Is there a Chinese AI safety community? Are there safety researchers working at leading Chinese AI labs?
According to Lao Mein‘s personal experience, “AI safety is almost non-existent in China”. Lao estimates that there are no more than 20 Chinese residents holding a job in the field of AI Safety, including one organization named Anyuan (安远), working on field-building.
Lao presents their assessment of the state of AI safety in China in the two posts on LessWrong (1, 2), which were written largely in the context of the discourse regarding AI-related regulations being drafted by the Chinese Communist Party, which sparked some enthusiasm regarding Western-Chinese collaboration in managing and navigating the development of powerful artificial intelligence (see: Does China have any AI regulations?).
On the other hand, according to James Marshall’s article published in WIRED, the president of the Beijing Academy of Artificial Intelligence (BAAI) Zhang Hongjiang sees the Chinese community of researchers as deeply concerned about risks posed by development of powerful AI, although historically, BAAI has been focused on more immediate, rather than existential risks. However, the concern itself is not a sufficient reason to motivate the action to slow down the development of AI, as many researchers (including Zhang Hongjiang himself) believe that smarter AI systems may actually be safer or easier to align.