Design By Algorithm; Will Robots Replace Fashion Designers?

As I welcome a new group of design students to my class this week, I find myself contemplating this question as the buzz around Artificial Intelligence heats up. StitchFix, a Silicon Valley-based subscription box service, already uses algorithms based on customer feedback to create the perfect blouse. Styles are broken down into variables––a sleeve, a lace trim, a body shape––that have proved successful, and these variables are reassembled a la carte into one desirable whole. The result of the kind of technology that has been common in the aviation industry, or that assembled the car in your garage, could soon be parked in a closet near you. So, do I need to send my students home, tell their parents to save on tuition?

The Darwinian Argument

“The machine algorithms are conceiving these styles,” Eric Colson, Stitch Fix’s algorithms officer explained to WWD. “The process is very much the same process Mother Nature uses, evolution by natural selection.” But he concedes that currently a human is still required to sign off on the final product.

If the customer-driven “see-now-buy-now” model eliminated many traditional steps of rolling out the latest styles, this ramps it up further, dropping the customer in the designer’s seat. “It’s really blending art and science,” said Colson, before loosely expanding,: “When I say art, I mean intuition; when I say science, I mean empirical decision-making. Either one has value on its own, the two combined are really interesting.”

First they came for our sewing jobs

The introduction of “Sewbots” from Chinese manufacture Tianyuan Garments Company, which is establishing a 20m dollar factory in Arkansas inside which a T-shirt will be manufactured every 22 seconds, and similarly, Sewbo, the Seattle-based start-up’s sewing robot, puts a whole new spin on bringing manufacturing home. Amid the nationwide struggle for minimum wage and the spotlight on unfair working conditions for underage, underpaid and exploited factory workers in the far east, these work-ready newcomers promise no need for bathroom breaks, days off, or heath coverage.

While the likes of Farfetch and Net-A-Porter have offered us a curated selection of the best of the season based on our browsing history, Stitch Fix’s hybrid assembly model offers us a more nuts and bolts approach and urges do-it-yourself. Currently it provokes as many questions as answers. Will it really lead to shopping nirvana? Will we look in the mirror after being dressed in a garment constructed from cold calculations and think, wow, I look fabulous? Dystopian images of a finger-wagging robot scolding, “you are not going out dressed like that” creep to mind. Then the voice of Scarlet Johannson from the movie Her, asking, “What were you thinking with that sleeve?” and, with an Edward Scissorhands swipe of a blade, I find myself bare-armed.

Style V Function; Man V Machine

For the customer who regularly shops fast fashion, who is not in the market for anything avant-garde, perhaps this service will satisfy their needs. They are being invited to reconfigure a template, and having a voice in the creation of their blouse may make them feel valued, catered to, in control. We could consider it a form of personalized uniform dressing. But there’s no accounting for taste in this efficient utilization of algorithms and data crunching, and what of that elusive but celebrated element we call “style”? Then there is human sentimentality for products imbued with character and flaws, like those 300-dollar jeans made from recycled denim that we love so much, that we’ve had mended multiple times, making them even more beautiful in our eyes. And who wants to think of themselves as so predictable that an entity with no capability to experience the emotion inspired by Dries Van Noten’s Fall runway spectacle can dress them? I hereby defy any robot to mimic Rei Kawakubo’s mercurial and brilliant grey matter.

“It’s all geometry, it’s curves, points and lines,” insists Francis Bitonti, chief executive officer of Studio Bitonti, another company developing machine-learning technology to apply to the world of apparel design. Yes, but it is the arrangement of those curves, points and lines that makes all the difference.

A Cut Above

As my students gaze back at me expectantly I choose to imagine the technology as a giant billion-dollar cookie cutter. Step forward, press a few buttons, and see it it dispense recognizable shapes in popular core ingredients while offering the sweet-toothed passerby an array of fixings of her choice from icing to sprinkles to jelly. Those online countertops will surely be stacked to toppling with “designs” created by this sticky-fingered insatiable clientele. But there is another shopper who will walk on by, barely glancing at the virtual line forming around the block, who is on a quest for the more thought-provoking fashion, the challenging untraditional shapes that are a result of drape and sculpture, the highly crafted textiles, the artisanal workmanship that will be increasingly spotlit simply because it has been touched by human hand and not materialized from bits and bytes. The genius of innovative cutting, in utter defiance of all that algorithms reveal, will continue to be lauded. And so I continue to push my students to fearlessly create what isn’t out there, and strive for the unexpected, the as-yet nonexistent. Revolutions in the classroom don’t tend to be shiny, mechanical, industry-shifting; they can simply amount to a new way to cut a sleeve.

By contributing guest editor Jackie Mallon, who is on the teaching faculty of several NYC fashion programmes and is the author of Silk for the Feed Dogs, a novel set in the international fashion industry.

Photos: Stitch Fix Facebook, Sewbo.com.