Abstract: [Purpose/Significance] This study aims to critically analyze the social philosophical roots of the digital intelligence divide from the perspective of machine functionalism. By uncovering the theoretical origins and generation pathways of the digital intelligence divide, countermeasures can be proposed. The research contributes to understanding the divide's impact on society and provides insights for promoting inclusive development of artificial intelligence (AI) technology. The study fills a gap in the literature by linking machine functionalism to the digital intelligence divide and offers a novel perspective on addressing the unequal use of AI technology. The findings have significant implications for policymakers, technology developers, and researchers in the fields of AI ethics, digital inequality, and social philosophy. [Method/Process] Using the theoretical lens of machine functionalism, this study examines the evolutionary pathways, generation mechanisms, and multiple risks of the digital intelligence divide. It draws on relevant theories, such as the extended mind thesis and the theory of technological determinism, to analyze how machine functionalism influences the design and application of AI technology. The study also draws on empirical evidence from case studies and surveys to illustrate the manifestation of the digital intelligence divide in different contexts. By synthesizing theoretical and empirical insights, the research proposes interventions that address the divide at different levels, from the philosophical underpinnings to the practical implementation of AI technology. [Results/Conclusions] The study shows that machine functionalism, which applies Turing machine principles to explain the mind and views the mind as a physically realized Turing machine. It has become the social philosophical foundation of AI technology. While breaking with the traditional biological essentialist view of the mind, machine functionalism inadvertently creates inequitable uses of AI through three main pathways: the mechanization of the mind, designer bias and algorithmic preference, and technological specialization and barriers to entry. This creates the digital intelligence divide and risks such as the evolution of information access inequality into social inequality and the weakening of information cocoons and public dialogue. The study argues that interventions are needed to mitigate these risks and promote a more equitable distribution of the benefits of AI technology. To bridge the digital intelligence divide, the study suggests a multi-pronged approach. First, future efforts should focus on promoting positive interaction between machines and humans through value-sensitive design, which incorporates ethical considerations into the development and deployment of AI systems. Second, developing ethical algorithms that eliminate designer bias and algorithmic preference is critical to ensuring fair and unbiased AI decision-making. Third, improving the digital intelligence skills of individuals and communities can help break down barriers to entry caused by technological specialization and enable more people to benefit from AI technology. Together, these policies can help break down the barriers of unequal technology use under machine functionalism. The study concludes by emphasizing the importance of a collaborative effort among policymakers, technology developers, researchers, and the public in addressing the digital intelligence divide. It calls for further research on the social implications of machine functionalism and the development of inclusive AI governance frameworks. The findings of this study serve as a foundation for future work to mitigate the risks of the digital intelligence divide and promote the responsible and equitable development of AI technology.

Key words: machine functionalism, digital intelligence gap, designer bias, algorithmic preference, information inequality, information behavior