Ai-Da is the youngest member of a highly diverse family of algorithms of influenced art. The humanoid robot character paints and draws in Oxford what she can scan with her camera eyes. Ai-Da owes its existence to the British company Engineered Arts, the name goes back to the pioneer of modern computer science, Ada Lovelace.
But is the art what the algorithm generates there with the help of a fine mechanism? And where does creativity start? Such questions are increasingly posed in exhibitions or inliers. With the triumphant advance of artificial intelligence (AI) new perspectives arise for both the viewer and the viewer. And the art market rumbles vigorously with great fanfare.
Copyright question has no digital origin
The thunderbolt came from New York. For a good $ 432,000 (380,000
Euro) changed “Edmond de Belamy” in the auction house Christie’s in October its owner. The slightly blurred computer print of a male figure goes back to the French collective Obvious, who fed the machine with 15,000 portraits from several centuries. Then they had two algorithms working against each other: the first designed images based on the stored portraits, the second rejected them when he suspected a machine behind the work – which made the former learn again and get better.
That’s how Edmond and a whole Belamy dynasty came to be. To the frustration of Robbie Barrat. The American artist and developer had put the algorithm as open source for free use on the net. The auction proceeds of Christie’s but went to the Obvious Tüfftlern to Paris. They had the work signed by a part of the algorithm: min G max D Ex [log (D (x))] + Ez [log (1-D (G (z)))]. However, the copyright question has no digital origin. Already in ready-mades such as Marcel Duchamp’s famous “Fontaine”, the question arose as to which rights still lie with the industrial manufacturer of a urinal declared to be art.
You might also be interested in this: Da Vinci would love KI
A “new” Rembrandt whom the master never painted
Change of location: The University of Delft created “The next Rembrandt”.
There, the computers with portraits of Rembrandt van Rijn (1606-1669) were fed. Even the brushstroke technique of the Dutchman was recorded. Three years ago, on the basis of 3D printing, a breathtaking “new” Rembrandt, which the master never painted, was purchased.
Thomas Christian Bächle, at the Berlin Humboldt Institute for Internet and Society, responsible for research on the “development of the digital society”, can not really convince “Belamy” or “Next Rembrandt”. “This is interesting in the way it is staged, a very neo-liberal market logic, Attention Economy and the scandal of the question of whether this is now a plagiarism and who owns it.” For mileage the media scientist lacks the idea of a break, something radically new behind it.
Bächle also avoids the AI concept in conversation, “because the expectation is awakened that there is a continuum between human and machine or artificial intelligence, as if they were expressions of one and the same object.” In the creative process, Bächle is trying to outsource human agency to the machine.
The AI world of art has many different buttons
But the scientist finds it problematic “when you take the term artificial intelligence at face value” and says, “There is creative intelligence, some consciousness behind it, and then something new happens.” No, some human artist has set up and installed this installation press the start button. “
The AI world of art has many different buttons: In Munich, Hell Gette works on her award-winning paper #digitalanalog with an app on her smartphone that lets her integrate emojis into landscapes. The digital works are processed analogously after printing.
It gets more complex at the Berlin-based artist Roman Lipski: He and his partner Florian Dohmann from the artists’ initiative YQP presented their collaboration during the recent Munich digital conference UBX as “a real partnership between a painter and artificial intelligence”.
Machine and painter inspire each other
Lipski gave the machine, which he now calls his “muse”, nine painted variants of a landscape in California. Dohmann’s computer threw out his ideas on this topic, which in turn inspired the painter to new interpretations. The ever new ideas of the computer can be seen online as a stream. The name of Lipskis own series corresponds to their goal: “unfinished”.
Similarly, the Aican network works on the basis of more than
100,000 works of art printed. The Munich artist Mario Klingemann has designed an AI mirror in which viewers appear and act as their own image.
Scientist Bächle sees in the AI development also one
Reconsideration: “The AI stamp ensures that classic aesthetic categories within the market-oriented art system are strengthened again by asking questions
like: what is creativity? What is the artist? Who is the author? The meaning that is assumed behind AI and art actually leads back exactly to these age-old questions. ”
The tulips of Anna Ridler and David Pfau also leave behind questions: the videorealistic flowers generated in the computer were auctioned off online. But like the role models, art also withers away – and has disappeared after a week in digital nirvana.
Also interesting: artificial intelligence in art: what is the legal situation?