Generative Pre-Trained Transformer for Design Concept Generation: An Exploration
This addresses the need for better generative tools in design innovation, but it is incremental as it applies existing GPT models to a new domain.
The paper tackled the problem of generating design concepts that are neither too abstract nor too detailed for early-phase exploration by exploring the use of GPT-2 and GPT-3 for natural language design concept generation, with both models showing reasonably good performance.
Novel concepts are essential for design innovation and can be generated with the aid of data stimuli and computers. However, current generative design algorithms focus on diagrammatic or spatial concepts that are either too abstract to understand or too detailed for early phase design exploration. This paper explores the uses of generative pre-trained transformers (GPT) for natural language design concept generation. Our experiments involve the use of GPT-2 and GPT-3 for different creative reasonings in design tasks. Both show reasonably good performance for verbal design concept generation.