Ai for Generative Design – as seen in 2019: The geometries presented, resembling familiar forms like a chair, were generated using a 3D Generative Adversarial Network (GAN). GANs operate through the interaction of two neural networks: a generator and a discriminator. The generator creates outputs, while the discriminator evaluates them, distinguishing real from generated data. Over time, the generator refines its designs, evolving from random noise to meaningful, structured forms. This process underscores the potential of AI in design by automating the iterative creation and refinement of shapes, enabling rapid prototyping and exploration of extensive design possibilities. While this research into GANs reflects the state of technology in 2019, more advanced generative algorithms have since emerged, offering unprecedented levels of realism, complexity, and practical applications in the field of 3D geometry.

University of Innsbruck, Department of Structure and Design (ISD), 2019