Algorithmic painters are starting to sell and be exhibited in museums.

Algorithmic painters are starting to sell and be exhibited in museums.

Surrounded by her paintings, in a wood-paneled room at Oxford University, Ai-Da slowly looked at them, examining them one by one. “Robotic art is a wonderful thing,” she says, speaking with pauses in which she chooses her words carefully. Like many other artists, she wants her work to provoke, to spark heated debate. But unlike them, his words do not feel passion, the burning creative fire. When he speaks, his face also remains expressionless, and his gaze is always glassy and seems to pass through the interlocutor. It does not betray any thoughts or feelings. But that’s normal for an android for now.

In the Shadow of the Sphinx

Built by a team of engineers specializing in robots with human-like features and using a neural network developed by two researchers at Oxford University, Ai-Da is presented as “the world’s first robot artist”, an implementation of artificial intelligence (AI). which blurs the line between the car and the creator of art; a vision of a future that suddenly becomes part of our present. Equipped with facial recognition technology, Ai-Da sees the world around it with the help of high-performance video cameras implanted in the eyes. The images taken by them are analyzed in real time by a series of algorithms, which then transmit instructions to his robotic arms, thus allowing him to create his own interpretation of the raw material of inspiration. His art includes drawings, paintings and sculptures and has sold over one million dollars in total so far.

Last month, one of his sculptures was to be featured in the first ever art exhibition at the Great Pyramid of Giza. The sculpture in question, a three-legged self-representation of the robot artist, is a modern interpretation of the ancient riddle of the Sphinx: – when he is a child, he crawls on all fours, then gets to his feet, so that in his old age he can walk with a stick.)

Ai-Da and his work were sent to Egypt by plane, stored in special containers. However, he was barred from entering the country, and his high-tech equipment aroused the suspicions of border guards, who saw a potential risk of espionage in his built-in modem and in the cameras in his eyes. Ai-Da was deactivated and then locked up for ten days in Cairo customs, which sparked a diplomatic crisis between Egypt and Britain. It was released just hours before the opening of the exhibition, after several interventions by the British Embassy.

“It simply came to our notice then. Ai-Da is not a spy. People are afraid of robots, they understand that. She is an artist. We are aware that the fictions of “1984” and “Wonderful New World” are now a reality. AI is growing rapidly. Supercomputers can use great algorithms and process huge amounts of data. In the coming years we will see tremendous progress, and Ai-Da is trying to use art to draw attention to this, ”says Aidan Meller, gallery director, who came up with the idea to create the android and is now its manager.

The art of mirage, the mirage of art

But Ai-Da is not the first artificial intelligence to create works of art. British computer scientist Simon Colton, for example, has been developing software since 2006 that can turn digital photos into paintings. Ten years later, in their Paris apartment, three young French artists laid the groundwork for the Obvious project, designed to “explain and democratize artificial intelligence through art.” Since then, they have trained an AI algorithm – created by Robbie Barrat, a 19-year-old programmer who makes his codes public online, through open-source licenses – with information about art history and painting and drawing techniques, learning it in the the latter how to create his own works. So far, the Obvious algorithm has painted 11 paintings, one of which, “Portrait of Edmond Belamy,” is a gilded canvas that depicts a “moving” figure of an 18th-century gentleman with the air of a gentleman. sold in October 2018 for $ 432,500 at an auction organized for the first time by Christie’s House in New York.

Obvious immediately became the new sensation of the moment, and the questions were not long in coming. Is this (a new form of) art? Who really is the artist? Are creative cars from now on? Valid dilemmas, all of them. But premature. The technology is not yet as advanced as those involved in the Obvious project have suggested, and much of the public is still fundamentally confused about what an AI algorithm is and what it is capable of doing.

In the particular case of the android Ai-Da, its human aspect brings in all this blurred landscape of neural networks and machine learning a plus as intriguing as it is strange. “Ai-Da has become a phenomenon because you can’t put it in one category; it is not just one thing, but more, and her work fully reflects that, ”says Meller.

