In this piece, I argue that China has neither the resources nor any interest in competing with the US in one specific sub-field of artificial intelligence: developing artificial general intelligence (AGI) via scaling large language models (LLMs) by training them on increasing numbers of graphic processing units (GPUs).