Abstract:

With the rise of Large Language Models, the field of opensource knowledge production will witness a new transformation. Large Language Models have positive impacts on opensource knowledge production while they also bring potential challenges. On the one hand, Large Language Models significantly improve the efficiency of opensource knowledge generation and dissemination by adapting to Cunningham’s Law of the opensource community, providing round-the-clock newcomer knowledge training support, and correcting the systematic bias of knowledge production through domain-based construction strategies; on the other hand, Large Language Models have brought about the phenomenon of hallucination, copyright risk, digital exploitation, and the intensification of the trend of “dead internet”, which poses a serious threat to the verification, legitimacy, values, and ecological environment of opensource knowledge. Thus, we should strengthen the core role of human cognitive experience in guiding the technology development of Large Language Models, and explore solutions through practice, in order to realize the harmonious coexistence and common progress of knowledge production of Large Language Models and opensource knowledge production.

Key words: Large Language Models, opensource knowledge production, systematic bias, illusionary phenomena, digital exploitation, Wikipedia