Subir Sarkar finds that AI was a hot topic among the 800 delegates who gathered in the Chinese capital last month
China has made astounding scientific progress over the past decade, with the country outperforming all others in the world according to a recent analysis by Nature. Based on data from 2025, its 2026 Research Leaders study puts China ahead of the US (in second), Germany (third) and the UK (fourth).
Across all disciplines, Nature reports that most (51) of the top 100 institutions in the world are in China, with only four in the UK. The picture is slightly different if you judge academic institutions based only on their work in the physical sciences, but China still has 30 of the top 100 institutions, with only seven from the UK.
Professional physicists may not care much for such league tables but they do reveal just how far China has come in recent years. The country’s scientific strength was particularly evident at the fourth International Congress of Basic Science (ICBS) in Beijing, which I recently attended.
Founded in 2023 by the Fields-medal-winning mathematician Shing-Tung Yau, this year’s ICBS ran from 9-21 August at the Beijing Institute of Mathematical Sciences and Applications (BIMSA), of which Yau is president. Its international character is evident from the fact that a total of 50 of its 140 full-time researchers are from outside China.
Awards and honours
Over 800 participants from more than 30 countries attended the congress, which had the theme “Advancing science for humanity”, with my participation supported by the organisers. The programme was huge, featuring over 500 academic sessions covering topics from generative AI reasoning to quantum information.
Nine scientists, including four women, were awarded an inaugural Basic Science Medal. They were Claire Voisin (algebraic gemoetry), Horng-Tzer Yau (probability theory), Shou-Wu Zhang (maths), Yifang Wang (particle physics), Xiao-Gang Wen (condensed matter), Andrea Liu (complex systems), Zhenan Bao (flexible electronics), Xiaowei Zhuang (imaging), and Feng Zhang (CRISPR co-discoverer).
Meanwhile, Frontiers of Science Awards were given to over 500 researchers from 20 countries who had done work of “excellent and of outstanding scholarly value” over the last 5-10 years in the three fields represented at the ICBS. Authors from many Chinese institutions were among the recipients, underscoring the growing international impact and vitality of China’s basic science research.
There were 20 plenary sessions, over 400 Frontiers of Science Awardee talks, plus events for entrepreneurs and school and university students. Especially engaging were the “nights of mathematics, physics and engineering”, enlivened by dance amd music performances by students. An exchange meeting was held too for research institutions in Beijing seeking to integrate basic science with frontier technologies
The closing ceremony saw the winners of this year’s ICBS Innovation Fellowships announced. They are all researchers under 50 who work full-time in China and have demonstrated potential in mathematics, physics, information sciences, and other fundamental and interdisciplinary fields. Each receives research funding and can work with a partner institution in China for an agreed period.
AI surges ahead
Artificial intelligence (AI) was one of the major themes of ICBS 2026. Fields medallist Martin Hairer from EPFL Lausanne, for example, said AI won’t end mathematics, but will transform how it’s carried out. It may speed up certain areas while others might fall away.
AI-led solutions of Erdős problems spark debate over the future of mathematics
Caucher Birkar from Tsinghua University – another Fields medallist – pointed out that AI is highly capable in computation, searching for examples, and completing proofs of mathematical lemmas, but it remains difficult for AI to generate truly original ideas.
As for Robert Endre Tarjan, a Turing award laureate from Princeton University, he encouraged young researchers to be curious and commit themselves to long-term scientific exploration. Researchers, he said, can train AI to be a powerful tool for scientific discovery, but they must also develop their own ability to identify meaningful questions and establish independent judgment.
The last word was left to congress chair Shing-Tung Yau, who emphasised that while AI will enhance scientific research, the essence of scientific discovery will always depend on human creativity, imagination and independent thinking.
