Novelty and subjectivity in architectural design using interactive genetic algorithms: assessing designers’ and assessors’ perceptions

In theory, Interactive Genetic Algorithms (IGAs) offer a unique opportunity to integrate human intuition and subjective judgment with algorithmic optimization, heightening creative potential in architectural design. However, despite claims linking IGAs to creativity, the relationship between the two has not been empirically tested. This research has two aims. First, to determine if, relative to the performance of a non-interactive genetic algorithm (GA), designers perceive IGAs as enhancing their level of engagement and alignment with their subjective aims. Second, to determine if, on average, IGAs support the production of more novel designs than GAs. To respond, this research reports the results of a two-part empirical study comparing designers’ and assessors’ perceptions of contextual novelty in morphologies produced using GAs and IGAs. The designers (n = 26) undertook a series of four design experiments and completed a post-experiment survey/interview about novelty and subjective engagement levels in their works. Their designs (n = 104) were independently evaluated by a panel of expert assessors (n = 68). The results compare designers’ and assessors’ perceptions of creativity indicators. The results, which suggest that interaction may contribute to generating novel design outcomes and support creative exploration, also reveal differences in perspectives and biases affecting design evaluations.