Creativity of Neural Networks: Risks and Opportunities for Modern Designers

Abstract

In connection with the large-scale development of artificial intelligence in all spheres of human activity, the article discusses the problems and prospects for the use of neural networks in the work of professional designers. The aim of the study is to identify risks and find optimal directions for the use of neural networks in graphic and fashion design based on the advantages of artificial intelligence. The objectives of the study include the analysis of the work and types of artificial neural networks, the search and selection of neural networks for this study, the development of queries for neural networks in accordance with the goal, the analysis and formulation of strengths and weaknesses in the work of neural networks for their application in the field of design. The basis of the functioning of neural networks is the principle of constructing a unique image based on the selection and generation of a large number of ready-made images uploaded to the database, created by professional designers from around the world. Since the neural network is an adapted biological model of the neural network, it is capable of remembering its own errors in the process of work and their subsequent correction, that is capable of learning. The neural network is trained using special machine learning methods. The following conclusions can be formulated as prospects for using neural networks for design. The lack of the ability of neural networks to cultivate their own author’s style and the impossibility of developing innovations in design does not allow to completely replace human labor. Work with the use of neural networks will be more efficient due to the intensification of solving typical design problems and freeing up time for the designer to create and improve new ideas.