A Chinese lab recently unveiled another astonishingly powerful and astonishingly cheap artificial intelligence model. GLM-5.2, produced by Z.ai, is almost as good as Anthropic’s latest model but runs at less than a tenth of its price.

中国的一家实验室最近推出了一款有着惊人的强大性能和低廉价格的人工智能模型。由智谱AI开发的GLM-5.2在性能上几乎可以媲美Anthropic的最新模型，但运行成本却不到其十分之一。

The A.I. race between the U.S. and China has been front of mind at the highest levels of government in both countries.

美中之间的人工智能竞赛一直受到两国政府最高层的高度关注。

But creating the most potent model is not the only — and maybe not even the best — way to measure A.I. success. Today I write about an area that’s had less attention, but where China might well have an edge: managing the human and political fallout from the A.I. revolution.

然而，开发出最强大的模型并不是衡量人工智能成功的唯一方式，甚至可能不是最好的方式。今天，我想谈谈一个较少受到关注、但中国很可能占据优势的领域：管理人工智能革命所带来的社会与政治影响。

Robotaxis and A.I. Marxism

无人驾驶出租车与AI马克思主义

If there is one thing China’s Communist Party fears, it’s a restive proletariat.

要说有什么是中国共产党格外害怕的，那就是躁动不安的无产阶级了。

In Wuhan, the world’s largest open-air laboratory for driverless cars, taxi drivers first complained about the growing fleet of robotaxis two years ago. Petitions were filed. Social media posts were hashtagged. The local outcry was noisy.

在全球最大的无人驾驶汽车露天试验场武汉，出租车司机们在两年前首次抱怨无人驾驶出租车车队规模的日益扩大。当时人们递交了请愿书，在社交媒体上建立了标签话题，当地的反对声音不绝于耳。

It focused minds in the party, which reacted swiftly to censor the protests online. But it also set off a bigger rethink of something that worries many in the West, too. How to avoid mass displacement of humans by A.I. in the labor market — and the political backlash that comes with it?

这引起了相关部门的高度重视，官方迅速采取措施平息了网络上的抗议。但这也引发了对一个同样令西方许多人感到担忧的大问题的重新思考：如何在劳动力市场中避免人工智能对人类的大规模替代，以及随之而来的社会反弹？

China has more experience than most countries with automating jobs. More than two million robots work in its factories. Driverless delivery vans roam many of its cities. Service robots attend to guests in hotels and restaurants. Parking-lot robots swap out dying E.V. batteries. Drones deliver lunch.

中国在工作自动化方面比大多数国家拥有更丰富的经验。其工厂里有超过200万台工业机器人正在工作。无人驾驶快递车在许多城市的街道上穿梭。服务机器人在酒店和餐厅里招待顾客。停车场里的机器人负责为电动汽车更换电池。无人机则忙着配送外卖。

So far the fallout has been mostly among blue-collar workers. But A.I. is primarily threatening college graduates. And for an authoritarian regime that fears political instability, that’s a group that has historically caused trouble.

迄今为止，这种影响主要集中在蓝领工人身上。但人工智能主要威胁到的却是大学毕业生。对于一个极其重视社会稳定的管理体制而言，高学历群体在历史上往往更需要稳妥的引导。

Which is why China’s goal to become the world’s A.I. superpower is now officially twinned with another one: keeping humans at the core of the A.I. economy. And over the past year, the government has started acting more decisively to make that happen.

这正是为什么中国将自己打造为全球人工智能超级大国的目标已经正式和另一个目标绑定在一起，那就是让普通人始终保持在人工智能经济的核心位置。而在过去一年里，政府已经开始采取更果断的行动来落实这一目标。

‘Liberating labor,’ one court case at a time

通过一个个法庭判例来“解放劳动力”

When the Communist Party wants to show it really means something, it puts it in a five-year plan. And on Page 72 of its current five-year plan, China commits to “comprehensively address” the impact of A.I. on employment.

当官方想要表达其对某件事动了真格时，通常会将其写入五年规划。在当前的五年规划中，中国承诺要“积极应对”人工智能对就业带来的冲击。

I spoke to Kyle Chan at the Brookings Institution, who studies China’s A.I. policy. He told me that China wants A.I. to augment humans — that is, make them more productive across old and new industries — not replace them. And in the process of transitioning to this A.I. economy, it says, it wants to cushion the impact to avoid social fallout.

我采访了布鲁金斯学会研究中国人工智能政策的专家陈凯欣(Kyle Chan)。他告诉我，中国希望人工智能能够增强人类的能力——也就是说，提高人类在传统和新兴行业中的生产力，而不是取代人类。他表示，在向这种人工智能经济过渡的过程中，官方希望能够缓冲其带来的冲击，以避免引发社会问题。

My colleague Catie Edmondson recently wrote about what this looks like. The Ministry of Human Resources and Social Security is promising “targeted employment support for key industries.” One member of the National People’s Congress is calling for an “A.I.-unemployment insurance program” as a safety net for displaced workers. Party officials have pushed for vocational training to help workers adapt to an A.I.-centric job market.

