Macro Crossroads | AI's K-Shaped Economy: How Can We Cushion the Impact?

Prof. Zhu Ning

Macro Crossroads is a column by Prof. Zhu Ning, Senior Macro Strategist at Primavera Capital Group. In this series, he offers perspectives on the macroeconomic trajectory of China and the global economy, technological and industrial developments, and key policy shifts.

China’s tech-heavy SSE STAR 50 index has surged by 65% in the first half of 2026, representing one of its strongest extended rallies ever. The performance far surpasses the 7.6% gain for the broad-based China market CSI 300 index, and highlights the rise of investor interest in the country’s LLM, semiconductors, optical modules, and other infrastructure and hardware firms involved in AI.

Traditional sectors are not the only ones where market performance has been lagging. Hong Kong’s Hang Seng Tech Index, with its heavy exposure to more established Chinese internet platform companies and limited direct exposure to AI companies, fell 19% during the same period.

Macro-level investment data in China appears to reflect these stock market trends. Investment in high-tech industries increased by 4.5% during the first five months of 2026, whereas national and non-governmental fixed asset investment dropped by 4.1% and 7.1% respectively.

On multiple levels, a two-speed growth pattern has emerged between the AI and other sectors, often referred to as K-shaped growth. This is extending beyond stock market performance and becoming increasingly evident in the pace of revenue growth, capital expenditure, wealth distribution, as well as broader societal views toward personal well-being.

Making things even more serious at a global level, the divergence in growth is occurring both within and between many different countries. At the same time that the U.S. and China are aiming to take the lead in AI development and deployment, most other nations lack independent AI capabilities as well as the level of investment needed to develop them domestically. Such drastic differences in AI capabilities will eventually lead not only to differences in productivity, but also deeper, long-term competitive concerns.

In recent years, a great deal of hope has surrounded AI as a great leveler by enabling and empowering anyone with the ability to do more. However, as seen with the direction of the world’s largest stock markets, AI has generated substantial prosperity for a concentrated group of top performers. In time, this may further aggravate instead of alleviating inequality and social tensions.

So how can modern society harness the greatest value from AI while it is still dealing with the wider K-shaped economic recovery from Covid? More focus is needed on building out improved social safety nets. Previous disruptive innovations have shown that new technology creates great wealth and displaces considerable employment at the same time. As these trends accelerate in tandem, so does broader resistance against the pace of technological advancement.

In parallel, societies must also strive to have greater AI readiness while ensuring the benefits that AI creates are shared among a broader base. One of the renewed arguments for universal basic income (UBI) is that the knowledge and data that enabled the AI revolution and corresponding wealth boom does not belong to any one individual or company.

Globally, international coordination and governance mechanisms are urgently required. This is not just to prevent and mediate rising inequality and potential societal disruptions among countries, but even more importantly, to avoid a two-speed world taking hold between humanity and AI itself.

——

Based on the author’s speech in “AI and the Two Speed Economy”, World Economic Forum AMNC 2026 in Dalian