The Masters and the Machine
See how design duo Marchesa created their first cognitive dress in partnership with IBM Watson.

A LOOK INSIDE MARCHESA
It’s so quiet at the Marchesa studio that you just might hear a pin drop — if, that is, anyone here were to actually drop a pin. The space is packed tight with designers, sewers, and pattern makers hard at work on the brand’s next breathtaking creation, and everyone is moving. The bustling workspace exudes a surprising calm. The manic energy we’re used to seeing in nearly every film and reality show about fashion — the mad dash, tense smoke breaks, and occasional tears brought on by impossible deadlines — none of that is here. These are professionals, and they don’t need the drama.



Georgina Chapman and Keren Craig lead the team. Since establishing the line in 2004, the duo has taken Marchesa into iconic territory on both the runway and the red carpet. Jaded fashion critics still let their jaws drop over Marchesa beading. A-list actresses turn to them in hopes of securing a spot on the “Best Dressed” lists. At this moment, though, Chapman and Craig are focused on creating a new type of dress that’s never been seen before, because it never could exist before. The piece that’s coming to life in their atelier today isn’t “only” a stunning garment, it’s a living, breathing, responsive work of art. It may be, quite simply, one of the smartest dresses ever made.



MEET IBM WATSON

Creating a very different kind of dress required a very different kind of partner. Enter IBM Watson, a cognitive system that uses natural language and machine learning to process “unstructured data.” In layman’s terms, unstructured data can be anything from scientific reports to social media posts, and Watson can churn through vast amounts of it to reveal otherwise hidden insights. According to IBM, the computers most of us are familiar with can only reasonably access and understand 20% of the world’s data. IBM Watson has us covered on the other 80%, and uses it to provide answers, ranked according to relevance, with supporting evidence.
You probably remember Watson from its 2011 Jeopardy! victory. Since then, it’s helped doctors and nurses make decisions about cancer treatment, partnered with apps like MusicGeek to improve music recommendations, worked as a recipe generator for Bon Appétit, and made it possible for customizable granola to be a thing. It even has thoughts on Game of Thrones character arcs. Now, Watson is turning its attention to fashion.
DATA MEETS DESIGN
When asked about meeting Watson for the first time, Chapman says, “It was a really thrilling experience and it lit up my mind with all the creative possibilities.” She half-jokes that after awhile “they literally had to ask me to leave!”
Excited by the new and endless possibilities, Georgina and Keren began to reimagine everything a dress could be. What if it incorporated fan feedback from past Marchesa designs? Better yet, what if the dress responded to fans in the moment? Soon, IBM Watson was providing Marchesa with insights on material, color, and something we don’t normally hear in the parlance of fashion, “fan voices.”
Chapman’s explanation of the process demonstrates what IBM Watson can do for all creatives: introduce new options and use data to suggest the best of those options.
What if the dress incorporated fan feedback? Better yet, what if the dress responded to fans in the moment?
“We fed [Watson] hundreds of images of Marchesa dresses, and it came back and gave us guidelines of what direction we should go in, which color stories we should take, what textures people respond to. So it was giving us guidelines, then using our expertise to know how to use what Watson gave us.”
To start, IBM Watson partnered with developer Inno360 to understand Marchesa’s criteria for fabric. Marchesa dreamed up the idea of a dress that communicates through pulsating light, so Watson processed data from millions of scientific reports on the compositions, amperage, weight and qualities of different fabrics — something that, let’s be honest, if designers had to do, would take the wind right out of fashion’s creative sails. The result? Watson found the material that best suited Marchesa’s needs.
To provide suggestions on color, Watson channeled its inner teenager and looked at hundreds of Instagram photos. It scanned posts from Marchesa’s feed to get a sense of the colors that captured the brand’s aesthetic, then helped Marchesa select five colors from their brand palette to be used for the responsive parts of the dress.
THE COLOR ANALYSIS


