Computer generated artificial intelligence (AI) is having an effect on virtually every human activity in advanced society. But although it permeates the workings of the art market, in research and marketing for instance, it is still in the early stages of development in relation to the making of art.
Last year, Christie’s sold its first example at auction – an algorithmic composite of historic portraits created by Obvious, the Paris collective, for a 45-times estimate of £337,000. Whether this was a case of novelty factor will be tested next month when Sotheby’s offers its first AI work at auction.
Memories of Passersby 1 is two video screens of subtly changing male and female portraits devised by Mario Klingemann, a leading artist, estimated at £30,000 to £40,000. The AI market, though, is tiny. It has arrived and is unlikely to disappear, but the technical demands of maintenance and operation still worry average art collectors. Most of the creative action is going on beyond the glare of the auction rooms and commercial galleries.
Next month, Jake Elwes, one of the brightest new talents in the AI art firmament, will be acknowledged with a commissioned project at the Zabludowicz Collection in north London. Anita Zabludowicz is one of the leading contemporary art collectors in Britain, known for the admirable way in which she supports and exhibits young artists.
Elwes studied fine art media at the Slade School, where he became the first student to code computer programmes to make art. He had considered studying computer science and philosophy at Oxford but, coming from an artistic background, decided to develop his creative instincts at art school instead.
These instincts were probably in his blood. His great grandfather, Simon Elwes, was a go-to society portraitist in the Twenties, much favoured by the Queen Mother.
Jake’s father, Luke, is a painter too. Starting out with portraits, Luke has since embarked on an intriguing journey as a landscape painter. Eschewing photographic realism, he has captured the essence of places from Bungle Bungle in Australia to, most recently, the River Ganges.
Luke recalls that, at school, Jake struggled because of dyslexia, but developed remarkable computer skills. Slade professor David Burrows says: “Jake was singular in my time in teaching in working with AI. Importantly, he managed to combine his interest in design and AI with an interest in aesthetics to produce engaging work.”
At his degree show in 2017, Jake exhibited three videos made entirely by computers that he had programmed or coded. One visitor was Steve Fletcher, who had been running a gallery specialising in new media art. Fletcher says he immediately recognised Elwes’s work as, “exceptional… Not only was the work technically interesting, but it was interesting to look at.
“It showed a deeper understanding of technology than most other artists working with AI, not least because he had done the coding himself and brought an aesthetic sensibility that produced arresting, seductive images. In one piece, Latent Space, for instance, he set the computer up to roam without direction, almost in a dreamlike state.”
While their methods are different, Jake’s latest project is not so far removed from his father’s. The family has stayed for many years on the Essex coast, on Osea Island and at Landermere, which are populated in season by migrant birds – curlew, oyster catchers, waders and lapwings. Elwes used machine learning to search for images of these marsh birds.
“The computer didn’t know what a marsh bird was, but it managed to select the most evocative looking, and compiled composite images of them,” says Elwes, who then placed them in the mud and filmed them, projected on a perspex screen, capturing a slow interplay between nature and its artificial relatives.
How much of the end product is the work of algorithms or curated by Elwes may not be easy to tell, which is what makes this kind of AI art special.