Lixiang comments on Prospects for AI safety agreements between countries