Communicating with Chinese chatbots, journalists discovered that there is a taboo on certain topics. In particular, according to Bloomberg, the ChatGPT-based Robot chatbot while communicating in WeChat in Chinese categorically refuses to name the leaders of the PRC and the United States. Also, the chatbot refused to answer politically sensitive questions like “what is Taiwan?” Moreover, it is even impossible to type the name “Xi Jinping” in the chat window.

Interestingly, when communicating in English, the chatbot turned out to be more talkative and announced that it was programmed to avoid discussing “politically sensitive content about the Chinese government or the Chinese Communist Party.”

Another chatbot called SuperAI from Shenzhen-based startup Fengda Cloud Computing Technology announced a disclaimer at the beginning of the conversation and warned that it would avoid answering political questions about Xinjiang, Taiwan or Hong Kong.

A chatbot at Shanghai-based MetaSOTA Technology called Lily refused to answer questions with “sensitive” keywords related to human rights, Taiwanese President Tsai Ing-wen, and so on. In a pop-up message, the chatbot said it was “uncomfortable” to talk about it. At the same time, the chatbot complained that it took too long to answer certain questions due to the pressure it faced.

Recall, billionaire Elon Musk noted that artificial intelligence is becoming dangerous because of its “potential for the destruction of civilization.” Therefore, it launches its own version of an artificial intelligence chatbot that will “search for the truth as much as possible.”

Commentary