What might appear like an unfinished art is in fact a portrait created by Artificial Intelligence, and it’s all set to become the first of its kind to be ever publically auctioned. It is to be sold at Christie’s in New York, marking the first time a leading auction house has dealt in art made by a computer software. The painting is the brainchild of a Paris-based trio of 25-year-olds named Hugo Caselles-Dupré, Pierre Fautrel and Gauthier Vernier, who are on a mission to show the public that algorithms can be creative. They used a two-part algorithm, called a Generative Adversarial Network. It was fed with a data set of 15,000 portraits painted between the 14th and 20th centuries, and the generator portion of the algorithm then started creating works.
“We wish to emphasize the parallel between the input parameters used for training an algorithm, and the expertise and influences that craft the style of an artist,” said Caselles-Dupré. “Most of all, we want the viewer to focus on the creative process: an algorithm usually functions by replicating human behavior, but it learns by using a path of its own.”
The portrait up for auction is of one of 11 members of the fictional Belamy family, who are named after Ian Goodfellow, a researcher who invented the GAN method in 2014. “Goodfellow” roughly translates to “bel ami” in French. One portrait in the series called “Le Comte de Belamy” was sold to Parisian collector Nicolas Laugero-Lassere earlier this year for around $12,000. The newest AI painting is estimated to sell for $10,000 when it goes under the hammer in October this year.