By Wendy K. Leigh
So what does a Seattle data scientist with a Ph.D. in Material Science and Engineering have in common with highly creative visual artists working in oils, mixed media, and watercolors? Quite a lot, as it turns.
Aaron Lichtner, a 2015 graduate of the University of Washington, has developed a program called Artorithmia that allows a user to browse hundreds of visual 2D artworks and then “like” specific pieces that appeal to them. Following that, the program recommends artwork based on that person’s unique and personal taste.
In a nutshell, Artorithmia software uses a highly selective system to build a profile of the user, based on five foundational categories: saturation, composition, technique, price, and content. The collection of algorithms pinpoints specific features about the “liked” art, and clusters them into groups using agnostic features.
Lichtner, who is also an artist and musician, uses his deep knowledge of statistics, analytics, data visualization and more to produce recommendation that are startling accurate for most users. To explain the process in terms that “the rest of us” can understand, he compares it to similar user interfaces already in mainstream use.
“The way I try to explain Artorithmia is usually that it’s Pandora for art,” he says. “The same way you tell Pandora the name of a band or song you like and it returns music based on your choice, Artorithmia does that for art.”
When you “like” a piece of art, he notes, it recommends other pieces based on how similar they are, and shows this by making the more similar pieces larger in size. The technical part comes in how Lichtner calculates similarity between the pieces. As far as practical application, he’s collaborated with another Seattle startup, Drizl, whose goal is to bring original art into everyday homes and to support the artists themselves.
Drizl believes that buying art should be fun, and that artists, galleries and buyers should be able to come together in more accessible ways – minus the traditionally intimidating and expensive gallery settings. Using custom visualization tools, they let anyone preview specific pieces of art in their own rooms by uploading photos of the rooms, walls and furniture in their home, then use measurement tools, filters and features that “place” the art on those same walls. It’s also possible to contact the artists directly through Drizl; invite them to suggest art for a specific room; and even commission original art.
Lichtner catalogued about 700 of the pieces of art in Drizl’s collection, and used his algorithms and software to explore the application in interactive art buying. A user can go to Artorithmia, explore the collection and find a piece that they love, and then get linked to Drizl where they can buy it, explains Lichtner. But his aspirations for the technology go even further.
“I would love to make the technology available for other recommendation systems. What if it could recommend wines or beers based on other flavors you like? Or suggest museum itineraries by you selecting artworks that you like the most?”
For now, both Artorithmia and Drizl are classic Seattle startups evolving in the midst of the city’s creative super-smashup of technology, inspiration and just plain boldness, with an undercurrent of social impact. They have in common the goal of making traditional, contemporary, organic, folk, local and international art accessible to everyone – while also ensuring that artists have the support they need to continue making art.
If innovators such as Lichtner want to extend that technology, drive and curiosity to the “art” of wine, beer and energizing the sometimes tedious museum experience, well, more power to them. And to us all.