“People don’t judge an artwork on its own merit,” said Aidan Meller, the director of a prestigious U.K. gallery called, well, the Aidan Meller Gallery. “They always do it through the lens of the artist. When you look at a Picasso or you look at Tracey Emin, you have a very different response based on the personalities that created [the work]. That’s no different with a robot.”
It’s not the kind of conversation you might expect to have with an art expert and dealer whose previous big splash of publicity came, in 2014, when he helped bring to light seven large cardboard boxes of drawings found in a loft, which turned out to be highly significant Pre-Raphaelite School sketches. But somewhere, it seems, Meller’s trajectory changed course.
“[A few years ago] I was playing with my son, and he was building a robot out of Lego,” he told Digital Trends. “It was an epiphany moment. I picked up his robot, and I thought, ‘I wonder — I just wonder — are we at the point where the creativity of technology is such that you could actually create an artist?”
The robot artist goes on tour
Jump forward several years and, working with computer scientists from the University of Oxford, a stone’s throw from his Oxford gallery, and roboticists from the U.K.’s University of Leeds, among others, Aida was born. Or, at least, switched on. Aida — or, to use the proper styling, Ai-Da; a portmanteau of A.I. and the name Ada, borrowed from the world’s first computer programmer — is a humanoid robot. It (or she?) is also a painter, sculptor, performance artist, and poet. These things, Meller insists, are not mutually exclusive.
“Art is too subjective and irrational to ever be a fully agreed upon measure.”
Aida has traveled around the world, exhibited work internationally on a scale that most conscious artists could never dream of, and, after a coronavirus-induced grounding for much of 2020, is currently gearing up for an impressive multicountry tour in 2020 — taking the robot from Taiwan to Istanbul to Abu Dhabi to New York to Silicon Valley to London, and probably a few more in between.
Unveiled in 2019, Aida’s original skillset involved painting abstract pictures. This is achieved by using a camera to look at real-life objects, which are then run through assorted machine learning algorithms, before the instructions to re-create them are fed to a mechanical arm. Since then, like an Amazon Echo receiving new skills, the number of artistic mediums Aida can work in has continued to blossom: Writing and reciting poetry, creating clay sculptures, the performance art of being, frankly, a robot doing things.
Meller makes no pretense of the fact that Aida has any consciousness or feels emotion. If the Pulitzer Prize-winning Tracy Kidder was talking about The Soul of a New Machine back in 1981 (with regards to the now-humble Data General Eclipse MV/8000 32-bit microcomputer), there is no suggestion of such a soul for Aida.
But that doesn’t mean Aida cannot be creative, Meller argued. He cites Professor Margaret Boden, one of the first, and certainly best, writers to discuss A.I. and creativity in-depth. “Creativity is something that is new, surprising, and of value,” Meller said.
The meaning of creativity
He defines it thusly: “When [we’re] doing the programming, it’s got to be new code, it can’t be library stuff. It’s got to be surprising; we don’t know exactly what Aida is doing, what Aida’s going to create. And of value? Well, in light of the 900 publications [that have been published about the project], and everything else that goes with it, we would say that people are putting a lot of value on this new area of creativity.”
“When you look at a human artist, consider the factors that are affecting their ability to paint: The teachers, the influences, the inspirations, the technical know-how, the parental background, the social context. All of these things go into making a human artist capable of creating an artwork in front of them. Aida is, on many levels, no different.”
All three points are, of course, open to scrutiny. Are the best artists always novel? Is being more novel of higher value than showing ability within a certain area? Is value equal to the number of articles published about the artist during their life or the amount that their work sells for? None of this is to say that Meller is wrong. But it does highlight one of the problematic aspects of creativity as a benchmark of artificial intelligence (not that this is necessarily what the endeavor is designed to prove.)
Simply put, art is too subjective and irrational to ever be a fully agreed upon measure. The pioneering filmmaker and theorist Sergei Eisenstein may have suggested that the “impression produced by art” could be turned into a useful quantifiable measure of its value (he suggested the term “attraction” as a single unit), but the reality is that this is never going to cut it. Two people can view a piece of art and come away having drawn entirely different conclusions about its value. Unless we are to live in a world where the measure of art’s value is its financial value at auction or the box office, there is simply no way to rank value in an objective manner that everyone can agree upon.
Meller also, correctly, brings up the thorny point about human creativity and the degree to which it ever exists in a vacuum. “When you look at a human artist, a conscious human artist, when they sit down to draw paint, [consider the factors] that are affecting their ability to paint: The teachers, the influences, the inspirations, the technical know-how, the parental background, the social context,” he said. “All of these things go into making a human artist capable of creating an artwork in front of them. Aida is, on many levels, no different.”
Asking the right questions
The point, though, is not to replace human artists or even, really, to answer questions about tricky areas like consciousness. It’s to raise questions, and open up new possibilities for the future of A.I. in an area that we have, over the years, viewed as quintessentially human. (An area which, as the astronomical sum paid for an A.I.-created painting a couple of years ago, is already being severely paradigmatically shifted.)
“We do see this as a new area within the art world,” Meller said. “Just like when the camera came in in the 1850s and people started to use it. I’ve got a whole number of newspaper articles that brought up how this would be the end of the artist, the end of painting. The camera [was] here, and it’s all going to be over. And obviously that didn’t happen; it became just another tool for the artist. We definitely see this new technology, specifically machine learning, as a new area of exploration for artists to grapple with.”
As to the response from the general public? Meller said that, while people are used to the idea of robots performing repetitive tasks, the notion of a machine creating something still takes many by surprise. “We’ve heard incredible responses,” he said. “Both ‘Aida is the future, it’s terribly exciting’ to ‘Oh, my gosh, destroy her. It’s awful. You shouldn’t be doing that.’”
If the goal of art is to provoke a reaction, then maybe Aida really is an artist, after all.
Editors' Recommendations
- ChatGPT AI chatbot can now be used without an account
- Reddit seals $60M deal with Google to boost AI tools, report claims
- AI chatbot goes rogue during customer service exchange
- OpenAI and Microsoft sued by NY Times for copyright infringement
- 2023 was the year of AI. Here were the 9 moments that defined it
