Despite these concerns, Hugging Face, the world's largest platform for AI models, continues to be a popular platform for Chinese AI companies to showcase their models. The company's CTO announced that the default model on HuggingChat is Qwen2.5-72B-Instruct, developed by Chinese e-commerce giant Alibaba. While this model does not censor questions about sensitive topics like the Tiananmen Square massacre, another model from Alibaba's Qwen family, QwQ-32B, does.
Key takeaways:
- China's open source AI models have been praised for their performance but criticized for censoring topics sensitive to the Chinese government, such as the Tiananmen Square massacre.
- Hugging Face CEO Clement Delangue expressed concerns about Western companies using Chinese AI, warning that it could lead to the spread of cultural aspects that the Western world may not want.
- Delangue also highlighted the need for AI to be distributed among all countries, not concentrated in one or two.
- Despite these concerns, Hugging Face's default model on HuggingChat, developed by Chinese company Alibaba, does not appear to censor questions about sensitive topics. However, another model from Alibaba does.