Autonomous Weapon Systems: Understanding Learning Algorithms and Bias
On 5 October 2017, there was a discussion about “Autonomous Weapons Systems: Learning Algorithms and Bias” at the United Nations Headquarters in New York.Kerstin Vignard, Chief of Operations and Deputy to the Director at UNIDIR raised a question about algorithm bias, or subjectiveness of the data. Cathy O’Neil said the defining what is a successful result of an algorithm is the most important. David Danks specified the topic into autonomous weapons systems. As the coding system did not reflect international laws and conventions that are related to war and human rights, it is not clear to define what is success in algorithms. The last pannal, Amandeep Gill, emphasized the importance of understanding the way an autonomous weapon is programmed and why the weapon responds and reacts in a particular way. It is because the user cannot surely expect a moral outcome of the weapon.
APPA UN NGO Intern & Young Research Fellow
University of Illinois-Urban Champaign