Modest Algorithmic Mediation can Maximize Topical Diversity in Hybrid Human-AI Systems
2026-07-27 • Social and Information Networks
Social and Information Networks
AI summaryⓘ
The authors studied how algorithms and social networks together affect what information people see and share on social media from 2014 to 2018. They found that after algorithms were introduced in 2016, the variety of topics people shared increased then leveled off, but some users shared much more diverse content than others. They created a model mixing social connections and algorithm recommendations and showed that moderate algorithm use can increase overall diversity and fairness, but too much can reduce diversity and make it uneven. Their analysis suggests algorithms can help widen exposure if used carefully. This work helps understand how human and AI systems combine to spread information.
information diffusionalgorithmic recommendationsocial mediatopical diversitysocial networksrecommender systemsinformation exposureinequalityhybrid human-AI modelalgorithmic mediation
Authors
Dini Wang, Ho-Chun Herbert Chang
Abstract
In the artificial intelligence (AI) era, the rise of algorithmic feeds has fundamentally transformed information diffusion on social media. While early platforms organized visibility through explicit social networks, contemporary systems mediate exposure through intelligent recommender algorithms that personalize attention. This paper examines how the social network and algorithmic architecture jointly shape the diversity of information sharing. Analysis of 18,076 users active throughout 2014--2018 shows that the topical diversity of sharing rose and then plateaued after the introduction of algorithmic ranking in 2016 while its inequality across users emerged alongside it. To this end, we introduce a hybrid human-AI information diffusion model in which information exposure is governed by a parameterized mixture of social propagation through the user-following network and algorithmic recommendation. Both qualitative analysis and simulations show that the effect of algorithmic mediation is non-monotonic. Modest mediation can raise average diversity and reduce inequality relative to a purely network-driven baseline, whereas strong mediation reduces diversity and concentrates it among fewer users. Fitting the model to four years of data yields a mediation share that increases from zero before 2016 to approximately 0.50 by 2018, a level that exceeds the compensation point of equality while remaining within the diversity-enhancing range. These results identify the conditions under which recommendation broadens rather than narrows exposure and provide a unified framework for information diffusion in hybrid human-AI systems.