“China Said to Fall Short of Matching US Advances in AI owing to ‘Many Challenges in Theory and Technologies’”, 2024-03-15 ():
Many Chinese-developed AI models have been built on Meta Platforms’ LLaMA system, according to a presentation made to Chinese Premier Li Qiang
The presentation at the Beijing Academy of Artificial Intelligence (BAAI) also pointed out issues related to computing infrastructure for training AI models
…China is falling short of matching the United States in artificial intelligence (AI) advances because the nation’s efforts are “littered with many essential challenges in theory and technologies”, according to a recent slide presentation made to Chinese Premier Li Qiang.
…The presentation said there is “a serious lack of self-sufficiency” in that area of Chinese AI development because most domestic LLMs are built on LLaMA. Facebook parent Meta in July last year made its open-source LLaMA-2 AI model free for research and commercial use.
…“Dozens of locally developed chips are different in terms of families and ecosystems”, making the 100-billion-parameter training for Chinese LLMs “very unstable”, the presentation said. US tech sanctions on China have restricted the mainland’s access to advanced semiconductors, made with American technology, for local AI development projects.
…The number of government-approved LLMs and related AI applications on the mainland currently total more than 40. But at present, there are more than 200 China-developed LLMs in the market.
Another major issue pointed out by the presentation at BAAI refers to control of AI-generated content.
It said the unique challenge faced by Chinese-developed LLMs is in generating “quality content that is in line with facts”, while also taking into account ideology and various emotions.