Abstract: This article points out that artificial intelligence is not artificial consciousness, and it is necessary to avoid using anthropomorphic terms such as “endowing machines with consciousness” and “machine values”; what human civilization truly needs to guard against is not “conscious AI”, but “unconscious intelligence” that is not constrained by human civilization but can continuously change the real world without “kill switch”; currently, almost all explorations of artificial intelligence are still within the framework of “computation”; before there is a fundamental breakthrough in the boundaries of computation, “simulating non-computation with computation” is the core paradigm for the development of artificial intelligence; the artificial intelligence that humans hope to develop is not another “mind” independent of humans, but a system that extends the will of human society; we must adhere to the “human-centered” development path of artificial intelligence.