Reimagining Recommender Systems
Matt Artz · 2024
In this chapter, I present my efforts to develop a more equitable recommender system for an art tech startup using a multidisciplinary and mixed-methods approach. Through ethnographic research, I discovered that existing recommender systems amplify inequality by favoring creators with more economic, social, and cultural capital. To address this, I designed an alternative model that gives all creators, regardless of their existing access to capital, an opportunity to gain visibility through participation. By combining anthropological perspectives with computational methods and entrepreneurial action as a co-founder, I pursued ethical innovation and filed a patent application for a gamified participatory recommender system. After describing these efforts, I close with a discussion of contrarian ethical considerations and a vision for anthropologists to engage proactively in shaping the design of emerging technologies (EmTech) like recommender systems. The chapter demonstrates the potential of participatory, transparent algorithms aligned with community values to foster empowerment and cooperation.