多转发——才是我们真正需要的。

But in the absence of that, you want to see exactly what you're talking about, Kevin, which is, like, the leaders of these companies being able to pick up the phone and call each other and think about, like, what is good for the rest of the world and not just what is gonna get a bunch of retweets.

是的

Yeah.

我真的希望达里奥、萨姆和德米斯能坐下来，好好谈一谈，把问题解决掉。

I just desperately need Dario and Sam and Demis to, like, all get together, take a little MDMA, work it out.

对

Mhmm.

多一点同理心吧。

Find some compassion.

是的

Yeah.

找到共同点。

Find some common ground.

对

Yep.

先把个人恩怨放一放。

You know, put the beef aside.

没错。

Yep.

我们都在同一条船上。

We're all in this together.

我发现有些企业领导者会忘记，我们每个人都是同一普遍意识的不同方面，正是这时他们才会陷入麻烦。

Some some what I found is sometimes business leaders will forget that we are all individual facets of the same universal consciousness, and that's when they get themselves into trouble.

回来后，我们将带来关于人工智能与就业最响亮的警告。

When we come back, the loudest warning yet on AI and Jobs.

这个

This

我是莫里斯·谢马，‘过去十二周’节目的主持人，这档新播客由Serial Productions、The Marshall Project和《纽约时报》联合制作。

is Maurice Schema, host of the last twelve weeks, a new podcast from Serial Productions, The Marshall Project, and The New York Times.

几年前，我收到一位辩护律师的邮件，他想请我写一写他的委托人。

A couple of years back, I got an email from a defense lawyer who wanted me to write about his client.

这位委托人叫大卫·伍德，关在德克萨斯州的死囚牢房里，已经三十多年了。

The client, David Wood, was on death row in Texas, had been there for more than thirty years.

律师写信是因为大卫·伍德的所有上诉都被驳回了。

The lawyer was writing because David Wood had lost all of his appeals.

他即将被执行死刑。

He was set to be executed.

律师的计划是试图证明一件事——三十年来没人成功过的事：得克萨斯州最臭名昭著的连环杀手其实清白无辜。

The lawyer's plan to stop the execution was to try and prove something that nobody had successfully done in three decades, That one of Texas' most notorious serial killers was actually innocent.

并不是律师们没理由可讲。

It wasn't that the lawyers didn't have a case to make.

我知道有两个人伪造证词，害得一个无辜的人被处决。

I know two people fabricated testimony to get a guy executed.

只是他们的时间实在太少了。

It's just that they had so little time to make it.

《最后的十二周》。

The last twelve weeks.

在你常用的播客平台收听。

Listen wherever you get your podcasts.

好的，凯文。

Alright, Kevin.

我们来聊聊人工智能和就业的问题。

Let's turn our attention to the question of AI and jobs.

过去几个月，我自己对这个话题特别感兴趣。

This is something that I have been fascinated by myself over the past few months.

我跟好几位经济学家和科技界人士聊过，讨论AI发展到一定程度后，雇主会不会突然觉得不再需要人类，只要花钱买算力就行。

I've been speaking to several economists and tech leaders about the short, medium, and long term risk of an AI getting so good that all of a sudden employers are going to feel like they don't need people anymore, like they can just spend tokens.

不过，目前很多人对这个观点还是持怀疑态度。

And there's frankly still a lot of skepticism on this point.

所以我特别留意到这周出现了一封新公开信，最初有将近两百位经济学家和研究者联名签署，提醒大家AI可能对经济造成的冲击。

And so that is why I was fascinated to see a new letter come out this week that had been signed by a group of nearly 200 economists and researchers initially, and they were warning about the impact that AI could have on the economy.

是的。

Yeah.

我也看到了。

I saw this too.

这是一份声明。

This was a statement.

很短，就三点。

It was very short, just three short points.

标题是：我们必须立即行动。

It was titled, we must act now.

其实，我们干脆把全文读一遍吧。

And basically, I mean, we can probably just read the whole thing.

读一遍吧。

Read the whole thing.

也不长。

It's not that long.

第一条：未来十年，人工智能可能会变得极其强大。

And it says, point one, AI may become radically more powerful over the next ten years.

第二条：这可能引发一场前所未有的经济变革，规模超过工业革命，但发生的时间却短得多。

Point two, this could drive an unprecedented transformation of our economy larger than the industrial revolution, but unfolding over a vastly shorter timeframe.

它可能带来风险，比如大规模的失业，同时也可能带来机遇，比如生活水平显著提升。

It could bring risks, including large scale job displacement, as well as opportunities such as major gains in living standards.

第三点，经济学家、政策制定者和科技领袖必须立即行动，深入理解变革性人工智能的经济影响，建立相应的激励机制、安全边界和制度体系，引导人工智能朝着增强人类能力、造福社会的方向发展。

Point three, economists, policymakers, and technology leaders must act now to understand the economics of transformative AI and to build the incentives, guardrails, and institutions needed to steer AI in a direction that complements humans and benefits society.

所以我觉得，这份声明本身其实挺温和的。

So the statement itself, I think, is pretty anodyne.

说白了就是：我们得认真对待人工智能，因为它真有可能带来这些影响。

It's basically like, we should take AI really seriously because it could do all these things.

它能带来一些好处。

It could do some good things.

也可能带来一些问题，所以我们得重视起来。

It could do some bad things, and we should pay attention to it.

我觉得更值得关注的是签署人名单，以前《硬分叉》节目的嘉宾，比如弗吉尼亚大学的安东·科罗尼克和Anthropic公司的人，都在上面，还有一些过去对AI可能引发大规模就业变革持怀疑态度的经济学家，这次也签了名。

I think what was more interesting was the list of signatories, former hard fork guests, Anton Koronick from the University of Virginia and Anthropic was on there, as well as, I think, some more skeptical economists who have previously been kind of dismissive of the idea that AI could cause large scale job transformation,

至少在短期内是这样。

at least in the short term.

是的。

Yes.

至少在短期内，签署者中包括几位诺贝尔奖得主，比如多罗娜·萨莫古卢——她一向对AI短期内导致大规模失业的说法持怀疑态度，还有麻省理工学院的大卫·奥托尔，以及其他许多知名经济学家和思想家，他们齐聚一堂发出这份声明，我觉得这特别有意思。

At least in the short term, It includes several Nobel laureates, including Dorona Samoglu, sort of a well known skeptic of the short term AI job loss hypothesis, David Autor at MIT, lots of other just sort of, like, well known economists and thinkers who are coming together to make this statement, and I thought this was really interesting.

对。

Yeah.

所以这周，我们想采访一位参与起草这份声明的人。

And so this week, we wanted to talk to a person who helped to put this statement together.

这个人就是埃里克·布林约尔松。

That person is Erik Brynjolfsson.

他是斯坦福大学的经济学家，协助组织了这份声明，而且他对这些潜在风险非常重视。

He is a Stanford economist who helped organize the statement, and he is somebody who takes these threats really, really seriously.

好。

Yeah.

我认识埃里克很多年了。

I've known Erik for years.

他是一位杰出的经济学家，很早就开始关注人工智能对就业和经济带来的威胁，比大多数经济学家都早，甚至早于ChatGPT的出现。

He's a great economist, and he's been writing about the danger that AI poses to jobs and the economy for a very long time, like longer than most economists, longer than ChatGPT has existed.