我的同事凯蒂·埃德蒙森(Catie Edmondson)最近撰文描述了这一现状。人力资源和社会保障部承诺将为“重点行业提供定向就业支持”。一位全国人大代表则呼吁建立一项“AI失业保险”制度，以此作为被转岗工人的安全网。官方也一直在推动职业培训，以帮助劳动者适应以人工智能为中心的工作市场。

Chinese scholars have even been developing a field they call A.I. Marxism — trying to apply a Marxist lens to questions like “Who or what creates value after the A.I. revolution?” (The machine? The human who invented it? The human who operates it?)

中国学者甚至一直在拓展一个他们称为“AI马克思主义”的领域——试图用马克思主义的视角来剖析诸如“在人工智能革命之后，究竟是谁或什么创造了价值？”（是机器？是发明机器的人？还是操作机器的人？）等问题。

Perhaps most strikingly, the government is leaning heavily on companies to avoid layoffs. And those who don’t fall in line might find themselves in court.

或许最引人注目的是，政府正在给企业施加巨大的压力以避免裁员。而那些未能严格执行的企业可能会发现自己被告上法庭。

There have already been several high-profile rulings siding with workers who were dismissed. In April, a court ruled that a tech company had illegally laid off a worker after replacing him with A.I. software. The ruling delivered an implicit warning to other employers.

目前已经出现了几起偏向被辞退员工的受热议判决。今年4月，一家法院裁定，某科技公司在用人工智能软件取代一名员工后将其解雇属于违法行为。这一判决向其他雇主发出了含蓄的警告。

“The development of artificial intelligence technology should be applied to liberating labor, promoting employment and improving people’s livelihood,” the Hangzhou Intermediate People’s Court wrote. “Labor law allows employers to undertake technological changes and upgrade their operations, but it should also take into account the protection of workers’ legitimate rights and interests.”

杭州市中级人民法院在一份判词中写道：“人工智能技术的发展本应用于解放劳动、促进就业、造福民生，劳动法允许用人单位承接技术变革进行更新转型，但亦应顾及保障劳动者的合法权益。”

Just how that will work in practice, and how far the government is actually prepared to go with companies that don’t comply, remains to be seen. But what these rulings underscore is how much China is thinking about the problem.

这种做法在实践中将如何运作，以及政府对不合规企业究竟准备采取多么严厉的措施，目前还有待观察。但这些判决凸显出中国在思考这一问题上投入了多少精力。

Two different A.I. visions

两种迥异的人工智能愿景

The U.S. is letting tech companies take the lead on A.I., and Silicon Valley is focused primarily on one thing: achieving superintelligent machines capable of displacing humans. It’s an approach the Trump administration seems to be broadly on board with, or at least isn’t willing to stop.

美国正让科技公司在人工智能领域占据主导地位，而硅谷主要专注于一件事：实现能够替代人类的超级智能机器。这是一种特朗普政府似乎广泛赞同、或者至少不打算加以阻止的发展模式。

China’s approach is different. China is imagining what it wants its economy and society to look like and how A.I. can help achieve that, Chan, the Brookings expert, says. It wants a self-reliant economy, and so it’s embedding A.I. in every industry — from flashy new businesses like robotics to uncool old industries like steel or cement — to turbocharge productivity so it can never be strategically vulnerable again.

布鲁金斯学会专家陈凯欣指出，中国的做法则不同。中国正在构想其期望的经济和社会版图，以及人工智能如何能协助实现这一目标。中国追求经济的自给自足，因此正将人工智能植入到每一个行业中——从机器人等时髦的新兴产业到钢铁或水泥等传统的老旧工业——以极大地提升生产力，从而确保自身在战略上不再面临脆弱性。

It also wants stability and so, even as it does so, it’s thinking about how to keep humans employed.

与此同时，中国追求稳定，因此即便在推进自动化的过程中，它也在思考如何保障人类的就业。

China’s vision for A.I. is state-driven, aimed at achieving government goals. America’s is company-led — companies like OpenAI are pursuing superintelligence because it fits with their own interests, not because of a broader U.S. strategy, for now.

中国对人工智能的愿景是由国家驱动的，旨在实现政府的目标。美国的愿景则是公司主导的——目前来看，像OpenAI这样的公司追求超级智能是为了符合自身的利益，而不是出于美国的某项宏观国家战略。

The lesson from China isn’t that countries should follow its specific approach to A.I. and jobs, Chan says — Chinese-style control over the tech industry isn’t feasible in most Western countries, for example. But China shows that policymakers have agency over this technology. They can influence its direction, rather than just letting the (A.I.) chips fall where they may.

陈凯欣表示，中国的经验并不是说其他国家应该完全照搬其对待人工智能和就业的具体做法——例如，大多数西方国家根本无法可行地复制中国对科技行业的控制方式。但中国表明，政策制定者对这项技术拥有主动干预的权力。他们能够引导其发展方向，而不是仅仅听任（人工智能对）命运的安排。

Human choices still matter. That’s one reason we’re already seeing two very different visions for the future of A.I., playing out in real time.

人类的选择依然至关重要。这也是为什么我们已经看到两种截然不同的人工智能未来愿景正在现实世界中实时上演。