Hey, it’s Marek.
Recording this one from Shanghai with the skyline lighting up behind me. Unusual setup: microphone in my glasses, cold weather, no studio. But I wanted to share what I’m seeing here as the Year of the Horse approaches.
Three patterns are emerging in China’s AI strategy:
Regulation ≠ the opposite of innovation. China leads the world in high-impact GenAI research while enforcing the strictest AI regulations. AI companions must warn users every two hours. Addiction monitoring is mandatory. And yet, they’re publishing more impactful research than anyone.
Smarter beats harder. Between 2023 and 2025, US tech spent $694 billion on AI infrastructure. China spent $123 billion, six times less. The result? Chinese models like MiniMax M2 now reach 90% of US capability. Architecture beats horsepower.
Fence locally, release globally. Chinese LLMs went from 1.2% global usage to nearly 30% in twelve months. Meanwhile, ChatGPT and Claude remain banned inside China. The world’s largest market blocks foreign models while flooding everyone else’s.
The Year of the Horse starts February 17. Wild horses didn’t transform civilisation; but their domestication did. Bridles, saddles, training protocols. The West treats AI regulation as a brake. China treats it as a saddle and reins.
Give it a listen/watch.
Stay curious!
