New Oriental founder Yu Minhong chose a slot at the 2026 Chongli Forum to argue that artificial intelligence could speed up income inequality, even as it creates new kinds of work. According to a November 12 report by Chinese technology outlet IThome, Yu said the AI wave is following the internet in generating opportunity, but it is also hitting workers who rely on repetitive labour hardest.
Drawing a comparison with the industrial revolution, Yu said the current shift may be more intense. AI, he said, is concentrating wealth in a very small group of people with unusual speed. He pointed to the market capitalisations of AI-related listed companies, which he put between the hundreds of billions and trillions of yuan, and noted that those gains are not flowing to ordinary households. In his view, the richest beneficiaries of AI may not create broad job opportunities for the public, nor is their money likely to circulate back to ordinary people, meaning many may never share in the industry's financial upside.
Jobs will change, not disappear
Yu was not arguing that AI will simply destroy employment. He said the internet once created countless jobs, and AI should do the same in the future. The catch, he added, is that the next round of jobs will require relearning and re-education. The people most exposed, he argued, are ordinary workers whose current income depends on repetitive tasks, precisely the kind of work AI automates first.
That puts the burden on individuals and institutions to adapt quickly. Yu's remarks reflect a concern heard in many countries: AI's productivity gains are real, but the distribution of those gains may be more uneven than in earlier technology booms. Where the internet lowered the cost of starting a business and created distributed opportunities, today's dominant AI platforms tend to require large capital investments in chips, data centres and research, which favours big companies and wealthy investors.
Signs AI is entering a new phase
The same IThome item carried other AI-related news that points to a maturing and more regulated industry. It summarised a Goldman Sachs view that focusing too much on whether an AI bubble will burst risks underestimating the value AI creates. It also noted that California had enacted the United States' first AI safety law, a sign that governments are starting to set rules. Another item reported that a US lawyer was fined US$5,000 after using ChatGPT to generate court documents containing fabricated police testimony, a reminder that AI tools can produce confident but false output.
For readers trying to make sense of AI's impact, Yu's warning is a useful counterweight to the hype. The technology is not only about better chatbots or faster code; it is also changing who captures economic value. In the auto industry, where software-defined vehicles rely on AI for driver assistance, voice control and manufacturing quality, the same concentration dynamics are at work. Companies that control models, data and computing power may keep most of the value, while suppliers and workers face pressure to adapt.
Yu did not propose a detailed solution in the reported remarks, but he stressed the need for people to relearn. Whether through education, new business models or policy, the challenge is to make sure that the people whose jobs are displaced have a path to the jobs AI creates.
Based on reporting by IT之家. Edited and published in English by geisou.
