Etické aspekty generativní umělé inteligence v oblasti syntézy obrazu

Generative models synthesizing images are currently extremely popular tools for an immediate and assisted creativity, not only among experts creating any form of visual representation, but also among the wider public, as their implementations such as DALL-E 2, Midjourney or Stable Diffusion are becoming more and more accessible. The rapid development of machine learning techniques and, consequently, the methods of generative artificial intelligence brings with it many ethical aspects and dilemmas that must be addressed in the interests of society and the shape of its future. The aim of this bachelor's thesis is to identify ethical aspects and risks of image synthesis as a subfield of generative artificial intelligence technology, and to answer research questions. Sub-goals are to define the term artificial intelligence and to clarify the principles of machine learning as a core method of generative artificial intelligence. The research problem of the practical part of this thesis is the existence of ethical risks associated with generative artificial intelligence in the field of image synthesis. The theoretical part is processed in the form of a literature review, and the practical part is done by the method of questioning, which results in the finding that the public is relatively well acquainted with the term generative artificial intelligence and the attitude towards existing ethical aspects and dilemmas is neutral / slightly concerned. The public perceives this technology as a potentially disruptive innovation, especially in the field of art and creative work.