A few days ago, in Colorado (USA), a person would have won an art competition by participating with an image generated by – needless to say – an “artificial intelligence”. It is certainly not the first case of “works of art” created without human intervention – just think of the Next Rembrandt – and the many examples of text-to-image used to demonstrate the efficiency of algorithms and platforms. Again, the comments were inspired by the Frankenstein syndrome, the fear that the “creature” will turn against his “master”. In other words, there are fears for the “death of creativity” and, more prosaically, for the loss of jobs in the graphics, photography and art sectors di Andrea Monti – Initially published in Italian by Italian Tech
The concern is undoubtedly well founded, but the nightmare it portends is not the fault of artificial intelligence.
Already software and plugins for digital painting and platforms that offer photos and illustrations in stock at low prices – or even free, as Unsplash does – have given the industry a big hit. Pictorial or quasi-pictorial images, stylized graphics or hyper-realistic photographs are already a consolidated presence in the (more or less) artistic image market.
And then, why pay for the photo taken by a professional to illustrate the news, when you can take the horrible vertical images taken with mobile phones that, absit injuria verbis, “rival the reflex cameras”, spread on some social profile? What good is an “author” if you are not willing to pay the fruit of a person’s commitment and sensitivity because the “creations” have become a commodity, and what matters is to accompany (or rather, mismatch) a text and an image?
Text-to-image applications fit perfectly into this context, where no technical skills or artistic culture are required to create the finished product. All it takes is a fervent imagination to describe paradoxical scenarios. Then, wait for the software to work out its muscles and finally disclose the outcome. Sooner or later, someone will find it “beautiful” and even pay some money to purchase it, or some journalist will use it to write yet another piece on how good or dangerous artificial intelligence is.
But, as should be clear by now, it’s not text-to-image applications that pose a danger to creativity. What puts it at risk, however, is the transformation of the creative act into a shelf product and the artist’s transformation into a producer. We need ready-to-use objects with a high rate of obsolescence and to be consumed quickly to replace them with the next on the list without getting too many problems or asking who knows what questions.
Anticipating the conclusion of this article, in reality, the industrialization of content is not a problem. If – as some critical currents theorize – the work of art is objectified with respect to its creator, what matters is only and exclusively the direct relationship between creation and the viewer. “How” work was created is irrelevant because the only thing that matters is whether it is liked by those who enjoy the painting, the music, or the text.
This approach, not without foundation, is the premise of the prevalence of copyright over copyright, that is, of the economic exploitation of an object over the reasons that prompted the artist to create it. Once the work has been detached from the reasons that motivated its creation, everything can go, even the images produced in a fully automated way.
Net of the intellectualistic and snobbish-chic exaggerations that lead to giving (economic) value to simple “crusts”, a work of art makes sense because it is the fruit of the artist’s inner struggle. These works will never be replaced by an automated procedure and will always have their own independent dignity with respect to mechanical and mechanized exercises. However, there is nothing wrong with recognizing the aesthetic value in conceptualizations or attributing meaning to an image that does not have any intrinsically.
If, in fact, what matters is not the work itself but the individual criteria based on which it is considered “art”, it is clear that, like “beauty”, the problem of text-to-image lies in the eye – or rather, in culture – of the beholder.