DO ANDROIDS ENVY CARAVAGGIO?

Robbie Barrat is among the youngest digital artists and has rightfully gained the title of the wonder child of the digital art world. Firstly, because of his young age (he started as a teenager) and secondly, because of the innovative techniques he employs in his works – over the last two years, Barrat became the first artist to use AI technologies to create visual works.


In October 2018 Christie’s New York made headlines for selling the first AI-generated painting during Prints and Multiples sale, for a whopping $432,500. The piece that fetched this price was Portrait of Edmond Belamy – a work referencing the style of an oil sketch by an Old Master. Somewhat blurry and pixelated, the face of the titular Edmond Belamy seemed to slowly disintegrate, making the specific facial features difficult to discern. The portrait was one of a series of works depicting a fictional family of Belamy, created by a Paris-based collective Obvious (Hugo Caselles-Dupré, Pierre Fautrel and Gauthier Vernier), which used an algorithm run by an artificial intelligence programme with a melodious name GAN (‘generative adversarial network’). The collective described their artistic process to Christie’s as ‘feeding’ the program with images of portraits coming from 14th to 20th century and then, GAN creating an image based on the selection seen, which is then compared to the man-made examples but as the program is far from perfect at this point, it is more prone to not noticing the differences which results in image distortions.


While the trio of the artists from the collective got immediate credit for the first work using AI technology sold at auction, the actual algorithm supporting the AI program GAN which they have used has been created, and shared for free use on the internet by Robbie Barrat, Dazed reported, who received no credit in the work. Portrait of Edmond Belamy contains an algorithmic equation in the bottom right corner but to a layman, it bears no obvious difference between a work of an AI and a joke meant to disguise the identity of the painting’s maker. Even though in an interview with Dazed Barratt claimed that the idea of open-source access to knowledge about artificial intelligence was how he learned about the possibilities of those programs in the first place, in 2020, in an interview with Ruby Boddington from Its Nice That, he was less enthusiastic - “The group that sold that work was not an artist group, they were a marketing group. I know that sounds like an attack but if you look at early articles that they've written, they refer to themselves as a start-up. Honestly, they really did take advantage of the project that I put up, even though it was open source.” Speaking to Its Nice That, Barrat was also very mindful of the misconceptions surrounding the images generated using AI technologies and how the narrative produced by Obvious contributed to the misconception that they are created ‘by a computer’, which seemed to completely remove the concept of artist’s agency but despite those experiences, he remains positive about the role of open-source tools for digital artists.


Nudes, generated with AI, Robbie Barrat, 2018

While digital art is commonly associated with the medium utilizing high resolution and crystal-clear colour contrasts, fully embracing the possibilities of the digital to produce sharp and saturated images. Barrat’s style is rather contradictory to that. One of his first-ever made series employing the AI, was a series based on thousands of images of landscapes sources from WikiArt, among which the experiments with distortions of the Post-Impressionistic works proving to be particularly visually pleasing. The next step was applying the same approach to the topic of the female nude. In this case, the varying proportions and ways of depicting the human body over the centuries have yielded interesting but rather surprising distortions, turning the body into an undefined, elongated, genderless form of flesh. Those works, which according to Barrat were produced very early in his career and were unrefined, were perhaps what stimulated Obvious to aim for similar results in relation to the distortion of the facial features. Now living and working in Paris, Barrat seems tired of being continuously associated only with the AI and has plans to branch out further, into sculptural work using 3-D holograms.