Confounding Complexity of Machine Action: A Hobbesian Account of Machine Responsibility

In this article, the core concepts in Thomas Hobbes’s framework of representation and responsibility are applied to the question of machine responsibility and the responsibility gap and the retribution gap. The method is philosophical analysis and involves the application of theories from political theory to the ethics of technology. A veil of complexity creates the illusion that machine actions belong to a mysterious and unpredictable domain, and some argue that this unpredictability absolves designers of responsibility. Such a move would create a moral hazard related to both (a) strategically increasing unpredictability and (b) taking more risk if responsible humans do not have to bear the costs of the risks they create. Hobbes’s theory allows for the clear and arguably fair attribution of action while allowing for necessary development and innovation. Innovation will be allowed as long as it is compatible with social order and provided the beneficial effects outweigh concerns about increased risk. Questions of responsibility are here considered to be political questions.publishedVersio