他一直都在研究这些问题。

He's been writing about this stuff.

他是斯坦福大学以人为中心的人工智能研究所的高级研究员。

He's a senior fellow at the Stanford Institute for Human Centered A.

我。

I.

同时还是斯坦福数字经济发展实验室的主任。

And the director of the Stanford Digital Economy Lab.

他长期致力于研究人工智能对劳动者和企业的影响。

He's been studying AI's impacts on human workers and firms for a very long time.

为了满足大家想向他提问的期待，我们现在请出埃里克·布林约尔松。

So to meet the demand for the questions that we have for him, we now supply Erik Brynjolfsson.

埃里克·布林约尔松，欢迎来到《硬核》节目。

Erik Brynjolfsson, welcome to Hard Fork.

很高兴又能跟你聊天，凯文。

Well, it's a pleasure to be talking to you again, Kevin.

你早就开始写文章、演讲、思考人工智能对经济的深远影响了，比我知道的几乎任何一位经济学家都要早。

So you have been writing and talking and thinking about the transformative potential of AI on the economy for a very long time, longer than, almost any economist I know.

我很好奇，是什么促使你和一群同行一起发起这封‘我们必须立即行动’的公开信？

And I'm curious what inspired you to organize this letter, this we must act now statement from a bunch of your peers and colleagues.

这件事其实早就在我心里酝酿很久了。

Well, this is something that's been brewing for a long time.

你说得对。

You're right.

我早就一直在想这个问题。

I've been thinking about it a long time.

早在1987年，我就开始教人工智能相关的课程了，那时候你们大部分听众可能还没出生。

I taught my first course about artificial intelligence back in 1987 before most of your listeners were probably born.

我一直深信，当时正在发生的事非同小可。

I I've just been very convinced about what was happening.

但你知道，指数曲线的特点就是，一开始增长很慢，然后突然就爆发了。

But, you know, the nature of exponential curves is they grow, you know, slowly at first and then suddenly.

而今年，尤其是现在，我们正好处在了这个突然爆发的阶段。

And and now, especially this year, we're in the that suddenly part of the curve.

技术发展得太快了，但我们的经济认知根本跟不上。

The technology is going so fast, but our economic understanding is nowhere close to keeping up.

所以我现在尽力做点什么，来改变这种状况。

And and so I'm doing what we can to to try to change that.

所以这封信的标题是‘我们必须立即行动’。

So the letter is titled, we must act now.

我觉得它并没有明确指出具体该怎么做。

I would say it doesn't really specify an action.

那你能说说写这封信的目标是什么吗？那些读到它、签了名的政策制定者和各界专家，你们希望他们采取什么行动？

So tell us about what your goals are with this letter and and what you think the policy makers who read it and all the the sort of esteemed signatories, what should they do?

嗯，这封信主要有两个目的。

Well, there's two main purposes.

首先是发生了一场根本性的转变，我们希望推动这种转变，让人们对人工智能的重视程度真正提升起来。

First is is that there's been a sea change, and we're hoping to catalyze that sea change in attitudes about how seriously we take AI.

就在一两年前，甚至更早之前，我接触的大多数经济学家都对人工智能没什么感觉。

As recently as a year or two ago, and certainly before that, most of the economists I talked to really were not very AI pilled.

他们觉得，嗨，这没什么大不了的。

They thought, hey, it's it's not a big deal.

我记得有一位财政部长曾经对我说：‘你觉得未来几年生产力可能会提升零点几个百分点吗？’

I remember one secretary of treasury saying to me, oh, do you think that maybe, you know, there'll be a couple of tenths of a percent of change in productivity over the coming years?

也就是说，他们想象的只是技术带来的一点点微小影响。

So you're sort of like imagining a a sort of a a very small technological impact.

但我个人认为，这将会是巨大的变革，所以我希望更多人能认真看待这件事，意识到这可能是经济学家最该关注的议题。

And I I personally think it's going to be huge, and I I wanted more people to take it seriously and realize that I think this is the most important thing economists should be working on.

不管他们正在研究什么其他课题，都应该用人工智能的视角去重新审视。

Whatever other projects they're working on, they should look at it through a lens of AI.

我们呼吁的其中一部分，就是让更多经济学家深入思考这个问题，推动更多相关研究。

Part of what we are calling for is is a better understanding, more economists thinking hard about this question, more research going into it.

不过，多伦·阿西莫格鲁和我，还有其他人，早就一直在讨论一个方向，那就是让AI辅助人类，而不是取代人类。

That said, there was the beginning of a direction that, you know, Doron Asimoglu and I and others have been talking about for a while, which is this idea of complementing humans rather than substituting for them.

具体该怎么实现，这问题确实更复杂，但如果我们能让几百位经济学家、不少诺贝尔奖得主，还有全球各地的AI研究者都认同这个方向，那已经是个不错的开端了。

Now fleshing out exactly how to do that is a is a bigger question, but at least if we could get, you know, a couple 100 economists and and a lot of Nobel Prize winners and and other AI researchers around the world to agree on on that direction, that's a good start.

我想跟你聊聊我跟其他科技领袖或经济学家讨论AI可能取代大量工作时，听到的一些反对意见。

Let me hit you with some of the objections that I have heard when I've talked to other tech leaders or economists about the idea that AI might take a lot of jobs.

我自己完全愿意相信，这里确实可能带来重大冲击，但我确实也听到不少怀疑的声音。

I myself, I'm totally prepared to believe that there could be a real disruption here, but I do hear some skepticism.

我常听到的说法是，工作岗位的流失几乎从不只有一个原因，从数据上看，根本不可能准确归因。

So one thing I hear is that job losses almost never have any one cause, and it's basically impossible to pinpoint when you're looking at the data.

这确实没错。

Well, it's certainly true.

经济中同时发生着太多事情。

There's lots of things going on in the economy.

AI并不是唯一的影响因素。

AI is not the only thing.

这是最重要的一件事，而且越来越重要。

It's one of the most important and increasingly the most important thing.

我们写过一篇叫《矿井中的金丝雀》的论文，你可能听说过。

You know, we wrote this paper, Canaries in the Coal Mine, that that you may be familiar with.

几天后我们会发布新版，正好赶得上你这期播客播出，里面也会探讨一些其他可能性。

We're gonna have a new version of it out in in a few days, probably in time for your your airing this podcast, where we look at some of the alternatives as well.

我们分析了利率的变化。

We look at what's going on with interest rates.

我们研究了远程办公的趋势。

We look at remote work.

我们考察了科技行业过度招聘的现象。

We look at the tech overhiring.

我们还看了教育领域的变革。

We look at changes in education.

这些因素都很关键，而且每一个都产生了影响。

All of those are important, and we find that all of them make a difference.

但即使你把它们一个个或同时放进回归模型里，试图理清到底发生了什么，人工智能依然是一个关键因素。

But even when you put them in individually or all at the same time into these regressions to try to to to parse out what's happening, AI continues to be an important factor.

它不只是在摧毁工作机会。

And it's not just destroying jobs.

人工智能也在创造新的工作，很多人都会这么讲。

AI is also creating jobs as as a lot of people will tell you.

