Meet Berenson, the Robot Art Critic

Could a bescarved, top hat-wearing robot do the job of a human art critic? That’s the question curators at Paris’s Musée du quai Branly pose with Berenson, a robot critic that roams the halls of the institution’s recently-opened exhibition “Persona: Oddly Human.”

While the museum specializes in African, Asian, and Oceanic artworks, the exhibition called upon contemporary anthropologists to explore the complicated relationships we have with objects.

Berenson—who is named for the late critic Bernard Berenson, once widely accepted as the foremost authority on Old Masters—was designed by robotics engineer Philippe Gaussier and anthropologist Denis Vidal. The well-dressed robot expresses his opinions about the art on display not in writing, but with a simple smile or frown—a decision he reaches based partially on the reactions of other people.

Berenson is able to analyze the facial expressions of fellow museumgoers, sort them into “negative” and “positive” categories, and feed the information into a neural network simulator called Prométhé. The program allows him to use this data to develop an overall impression about the object. If it is positive, he smiles and moves towards it. If it is negative, he frowns and moves away.

It’s a far cry, of course, from the kind of nuanced analyses made by most art critics, who strive not to regurgitate popular opinion, but do just the opposite. The bot is, nevertheless, a fascinating (and frankly, pretty adorable) experiment.

Perhaps in the future, robot art critics will judge robot art, much like the sketching robot critic Jonathan Jones mocked in a recent column in the Guardian. After all, neither robot art nor robot art criticism seems to make much sense to us humans, at least for now.