Speaker
Prof.
Emma Ruttkamp-Bloem
(University of Pretoria)
Description
The first of the 2018 EU guidelines for ethical AI states “AI should not trample on human autonomy. People should not be manipulated or coerced by AI systems, and humans should be able to intervene or oversee every decision that the software makes”. The thinking behind this guideline is the Kantian notion that humans have intrinsic worth, and should therefore always be the best final arbiters in considering the impact of decisions on humans. Are such views justified in an age where humans are not necessarily the only rational beings anymore, and also do not have the best history of beneficial moral decision-making? In the age of 4IR, we need to rethink our reasons for framing guidelines such as the above, especially, I argue, if they are built on assumptions of notions of personal identity giving humans the upper hand. In order to commence this ‘rethinking of humanity’, I consider the arguments of writers such as Asaro, Moor, Allen and Wallach for identifying levels of artificial morality as a way in which to formulate a positive contemporary response to the well-known 1991 article by Cole in which he defends the potential mental abilities of algorithms with his notion of ‘virtual persons’.
Primary author
Prof.
Emma Ruttkamp-Bloem
(University of Pretoria)
