Berto Jongman: Melanie Swan on Machine Ethics

Ethics
Berto Jongman
Berto Jongman

Machine Ethics Interfaces

Melanie Swan

Machine ethics is a term used in different ways. The basic use is in the sense of people attempting to instill some sort of human-centric ethics or morality in the machines we build like robots, self-driving vehicles, and artificial intelligence (Wallach 2010) so that machines do not harm humans either maliciously or unintentionally.

This trend may have begun with Asimov’s Three Laws of Robotics. However, there are many different philosophical and other issues with this definition of machine ethics, including the lack of grounds for anthropomorphically assuming that a human ethics would be appropriate for a machine ethics, beyond the context of human-machine interaction.

There is another broader sense of the term machine ethics which means any issue pertaining to machines and ethics, including how a machine ethics could be articulated by observing machine behavior, and (in a Simondonian sense (French philosopher Gilbert Simondon)) how different machine classes might evolve their own ethics as they themselves develop over time.

There is yet a third sense of the term machine ethics – to contemplate human-machine hybrids, specifically how humans augmented with nanocognition machines might trigger the development of new human ethical paradigms, for example an ethics of immanence that is completely unlike traditional ethical paradigms and allows for a greater realization of human capacity.

Machine ethics interfaces then, are interfaces (software modules for communication between users and technologies (machines, devices, software, nanorobots)) with ethical aspects deliberately designed into them. This could mean communication about ethical issues, user selection of ethically-related parameters, ethical issues regarding machine behavior, and ethical dimensions transparently built into the technology (like a kill switch in the case of malfunction). Machine ethics interfaces are the modules within machines that interact with living beings regarding ethical issues, pertaining to the ethics of machine behavior or the ethics of human behavior

Definitions:
Machine Ethics: 1) (conventional) technology designers attempting to incorporate models of human-centric morality into machines like robots, self-driving vehicles, and artificial intelligence to prevent humans from being harmed either maliciously or unintentionally, 2) any issue pertaining to machines and ethics, 3) the possibility of new ethical paradigms arising from human augmentation and human-machine hybrids.

Machine Ethics Interfaces: Interfaces (software modules for communication between users and technologies (machines, devices, software, nanorobots)) with ethical aspects deliberately designed into them. This could mean communication about ethical issues, user selection of ethically-related parameters, and ethical dimensions transparently built into the technology (like a kill switch in the case of malfunction).

Reference: 
Wallach, W. (2010). Moral Machines: Teaching Robots Right from Wrong. Oxford, UK: Oxford University Press.