[ESSEC Knowledge] by Guillaume Chevillon - Professor at ESSEC Business School ,Director of the ESSEC-CentraleSupélec Master in Data Sciences & Business Analytics.
Abstract
This article delves into the debate surrounding the regulation and control of algorithms, particularly those used by social media platforms, and offers an alternative approach to addressing their potentially harmful effects.
Key points include:
- Differing Perspectives: The article notes that there are varying opinions on how to address the concerns related to algorithms. Some, like Elizabeth Warren and Bernie Sanders, call for dismantling big tech monopolies, while others, such as Margrethe Vestager and Mark Zuckerberg, advocate for more regulation. The question is how to strike a balance.
- Proposal for a Regulatory Authority for Algorithms: The author suggests creating a Regulatory Authority for Algorithms, inspired by the principles of the U.S. Food and Drug Administration. This authority would evaluate algorithms, primarily those used by social networks, to assess potential risks before they are commercialized.
- Algorithmic Impact on User Behavior: Algorithms used by social media platforms aim to maximize user engagement, often prioritizing sensational content over quality. This can lead to information bubbles, reinforcing existing beliefs and limiting exposure to diverse perspectives.
- Limitations of Ex-Ante Control: The article argues that assessing algorithms before their use is problematic since algorithms affect and are influenced by human behavior in a dynamic and unpredictable way. Ex-ante control may be insufficient to address medium-term consequences.
- A Central Bank of Algorithms: The proposed solution is to establish an independent authority, referred to as a "Central Bank of Algorithms." This agency would supervise AI companies, modify essential algorithms, and monitor their impact. It would prioritize societal objectives and the evolution of technology in the medium term.
- Building Systemic Trust: By providing oversight and independent control, the Central Bank of Algorithms would aim to enhance trust in AI systems, better anticipate individual reactions, address information bubbles, and mitigate misinformation.
- Long-Term Sustainability: The article emphasizes that the proposed approach may slightly slow the development of artificial intelligence but would focus on public interest and long-term sustainability, similar to how financial regulations aim to prevent economic crises.
In summary, the article explores an alternative approach to the regulation of algorithms, focusing on systemic oversight by an independent authority to balance innovation with societal interests and sustainability.
[To read the full article please follow this link.]
