Almost a year since generative artificial intelligence (AI) models became popular, more questions than answers persist about their risks and benefits. But there is one conclusion that is becoming clear: the key to taking advantage of the advantages that this technological revolution brings is precisely the ability to ask well. It’s not just any quality. Formulating the right question has always been one of the foundations of science or philosophy, fields in which there is rarely a definitive answer.
The question and, more importantly, the well-formulated re-question allow us to obtain the maximum benefit from tools that are conceived as an extension of our own neural network. This has been verified by anyone who has interacted with ChatGPT or other similar models. The machine is not intelligent, but has an enormous capacity to take our natural intelligence beyond unsuspected limits.
The basic question gets very apparent, but simple answers. The sophisticated question, on the other hand, fosters dialogues that expand our capacities to levels that were unimaginable until recently.
This reasoning also applies to tools that generate images from questions. Anyone with a bit of ingenuity can get very colorful answers, upload them to Instagram and get rich from likes. Ordinary creators will achieve equally trivial, though much more apparent, results. But only the artist with dedication, talent, training and critical spirit who already explores his limits in analog contexts can develop the potential of this expanded intelligence.
These are some of the reflections raised by Jardins de pols, the work that Joan Fontcuberta is showing until November 4 at Àngels Barcelona, ??an exhibition integrated into the Barcelona Gallery Weekend that ends today. There is a first part (Cultiu de pols) in which Fontcuberta places a microscopic focus on the fungi that devour photographs of alpine landscapes taken by an Italian prince more than a century ago. And a second, De rerum natura, in which all the author’s background as an observer of nature, channeled through the AI ??tool Stable Diffusion, gives rise to a hypnotic succession of portraits of plants and flowers never before never seen before in the real world. It can be said, without missing the truth, that a surprising artificial garden has sprung up in the middle of Raval these days.
In line with other works by Fontcuberta, his flowers and plants are situated on the blurred border between what is real and what is apocryphal. Observed from a meter away, what we see in the photos of the exhibition are rare botanical specimens of luminous beauty. But a closer look allows us to see the plots, the unnatural features that reveal the deception. As usual, Fontcuberta leaves clues for the curious viewer to discover the ruse and join the game.
This pairing between what is natural and what is artificial, played on the terrain of AI, arises from the questions, suggestions, requests that the artist has made to the machine. Questions that Fontcuberta has been asking since its inception and which are now finding new answers in this festival of algorithms that is AI.
The proof that the author is not playing with a gift he has just received, but is using a new technology to reformulate a discourse he has been holding for decades, is in the exhibition itself. In a showcase is his book Herbarium, which dates from 1982. There appears a Fontcuberta who combined real plants with scrap that he found wandering around Barcelona neighborhoods, such as the Zona Franca, and gave rise to shocking species of artificial nature, such as the so-called Victorious dentrite, because it is topped by two V-shaped stalks.
This is what we talk about when we talk about the potential that AI has to expand the boundaries of art.
