Abstract:

For Hans Jonas, the „serious side of science fiction“ lies in its ability to encourage reflections about future realistic and ethically relevant scenarios [1]. According to him, science fiction narratives are thought experiments, that can initiate and guide ethical reflections, so that based on them, (potential) problems of a specific field of action can be anticipated and considered ex ante. This hypothesis is applicable to the field of artificial intelligence (AI). Following Jonas, our assertion, therefore, is that science fiction narratives can and should be used in the education of persons who are responsible for AI developments. The reason is twofold.

First of all, in media offerings of AI, problems regarding these technologies are illustrated. This demonstration enables the viewers to critically reflect possible ethical problems even before people have to deal with in reality. In contrast to, for example, the combustion engine, whose negative consequences for the environment could not be reflected before its introduction, science fiction narratives allow us to reflect the consequences of AI critically beforehand [2]. This concerns, among others, the interaction between humans and AI, possible manipulations of AI, or the issue of responsibility for different operations of AI. In other words: science fiction enables us to move from a technical Epimetheus to the position of a Prometheus.

Second, medial presentations of AI that surround us can be defined as „world view generators" [3] which decisively shape our image of such technologies. This can lead to unrealistic opinions of such technologies because, from a scientific point of view, presentations of AI in science fiction narratives often contain certain flaws, which nevertheless influence people‘s mindsets: dystopic scenarios, anthropomorphizations of AI and unrealistic depictions of AI (at the moment or soon). That is why some take the view that science fiction narratives „won‘t save us“ [4]. While this may seem to be a reason not to use fictive narratives for ethical reflections, we support the opposite: Such depictions can be used in scientific contexts to enable students to reflect their concepts and expectations of AI critically. Science fiction narratives can, hence, function as counter-images of real AI developments to get a more adequate and realistic picture of reality and fantasy. This irritation then could be used as basis of several ethical reflections.

In our presentation, we want to elaborate on these ideas using the digital game "Detroit: Become Human." We use the game to exemplarily show which role science fiction can play in the training of future developers of AI (or people responsible for it) concerning the reflection of the current and future state of AI and the potential ethical problems resulting from it.

References:
[1] H. Jonas, The Imperative of Responsibility. In Search of an Ethics for the Technological Age. Chicago/Illinois: The University of Chicago Press, 1985. P. 30.
[2] T. Ramge, Mensch und Maschine. Wie Künstliche Intelligenz und Roboter unser Leben verändern, Stuttgart: Reclam, 2018. P. 81-82.
[3] M. Rath, „Medienwirkungsforschung in Deutschland – eine Annäherung“ in Buchwissenschaft und Buchwirkungsforschung (D. Kerlen and I. Kirste, eds), pp. 89-98, Leipzig: Institut für Kommunikations- und Medienwissenschaft, 2000.
[4] I. Hermann: „Terminator won't save us." https://www.hiig.de/en/terminator-wont-save-us/, 2018.

Keywords:

Digital games, media ethics, AI, artificial intelligence, media pedagogy.