REFLECTIONS ON ARTIFICIAL INTELLIGENCE ALIGNMENT WITH HUMAN VALUES: A PHENOMENOLOGICAL PERSPECTIVE

Authors

Shengnan Han, Stockholm UniversityFollow
Eugene Kelly, New York Institute of TechnologyFollow
Shahrokh Nikou, Åbo Akademi UniversityFollow
Eric-Oluf Svee, Stockholm UniversityFollow

Abstract

The need for a systematic approach to work with artificial intelligence (AI) is current and rapidly growing. It is important that Information Systems researchers get ahead of public sentiment and be able to provide proactive commentary about the current state-of-the-art, as well as solutions for future systems. One critical question is how can we ensure value alignment between AI and human values through AI operations from design to use? For the purposes of this discussion, we adopt the phenomenological theories of material values and technological mediation to be that beginning step. In this paper, we firstly analyze the AI phenomenon from selected resources from the top IS research outlets (basket of 8 journals and 5 AI journals in IS). Secondly, we briefly present what are material values and technological mediation and reflect on the AI value alignment principle through the lenses of these theories. Supported by these new understandings and reflections, we propose to build a common principle of human values to understand the AI value alignment principle through phenomenological theories. The paper contributes the unique aspect of material values to the discourse within current AI research.

Recommended Citation

Han, Shengnan; Kelly, Eugene; Nikou, Shahrokh; and Svee, Eric-Oluf, "REFLECTIONS ON ARTIFICIAL INTELLIGENCE ALIGNMENT WITH HUMAN VALUES: A PHENOMENOLOGICAL PERSPECTIVE" (2020). In Proceedings of the 28th European Conference on Information Systems (ECIS), An Online AIS Conference, June 15-17, 2020.
https://aisel.aisnet.org/ecis2020_rp/92