Unlike human judgment, machines are not affected by emotions in their decision making. This can make them seem superior in the sense of being more logical or rational, which is often seen as more accurate or reliable than intuition or “gut feeling”, especially when making important decisions in business or public policy, for instance. However, while it appears to be deceptively easy to delegate the decision process as much as possible to machines, several issues suggest that this is not advisable. Besides the matter of obtaining accurate, or even decent, data to produce useful algorithms (a discussion we will leave for another time), considerable challenges arise when delegating the responsibility for a decision to a machine.
Arguably, AI offers great potential in several areas ranging from health care and logistics to sustainability, to name a few. But how are we to think about the ethical challenges concerning AI and responsibility? I invite you to listen to some of my thoughts on the topic in this podcast (in Swedish), where I am interviewed alongside Fredrika Ling, Head of Design at TietoEVRY and Therese Svensson, Data Science & AI Ethics Solution Specialist at IBM.