“Portrait of Edmond De Belamy”, 2018.

An artwork made by an algorithm has been sold for almost half a million dollars. Oddity of art? No, and there’s more here to say.

On October 25th, Christie’s auctioned a very special artwork, because it was made by artificial intelligence. Initially expected to be sold for about $ 10,000, the work of the French collective Obvious titled «Portrait of Edmond De Belamy» closed for a much higher price: $ 432,000. What does this mean? Let’s try to see the big picture with one assumption: the art world is unpredictable and trying to decode it with a rational approach is not only impossible, but also intrinsically wrong.


The boundaries between art and technology are fading.

Technology transforms the elements we know and recombines them in ways that transcend human capabilities. In this case, it’s the quantity and speed of information processed to make the work interesting, together with the curious ways in which this picture (in reality it’s a print) has been created and placed in the complex art world. Is the price a quality index? Difficult to say, quality in art is subjective. What makes technology so fascinating for art is the path it makes, which is different from the human one.

Who’s the real artist? (ah, by the way, Edmond De Belamy does not exist).

The artificial creative process developed by Obvious is based on a technique that researchers in the Artificial Intelligence field know very well: the GAN (Generative Adversarial Network). Two algorithms, a «generator» and a «discriminator», face each other in the attempt to prevail each over the other. The generator creates an artificial artwork starting from a database of 15,000 images of real paintings. To the discriminator the role of judge, to decide whether the new work is real or artificial. The process ends when the discriminator fails to recognize artificial from real.

Who is the painter then? The generator? The discriminator? The author of the code? The authors of the 15,000 original images that fed the algorithms? All and nobody, in a sense. Or – and this is my personal view – authorship is in the very execution of the code by the specific machine. The code defines a process that has its own originality, but which is void until it’s executed.

The execution determines one of the infinite possible instances of a model and then realizes the purpose of the code. The end result is therefore the inseparable product of man and machine, of idea and execution. This is a concept that often comes back when analyzing artificial intelligence.

The value of the code is not the value of the work.

Some say that the code used by Obvious isn’t so original and that the true code’s author hasn’t been given proper attribution. Beyond the (understandable) diatribes regarding innovation and intellectual property, the work has nevertheless made its way through the art’s rooms. Someone gave it a tangible value: he/she paid it.

The discussion about the intrinsic innovative value of the algorithm makes sense, but this discussion is not relevant from an artistic point of view. The only thing that matters to art is the object in relation to the creative process that made it possible. The buyer, as well as those who simply look at the picture on the Internet, sees something waiting to be processed. In this sense art is code and code is art. Code is poetry. Code is Law. Code is everything.

One, none, one hundred thousands code-painters.

Here, how many code-painters will come out now? After the first one reaching audience, usually others arrive, in a kind of gold rush that is already over. Those who paid 432,500 $ for a print created by an algorithm made it based on a criteria of uniqueness and privilege, in the name of recognition and possession that can not be copied. The process that gave birth to «Portrait of Edmond De Belamy» now continues his digital life by creating a series of family portraits, as if to describe the artificial family tree of someone who, as we know, never existed.

The «De Belamy» according to the algorithm.

However, it’s clear that, along with the path it is doing with this algorithm, Obvious must now raise the bar and look forward. The risk is end up as those who made a great debut album and then disappear in their own shadow. Good luck, Obvious.

Art in the era of APIs.

Recombination. Processing. Art has always been a territory of comparison on relative values, someone sees it as exploration, others as introspection. Everything is fine, provided that you create something worth looking at, heard, touched and discussed. In its relationship with the technologies of artificial intelligence, art is building a whole new «tube» in which new values ​​are flowing towards unpredictable destinations.

No matter the interpretation of the public, or better, there is room for any interpretation. At the intersection of technology and art there’s above all the need to experiment and investigate the nuances between technological optimism and pessimism. To provoke, contest, overturn and amaze with the language of code is one of the ability that society must develop to reach a digital maturity.

«Boil the Ocean, Cool the Books», Carlo Zanni, 2018.

For example, Carlo Zanni, the talented Italian artist, has created a work-coin on Ethereum’s blockchain doing what most companies have so far left written only in some powerpoint chart: study, apply, learn and go further.

Conclusion (wrong?).

What then remains after the «Portrait of Edmond De Belamy»? Surely a check for $ 432,500 in Christie’s coffers and a certainty: art is tight with new technologies because both are the expression of humanity. As was also written in the White Paper on Artificial Intelligence, art (such as design and the humanities) plays a key role in the development of technology, in the awareness of its potential and its limits.

This isn’t just a technological path, also because in the digital society the limits of technology are our limits. At the same time, technology, in particular Artificial Intelligence, is rapidly inheriting our flaws and transforming them into bias. What remains is the need to experiment and improve. Also in art. It’s a process that involves everyone, not just a handful of well-paid engineers.

Are we really creating intelligence through technology? Someone prefers to call it «Artificial Rationality» or «Augmented Intelligence». Whatever its name, this is a form of our ability to express progress, a creative step in a precise direction. The future is an expression of «what» and «how» we advance.

Edmond De Belamy is a creation of the present that speaks of the future with a form inspired by the past. In its traits we can see a human figure, distorted as in an interpretation by Francis Bacon, but less scary. Inside Edmond De Belamy there’s a possibility for everyone: take what is there and create something new, even if artificial.

-Riccardo Zanardelli

(This piece was originally published in November 2018 by agendadigitale.eu)


Riccardo Zanardelli is Beretta’s Digital Business Development Manager. Graduated in Engineering, he obtained a Masters in Business Administration and has done most of his professional career in B2B and B2C marketing. Since 2016, he deals with business transformation and digital services. Passionate about the digital economy and informational privacy, in 2018 he published “OPAL and Code-Contract: a model of responsible and efficient data ownership for citizens and businesses”. He is a member of the advisory board of “Quota 8000 – Service Innovation Hub” at TEH Ambrosetti. Since 2000 he deals with digital art as an independent researcher. Some of his projects have been acquired from the permanent ArtBase collection of Rhizome.org – NY (2002) and exhibited at the Montreal Biennial of Contemporary Art (2004), as well as at Interface Monthly (London, 2016, by The Trampery and Barbican). In 2015, he released FAC3, one of the first artworks in the world to experiment the use of artificial intelligence. He is married and father of two.