“Last year China installed more industrial robots than the rest of the world put together; it also makes most of the world’s humanoids.” FT, 25.06.
Today’s FT article is especially interesting for those of us tracking the paradigm shift playing out in China, as robots become capable of an increasing range of tasks – some going far beyond typical expectations.
Image: Humanoid robots perform during the opening ceremony of the 2026 Tianjin Fashion Week on June 12 © Tong Yu/China News Service/VCG/Reuters
- Intricate industrial processes, such as factory quality control.
- Precision manufacturing tasks, such as screw-driving, glue application or bespoke garment making.
- Elite service industry jobs, such as preparing Michelin star-level meals.
- Logistics, such as the 700,000 delivery workers currently employed at JD.com.
- Hospitality, including handling luggage, food deliveries and some room cleaning in hotels.
The scale and pace of adoption is boosted by the nature of China’s state-directed economy.
State-owned enterprises and local government were just ordered to include “embodied artificial intelligence” in manufacturing, logistics, retail and healthcare industries – with a goal of establishing 10,000 AI-powered robots in commercial settings this year.
Humans wearing cameras and sensors are currently training the new army of robotic workers on a variety of tasks. And the impact on employment is set to be cataclysmic.
‘A large hotel chain has already seen staff-to-room ratios reduced from the industry benchmark of 30 to 80 workers per 100 rooms, to just 10.’
A large hotel chain has already seen staff-to-room ratios reduced from the industry benchmark of 30 to 80 workers per 100 rooms, to just 10.
In Europe we have less centralised state control, and stronger worker protection. But if companies are to remain globally competitive – especially in sectors like automotive which is already being disrupted by China – increasing adoption of embodied AI on our own shores is a necessity.
One of the key policy tasks of the next decade will be shaping what an acceptable balance of human-machine collaboration looks like inside factories, hotels, shops, restaurants and care homes.
Germany’s bind is that the design and manufacture of embodied AI units domestically could – with the right support – grow to replace some of the decline in automotive revenues (worth up to 6% of GDP.) But that very capability will simultaneously reduce local employment, even as it restores Germany’s export-led economic strength.
It feels like sci-fi, it feels far away in time and distance. But it will be one of the most strategic and far-reaching decisions of the next 10 years. Maybe even the next 50.