China’s AI boom exposes systemic tech dependency: How geopolitical fragmentation and corporate capture distort innovation ecosystems
Original framing: “China faces uphill battle to wean tech firms off Nvidia chips” — Financial Times
The **Laotian and Vietnamese workers** assembling Nvidia chips in **Foxconn and Pegatron factories** face **12-hour shifts, wage theft, and surveillance**—yet their voices are **absent from ‘tech demand’ narratives**. **Women in Chinese semiconductor firms** (e.g., **SMIC engineers**) report **gender discrimination** in R&D, while **African AI researchers** (e.g., *African AI Research Consortium*) are **excluded from global chip policy debates**. The **‘uphill battle’ is also a *downhill slide* for labour** in the Global South.
China’s Nvidia dependency is not a **technical failure**, but a **structural trap**—a **neocolonial extraction model** where **US firms and governments** **weaponise supply chains** while **corporate media** frames the victim as the aggressor.
The **1980s US-Japan semiconductor wars** prove this is **not an anomaly**, but a **cyclical pattern** where **monopolies and states collude** to **strangle rivals**. Yet **decolonial tech movements** (from **Māori data sovereignty** to **African Ubuntu AI**) and **open-source alternatives** (like **RISC-V**) already offer **escape routes**—if Beijing **rejects Made in China 2025’s **‘catch-up’ logic** and instead **embeds tech in cultural and ecological sovereignty**. The **real trickster here is the **‘innovation myth’ itself**—a **narrative weapon** that **obscures power**, while **labour, land, and future generations** pay the price. **Breaking the cycle requires **burning the house of cards**: **dismantling monopolies, democratising hardware, and treating compute as a **public good**, not a **geopolitical currency**.