Mondal, Manuel Lalanne, Denis 2021-10-01 The objective of this document is to provide a non-technical introduction to the field of Artificial Intelligence. It is intended for anyone outside of the data science community curi-ous about the subject, such as researchers from other areas looking for new instruments, legal professionals confronted with automated decision making algorithms or legal tech tools, en-trepreneurs interested in expanding their products into new areas, etc. No prior knowledge of the domain is required for this introduction. External resources with technical details and additional examples are provided in the footnotes for the interested reader. In the following sections, we introduce the fundamental concepts in the field, aim to clarify some commonly used terms and explain some limitations of modern AI techniques. First, we distinguish two types of AI: the symbolic approach, consisting of systems encoding rules and explicit knowledge, and the machine learning approach, which uses statistical methods and models to automatically infer relations between data and tasks to solve. In the subsequent section, we distinguish two complexity levels of these tasks. Finally, we introduce the subfield of “Explainable Artificial Intelligence” and distinguish three classes of explanations. application/pdf https://folia.unifr.ch/global/documents/309537 https://folia.unifr.ch/documents/309537/files/IWP21_02_diuf.pdf eng http://doc.rero.ch/record/333278 info:eu-repo/semantics/openAccess License undefined Artificial Intelligence info:eu-repo/classification/udc/004 Demystifying Artificial Intelligence http://purl.org/coar/resource_type/c_18ws