Meet the Indians using artificial intelligence to create art
They feed data sets into special algorithms to create canvases; they’re showing at art events around the world. Elsewhere, AI art is even ending up at auction.Updated: Mar 31, 2019 17:44 IST
When Rembrandt was 26, he created one of his first masterpieces, The Anatomy Lesson of Dr Nicolaes Tulp, an oil-on-canvas that showed medical professionals studying the human body, at a time (the 17th century) when medical science was still in a nascent stage.
In 2018, Bengaluru’s Harshit Agrawal, also 26, was inspired by Rembrandt’s painting to create something path-breaking of his own. He fed images of surgeries into a software algorithm and used artificial intelligence (AI) to create art based on the images. The result was abstract crimson canvases that could be interpreted as clouds, flowers or fibroids.
“While Rembrandt was addressing the power of medical science in his art I wanted to talk about the fascinating world of AI,” says the artist and human-computer interaction researcher. He first began tinkering with AI while studying at the Massachusetts Institute of Technology (MIT) in 2015. This series of works titled The Anatomy Lesson of Dr Algorithm, and was showcased at Gradient Descent, a first-of-its-kind AI art exhibition held at Delhi’s Nature Morte gallery in August.
The work went on display again, yesterday, for a second edition of Gradient Descent, in Bengaluru, organised by Nature Morte and curated by an art collective called 64/1. Also on display are the works of AI artists Mario Klingemann from Germany, Jake Elwes and Anna Ridler from the UK and Memo Akten from Turkey.
In January, artist and sculptor Sahej Rahal, 30, displayed his latest creation, Juggernaut, at Mumbai’s Chatterjee & Lal. It’s an AI-driven audio-visual piece inspired by video games. The video traces the journey of a droid as it floats through an alternative civilisation, constantly evolving and building itself up by picking up totems and ruins of pillars to self-create a multi-limbed sculptural installation.
Around the world, artists are exploring the interface between art and artificial intelligence, to create works, some of which are even being auctioned. The New York auction house Christie’s sold the AI-generated print, Portrait of Edmond de Belamy, by Paris-based art collective Obvious, for $432,500, last year. Sotheby’s followed suit and sold the two-screen installation Memories of Passerby I by Mario Klingemann for £32,000 earlier this month.
“We have only just begun exploring AI-driven art as a genre in India and there is much scope for it to grow,” says artist Raghava KK, who is also co-founder of 64/1. “I have always thought of myself as a cyborg and there are a couple of others like me who find working with AI an extremely exciting prospect.”
Since 2014, Raghava has been using different forms of technology, from drawing on iPads and tapping into brainwaves, to create art. “Now the computer science department of Ashoka University also plans to include AI-driven art in their curriculum,” he says.
Agrawal’s first tech-based art project was The Flying Pantograph in 2015, which used a pen-carrying drone to draw on a vertical wall. “The motion dynamics of the drone and the software intelligence added a new visual language to the art,” he explains. The art from that first experiment ended up being, essentially, a scribble. But this got him interested in how human and machine could come together for creative expression, which prompted him to design a software program called Tandem in 2016.
Tandem uses an algorithm that lets the computer imagine and suggest a visual output. A roughly drawn figure of trees can be interpreted as dogs; the human artist interfacing with the AI can also add filters of emotions like happy, sad or dreamy, which the computer will interpret in the form of hues. Agrawal’s AI art created through Tandem has been exhibited at AI art exhibitions in South Korea, Germany, Austria, the US and Australia.
His latest work, The Anatomy Lesson…, was created after he fed 60,000 images of surgical dissections into a similar algorithm. “The feedback I’ve got for this work has been as exciting as the creative process,” he says. One observer told him a painting reminded him of the interiors of a cave; a woman said she felt it celebrated womanhood.
ONES & ZEROES
Artist Baiju Parthan says one thing holding back AI art in India is the large expense involved in creating your own algorithm.
While Agrawal had considerable assistance from the MIT Media Lab, Rahal used his rudimentary programming knowledge and a free game development program called Unity to create Juggernaut. “After several messy rounds of programming, I managed to write a code that lets the droid perceive its space, choose objects around it, overcome obstacles and create an end product,” he says. The programme produces new sculptural forms every time the simulation is run. “My art is heavily inspired by the video games I’ve played all my life. This has a videogame feel but a deeper subtext of evolving time, history and cultures.”
Artists have been using technology for a long time, and it was inevitable that AI would be part of this process as well, says art critic and curator Girish Shahane. Artist and curator Bose Krishnamachari, co-founder of the Kochi-Muziris Biennale, disagrees.
“I would like to think that AI-driven art is a temporary phase. Internationally, it has become a big trend and while I appreciate it, I don’t see it as something that is here to stay or is a big game-changer,” he says.
Meanwhile, with a degree from IIT-Guwahati and a Masters in media, arts and science from MIT, Agrawal is keen on researching ways of merging science and art, but his mother is struggling to understand how her engineer son can also be an artist. “She was taken aback by the surgery images and it took her a while to understand how tech programs can help create art,” he says, laughing. “But she is slowly opening up to how both the fields can meet and I hope so can rest of the art world.”