I love to shop as much as the next girl, but a recent hunt for a dress to wear to a friend’s downtown-L.A. warehouse wedding left me spent. I was thinking something midi length, fitted without being too revealing, maybe floral? Glassy-eyed, I scrolled through page after page of options online, then I stormed laps through the mall. All the while I thought: This shouldn’t be so hard.
In the future it won’t be. Within a decade artificial intelligence is poised to reinvent the shopping landscape. Think of a computer or device that knows not only your budget but also your measurements, what’s already in your closet, what events are on your iCalendar, and what’s trending on the runways—and uses that information to formulate the kind of nuanced suggestions that we didn’t think a machine would be capable of. In other words, it might very well be able to pinpoint what you should wear better than you could.
We bring you the trends. You make them your own. Sign up for our daily newsletter to find the best fashion for YOU.
Stitch Fix already has an algorithm shopping model that selects clothes for customers (with the help of a team of human stylists) sight unseen. (While the company doesn’t disclose the success rate of its customer-to-clothing matches, the fact that it has 2.7 million customers should be some indication of the quality of its service.) And Amazon’s Echo Look uses a mix of humans and AI to power its styling services. The device has a hands-free camera that takes pictures of your outfit, blurs out the background, and, in the accompanying smartphone app, can catalog your entire wardrobe into style categories (it placed a chambray dress of mine in a “fancy casual” collection).
The app’s Style Check feature also allows users to submit images of two outfits and ask the Look which is better. Right now the verdict comes in about a minute from Amazon’s algorithms and a team of fashion experts, with a brief explanation. But Linda Ranz, director of Echo product management, says that will shift: “Over time, the machine will get smarter and smarter.” When I submitted a question, I was told that one outfit was the winner because “the colors work better together.” Frustratingly vague, but you can see the potential.
There are some areas in which computer precision could quickly best humans, namely when it comes to sizing. Lumping customers into groups—medium and large, say, or even a 10 or a 12—has always been a flawed system. Stitch Fix data places people on a more exact spectrum (maybe you are a smaller-chested 12 or a long-legged 8), and Stitch Fix chief algorithms officer Eric Colson envisions a day when we will have our dimensions on a scannable chip, allowing us to find the perfect fit every time.
Still, there are some instances where it’s hard to imagine a computer having an advantage over the human mind, in all its complexities. For example, a machine could understand that you need a dress to wear to a wedding but not the subtlety of what kind of dress you’d want to wear to your ex-boyfriend’s wedding. Only a human would know what that means, Colson says: “She wants to look good.”
There are still plenty of skeptics about this technology, who (rightly) point out that fashion is an art form and also a means of self-expression—things no computer can replace. As with all technological advancements, the key is figuring out how to use it in our lives. For instance: when you’re frantically looking for a dress for a wedding and your ex-boyfriend isn’t the groom.
Elizabeth Holmes, formerly a style reporter for The Wall Street Journal, has also written for The New York Times, InStyle, and Elle.