AI, Norms, Big Data, and the Law

This is an overview article regarding artificial intelligence (AI) and its potential normative implications. Technology has always had inherent normative consequences not least due to AI and the use of algorithms. There is a crucial difference between algorithms in a technical sense and from a social-science perspective. It is a question of different orders of normativity - the first related to the algorithm as a technical instruction and the second to the consequences springing from the first order. I call these last-mentioned norms algo norms. These are embedded in the technology and determined by the design of the AI. The outcome is an empirical question. AI and algo norms are moving targets, which call for a novel scientific approach that relates to advanced practice. Law actualizes primarily for preventive reasons in relation to negative aspects of the new technology. No major regulatory scheme for AI exists. In the article, I point out some areas that raise the need for legal regulation. Finally, I comment on three main challenges for the digital development in relation to AI: (1) the energy costs; (2) the singularity point; and (3) the governance problems.