Wujie Frontier · Information as of October 1, 2026

Trump takes questions from reporters after the AI luncheon. Screenshot from the original page video, source: Trump Archive, event date September 29, 2026, US time.

On September 29, Trump was asked whether China's AI development could surpass that of the United States. He first said he did not think it would happen, then added that unless the US does something stupid and drives these people to other countries, and that this kind of situation could indeed happen.

From China's perspective, the question worth pursuing in this passage is: even if talent and technology flow over, what do we rely on to keep them here and turn them into products that can be delivered sustainably?

A recently released data set provides a more concrete reason for optimism. The International Federation of Robotics disclosed on September 24 that China actually installed 354,000 industrial robots in 2025, accounting for 59% of global installations that year. China already has a massive production floor where AI teams can find problems, validate technology, and face customer scrutiny.

First, Put Trump's Remark Back into Context

This Q&A appeared around the 28-minute-10-second mark of the press conference after the September 29 luncheon. The English original used "these people," did not explicitly say "Chinese researchers," and did not announce that US talent is flowing back to China.

The day's Q&A as archived by Trump Archive. At 28 minutes 20 seconds, he talks about "driving these people to other countries," then gives an example of someone going to Europe to build a project because it is easier there. Source date: September 29, 2026.

Immediately after, Trump gave an example, saying that a person present was building a project in Europe because it was easier there. He wanted the project built in the United States. Combined with the preceding discussion about data center construction, this is a passage expressing concern about the flow of people and investment. It is not enough to prove that China has already acquired this group of talent.

But this passage reminds us that AI competition also depends on whether a place can enable people to get things done. Research conditions, construction costs, and customer demand can all influence where teams and projects go. This is the judgment of this article and cannot be treated as Trump endorsing China.

China's talent base has supporting evidence. MacroPolo's Global AI Talent Tracker project, using authors of NeurIPS 2022 papers as a sample, found that 47% of researchers did their undergraduate studies in China, up from 29% in the 2019 sample. The project released this update in 2024, and the measure is the location of undergraduate education, which is neither equivalent to nationality nor usable to count the number of returnees in 2026.

China Already Has Many Places That Can Absorb Technology

For physical AI, the production floor is especially important. Physical AI means letting models participate in real-world tasks through cameras, sensors, and machine actions. For example, recognizing parts, grasping objects, or enabling vehicles to understand road conditions. It has to deal with reflective surfaces, misaligned workpieces, and downtime that customers are unwilling to accept.

Screenshot of the original page of the International Federation of Robotics Chinese press release, published September 24, 2026. Both 354,000 units and 59% refer to new industrial robot installations in 2025, not the robot stock, and not humanoid robot sales.

The value of 354,000 installations lies in the large number of already-operating equipment, buyers, and maintenance needs behind it. Chinese teams have the opportunity to improve perception, control, and deployment methods in these specific tasks, then hand the solution to the next customer. This is closer to a business than completing a single action in a demo venue.

This is an industry judgment based on installation data. Traditional industrial robots are not inherently equivalent to embodied intelligence, and factory data will not automatically become usable training data. Whether companies are willing to open their data, whether old equipment can be connected, and whether reliability holds when a workpiece is changed all need to be verified item by item.

The same IFR statistics also show that in 2025, Chinese domestic brands accounted for 55% of new domestic installations, with domestic brand installations growing 15% year-over-year. This shows that domestic suppliers have already participated in actual deployment at considerable scale. However, the 55% share is lower than 57% in 2024, so it cannot be written as domestic share rising continuously.

The basis for optimism is that China already has a site where customers, equipment, and suppliers solve problems together. Whether it can translate into AI advantages depends on whether new algorithms can reduce rework and downtime for customers and remain affordable in procurement and maintenance costs.

After Someone Pays, It Still Depends on Whether Delivery Can Be Sustained

Baidu's second-quarter 2026 results released on August 18 provide a business case that can be tracked. According to company disclosure, AI Cloud Infra revenue from April to June was 7.3 billion yuan, up 50% year-over-year; AI Applications revenue was 2.5 billion yuan, up 3% year-over-year. These business category figures come from unaudited management accounts.

This at least shows that part of AI demand has already entered revenue, rather than remaining only in trials and experiences. The growth difference between infrastructure and applications is also worth noting, but a single quarter and different business categories alone are not enough to judge the overall profitability of China's AI applications.

In the same disclosure, Baidu said that Apollo Go has launched fully driverless commercial operations in Dubai, where passengers can hail rides through Uber and its own app; London is open-road testing, and Hong Kong Airport Island is also testing. Commercial operations and testing are different stages and cannot be merged into a single list of "globally deployed" cities.

Autonomous driving is a real application of physical AI. Chinese teams beginning to deliver services overseas is a positive signal. However, the above progress is company disclosure; the announcement did not separately disclose Apollo Go's revenue, profit, or per-ride cost, so it cannot be used to declare that the business model is already profitable.

For readers focused on hard-tech investment, the more valuable evidence next is whether paying customers renew, whether deployments are replicated at a second site, and whether gross margin and cash recovery improve as revenue rises. Technology being able to complete a task only answers part of the question of whether customers will buy.

Optimism Can Be Built on What Has Already Happened

China also still has [needs]