en Algorithmic Societies, Medium

Ethics in AI is worthless without action

Publicado originalmente en Medium – Manuchis

Last week I went to Lyon (France) to teach at the EM Lyon Business School. The course we created with David Solans about Discrimination and Bias in was for the MSc in Health Management & Data Intelligence. The course covers not only basic knowledge on discrimination concepts, EU regulation and data protection, but also algorithmic fairness, critical theory such as sociotechnical analysis, decoloniality and intersectionality, as well as basic knowledge of human-computing interaction to tackle human oversight.

This integral and multidisciplinary course was a great opportunity to give students a broad view beyond Ethics on AI. Students from this course came from different disciplines, from engineering, medicine or business, and all are expected to design, develop, manage, or implement AI systems in the context of healthcare.

The problem of current ethical approach to AI is twofold. First, creates a distance from real use of algorithmic systems and the conceptual discussion about what is ethical for society. This does not help to bring solutions to discriminatory practices or concrete complex systems where algorithms are inserted. Second, does not contest their own epistemologies, leaving discrimination and privilege as something objective or inexistent from themselves.

Our course was meant to not only assess an algorithmic system but to take a course of action that will help approach it with a critical point of view and with the necessary technical and non-technical tools to mitigate the harms of a system.

Ethics in AI is a good approach to discuss the values and beliefs from society regarding this (disrupting?) technology. Nevertheless, I consider that any developer, designer, manager or researcher should have more tools to work with. Again, it is like discussing autonomous driving with the trolley problem. It is a nice exercise, but it is far from reality.

Raise awareness and teaching what todo, why to do and how to do, was one of the founding principles of Algorithmic Societies. We are glad to expand such contents to the whole society. We are keen to present our material and share our courses to everyone.

You can read the syllabus and materials will be published soon.

Escribe un comentario

Comentario

Este sitio usa Akismet para reducir el spam. Aprende cómo se procesan los datos de tus comentarios.

  • Contenido relacionado por Etiqueta