Humanoid Models, Smarter Merchandising, and AI With Taste
Noetix’s robot takes to the catwalk in Paris; AI tools tackling fashion’s $140B overproduction problem; Zelig’s styling engine blending data with intuition.
Noetix’s Humanoid Robot Walks the Paris Runway
Beijing-based Noetix Robotics sent its humanoid robot, N2, down the catwalk at the UNESCO venue in Paris on October 8, marking the first appearance of its kind outside China. Originally slated to collaborate with a Chinese designer during Paris Fashion Week, the partnership collapsed over funding, so N2 instead modeled vintage looks while performing acrobatics for the crowd. The showcase follows a surge of humanoid robotics activity in China, fueled by a $137 billion national tech fund and rapid industrial development across chips, sensors, and AI.
Why it matters: Though N2’s performance might look like another fashion stunt, it points to a deeper convergence between fashion and robotics. For humanoids to evolve beyond showpieces, they need large volumes of “3D” embodied data, precisely the kind that movement-heavy industries like fashion could generate. As AI blurs into physical form, fashion’s flair for spectacle could inadvertently help train the next generation of robots.
Charlie Campbell
Fashionista ask if AI can Tackle Fashion’s $140B Overproduction Problem
A new wave of AI tools is promising to fix one of fashion’s most stubborn issues: excess inventory. From WGSN’s Opportunity Calculator and Assortment Builder, which help brands like Coach and Adidas predict demand and optimize buys, to startups like Flagship and Depict, which tailor store layouts and search experiences in real time, AI is being woven deep into fashion’s planning and merchandising backbone. Meanwhile, designers from Norma Kamali to Moncler are testing generative models to refine aesthetics and anticipate audience tastes, as fast-fashion giants like Shein push automation to dizzying scale.
Why it matters: AI could finally help brands produce what people actually want - no more, no less. If used well, it promises fewer markdowns, smarter assortments, and a measurable dent in the $70-$140 billion of unsold stock clogging warehouses and landfills. But, it also risks supercharging the same overproduction it’s meant to cure, just faster and with better data!
Zelig Teaches AI the Rules of Fashion
Zelig, founded by MadaLuxe co-founder Sandy Sholl, is reshaping fashion e-commerce with an AI platform that can generate thousands of mix-and-match outfit combinations in seconds. Designed as a “Build a Look” experience for retailers, Zelig simulates how fabrics drape and garments layer in real time, creating high-fidelity virtual outfits that mirror real-world fit. Its first major rollout with REVOLVE tripled session times and conversions while cutting returns by double digits—proof, says Sholl, that “we’re making progress on one of fashion’s biggest sustainability issues.”
Courtesy of Zelig
Why it matters: Unlike most AI styling tools built for consumers, Zelig is a B2B platform that merges taste and data at scale. Human stylists train its large language model to understand proportion, texture, and occasion, turning “AI with taste” into a retail intelligence engine that decodes not only what people buy, but how they 'style'. For fashion’s digital future, that blend of creative intuition and machine precision could be the key to personalization.