AI Machine Learning in Creative Architectural Design Processes

An increasing number of designers, architects, and artists are using new technologies and artificial intelligence as mediums, aware of their contemporary relevance. Architectural design is based on technical knowledge, accumulated experience, and an intuitive “creative” component able to express the designer's unique talent. Artificial Intelligence provides suggestions and ideas to designers, acting as a collaborative partner in their creative phase. From this perspective, AI has been identified as a good support, able to suggest ideas during conceptualisation, one of the most critical phases of the Design process. This paper explores the integration of AI and Machine Learning style transfer in architectural education to enhance creativity and design inspiration. It comes as the result of three years of experimentation in an academic environment where Master's students in Architecture have been asked to apply and understand Neural Networks and methods to train them to perform specific given tasks. During the numerous experimentations, an interesting evolution has been observed across the three years when users approached this innovative Design tool in architecture. AI-informed design processes have been a good practice of “learning by doing”, breaking the rule about having just one solution in architectural design processes. Machine Learning applied to creativity requires a correct understanding of the input and a good level of abstraction during the phase of interpreting the results. This paper also aims to inform the audience about the workflow of creative AI applications and the philosophy of “Creating with Artificial Intelligence.”