Padma De Pana
Three years ago, the launch of ChatGPT sent China's tech sector into a panic, scrambling to close a gap of 2-3 years with the US. Today, that timeline has collapsed. The question "Are we losing the AI race?" now echoes in Washington and Silicon Valley—the very same question once asked in Beijing.
The Turning Point:
The game changed with DeepSeek's late 2024/early 2025 releases. Its V3 and R1 models proved to be on par with top Western models like GPT and Llama but at a fraction of the cost. This breakthrough sparked a strategic shift towards open-source dominance.
Chinese open-source models, led by Alibaba's Qwen and DeepSeek, now account for ~30% of global use, winning trust from developers worldwide—even US firms like Airbnb and Meta use them. Analysts estimate the tech gap has shrunk from over a year to just ~3 months, fueled by rapid iteration and abundant talent.
Divergent Paths:
As Huawei's Ren Zhengfei notes, the US and China are now on different tracks: the US chases supercomputing for AGI, while China focuses AI on solving practical, real-world problems.
The Bottom Line:
The panic is over. China's AI sector has moved from defensive isolation to open-source challenger, fundamentally reshaping the global competitive landscape. The race is now on China's terms.
Author: Saikat Bhattacharya
