The US training humans and robots making AI

Hello from Yifan in Silicon Valley, your #techAsia host this week.

It's good to be back in sunny California after spending the past months reporting across the U.S., tracking how different states and companies are preparing for the "reindustrialization" of America, or to put it in plain terms, bringing back manufacturing to U.S. soil.

The plan has been in motion since the pandemic-era supply chain chaos but has been kicked into a higher gear by President Trump's tariffs and other policy changes. Yet from the very beginning, there has been a piece of the puzzle missing: Who will work those manufacturing jobs once they get here?

If you ask members of Gen Z what their dream job is, working in a fab will probably rank pretty near the bottom of the list (definitely lower than being an influencer). But Washington along with state governments, companies and education systems are working hard to change that.

When I first visited Arizona back in 2022, few people there knew what a fab was or had ever heard of TSMC. Now the Taiwanese company seems to be almost a household name in the Phoenix area, especially in community colleges and universities where chipmakers have set up semiconductor training programs. A small gift bag I received when I visited SEMI Quest, a semiconductor-themed playground at the Arizona Science Center, included a cute coloring book to illustrate what it is like to work in a fab. Well, if that doesn't get kids interested, I don't know what will.

But as I watched children playing with mock silicon wafers at the park, I couldn't help but wonder what kind of jobs will be left for them to do once they are grown up.

As a reporter based in Silicon Valley, I've had a front row view to how quickly AI has impacted the entire tech ecosystem, including the massive layoffs happening at Amazon and other companies.

My latest trip brought me to Washington D.C. last week, where I watched Nvidia CEO Jensen Huang stand in front of a row of humanoids while describing a future of robots making robots factories.

While Huang later added that such robots will only be a companion to human workers, "jobless growth" -- when the economy expands without adding more employment -- fueled by AI seems to be the new norm.

It would be nice if those working on a factory line today can be reskilled and move to better jobs, as some CEOs and technologists say. But the question remains: What jobs?

In other news, we will be hosting a webinar on Nov. 13 exclusively for our subscribers: "Nikkei Asia Live: Our editors' predictions for 2026." If you're not yet a subscriber, you can join the webinar by subscribing for just $1 here. If you're already a paid subscriber, no registration is required.

Mind the gap

With hundreds of billions of investment dollars pouring into manufacturing in the U.S., led by Apple's $600 billion commitment, the country is expected to have a worker shortfall of 67,000 technicians, computer scientists and engineers in the chip industry alone by 2030, and a gap of 1.4 million for the overall economy by some projections.

America's newfound enthusiasm for high-tech manufacturing comes against a backdrop of competition with China and other countries for tech talent and leadership in artificial intelligence. But as states like Arizona are rushing to train their residents and fill the gap in human resources, the rise of artificial intelligence threatens to make these highly in-demand jobs obsolete before long, Nikkei Asia's Yifan Yu writes.

In the short term, however, building a domestic workforce that can fill the rising number of American factories quickly is becoming even more urgent as the Trump administration makes it increasingly difficult for foreign workers, who have been a vital pillar of the advanced manufacturing industry, to stay in the U.S., with new immigration policies such as a $100,000 fee for H1B visas.

Power-ups

China is increasing electricity subsidies to some of the nation's largest data centers that use domestic chips, in a bid to accelerate the development of its semiconductor industry and compete against the U.S., write the Financial Times' Zijing Wu and Eleanor Olcott.

The initiative comes in response to a ban on purchasing advanced, energy-efficient Nvidia AI chips, which has forced Chinese tech giants like ByteDance, Alibaba and Tencent to use less efficient domestic alternatives from companies like Huawei and Cambricon.

These Chinese chips typically require more electricity to perform the same amount of computing tasks as their Nvidia counterparts, significantly increasing operational costs.

To mitigate this financial impact, provincial governments in data center hubs such as Gansu, Guizhou and Inner Mongolia are offering subsidies that slash electricity bills by up to 50% -- but only for facilities powered by Chinese chips. Data centers using overseas chips do not qualify for such entitlements.

This targeted support helps level the playing field for Chinese tech companies, making it more viable to break their reliance on foreign technology. Furthermore, it leverages China's centralized energy grid, which provides cheaper and greener power than the fragmented U.S. network.

Eyes on India

The world's most populous country is getting more attention from outsiders as it grows both its tech industry and capital market.

South Korean conglomerates led by LG and Hyundai Motor have rushed to India's stock market to raise capital to beef up business operations in the country by listing their local subsidiaries, Nikkei Asia's Kim Jaewon and Soumyajit Saha report.

On the listing day, shares of the Indian unit of LG Electronics jumped 50%. While the IPO had valued the unit at 774 billion rupees ($8.7 billion), the company's valuation had jumped to 1.13 trillion rupees ($12.7 billion) as of Friday. This is even bigger than the parent company's value of 14.5 trillion won ($10.1 billion) on Seoul's stock market.

Meanwhile, foreign AI companies are hopeful they have found a way to crack this increasingly important yet complex market. Google and Perplexity are tying up with India's telecom giants to offer free services, as the biggest names in artificial intelligence up their efforts to win over the country's billion internet users, Saha writes.

Such moves -- which OpenAI has matched by making its intermediate-level services free since this Tuesday -- are replicating a strategy that helped boost digital media distribution companies and spending on mobile data usage.

Robot-made AI

Foxconn will deploy humanoid robots to make AI servers in Texas within months as the world's largest contract electronics manufacturer continues to expand aggressively in the U.S., Nikkei Asia's Lauly Li reports.

"Within the next six months or so, we will start to see humanoid robots [in our factory]," Chairman and CEO Young Liu said. "It will be AI humanoid robots making AI servers."

The move will mark the first time in its more than 50-year history that Foxconn will use humanoid robots on its production lines. The move is expected to boost the efficiency and output of AI server production. "Speed is very critical for high technology like AI," Liu said.

Long known as a key Apple supplier, Foxconn also has a close relationship with Nvidia. In North America, it has AI server production capacity in Texas, California and Wisconsin, as well as Guadalajara, Mexico. It also plans to start making them in Ohio as part of the Stargate AI infrastructure project.

Podcast: Tech Latest

Where Samsung stands in the AI race

Welcome to the Tech Latest podcast. Hosted by our tech coverage veterans, Katey Creel and Shotaro Tani, every Tuesday we deliver the hottest trends and news from the sector.

In this episode, our host Katey speaks with Seoul correspondent Kim Jaewon about Samsung's efforts to make up lost ground in the AI supply chain and innovation frontlines after falling behind its rivals.

For more great stories like this delivered to your inbox every week, sign up to our #techAsia newsletter. Current subscribers, click here to update your newsletters preferences.