Abstract：Artificial intelligence generated content (AIGC) marks the rise of the knowledge structure of the “world brain.” Berger and Luckmann regard “knowledge” as the foundation for reality construction. On this basis， Couldry and Hepp consider media technology as a key variable， pointing out that the previous knowledge production model based on symbolic interaction has now been replaced by a data-driven distribution model， marking a “media turn” in the sociology of knowledge. This article takes this opportunity to gain insights into the transformation of knowledge production models under the waves of mediatization and proposes that AIGC possesses an inherent nature of “engineering sociality，” which has evolved along with the strengthening of the encoding and antrhopomophization of knowledge subjects by technology. AIGC exists in its relationship with society， initially driven by the input of human knowledge. It becomes a technological entity in the process of “data materialization，” but its automation ethics inevitably determine what knowledge is accepted and what is rejected. The root cause of the “imperfection” of AIGC does not merely lie in technical flaws， but rather in the structural asymmetry between machine value and human value within the coupling relationship. This article calls for the construction of a “society-in-the-loop” knowledge ecosystem to ensure that the purposiveness of human knowledge practice is reflected in the domestication of AIGC.
陈 龙 丁 冉. 工程社会性趋势下AIGC知识生产及其新结构转型[J]. 华中师范大学学报（人文社会科学版）, 2025, 64(6): 47-56.
Chen Long Ding Ran. Knowledge Production and Its New Structural Transformation of AIGC under the Engineering Sociality Trends. journal1, 2025, 64(6): 47-56.
2025, Vol. 64 
