IMPLEMENTING ETHICS DIALECT:A NEURO-FUZZY SYSTEM FOR FINE-TUNING ETHICS

摘要

It is a common practice to regard machine/thing/device/software functioning disconnected from the moral/ethical/religious/legal/cultural/regional parameters that may also be intertwined with its functioning. However, such a standpoint on machine functioning may lead to situations where core human values could be compromised e.g. serious physical, mental, emotional or psychological harm to human beings or even loss of life. This is particularly true in case of modern pervasive technologies like the Internet of Things that increasingly make use of embedded artificial intelligence. To address these concerns, we propose a neuro-fuzzy system that represents ethics (moral/ethical/religious/legal/cultural/ regional/management policies) relevant in the context of a given technology (machine/thing/ device/software operation) and also adapts (fine tunes) them for appropriate ethical behaviour. References [1] W. Wallach and C. Allen, Moral machines : teaching robots right from wrong. Oxford University Press, 2009. [2] S. Sholla, R. N. Mir, and M. A. Chishti, Teaching Manners to Machines : A method for ethics-aware technology, Patent App No. 201711001967, Indian Patent Office, 2017. [3] S. Sholla, R. N. Mir, and M. A. Chishti, Taming technology: A fuzzy logic based method for incorporating ethics in the Internet of Things, Patent App No. 201711008760, Indian Patent Office, 2017. 4.E. H. Mamdani and S. Assilian, An experiment in linguistic synthesis with a fuzzy logic controller,///. J. Man. Mack Stud, vol. 7, no. 1, pp. 1-13, 1975. 5.C. von Altrock, Fuzzy Logic and NeuroFuzzy Technologies in Appliances, Ing. Udea, 2001.