我们这些人的责任，像我这样的经济学家，就是要尽早、尽可能细致地弄清楚哪些方面真正重要。

And it's the job of us, people like me, like economists, to try to sort that out and see which parts are important as early as possible and as fine grained as possible.

正因为不容易，我们才更应该努力去理清它。

Just because it's not easy doesn't mean we shouldn't try to sort it out.

对。

Yeah.

说实话，我对人工智能变得更强大后经济会如何发展，曾经的预期是错的。

I I mean, I confess that I have been wrong about the economic expectations I had for what would happen when AI systems got more powerful.

如果你能穿越回2020年，给我看看Claude Fable或者GPT 5.6，然后告诉我，这种技术会在2026年出现，我会怎么想呢？

I think if you had time traveled back to 2020 and showed me like Claude Fable or GPT 5.6, and then you asked me, like, this technology will exist in 2026.

当前的失业率是多少？

What is the headline unemployment rate?

我可能猜会是百分之十到二十。

I probably would have said like 10 or 20%.

但实际情况是，现在的就业率和几年前基本持平。

And what we see instead is that employment rates today are basically the same as they were a few years ago.

官方失业率一直维持在4%左右。

Headline unemployment is hovering around 4%.

你是不是也和我一样，对这项技术如此强大却没带来更多失业感到惊讶？

So are you as surprised as I am that this technology becoming so powerful has not created more unemployment?

不惊讶。

No.

一点都没想到会这样。

Not at all.

我写过很多论文，讨论生产力的J型曲线，以及AI如何与经济变革相互配合。

So I've I've written many papers on this idea of the productivity j curve and on the way that AI complements changes.

你可能听说过，我再讲一遍这个故事：当年电力刚出现的时候，足足花了二三十年，才彻底改变美国工厂的工作方式。

You may have heard I'll I'll tell I'll retell the story about how when electricity was first introduced, it took about twenty to thirty years before it fundamentally changed the way work was in American factories.

一开始，生产率根本没什么提升。

At first, there was little or no productivity change.

就业情况也没啥变化，因为人们只是在老路上修修补补。

There's little or no employment change because people were just kind of paving the cow paths.

他们还是做以前那些事，只不过换成了电动机，但本质一点没变。

They were redoing the same things they used to do, but now with electric motors and nothing fundamentally changed.

直到工厂彻底重新设计，最终整个经济结构也跟着重构，才真正释放出巨大的变革力量。

It took a lot of reinvention of those factories and ultimately the whole economy to unleash a big change.

而这种重构意味着数以百万计的工作岗位被取消。

And that reinvention meant eliminating millions of jobs.

同时也意味着数以百万计的新岗位被创造出来。

It meant creating millions of jobs.

就像我说的，电力带来的这一切，花了二三十年才完成。

All of that in the case of electricity, like I said, took twenty to thirty years.

我不觉得这次会花上二三十年，但也不可能在二三十个月内就完成。

I don't expect it to take twenty to thirty years this time around, but it's not the kind of thing that happens in in twenty to thirty months either.

这事儿得花上好几年才能慢慢显现出来。

It's something that plays out over over a period of years.

我觉得我们现在还处在变革的初期阶段。

I think we're in the early stages of that transformation.

我毫不怀疑，会有很多岗位被取代，大量工作消失。

I have no doubt there's going to be a lot of disruption and a lot of jobs destroyed.

同时也会涌现出大量新岗位。

There's also gonna be a lot of jobs created.

目前我们才刚刚起步，但接下来规模会越来越大。

We're just in the earliest stages of that right now, but it's going to get bigger.

你提到过斯坦福的金丝雀仪表盘，我一直很关注这个数据。

You mentioned the Stanford Canaries dashboard, which I've been following with great interest.

数据显示，初级岗位同比减少了2.7%，而中期岗位却增长了1.6%。

It showed that early career jobs shrank 2.7% year over year, while mid career jobs grew 1.6% year over year.

经济学家们是否越来越认同这是由人工智能带来的影响？

Is there growing consensus among economists that this is AI driven impact?

我们该如何应对这些数据？

And how should we respond to those numbers?

我不能代表整个经济学界发言。

I can't speak for the economics profession as a whole.

我是相信这些数据的，我花了很多时间研究它们。

I mean, I believe the numbers, I've I've spent a lot of time looking at them.

随着时间推移，这一点会越来越清楚。

I think over time, it'll become more and more obvious.

除了我之前提到的，我们测试了所有其他可能的解释，但都没能推翻人工智能这个结论，这可能是最有说服力的一点。

Maybe one of the most convincing things in addition to I mentioned we we tested all these other hypotheses and and they didn't knock out the AI result.

另外，我们最早在2025年8月就发布过这个版本，当时就看到了几乎一模一样的趋势。

But the other thing just to you know, we first published a version of this back in August 2025, and we saw, you know, basically the same pattern.

当时人们说，这不过是暂时的波动。

And people said, oh, it's a blip.

这跟利率什么的有关系。

It's due to, you know, interest rates or whatever.

这股趋势是慢慢积累起来的。

It's just grown over time.

你可以去看看加那利群岛的数据。

You can go to the canaries.

你可以去斯坦福的AI经济指标仪表盘看看。

You can go to the dashboard, AI economic indicators at Stanford.

你会看到，这些趋势一直在持续上升。

And what you'll see is that these trends have just been continuing to grow over time.

这可不是因为新冠疫情或者其他临时因素造成的。

It's not something that's just due to COVID or something else.

我再问一个问题，埃里克，关于这封信和这份声明。

I wanna ask one more question, Erik, about this letter, this this statement.

签署这封信的联盟阵容太强大了，汇集了众多经济学家、思想家和科技高管，堪称业界顶尖人物。

Obviously, the coalition that signed the letter, like, hugely impressive group of economists and thinkers, tech executives, a real who's who of the field.

但确实有些人不喜欢这封信，也没有签名。

But there were some people who didn't like this letter and and didn't sign it.

他们觉得这封信太模糊，或者没有提出足够具体的政策建议。

Either they thought it was too vague or it didn't make enough concrete policy recommendations.

我想特别提一下诺亚·史密斯的反应，他是一位经济学作家和博主。

But I wanted to single out one reaction from Noah Smith, who's an an economics writer and blogger.

他写道，自己没有签署这份声明，原因之一是他对‘引导人工智能朝着与人类互补、造福社会’这种说法持高度怀疑态度。

He wrote that he didn't sign this statement because, among other reasons, he was just very skeptical of this line about trying to steer AI in a direction that complements humans and benefits society.

他基本上说，这根本不可能做到。

He basically wrote, like, this is not possible.

我们根本不知道是否真能引导这项技术朝着与人类互补的方向发展。

We we don't know that we can steer this technology toward being complementary.

他问，企业家和发明家到底该怎么引导人工智能，让它更好地配合人类劳动者呢？

He writes, how the heck are business people and inventors supposed to steer AI toward being complementary to human workers?

他们连上一轮人工智能对劳动力市场的影响都预测不了，至少在短期内做不到。

They obviously couldn't predict the labor market effects of the last round of AI, at least in the short term.

那为什么有人会相信，技术专家有能力有意识地创造出不同劳动市场影响的AI形式呢？

So why should anyone believe the technologists have the ability to purposefully invent different forms of AI with different labor market effects.

