Dima Durah
Oct 7, 2024
Algorithmic Knockoffs and Digital Déjà Vu: GEN AI's Fast Fashion Moment
We live in a world where every creative spark risks instant cloning and mass production. Designs are copied and mass-produced off the runway within hours. AI content generation mirrors every aspect of the fast fashion frenzy. Pixel and Large Language models flood the internet with imitations, endlessly replicating trends. The result? A tsunami of low-quality outputs designed for fleeting or little to no consumption—filling up the landfills. Originality is the first casualty of this hydra-beast.
Fast fashion has been criticized for its focus on quantity over quality, exploiting creativity without fairly compensating those behind it. AI is walking the same path, taking inspiration from artists, writers, and musicians without securing their consent or giving them credit, producing for the sake of production. The industry has been so focused on rapid development that it’s neglected the crucial aspects of creation—the creator and intended use!
The "3Cs": Consent, Credit, and Compensation. Established by the Cultural Intellectual Property Rights Initiative® (CIPRI) in 2017 provide a framework that should inspire AI devs. It’s simple: creators should be asked for permission (Consent), recognized for their contributions (Credit), and fairly rewarded for the use of their work (Compensation). Building on in a recent article Karen Smiley outlines how the Algorithmic Justice League (AJL) introduced a fourth "C": Control. This addition emphasizes that creators should have the power to decide how their works are used in AI models. The "4Cs"—Consent, Credit, Compensation, and Control—offer a path for AI devs to shift away from the fast fashion mindfuck.
So why should AI learn from the fashion world? Because the numbers show consumers are increasingly choosing quality over quantity. Despite the economic downturn, data shows that luxury fashion brands have seen a 6% larger growth in orders year over year compared to fast fashion. In Q1 2023, while overall fast fashion sales fell by 2%, the average order value for luxury fashion increased by 8%, driven by a preference for high-end, well-crafted pieces. READ THAT AGAIN.
The takeaway is clear: people are willing to invest in things that are thoughtfully made. AI can learn from this shift. Adopting the "4Cs," means the AI industry can produce tools to generate content that is crafted with care, intention, and ethics at its core.
In part one of this series, we talked about how Prada’s latest show at Milan Fashion Week exemplifies this shift. Prada dove deep into their archives and reimagined old pieces, demonstrating the importance of the mark of the creator—something algorithms struggle to replicate. This celebration of authorship and the past’s influence on the present is a good blueprint for how AI can evolve: by valuing provenance and the mark of the creator.
I’m not proposing making “luxury AI tools”, what I am proposing/encouraging/urging AI devs to think about the is impact of their creations and the cost to the commons. Afterall, the endless flood of AI generated content is poisoning the collective well and AI imodels themselves in the process.
Living Assets is at the forefront of this movement, we help creators preserve their mark on their work. We are living in a time where AI can generate endless content. so the question becomes how can the original work stand out? I think we have a golden opportunity to learn from the pitfalls of fast fashion and pivot towards a future where quality, originality, and ethical standards are hardcoded into our tech.
Maybe instead of cutting off its head, this hydra beast can be tamed? If the fashion industry can evolve and leave its "fast" roots behind, so can GEN AI.
