[Featured image: SS24 Copenhagen Fashion Week, Ganni backstage, shot by and belonging to @thestreetland.]
Sign up to Interline Insiders to receive our weekly analysis in your inbox every Friday.
- AI has made its way into fashion weeks, but not in the form many expected – with notable brands holding back from using AI tools to design collections, and instead using bespoke models to provoke discussion.
- Despite this, the trend is likely going to be towards more widespread use of generative AI in creative areas, with a strong commercial mandate for fashion to move beyond back-of-house intelligence and efficiency.
- Instead of directly challenging the nature of human creativity, or established copyright law, fashion must now take a more pragmatic, sensitive, and considered approach to bringing AI and creativity closer together.
At the final show of Copenhagen Fashion Week Spring/Summer 2024 last week, which The Interline was on-site for, Danish brand Ganni explored how generative artificial intelligence (AI) could be used in creating immersive experiences. The brand presented an AI-infused runway, where guests had the opportunity to engage with Ganni-trained AI chatbots, nestled in living tree installations. The trees (all native to Denmark) were scattered throughout the location, encouraging attendees to speak to them and ask questions.
A major theme was the connection between the virtual and the natural world, as captured in the “HELLO WORLD” title of the show. This plays on “Hello, world!” which is the common refrain programmers use when learning a language or an environment – printing that text snippet is often task one in coding handbooks – while also hinting at Ganni’s core mission of innovation and sustainability. To note, Copenhagen Fashion Week currently has 18 minimum sustainability requirements that must be met in order for designers to showcase, including that each collection is “at least 50% either certified, made of preferred materials or new generation sustainable materials, upcycled, recycled or made of deadstock” and that “set design and show production is zero waste.”
With sustainability already firmly on the agenda for the event, Ganni co-founder and creative director Ditte Reffstrup set out to create “a kinder, more thoughtful AI” because “recently, it feels like something else to panic about.” This is perhaps a glib summary of a wide-ranging panel of existential problems, but it’s still an accurate one: at a time of fear and uncertainty, AI can indeed feel more like a threat, or the latest volley in the ongoing war between capital holders and workers, than an opportunity.
So, with the help of digital artist Cecilie Waagner Falkenstrøm, the show embodied “the spirit of the AI” in a physical way using something almost antithetical – nature. Reffstrup was fascinated by the idea of incorporating AI into the show and anthropomorphising it. Notably, the AI chatbot was trained on data that Ganni gathered from its community, making it proprietary and more likely to remain on-brand than an off-the-shelf language model would have been, but a move that was no doubt a substantial financial investment to create and then to maintain.
Crucially, it was also never the intention for Ganni’s actual collection to be generated by AI – something that continues to raise problems around originality and copyright. This is an important milestone, since it directly addresses one of the fashion industry’s biggest apprehensions when it comes to AI: the idea that it will be used to ingest, remix, and dilute human creativity, de-emphasising the importance of originality in design.
This is something The Interline has written about before, around the time that the industry appeared to be blithely running towards “AI Fashion Weeks,” and Reffstrup’s statement indicates that the same concern has now worked its way into the wider fashion industry. Right now, the overwhelming sentiment around AI seems to still be one of apprehension, even fear, in the fashion industry and so far, not many brands have used it in their final, physical product.
This combination of the fear of diluting human creativity, and the extremely murky copyright situation around AI training and authorship, has meant that most brands (some notable exceptions excluded) have stuck to using AI in more prosaic ways – as a tool for deriving insights from large amounts of data, fuelling new approaches to trend forecasting, and targeting hyper-personalisation.
But AI is one buzzword from the cloud of new terms floated in the last year or two that won’t go away anytime soon, possibly unlike its peers, the NFT and the metaverse. This week BoF reported on the high frequency of the terms “AI”, “artificial intelligence”, and “machine learning” on transcripts from 2022 and 2023 of more than 30 public fashion, activewear, and retail companies. During earnings calls, fashion company leaders have increasingly been drawing attention to AI technology, indicating its growing adoption rate, or at least its position as a long-term player that survived its current hype-cycle. And why wouldn’t it – AI means money, right? Making it and saving it. According to McKinsey, the application of AI technology has the potential to elevate operating profits within the fashion and luxury sectors by a substantial $275 billion within the forthcoming years.
The story continues: “Farfetch and Revolve have more recently joined in and been some of the most likely to mention AI. On its May 2023 call, Revolve executives referenced “AI” two dozen times, discussing how they’re using it to improve their product recommendations and search capabilities but also its use for an AI-generated billboard campaign and the company’s involvement in the first-ever AI fashion week.”
These, notably, are very different ways of framing the use of AI than the chatbots and backend insights that characterise the way it’s currently being deployed. These go straight for the jugular of creativity, in fact.
But in a way, this sort of head-on challenge to fashion’s fear of AI might be necessary. For the industry to really change its perceptions around AI, something more visible and on a broader scale will need to change. Something that doesn’t position it to only mean AI-generated clothing, or; the cold and distant optimisation sense that executives have come to extol.
Instead, fashion is likely going to start using AI for displays of fun, enjoyment and immersion, showing that it isn’t just a spectre, lurking out of sight, waiting to steal the next tranche of jobs. AI can fit in with the established practices around creativity in the fashion industry, without letting it take over entirely. While there are undeniably risks associated with AI in terms of encroaching on artistic liberty and capacities – when managed adeptly, it becomes a complementary tool that can be used to understand a brand on a deeper level.
And this is especially true of brands that are not usually featured in the same sentence as AI or any technology, who might feel like it is a step away from their ethos to incorporate into their collections or messaging. For AI to be appearing on the catwalk in what is one of the most well known fashion weeks (behind Paris, Milan, New York, and London) is an indication that the gap between digital and physical fashion in the mainstream is closing, and we could see more shows and presentations to come looking like Digital Fashion Week New York. The event, which has grown considerably since its inception in 2021, invites guests to: “take in a digital fashion exhibition, enjoy a layered runway show with digital and physical fashion, and dance to a virtual DJ” for its September 2023 offering.
And in the advent of AI going mainstream, a new set of challenges inevitably arises. Currently, AI-generated images cannot be directly passed on to the manufacturing line. Designers and model-makers, who are knowledgeable about fabrics and patterning, must find a way around this in the years to come.
If this concept continues to evoke apprehension, be reassured: AI is far from being a perfect substitute for strong design. If anything, it will grow to be an essential component within the creative toolkit: allowing for more abd better moodboards for amping up creativity, and pushing those with vision and conviction towards excellence. It’s a new creative field that is opening up. And in the same way photography was seen to signal the demise of painting, in reality, all it did was steer painting into new directions. Some of them made people a lot of money; others helped people make better art.
The best from The Interline:
Mark Harrop asks whether true digital twins could be the key to unlocking sustainability, customer experience, streamlined product design, and supply chain transparency and efficiency.
We talk to Helene Behrenfeldt of Infor about the changing definition of PLM and how broad supply chain visibility can be focused into product-level impact assessment and management.