你对这个怎么看？

What's your response to that?

我其实特别欣赏诺亚。

Well, I love Noah.

他非常聪明。

He's super smart.

他是我的朋友。

He's a friend of mine.

我们偶尔会在湾区碰面。

We hang out in the Bay Area from time to time.

我很高兴他认真回应了这封信，也很高兴不是所有人都盲目同意。

I'm so glad he he engaged with the letter, and I'm also glad that not everybody just reflexively agreed with it.

我觉得信里有那么几个人提出不同意见，反而让内容更有意思了。

I think we had just the right amount of in there, a few people disagreeing just to make it interesting.

如果大家都只是附和，那可就太无趣了。

It would have been boring if everybody just said, yeah.

对。

Yeah.

这很明显。

This is obvious.

不过，说不的人在这方面是错的。

That said, no is wrong about this.

你明白吧？

You know?

是啊。

Uh-huh.

历史上有很多例子，说明技术的发展方向是可以被引导的。

Have lots of examples in history of technology being directed in different ways.

碳税这个想法的背后，正是希望我们能做到这一点。

One of the ideas behind carbon taxes is that you do wanna do that.

而且有研究显示，那些征收更高税收的国家，其发动机技术的发展路径也有所不同。

And there's research that shows that, you know, countries that impose more taxes, engine technology evolved in different ways.

就计算机而言，长期以来一直存在两种不同的方向。

For computers specifically, there's long been two strains.

你知道，艾伦·图灵提出过图灵测试，就是让机器模仿甚至取代人类。

You know, Alan Turing talked about the Turing test, imitating, replacing humans.

还有一种思路，比如道格·恩格尔巴特等人主张，人工智能应该增强人的能力。

There's another stream, Doug Engelbart and others who talked about AI amplifying people.

我经常花很多时间跟企业首席执行官们聊天。

And one of the things I do, I spend a lot of time talking to CEOs.

让我给你讲个具体的例子。

Let me tell you a very concrete story.

我见过一位首席财务官，她说她想用人工智能来提升投资回报率。

I I met with the CFO, and she told me that she wanted to use AI and get higher ROI for it.

我说，这很好。

And I said, that's great.

这正是我们该做的。

That's what we should be doing.

然后她说，因此我们现在的做法是，要求每个部门统计因人工智能而减少的员工人数。

And then she said, and therefore, what we're doing is requiring measuring in every department how much headcount reduction they have, thanks to AI.

我一听，天哪。

And I said, woah.

等等，你先别急。

Hold it just a minute there.

把人工智能当作取代员工的工具来衡量，这只是其中一种方式。

That's one way to measure AI is replacing workers.

有时候，这确实也不错。

And, you know, and sometimes that's a good thing.

但你可以做的远不止这些。

But there's so many other things you can be doing.

你是首席财务官。

You're a CFO.

你拿的可是几百万的薪水。

You're being paid millions of dollars.

我其实没对她这么说，但心里就是这么想的。

I didn't actually say that to her, but that was what I was thinking.

你得更有创意一点。

You should be more creative.

咱们来看看能带来什么好处。

Like, let's look at the upside.

能推出哪些新产品或新服务？

What kinds of new products and services?

怎么提升客户服务的质量？

What kind of better customer service?

怎么降低员工流失率？

What kind of, you know, reduced employee turnover?

还有很多其他指标，一个称职的管理者都该关注，这些才能真正创造更多价值。

There are all these other metrics that a good manager can look at that will increase the upside.

如果他们真能这么想，改变心态，我告诉你，他们一定能主导这项技术的发展。

And if they think that way, if they change their mentality, believe me, they're gonna steer the technology.

到时候，结果会是让人更多地参与进来，技术更多地辅助人，而不是简单地取代人。

You're gonna have outcomes that involve people more, that complement people more rather than just replace them.

是的。

Yeah.

我对这一点没你那么乐观，稍微怀疑一点。

I I'm I'm just a little more cynical about that than than maybe you are.

我真希望你说得对，希望那些有远见的企业领袖会说：我们真的不希望技术取代人类，因为人太宝贵了，而且有很多方式能让技术帮助人。

I I wanna believe that you're right and that these enlightened corporate leaders will say, well, you know, we we really don't want the technology to replace humans because humans are so valuable, and and there's all these ways that they can be, you know, be complemented by a family

对我们说。

to at us.

橄榄园餐厅。

The Olive Garden.

对。

Yeah.

对。

Right.

但让我直说吧，做这件事有三个原因。

But, like Well, let let me just say this you know, there there's three reasons for doing this.

好的。

Okay.

第一，我觉得这对社会是对的，不过咱们先放一放这个理由。

One is that I do think it's the right thing for society, but let's just set that one aside for now.

第二，我认为从企业自身的长远利益出发，他们也应该这么做。我在商学院教企业战略时总会说。

The second reason is that I think it's in their enlightened self interest that they should that the more I'm I I could teach at business schools a corporate strategy.

更明智、更可持续的战略是开发新产品和服务、建立进入壁垒，而不是一味地削减成本。

The the the better strategy, the more defensible strategy is creating new products and services and barriers to entry, not just cutting costs.

第三点，确实如此。

And then the third thing I'll say is it's true.

现在有很多税收和其他激励政策，让企业更倾向于投资资本，而不是人力。

There are a lot of tax and other incentives right now that skew companies disproportionately towards investing in capital versus labor.

比如，资本的边际税率比劳动的边际税率要低。

For instance, marginal tax rates on capital lower than than than marginal tax rates on labor.

所以，如果你的策略是用机器取代大量员工，最终缴的税就会更少。

And so so if if you have a strategy that replaces a lot of workers with machines, you end up paying lower taxes.

我觉得，这是税收体系的一个漏洞。

That's the flaw, I think, in the tax system.

这个问题是可以解决的，我们可以创造更公平的激励机制，让管理者更均衡地对待人力投入。

That's something we can fix and create better incentives for work for for managers to cheat them more evenly.

最后一点，我也相信很多人会支持这一点，因为这确实对世界有利，而且员工也会更愿意在这样的公司工作——公司如果善待他们，他们自然会更认同。

And then last but not least, I do think there are a lot of people who will will embrace this because it is the right thing for the world, and and I think their workforce is gonna like working at that company if it's a company that does right by them.

你。

You

你曾对《纽约时报》的本·卡尔施说过，经济学家需要更可靠的数据。

told Ben Kalshi at the New York Times that economists need more reliable data.

我们现在用的数据有哪些问题？你希望看到哪些改进？

What are some of the problems with the data that we have right now, and what would you like to see more of?

说实话，这正是我目前最关注的问题。

I mean, this is something I'm focused on right now.

遗憾的是，就在经济史上最具动荡的时期，我们却在毁掉自己的仪表盘。

Sadly, as we're going to one of the most turbulent periods in all of economic history, we are destroying our instrument panel.

政府正在削减数据收集，削弱了不少统计机构。

The government is cutting back on data collection, eviscerating some of the statistical agencies.

我真的很理解那些还在坚持工作的人，现在恰恰是最需要加大投入、提升数据透明度的时候。

I feel for those people as they're trying to this is the time, if anything, we should be doubling and, you know, investing more in visibility.

