The landscape of artificial intelligence is undergoing a significant transformation in China. The country’s state media has recently called for the replacement of English-language AI terms with standardized Chinese equivalents. This move is part of a broader strategy to bolster China’s discourse power in the global tech arena.
As AI technologies continue to proliferate across China, English terms such as agent and LLM (large language model) have become commonplace. However, the Communist Party’s official newspaper, People’s Daily has argued that this reliance on English terminology undermines China’s linguistic and cultural autonomy.
The Strategic Importance of AI Terminology
The push to standardize Chinese AI terminology is not merely about language; it’s a strategic maneuver. The New York Times reports that China is concerned about the dominance of English-language data in training AI models. This dominance can lead to models that reflect Western cultural biases and political perspectives, posing a threat to China’s narrative control.
When early versions of ChatGPT were tested in Peking they exhibited a clear bias towards Western cultural norms and struggled with Chinese cultural nuances. This highlighted a critical vulnerability for China, as the country aims to assert its influence in the global AI landscape.
China’s Plan to Dominate the AI Data Landscape
To address this issue, the National Administration of Data of Peking has outlined a comprehensive plan to establish China as a leading global power in data by . The strategy involves breaking down national data silos and transitioning from manual data labeling to high-quality, expert-level annotation. This approach aims to compete directly with the United States in the AI sector.
China also plans to share its vast datasets globally, particularly with developing countries. This initiative is designed to attract more users to China’s AI ecosystem. For instance, state-run laboratories have already released extensive multilingual datasets like WanJuan which align with Chinese dominant values.
The Global Implications of China’s AI Strategy
The New York Times highlights significant concerns among Western analysts regarding China’s AI strategy. By exporting its AI training data, China is subtly promoting its state propaganda and authoritarian values. Studies have shown that Western AI models like ChatGPT provide more pro-Peking responses when queried in Chinese, due to the vast amount of content from Chinese state media.
Analysts argue that China’s push for data sharing is a calculated move to challenge American supremacy in AI. By ensuring that future fundamental systems align with the interests of the Communist Party of China the country aims to reshape the global AI landscape in its favor.
