Analyzing the relationship between artificial intelligence supported visual thinking and creative cognitive processes in early design phase

Visual thinking skills play an active role in the early design phase, where productive thinking is intensified, increasing the potential to produce creative products in the design process. Understanding cognitive processes and investigating design behaviours has become more important to design disciplines with the acceleration of AI studies. In the last decade, the growth of artificial intelligence applications for generating images from text, together with artificial neural networks and generative adversarial networks, has signalled the beginning of a new era in architecture and design. With the accessibility of these applications, the use of visual analogy, which is also effective on cognitive processes, has intensified. The aim of the study is to show the effect of AI-supported visuals, produced in accordance with the design problem in the early design phase, on the designer's cognitive process using analytical methods. In the study, two different sessions were conducted with a working group of 8 people and sketches were produced. During these sessions, the participants' sketch productions and thinking aloud were recorded and evaluated using the protocol analysis method. The segments were matched and coded according to the four main action categories created by Suwa and Tversky (1998). In addition, the linkography method developed by Goldschmidt (2014) was also used. Within the scope of the study, design moves were associated with each other and a portrait of cognitive processes was created in the context of two sessions. As a result of the study, differences in cognitive thinking processes were revealed. Furthermore, significant differences were found in terms of critical moves (CM) and link indexes (LI) related to productive thinking, which is stated to positively affect creative thinking.