China’s Agibot Robotics livestreamed the first continuous deployment of humanoid robots on a commercial production line in Nanchang, where eight G2 models operated for 64 hours, completing 64,828 tasks with a 99.99% success rate and producing 17,625 tablets. The six-day demonstration, using a wheeled manipulator design for industrial handling and inspection, signals a move from laboratory prototypes to durable shop-floor automation. Agibot has since delivered its 15,000th robot, having cut the time needed to add 5,000 units from roughly a year to three months.

The Chinese approach has been forged by necessity. US export restrictions on advanced graphics processors pushed engineers to build the LineShine supercomputer without them, and it now ranks as the world’s most powerful. AI models from Zhipu and DeepSeek achieve competitive performance in certain benchmarks while using far less computing power—a pattern of efficiency now extending into physical automation. Viewed from Beijing, robots are seen not merely as productivity tools but as a demographic safety valve; official projections show the working-age population contracting from 1 billion to 300 million by century’s end, and companies such as JD.com already predict wholesale replacement of delivery workers.

This push for AI-driven efficiency is reshaping creative sectors unevenly. Netflix’s use of generative AI for visual effects in the Argentine series “El Eternauta” cut a collapse scene’s production time tenfold, while Hollywood’s output of films with budgets above $40 million fell 37% in five years, to 159 releases. Amazon and Disney are committing hundreds of millions of dollars to proprietary AI tools, yet labour unions in Los Angeles accuse studios and even directors such as Martin Scorsese of betraying collaborative craft. In parallel, a copyright conflict is intensifying: EU legislation requires transparency on training data and allows rights holders to opt out, while US courts hear claims that AI firms built lucrative models on copyrighted works without compensation.

Infrastructure demands are spreading beyond traditional tech hubs. Mexico is experiencing a data-centre boom, with Microsoft, Amazon Web Services and Google investing heavily, but analysts in Mexico City warn that the strain on electricity and water could offset gains, especially as government estimates suggest 22 million jobs may be affected within five years. The next milestones to watch are US federal rulings on AI copyright liability and the enforcement of the EU AI Act’s training-data transparency provisions, both of which will determine how the costs and benefits of AI’s industrialization are shared globally.