Refashioning research through AI: Even faster fashion

Fashion is predicated on perpetual change, requiring infinite designs to feed an insatiable appetite for consumption. Traditional research methods of garnering visual trends data require labour-intensive, analogue approaches-[1]. Outcomes predict a visual 'snapshot' limited to a point in time, for designers to make decisions about future product. Informed guesses are open to misinterpretation, exposing the industry to overstock, deadstock, poorly designed garments, ending up in outlets or landfill-[2]. AI analysis of trends research enables designers, manufacturers and consumers to make real-time economic decisions and sound ethical choices, whilst improving sustainability. In wider academic contexts AI can open new opportunities for fashion research in real-world contexts around inclusive design, healthcare, and impact of virtual try-ons on customer satisfaction-[3] and body esteem-[4]. GenAI in the fashion industry is expected to contribute $275b to operating profits in the next five years-[5]. Its image and animation-based capability from learned patterns offers an attractive, creative platform to the industry. If channelled as a co-design tool, it can facilitate research processes between individual's ideas and machine learning to accelerate design and production methods. Universities are beginning to explore the potential of GenAI in textile design to create new archetypes for students to harness AI responsibly. Norwich-Nottingham Trent-UAL collaboration with Liberty typifies a new research paradigm-[6]. GenAI tools such as Firefly, NewArc and MidJourney can be leveraged by designers as high-fidelity prototyping tools and create knowledge-enhanced models. However, they raise issues around intellectual property, copyright infringement and potentially foster homogenisation which stifles design originality-[7] which could be counter-intuitive to intention. This presentation discusses AI/genAI in fashion contexts, its potential and limitations as a design-research tool. It demonstrates a PhD student's early 'AI fashion generator' prototype which explores how innovative implementation of AI technology can enhance research effectiveness of trends, decision-making and consumer preferences and invites audience comment.