这是一项回报率极高的投入。

It's a really high ROI activity.

花不了多少预算，却能帮助我们——无论是政策制定者还是企业管理者——做出更明智的决策。

It just costs a tiny bit of the budget and will help us make better decisions not only as policymakers, but also as managers.

话虽如此，联邦预算的事，我能做的实在有限。

That said, you know, there's only so much I can do about the federal budget.

但我能做的，是在数字经济发展实验室里，开发人工智能经济指标，跟私营机构合作。

What I can do is at the Digital Economy Lab create the AI economic indicators, partner with private organizations.

我们特别幸运，因为很多公司本来就在收集实时数据，比如RAMP、Revelio、ADP这些。

We're incredibly lucky that there's a bunch of real time data that's being just collected by companies anyways, you know, RAMP, Revelio, ADP.

所以我们能接入这些数据源，获得一些有价值的洞察。

And so we can tap into these data sources and get some kinds of visibility.

这些数据并不总是为了代表整体而收集的，因此我们需要花功夫去标准化和校准它们。

They're not always collected for the purpose of being representative, so you have to do some work to kind of normalize them and and measure them.

但只要建立起一套能提升洞察力的基础设施，我们就能做更多事。

But by creating an infrastructure to get more visibility, we can do more.

我们正在构建一套全新的经济任务分类体系。

And we're creating a new a new taxonomy of the kinds of tasks that are being done in the economy.

美国有个叫O*NET的系统，列出了劳动者在经济中做的各种工作，我们正在更新这个体系。

There's something called O Net that lists all the tasks that workers do in the American economy, and we're creating an updated version of that.

我觉得在未来几个月甚至几年里，我们能获得更实时、更精细的经济动态洞察，从而更好地感知并应对变化。

I think, you know, over the coming months and years, we'll be able to get much more real time visibility and a much more fine grained way of what's happening in the economy, and that will allow us to sense and respond to what's doing.

它当然没法完全预测未来会发生什么，但至少能快速衡量现状，让人及时做出反应。

I mean, it can only do so much about predicting what's gonna happen next, but it's sometimes easier to at least measure it quickly and then allow people to respond to that.

埃里克，我想听听你对最近几周热议的一个政策想法的看法，就是由人工智能公司出资设立一个主权财富基金。

Erik, I wanna get your take on a policy idea that's been making a lot of waves in recent weeks, which is this idea of a some sort of a sovereign wealth fund that would be sort of paid for by the AI companies.

换句话说，每个美国人都能间接拥有前沿AI公司的股份，通过这种方式把部分收益分给美国民众。

Like, every American would get a stake effectively in the frontier AI companies, and you would kind of use that to capture some of the value and redistribute it to American citizens.

从经济学家的角度来看，你支持这个想法吗？还是说，有其他更好的方式能让美国人分享这个蓬勃行业的红利？

From an economist perspective, do you like that idea, or is there some other way that we should be getting Americans some share of the proceeds from this booming industry?

这并不是我最优先的选择，但我认为我们需要采取多种措施并行推进。

It's not my first choice, but I think we need sort of an all of the above solution.

我希望我们能更多地思考如何实现这种收益共享的机制。

So I'd like to see us thinking more about ways that we implement those kinds of of ways of of sharing the gains.

我担心的是，如果只靠主权财富基金直接发钱，那谁来决定谁该拿、谁不该拿？

My concerns with with just a a payout in the sovereign wealth fund is that, like, who decides who gets the money and who doesn't?

如果你不同意这个分配方式，你又有什么办法去影响它呢？

And and if you don't agree with it, what what leverage do you have?

如果你不给农民报酬，他们就不会给你粮食；不给工厂工人发工资，他们也不会付出应有的劳动。

If you don't pay a farmer, they're not gonna give you the food or you don't pay a factory worker, they're not gonna give you the the the the effort that they put into it.

但如果你只是领救济金，你就没什么办法去施加影响了。

But they if you're just getting a dole, you don't really have any leverage to to step back.

我的意思是，你或许可以指望通过民选代表来表达意见，但说实话，单纯靠发钱这种方式，社会平衡会比较脆弱。

I mean, you hope you can work through your representatives, but, yeah, I I think it's it's a less stable equilibrium to just have people getting a check that way.

不过，就像我一开始说的，我们还是应该探索多种不同的方案。

That said, as I said at the beginning, we should, you know, explore a variety of approaches.

对。

Yeah.

我理解的新救济机制是，如果你对它有意见，就去给大卫·萨克斯发条推文。

My understanding about how, like, the new Dole would work is that if you had some sort of issue with it, you would tweet at David Sachs.

只要你的推文获得足够多的转发，他就会回复你。

And if your tweet got enough retweets, then he would respond to you.

所以，

So I

这倒也不是不行。

mean That could work.

是的。

Yeah.

我们大概能设计出一个更好的系统。

We could probably come up with a better system.

我来提几个目前在这个领域里被讨论的其他想法，埃里克。

Let me throw a couple other ideas up on the board that are are getting kicked around in in this space, Erik.

你觉得工资保险怎么样？比如，如果你因为失业不得不找一份收入更低的工作。

What do you think about, like, wage insurance if, you know, you have to take another job because your job goes away, and and you're you have to you're forced to work a lower paying one?

对。

Yeah.

工资保险确实是另一个方向。

Wage insurance is is another thing.

几十年来，经济学家一直在讨论这个，它有更大的潜力。

Economists have been talking about this for for decades that there's more potential.

这就像一个缺失的市场，就像有房屋保险、火灾保险和健康保险一样。

It's kind of a missing market there just like you have housing insurance and fire insurance and health insurance.

这有点棘手，因为你提到的道德风险问题——我们还是得让人有动力去工作。

It's tricky because of moral hazard as you call it that like, okay, you you still wanna have incentives for people to work.

但我们能用更好的工具来分析，哪些工作是受外部因素影响，哪些不是。

But we could have better tools to understand which jobs are being affected and how much of that is due to exogenous forces and how many aren't.

我很支持再深入探索一下这个方向。

I'm I'm all for exploring that a little bit further.

再说一遍，我觉得这解决不了根本问题，但可能是拼图中的一块。

Again, I I don't think it's gonna solve the problem, but it could be one piece of the puzzle.

那政府要不要给雇佣年轻人的雇主提供激励呢？

What about government incentives to hire young workers?

我觉得这部分其实也有道理。

I think some of that actually makes sense as well.

原因在于，如果没有年轻人先入行，就不会有后来的中年员工。

And the reason for that is that you don't get mid career workers unless they first start as young workers.

所以这里存在一个市场失灵，需要有人来介入。

So there's kind of a market failure there that needs to be stepped into.

历史上，教育和培训在某种程度上属于公共产品，其价值超越了直接付费的人群。

Historically, education training have been, to some extent, public goods where there's value that goes beyond the people who are paying for them.

所以几乎每个国家，政府和非营利组织都参与其中，努力提供支持。

And that's why in almost every country, government and nonprofits have had a role in trying to support them.

几个世纪以来，美国一直全球领先，在教育和培训方面投入最多。

The United States has been, for many for many centuries, the number one in the world at making investments in education training.

