JWT and Photographer Jimmy Nelson Are Fighting Cultural Homogenization With AI

The Preservation Robot debuted at SXSW

The Preservation Robot will capture images of indigenous groups and help promote them in search results. JWT
Headshot of Marty Swant

The internet has rapidly proliferated western culture around the world, but one photographer wants to use artificial intelligence to heighten the online presence of some of Earth’s most remote people groups.

On Saturday at South by Southwest, photographer Jimmy Nelson and JWT-Amsterdam unveiled a technology that will use AI to both aggregate—and disseminate—photos of indigenous groups in hopes that they will be better represented online. The project, called The Preservation Robot, has an ambitious mission: to reverse the homogenization of humanity on social networks and search engines.

Using the Preservation Robot, Nelson and the team at JWT-Amsterdam have begun to collect and upload images from professional and amateur photographers from around the world. Using a software script heavily influenced by search optimization efforts, the robot will be able to improve search rank of the images based on key words and new events while uploading them to social networks like Reddit, Twitter and Imgur.

Having spent decades traveling around the world photographing people groups such as the Marquesans in French Polynesia, the Sadhus in India and the Q’eros in Peru, Nelson said he wants to improve cultural diversity. Along with professional images, he said teams will distribute analog cameras to indigenous groups themselves in order to provide an opportunity to have a more active role in documenting themselves.

Nelson said indigenous people should be better celebrated, adding that the cultures that are lesser known add just as much aesthetic value as western celebrities and models with money.

“It’s about a balancing of knowledge and balancing of cultures,” Nelson said. “I’m not saying these indigenous cultures should remain isolated and naked. I’m saying bring your heritage with you into the future, don’t abandon it. If you could imagine a yin-yang-esque balance of our technology, our advanced way of living, but it’s not the only way of living.”

Housed in black box with four screens and rounded edges, The Preservation Robot measures just 12 inches tall, 10.5 inches deep and 10.5 inches wide. On the top is a speaker, which enables a voice-activation tool similar to mainstream voice assistance devices such as Amazon’s Alexa or Google’s Assistant. Upon command, the box will be able to show where images are being uploaded around the world.

The project in some ways an evolution of a previous initiative called The Next Rembrandt, which JTW-Amsterdam created several years ago to teach an AI to paint in the style of the renowned Dutch painter Rembrandt van Rijn. The project was a high-profile example of the intersection of creativity and technology can intersect, and as a result won the Grand Prix award at the Cannes Lions festival in 2016.

Bas Korsten, who created The Next Rembrandt, is also leading development of The Preservation Robot. He said JWT wanted to create a way that would use the power of the internet by using APIs to automate what would otherwise be a tedious process. And while the project isn’t client work, it has helped to inform the way they think about innovation within the agency.

“The main learning is to do stuff that nobody asked for that take you to places you haven’t gone before,” he said. “Otherwise you’ll stay in the realm of what people expect from you…You need to push yourself as an agency to adopt that and to work with those technologies and to learn from them to see if they can be relevant.”

But how do JWT and Nelson plan to project The Preservation Robot from being used by hackers or trolls that might find a way to game the AI to upload and spread malicious images such as fake news or obscenities—much like what happened several years ago when Microsoft debuted a chatbot on Twitter that started off well-meaning before becoming a sponge for hate speech? Korsten said the team—which is also working with adding facial recognition—is starting slowly and manually.

“It’s up to now to us to develop and build this and to make it smarter and to work with facial recognition so we grow that group of people contributing to this,” he said. “This is a good starting point.”

So why make a physical “robot” in the first place in a box when everything exists in the cloud? Kev Mayo, chief technologist at Amsterdam-based production company SuperHero CheeseCake, said the tangible device helps people to understand how it works.

“The concept is so grand, and dealing with the internet and storage is a little bit ethereal,” Mayo said. “Really having a manifestation of the concept was really important creatively, and I think from day 1 we always had this idea of a sculptural, tangible presence for the project. And then also, it’s sort of the anti-robot. It has no dance moves, no moving parts. It’s meant to sort of embody some sort of working intelligence—the idea that there is something working behind the scenes.”

@martyswant martin.swant@adweek.com Marty Swant is a former technology staff writer for Adweek.