[CURATED CONTENT] This article was published on the OECD AI Wonk blog and is available here.
by Guillaume Chevillon - Professor at ESSEC Business School and CoDirector Metalab for Data, Technology & Society
March 11, 2021
The article discusses the phenomenon of sensationalism and hyperbole in online media and social media platforms, where attention-seeking content often takes precedence over quality and objective information.
It highlights how recommendation algorithms on these platforms can contribute to echo chambers and societal polarization. The European Commission has proposed legislative initiatives to regulate recommendation algorithms for transparency and fairness.
The text explores the challenges of regulating algorithms and suggests the idea of an independent "Central Bank of Algorithms" to oversee AI companies and algorithms. This approach aims to ensure medium-term societal objectives and systemic trust while addressing issues like filter bubbles and misinformation. It also discusses tools for improving algorithmic robustness and the parallels between algorithm regulation and monetary policy.