The launch of DeepSeek, a new Chinese chatbot designed to compete with OpenAI, Google, and Meta, has created significant ripples in the AI industry and the US stock market. This chatbot is claimed to be more efficient and cost-effective than its competitors, leading to a dramatic decline in Nvidia’s stock, which lost $938 billion in value in just one day.
Initial tests of DeepSeek raised immediate concerns about censorship, particularly its refusal to address sensitive topics in China, such as the Tiananmen Square massacre—a limitation I encountered during my first interaction with the chatbot.
In a video demonstration, DeepSeek’s real-time self-censorship is showcased. When I posed unexpected questions, the chatbot did provide answers, only to erase them shortly after.
The video explores DeepSeek’s responses regarding China’s history of human rights violations, pro-democracy protests in Hong Kong, and even a positive critique of Martin Scorsese’s “visually stunning” biopic of the Dalai Lama, Kundun, which is banned in China. However, these responses were abruptly removed.
