ARTIFICIAL INTELLIGENCE IN ARCHITECTURE AND DESIGN: NEW DESIGN METHODS AND THE TRANSFORMATION OF THE PROFESSION

Abstract

The article examines the impact of contemporary artificial intelligence (AI) tools on architectural and design workflows, as well as the transformation of creative and educational methods resulting from their integration into professional practice. Special attention is given to the interaction between the designer and generative, parametric, and language-oriented algorithms. Purpose. To analyse how AI tools transform design methodologies and educa- tional practices, and to determine their influence on decision-making and the devel- opment of spatial thinking. The study focuses on the results of an educational exper- iment involving the Finch 3D system. Methodology. The research is based on an analysis of scientific publications, a comparison of widely used AI platforms, and the implementation and evaluation of a real educational experiment. The methodology combines practical testing, compara- tive analysis, and pedagogical observation. Results. Generative models proved effective at the conceptual stage, enabling rapid verification of ideas and exploration of multiple visual and compositional solu- tions. Parametric systems, particularly Finch 3D, provide automated generation of spatial layouts, accelerating work with variations and helping students better under- stand spatial logic. The educational experiment demonstrated that comparing stu- dents’ own solutions with algorithmic ones enhances critical thinking and improves the quality of design decisions. The study also highlights the growing shift toward language-based interfaces (“vibe-coding”), where interaction with the system occurs through the articulation of intentions and descriptive prompts. Scientific novelty. The study clarifies the role of AI in shaping a new mul- ti-variant model of design interaction, in which the designer works not with a single solution but with numerous algorithmically generated alternatives. The concept of “vibe-coding” is proposed as a promising form of communication between the de- signer and the system, based on semantic formulations rather than technical com- mands. Practical relevance. The findings may be applied to updating educational cur- ricula, developing recommendations for integrating AI into professional practice, and creating effective methodologies for working with AI-based design tools.