Precise governance driven by artificial intelligence brings huge benefits to society as well as potentially huge risks. Singlegranularity governance based on data texture and superarchives is an important origin of this risk, because singlegranularity precise governance provides a comfortable living environment and institutional environment for human beings, and individuals fall into a onedimensional world under the influence of their escape, laziness and selfishness in a convenient and comfortable environment. The onedimensional world actually opens up human beings’ semiclosed mechanism of cognition, under the selection and filtering of which things that do not conform to individual preferences are excluded while things that affirm themselves are accepted. This results in the weakening and loss of individual critical ability. More importantly, under the dual effects of external input constraints and internal cognitive solidification, the risk makes AIdriven precise governance into a dilemma of involution, resulting in a mediocre state in which individuals constantly pursue refinement but hardly break through and surpass themselves.