从19世纪开始，我们在基础教育、中等教育，再到高等教育上的投入，都远远领先于欧洲和其他地区。

For, you know, the eighteen hundreds, we were way ahead of Europe or other country other places in the world at investing in primary education, then secondary education, then college education.

现在情况不一样了，但当年我们这么做的时候，正是美国生产力最高、中产阶级规模最大的重要原因。

It's not true anymore, but but when we did, that was a big reason why The United States not only had the highest productivity, but also one of the biggest middle classes in the world.

埃里克，我知道时间差不多了，但我还想问最后一个问题是。

Erik, I know we're almost out of time, but I wanted to ask one final question.

埃里克，你的领域——经济学，正经历一些根本性的变化。

Erik, your field economics has been going through some pretty radical changes.

我注意到最明显的一点是，很多学术界的同行都跑去AI实验室工作了。

And the one that I see the most is just that a lot of your colleagues in academia are going to take jobs at AI labs.

好像每周都能在网上看到，有位德高望重的经济学家辞掉大学的终身教职，去一家前沿AI公司上班。

It seems like every week, I'm I read on the Internet that some esteemed economist is taking a leave or leaving their tenured position at some university to go work for a frontier AI company.

你担心这样会掏空学术界经济学系的队伍吗？

Are you worried that that is going to hollow out the ranks of academic economics departments?

要是有前沿AI公司想挖你，你开价多少？

And what is your poaching price for any AI frontier lab that is interested in hiring you?

我到目前为止还撑着没走，但确实看到我的同事们——隔壁办公室的朋友、学生，好多人已经去了前沿实验室。

I've been resisting so far, but I do see like my colleagues, yeah, in the offices next door, my friends, my students, a lot of them have gone to take jobs at Frontier Labs.

说实话，这些前沿岗位确实做得不错，能接触到顶尖的数据，这真的很吸引人。

And I gotta say, by and large, you know, these frontier jobs, they're doing great work and they get access to amazing data, And so that's exciting.

但我觉得，还是需要一些人，他们没有老板天天盯着要为公司赚钱。

At the same time, I think there's a role for people that, you know, don't have a boss who who, you know, is looking to make profits for the Frontier Labs.

就算他们实际上很客观，外界也会觉得他们是在替公司说话，所以我们需要一些不太受这种影响的人。

And, you know, even if the reality is that they're unbiased, there's gonna be a perception that they're speaking for that company, and we need people who who aren't as affected by that.

我也没说学术界就完全不受影响。

I'm not saying academia is completely unaffected either.

所以我还是希望有人能留下来。

So, you know, I do hope people will will stay.

很多聪明人都会留在学术界，继续做我这样的工作。

A good a lot of smart people will stay in academia and continue to do what I'm doing.

我可是下定决心不走了。

I've sat down very hard.

我确实收到过前沿实验室的邀请，也认真考虑过。

I've I've had offers from Frontier Labs, and and I've thought about it.

对我来说，再多的钱带来的额外价值并不大，所以我现在这份工作挺合我心意的。

And for me, the the marginal utility of more money isn't that high, so I I just sort of like what I'm doing right now.

斯坦福对我一直很好。

Stanford's been great to me.

我喜欢能有个地方，自由地做我想做的研究。

And I like just being able to have a perch where I can I can do the research that I'm doing?

所以

So

埃里克，我听你的意思，除非是像Meta超级智能实验室那种级别的邀约，比如十亿美元的报价，才可能让你离开现在的位子去Frontier工作。

What I what I'm hearing you say, Erik, is that it would take some kind of, like, Meta Super Intelligence Lab style offer, Like, maybe a billion dollars to get you out of your chair and and go work I for Frontier

我不确定。

don't know.

也许是我自己有问题，但我真不知道，去了能干嘛呢？

I I maybe there's something wrong with me, but I I feel like, like, what am I gonna do?

买个更大的房子？

Get a bigger house?

多买几辆车？

More cars?

我真的不知道。

I don't know.

我的意思是，你懂的？

I mean, I I you know?

所以我觉得，我这份工作简直是世上最好的。

So so I kind of I feel like I have the greatest job.

你知道吗，别告诉我院长，也别去查我的工资，但我真的觉得，我现在做的事让我特别满足。

You know, don't tell my dean or look up my salary, but but I feel like I'm I'm, like, basically very happy with what I'm doing.

不管挣多少钱，我都会做现在这份工作。

Whatever amount of money I had, I would be doing basically what I'm doing right now.

这一点上，我挺幸运的。

So I'm lucky in that.

不是每个人都能有这么一份工作。

Not everybody's lucky enough to have a job like that.

也许你们就有。

Maybe you guys do.

你们看起来玩得很开心。

You guys look like you're having fun.

希望将来能有越来越多的人，像我一样，做自己真正热爱的事，而不是为了钱才选工作。

And, hopefully, we'll be entering a world in the future where more and more people can just be doing what they really enjoy doing like I do, and and money won't be the main reason that they pick one job versus the other.

那多好啊。

Oh, wouldn't that be nice?

我喜欢这个。

I like it.

我们真的非常感谢你所做的工作。

Well well, we really do appreciate the work that you do.

能有这样扎实、独立的研究和数据作为我们工作的依据，真的特别好。

It's great to have, you know, strong independent research and data about this that we can, you know, rely on as we do our work.

非常感谢你，埃里克。

So thank you very much, Erik.

谢谢你，埃里克。

Thanks, Erik.

谢谢。

Thank you.

和你们聊天真是太开心了。

Such fun talking to you guys.

回来之后，我们会聊聊布朗大学的作弊事件，还有本周《HatGPT》的其他故事。

When we come back, there's cheating at Brown University and other stories from this week's installment of HatGPT.

我是《纽约时报》的记者丹妮·布鲁姆。

I'm Dani Blum, a journalist at The New York Times.

我大部分时间都在报道GLP-1类药物。

I spend a lot of my time reporting on GLP-one medications.

人们管它们叫减肥药。

People call them weight loss drugs.

但它们在体内发挥的作用远不止于此，很多人对它们寄予厚望。

They are doing so much more in the body, and there's a lot of hope pinned on them.

我几乎每天上班都在跟进最新的科研进展。

I am spending part of truly every single workday making sure that I am up to date on the latest science.

我特别希望接触到顶尖的研究人员，关注临床试验结果，也和正在使用这些药物的人交流，把那些复杂、模糊、不断变化的科学信息，转化成大家能听懂的内容。

I really wanna make sure that I am talking to the leading researchers, keeping an eye on trial results, and that I'm talking to people who are on these medications to learn more from them, to translate the really messy and murky and shifting science into a way that people can understand.

对于人们关心的问题，尤其是关于他们正在服用的药物，提供有科学依据的答案非常重要。

It's really important to have evidence based answers to some of the questions that people have, especially about something as vital as a medication they're taking.

这就是订阅《纽约时报》能获得的内容。

That's what you get when you subscribe to The New York Times.

如果你已经订阅了，谢谢。

If you already subscribe, thanks.

如果想订阅，请访问 mytimes.com/subscribe。

If you'd like to, go to mytimes.com/subscribe.

好了，凯西，这周新闻太多，我们没法一一深入聊，但可以来一轮‘帽子GPT’。

Well, Casey, so much news this week that we can't talk about it all in-depth, but we can do a round of HatGPT.