Not a few of us have been and continue to be fascinated by the cross between man and machine and the potential results of their combination. Recent examples include a bestseller by Yuval Noah Harari, “Homo Deus: A Brief History of Tomorrow” – an exploration of the impact of biotechnology on humans – or the viral sensation Sophia, the humanoid robot designed to look like Audrey Hepburn, who convinced them (read in fact deceived) many that general artificial intelligence (the ability to understand or learn an intellectual task that a human being can perform) is just around the corner.

Not above the sand

“What’s intriguing is that humans are supercaptured by a human-looking robot,” says Marcus du Sautoy, a professor of mathematics at Oxford University and author of “The Code of Creativity: How Artificial Intelligence Learns to Write, Paint, and Think.” , telling that an art critic was so pleased with the lips and eyes of the android Ai-Da that he confessed that, if he could, he would have left his phone number written on her robotic hand. “I haven’t seen such lips in a long time since I interviewed Liv Ullmann (Swedish director Ingmar Bergman’s actress-fetish no.),” The episode’s hero, Waldemar Januszczak, would write.

On the other hand, some believe that all this fascination with the appearance of the android Ai-Da could distract from the substance and mastery of his art. However, Du Sautoy – like the creators of Android – is optimistic about the combination of AI and creativity, especially when it adds value and is surprising. “Machine learning means being able to take advantage of unique features to produce more or take things in a new direction. The challenge is not to create things that look alike. The most impressive cases are those in which AI pushes us, the people, towards us “, he adds.

An AI that transcends the human boundaries of creativity and helps us discover new things, exemplifies Du Sautoy, is also Continuator, a musical instrument developed to interact with users. Trained by French jazz musician Bernard Lubat to sing in his style, the algorithm sang alongside him in an improvisation session with the audience, leaving the audience unable to tell the difference between car and man. “It’s really fascinating that Lubat confessed that when he improvised with his algorithm, it stimulated him to musically explore directions he had never thought of,” adds Du Sautoy.

But where does human influence in the form of programming end and where does the algorithm actually unfold? It’s a question that will remain open for a long time to come and that has led to a lot of contradictory discussions, say the creators of the android Ai-Da. “Some people think it’s the worst thing that ever happened and they feel threatened by it. Others, on the other hand, are very excited. Its very existence is somehow wrong and we are aware of it “, Maller explained.

Considered by many to be a work of art in its own right, Ai-Da practically questions the traditional belief that art is a fundamental human concept. “I like to be someone who challenges people to think. I think art needs more than just drawing something; art means conveying something in a way that allows you to resonate with that something, ”Ai-Da said in a BBC interview.

The beauty in the eye of the beholder

In the (still) narrow circle of artists working with neural networks, the idea that an AI could be creative in the generally accepted sense is rather rejected and put on its knees on walnut shells. They certainly create things, sometimes in new and effective ways, but they do it without any intention and without the meaning of what is relevant, they say.

“The car has no intention of creating anything. Make a fire and you will see that it produces interesting shapes. But in the end, fire is not creative – you are the one who identifies shapes and sees patterns. AI is a glorified campfire, ”says Mario Klingemann, one of the pioneers of using artificial intelligence in art. For him and others like him, an AI is nothing more than a tool to help him evolve, to overcome himself and, why not, to glimpse the answers to the “big questions”.

“It’s a chance to reflect first and foremost on what it means to be human and what it means to be intelligent. By building these algorithms that mimic our own intelligence, we have the chance to answer questions such as “What does it mean to be creative?”, “Why is art good or bad?”, “Why do we relate to it?” “How important is authorship – when I listen to a really good song, does it matter if it was composed by an algorithm or a human being?” Says Kyle McDonald, an artist who uses artificial intelligence in dance.

Douglas Hofstadter, an American physicist and computer scientist and professor of cognitive science at Indiana University, once wrote that “every new step toward AI, instead of producing something that everyone agrees is real intelligence. , reveals only what is not real intelligence “. The same could be said about creativity: the more successful cars are, the higher the standards and the better we understand human creativity. After all, competition – wherever and in whatever form it comes from – always forces us to be better.

What art historians of the future will say about the work of the android Ai-Da and Obvious or the impact of AI as a working tool for artists remains to be seen. For now, we are left with the instinctive reaction of Frédérique Baumgartner, a professor of art history at Columbia University, to the Obvious portrait sold at Christie’s auction: “It’s just so weird.”

PHOTO: Getty