“AI: Trust as a Challenge of the Digital Age” – ConsiousCoders at the UN in Geneva
A delegation of ConsciousCoders hosted a side event at the Group of Governmental Experts on Emerging Technologies in the Area of Lethal Autonomous Weapons Systems at the United Nations in Geneva.
Our contribution to the GGE 2019 deals with the topic of ‘Trust in the digital age’ and aims to highlight various facets of trust in emerging machine learning and artificial intelligence systems from an interdisciplinary and ‘young generation’s perspective. Three presentations from members of ConsciousCoders explored whether and how our norms of trust change with the recent development of AI systems.
1. Integrity of cyber intelligent systems (Florian von Keller): This talk deals with the threat of external manipulation the integrity of cyber intelligent through its underlying training data. This process of “data poisoning” can be used to change the expected results and to undermine the reliability of systems as well as the trust we put into them.
2. Agency of AI systems (Louis Longin): This talk engages with the question of whether AI systems should be understood as self-sustaining agents which informs the adequate ethical and societal treatment of such systems. The talk highlights various philosophical viewpoints from a narrow to a broad notion of agency and finally argues for a legal-inspired middle ground treatment of AI systems.
3. The (cyber)threat of public misinformation (Eva Charlotte Mayer): This talk addresses the ongoing application of tools to manipulate online information which could allow for dissemination of false information. It examines how propaganda and misinformation can be spread and outlines research in the field of techniques to more efficiently persuade people over social media websites.