没错。

That's right.

‘帽子GPT’是我们的一个环节，把本周的新闻全扔进帽子里，然后一个个抽出来，边聊边讨论，直到有一个人觉得烦了，说‘别再生成了’。

HatGPT, of course, our segment where we toss the week's news into a hat, pull out items one by one, and discuss them until one

我们当中有个人觉得无聊了，就说‘停吧’。

of us gets bored and says, stop generating.

好。

Okay.

凯西，这周帽子里装了不少东西。

Casey, we got a bunch of stuff in the hat this week.

你先来。

You go first.

好的。

Alright.

我们看看，凯文。

Let's see, Kevin.

这条新闻来自《纽约时报》和我们的朋友迈克·艾萨克。

This first story comes to us from the times and our friend, Mike Isaac.

OpenAI 在 ChatGPT 里展示了 Kalshi 的世界杯赔率，凯文。

OpenAI is showing Kalshi's World Cup odds in ChatGPT, Kevin.

作为两家公司合作的一部分，当用户询问各种世界杯比赛时，ChatGPT 会自动显示 Kalshi 的胜率预测。

As part of a deal between the two companies, ChatGPT started displaying the Kalshi odds of victory in response to user queries about various World Cup matches.

比如在法国对西班牙那场比赛中，页面上显示预测市场给法国队的胜率是 60%，这信息其实挺关键的，毕竟后来法国队可是 0 比 2 输了。

For example, in the case of France and Spain, there was a graphic showing that the prediction market had assigned the French a 60% chance of winning, which was a very important information to have before France would go on to lose two zero.

所以我觉得这个

So I think this is

这真是个绝妙的故事，因为现在ChatGPT里有了能让你亏钱的东西，而且还不用靠你每月付给OpenAI的订阅费。

a perfect story because now there is something in ChatGPT that can help you lose money in a way that does not involve your monthly subscription payment to OpenAI.

太妙了。

It's beautiful.

我正琢磨一个理论，其实跟这个故事直接相关：当今世界最重要的两种力量，一是AI能力，二是赌博成瘾。

I am I am circling a theory that actually has to do with this particular story, which is that the two most important forces operating in the world today are one, AI capabilities and two, gambling addiction.

对。

Yes.

这两者结合在一起，就像拼成了一个社会版的沃顿机器人，恐怕很快就会势不可挡。

This combines them both in a into a sort of like societal Voltron that is probably going to become unstoppable.

没错。

Yeah.

这个特别有意思，因为预测市场本来就跟理性主义者关系密切，而他们正是推动当前AI浪潮的核心力量。

And this one is a really funny one because, of course, like, prediction markets are so closely associated with the rationalists who were a driving force in, like, bringing about this current AI moment.

你懂的。

You know?

所以这看起来就像是ChatGPT里一个深深带着它诞生地文化烙印的功能。

So it sort of looks like, you know, maybe a feature in ChatGPT that is, like, steeped in the cultural traditions of the place that birthed it.

但另一方面，这纯粹就是赌博。

But on the other hand, it is just gambling.

作为ChatGPT的付费用户，我实在受不了在使用时看到这种东西。

And I, a, you know, like, a paying subscriber to ChatGPT, it it does offend me that

我现在在回复里都要看到赌博广告了。

I have to see gambling advertisements in my responses now.

对啊。

Yeah.

除非有监管介入，不然这种事迟早无处不在。

I I just think this is going to be sort of everywhere unless there's some sort of regulatory response.

我甚至觉得，过一两年，打开一盒麦片，它都可能给我推送个Kalshi的赔率。

Like, I expect to, like, open a box of cereal, and it should just give me some Kalshi odds for something Yeah.

真的，用不了多久就会这样。

Like, in a year or two.

我觉得，以后任何我们注意力能触及的地方，都会被赌博广告占据。

I just think there's going to be no surface for our attention that does not in some way have a gambling ad attached to it.

好吧。

Alright.

别再生成了。

Stop generating.

明白。

Okay.

下一个抽出来的是。

Next out of the hat.

纽约州出台了全美首个针对数据中心的 statewide 暂停令。

New York enacts nation's first statewide moratorium on data centers.

州长凯西·霍楚尔——这位前《硬分叉》节目的嘉宾——最近签署了一项行政命令，暂停大型数据中心的建设一年，以便评估它们对环境和能源消耗的影响。

Governor Kathy Hochul, former hard fork guest, signed an executive order recently pausing the construction of the largest data centers for one year while the state assesses their effects on the environment and energy use.

这项暂停措施针对的是耗电50兆瓦以上的超大规模数据中心。

This is a temporary pause on hyperscale data centers, which use 50 or more megawatts of power to operate.

特朗普总统在真实社交平台上对此作出回应，称这是一个糟糕的决定，并表示绝不能让激进派民主党人把数据中心、人工智能以及所有这些令人惊叹的新科技拱手让给中国和其他国家。

And president Trump responded to this on True Social, calling it a terrible decision and saying, the radical left Democrats must not be allowed to cause us to lose data centers, AI, and all this incredible new technology to China and other countries.

凯西，你怎么看纽约州这项全州范围的暂停令？

Casey, what do you make of the statewide moratorium in New York?

说实话，我对这件事挺同情的。

So I I I I have to say I'm quite sympathetic here.

州议会之前通过了一项暂停令，其限制力度比现在这项还要大。

The state legislature had passed a moratorium that would have been even more restrictive than this.

所以我的理解是，州长想走一条中间路线，给那些担心数据中心建在自家社区的人一点安抚，但又不像立法机构里某些人那样做得那么绝。

And so my belief here is that the governor wanted to find a kind of middle path and throw a sop to people who are worried about data centers in their communities without sort of going as far as some of the the legislature did.

这算是

It was sort

一个‘缓建令’吧。

of a a lessatorium.

对。

Yes.

确实如此。

It really was.

很好。

Very good.

谢谢。

Thank you.

所以，我觉得有理由说，这些数据中心的影响确实存在很多疑问，我们确实需要好好调查一下。

And so, look, I think it is reasonable to say that there are a lot of questions about the impact of these data centers, and we want to investigate it.

我的希望是，明年他们能像这里说的那样，搞清楚这些数据中心到底对普通用户的电费有多大影响。

Now, you know, my hope is that during the next year, they do what they're saying here, and they get a handle on what is the actual impact to, like, the average rate payer.

对吧？

Right?

我们能不能真正解决那些摆在明处的环境担忧？

And can we address some of these environmental concerns that are out there meaningfully?

以前我们在节目里聊过，也许一些关于用水量的说法被夸大了，我确实这么认为。

You know, we we've talked on the show in the past about, well, maybe some of the, you know, water use claims are are are overstated here, and I do believe that.

但我看到其他一些说法，说数据中心会导致城市热岛效应，或者污染水源之类的。

But I'm reading other stuff about data centers contributing to, like, an urban heat island effect, for example, or like a like like, sort of polluting the water supply.

所以这些正是我觉得值得认真探讨的问题。

So these are just the sorts of things that I think it's reasonable to say.

在我们州再建十个这样的设施之前，我们得先把这些问题搞清楚。

We want to look into this before we have another 10 of these things go up in the state.