“Keren and I, we pain over color stories,” Chapman said. “We have millions of them in the room, so I tell you, that was something we were so thrilled about. When Watson was like ‘Here is a color story,’ we were like ‘Hooray.’”
“We always feel like we have quite a neutral color palette with a few pop colors,” Craig added. “But what Watson came back with is a collaboration that we wouldn’t have necessarily thought was quintessentially Marchesa. So, that felt like, ‘Oh, we can broaden our horizons.’”
DREAMS MADE POSSIBLE
Each year, the exhibit that takes over The Met’s Costume Institute sets the tone for the corresponding Met Gala, and this year’s theme is “Manus x Machina: Fashion in an Age of Technology.” There can be no more apropos an occasion for Marchesa and IBM Watson to debut their dress to the world, which they will do on the Met Gala’s red carpet.
“Manus x Machina” explores the two sides of fashion that exist today, man and machine. In one corner, we have the handmade detail of couture. In the other, we have clothes efficiently manufactured by machinery. But these two corners are moving ever closer together. Machines can now contribute to couture-quality design, with artful capabilities to laser-cut, splice, and embellish. The dress from Marchesa and Watson perfectly embodies this dichotomy. Instead of competing, they’re joining forces.

“Watson gives you information, but Watson itself isn’t designing the dress. And that’s what I liked,” says Chapman. “It demonstrated to me that this wasn’t removing my design process, it was enhancing my design process. It was cutting out the mistakes perhaps I would have made, and questions that would have taken me longer to answer.”
That assistance opens up new avenues of creativity. She continues, “It’s a way of using [technology] to enhance what we do. It’s being able to dream something. It brings a childlike quality back to it. When you were young and you’d imagine a dress that would do things that were a little otherworldly — suddenly it becomes a possibility.”
[Watson] wasn’t removing my design process, it was enhancing my design process. It was cutting out the mistakes perhaps I would have made, and questions that would have taken me longer to answer. — GEORGINA CHAPMAN
VOICES & LIGHT
When the dress makes its debut at the Met Gala, the magic that will get people talking, literally, is the incorporation of “fan voices.”
One of the most intricate steps of this piece’s production is the cutting and shaping of nearly 120 fabric flowers, each embedded with color-changing LED lights, all connected to IBM Watson. As the dress makes its way down the carpet, Watson will monitor Twitter and change the color of the dress to match the emotions being expressed by fans at any given moment, effectively making it a cognitive dress: the ultimate pairing of man and machine, of human ideas and computer insight coming together in a piece of clothing that can respond to the reactions it’s creating. For the first time, a dress can understand its audience.
Harnessing social media reaction to a design is especially relevant in a time when heavy-hitters like Proenza Schouler and Alexander Wang admit to designing for Instagram, and when collections go global via Twitter, Instagram, and Snapchat the second they hit the runway.
“In the world we’re in right now, it’s a much more democratic point of view to fashion,” Chapman said. “So to be able to get the feedback from our consumer and to aggregate that and get a sense of what people respond to, I think that’s incredible.”
BRINGING MAGIC TO LIFE
It could be said that there were actually three contributors to the creation of this dress: Marchesa, IBM Watson, and Marchesa’s fans. In addition to providing insights, Watson is bringing the brand’s customers into the fold from the very beginning — something that designers and marketing teams, until now, could only dream about.
Maybe most exciting for Marchesa, though, is that otherworldly quality that Watson enabled for the design of the dress.
“The idea that fabrics could change colors, or that lights can come on, it has this whole magical component and that’s what the aesthetic of Marchesa is about,” Chapman said. “It’s about dreams. It’s about magic. And incorporating this technology really brings that to life.”





THEBIG NIGHT
On the eve of the gala, amidst a blinding volley of flashing bulbs, model Karolina Kurkova lit up the red carpet wearing the long-awaited cognitive dress. Social media buzzed feverishly.



As Watson analyzed the reaction, the dress glowed with hues of rose, aqua, and lavender, reflecting the emotions of joy, excitement, and curiosity. It was unlike anything else seen at the gala, the grand finale to a parade of the world’s best dressed, and a garment that could only exist thanks to a groundbreaking collaboration. If this is a vision of the future, the future looks bright.
.
What Is a Sponsor Story?