我理解您希望获得更多信息的初衷。

What do you I am sympathetic to wanting to have more information.

我只是希望送到立法机构和州长办公室的信息，是准确、高质量的。

I just hope that the information that makes it to the legislature and to the governor's office will be good, high quality information.

之前关于数据中心，尤其是用水和环境影响的很多说法，质量都很差。

I think there's been a lot of, like, shoddy information out there about the especially the environmental effects of data centers, the water use.

很多相关数据在我看来都靠不住。

Like, a lot of that data seems pretty suspect to me.

所以我只希望，能有一场扎实、严谨的调查，弄清楚这些设施到底造成了什么实际影响。

So I just hope that there is, like, a good robust inquiry into, like, what is the actual effect on these?

我认为短期影响是，如果你是一家打算在纽约建数据中心的科技公司，其实你根本就不会考虑在纽约建，因为大家都知道，大型AI公司和超大规模云服务商一般都会选择得克萨斯州或其他更支持这类项目的红州。

And I think the short term effect is that if you were a tech company that was planning to build a data center in New York first of all, you probably weren't planning to build a data center in New York because I think that's this you know, it's generally known among the big AI companies and the hyperscalers that you're gonna build something, you wanna build it in either Texas or some other red state that is likely to look more favorably on your project.

但我认为，这种情况还会继续下去——这些数据中心还是会主要建在红州，那里民主党人少，反对的声音也弱。

But I think that is going to continue to happen where these data centers will open in largely red states where, you know, there are fewer Democrats to oppose them.

可惜的是，这个问题正变得越来越两极分化。

I think this is unfortunately becoming quite a polarized issue.

最终会怎样，还有待观察：那些欢迎数据中心的州，到底能从中获得多少税收，真正改善居民生活？还是说，这些设施最终只会成为负担，让当地居民后悔当初放行？

And it will remain to be seen whether those states that want the data centers end up getting a bunch of tax revenue from that that makes their citizens' lives better or whether it is just a blight and the they will regret having let these things in.

是的。

Yeah.

我的意思是，我不太确定。

I mean, I I don't know.

我觉得这个问题可能没有有些人想的那么对立。

I I think that this this issue is maybe a little bit less polarized than some people might think.

就在不久前，犹他州——一个典型的红州——也爆发了一场大规模的数据中心争议。

There was just a massive data center fight in Utah, a very red state.

你知道吗，有些市议员就因为支持建数据中心，结果丢了职位。

You know, we've seen, you know, members of city councils, like, lose their positions over endorsing data centers.

所以这事儿不是简单的民主党讨厌、共和党喜欢。

So this is not just a pure, like, Democrats hate data centers and Republicans love them.

共和党也不见得就喜欢数据中心。

Republicans do not love data centers either.

所以我觉得，想建数据中心的公司得拿出真正有说服力的方案，得让人相信这事儿不会推高电费，也不会对环境造成灾难。

And so I think the message to companies that want to build them is, like, you have to make a really solid pitch, and that includes convincing people that this thing is not gonna drive up their electricity prices and that it's not gonna be horrible for the environment.

你知道犹他州临时禁止新建数据中心叫什么吗？

Do you know what they call it when they put a temporary ban on data center construction in Utah?

叫什么？

What's that?

叫‘摩门冻结期’。

A Mormontorium.

别再建了。

Stop generating.

这个笑话是说给惠特尼听的。

That joke is for Whitney.

给大家打个招呼。

Shout out.

好的。

Okay.

凯西，该你了。

Casey, you're up.

行吧。

Alright.

这是一个

It's a

这儿有个重头戏。

thick one here.

凯夫一直想以布朗大学毕业生的身份问问你这件事。

Kev had been wanting to ask you about this as a Brown University graduate.

这个故事来自高等教育圈内。

So this one comes to us from inside higher ed.

布朗大学的经济学教授罗伯托·塞拉诺今年春天给学生布置了一次开卷期中考试，结果全班平均分高达96%，远超正常水平。

A Brown University economics professor, Roberto Serrano, had given his students a take home midterm exam in the spring, and students scored a whopping 96% on that midterm, which is way higher than normal.

他说，平时的平均分一般在65到80分之间。

He said the average grade is usually between 6580%.

到了期末考试的时候，他改成了现场闭卷考试。

And so then it came time to do the final exam, which he gave in person.

凯文，期末考试的平均分只有48.6%，是这门课历史上最低的分数。

And the final exam score, Kevin, was 48.6%, an all time low for the course.

他公布的数据显示了很多有趣的现象，但我最喜欢的是，有个学生期中考了55分，期末反而提高到了59分。

There were many funny things in the data that he released, but my very favorite piece of data was that there was one student who scored 55% on the midterm and improved to 59 on the final.

这位绝对是布朗大学最诚实的学生。

This is the most honest person at Brown University.

如果你在听，我们向你致敬。

And if you're listening, we salute you.

我们向你致敬。

We salute you.

是的。

Yeah.

这个故事太精彩了，我特别感兴趣，原因嘛，你懂的。

This was an amazing story that I was very interested in for, you know, obvious reasons.

我当年在布朗大学上学。

I I attended Brown University.

我其实选过罗伯托·塞拉诺教授的经济学课，待了几个星期，后来因为太难就退了。

I actually spent a couple weeks in professor Roberto Serrano's econ class before I dropped it because it was too hard.

我敢打赌，我当年肯定也是那种在期中考试时靠ChatGPT或者其他AI工具作弊，然后期末考试直接崩盘的学生之一，没错。

And I'm sure that I would have been one of these students who has sort of, you know, cheated on the midterm using ChatGPT or some other AI system and then Yeah.

期末考考得一塌糊涂。

Absolutely bombed the final.

值得一提的是，他之所以给期中考试安排成开卷，其实特别体谅学生。

It's worth saying, like, the reason that he gave the take home midterm was particularly generous.

就在那之前，布朗大学校园里刚发生了一起校园枪击事件。

Like, they had just had a school shooting at on the campus of Brown.

太可怕了。

Terrible.

所以他才说，鉴于这件事，你们就回家做期中考试吧。

So he says, you know, in the wake of that, like, do your take home exam.

没有时间限制。

No time limit.

我觉得这位教授这么做真是非常体贴、充满善意。

Like, I think a very generous and and compassionate move by a professor.

可学生们是怎么回报他的呢？竟然是集体作弊。

And how do his students repay him but by cheating en masse?

要说所有人期中考试平均分高达96%，但现场期末考试却只有48.6%，这根本不可能有别的解释。

I mean, there is no other explanation for how everyone would score an average of 96% on the midterm and then 48.6% on the in person final.

那篇报道附带的分数分布图简直让人笑出声。

The score distribution chart that was made to accompany this article was hilarious.

确实如此。

Truly.

当时就是。

It was

简直让人觉得，要么大家在期中和期末之间都得了严重的脑损伤，要么就是全用AI作弊了。

it was just like, oh, everyone has had either had severe brain damage in the time between the midterm and the final or everyone was cheating using AI.

是的。

Yeah.

我的体会是，作弊不好，千万别作弊。

My takeaway here is that cheating is bad and you shouldn't cheat.

不过呢，这一点也确实如此。

So Well, there's that too.

好吧。

Alright.

别再生成了。

Stop generating.