Artificial Intelligence (AI) is everywhere, challenging the limits of what we thought possible. Thanks to it, a comedy considered anonymous has recently been identified as the work of Lope de Vega. To date, NASA is training a future generation of fully autonomous robots that in the future will be in charge of searching for life on other planets, a group of Biologists have reconstructed the smell of a flowering tree that grew in the lava fields of Hawaii that is now extinct and supercomputing centers are advancing in the search for solutions for building healthier cities or fighting cancer.
So, what is the point of this irrational (or not so irrational) fear of intelligent machines? “Artificial intelligence is a game of mirrors that is constantly returning our own image to us. And questions about AI are, therefore, and necessarily, questions about us, our culture, our knowledge and our ways of feeling,” says researcher Lluís Nacenta, curator of the exhibition AI: Artificial Intelligence, with which the CCCB opens a great public debate around the uncertainties and ethical and legislative challenges it generates.
AI has totally infiltrated our lives and is in the media every day, but our understanding of it is still limited. Hence, for the creation of the exhibition, which takes as its starting point the one held at the Barbican in London, the CCCB has partnered with the Barcelona Supercomputing Center-National Supercomputing Centre, “a meeting between different people”, in the words of the associate director of the center Josep M. Martorell, for whom “it is essential for us to transfer the knowledge we develop to society and for people to leave here knowing more than they knew before entering. Because more than ever, critical knowledge is necessary that allows us to have complex discussions based on knowledge, not on a series of preconceived ideas.”
We are not facing something new. AI has been with us for at least seventy years, although science fiction did not begin to be seen as a reality until 2016, when a machine beat the best player in the world in the strategy game Go, which until then was said to be too intuitive. for a computer. AlphaGo's victory was doubly surprising because it learned to use Go through trial and error. Computers could learn. A milestone that paled in May 2020 with the arrival of GPT chat, generative artificial intelligence: the machine no longer only learns but is capable of generating language and perhaps having consciousness?
The exhibition travels even further back to locate the beginnings of our nightmares using the example of the Jewish myth of the golem, a clay monster that comes to life by magic, or Mary Shelley's warning against science's attempts to create life. in Frankenstein. Nothing to do with Eastern thought, for which inanimate artifacts that come to life have not only never been a threat, but rather a reason for rejoicing.
“But the blood that runs through the veins of AI is data. They do not have the capacity to look at the world or they do so through data,” says Nacenta, who has placed right at the entrance to the exhibition an installation by Eduard Escofet, Eco y el oracle, a metaphor for this exercise of repetition continued in which a synthetic model of your own voice reads news that is constantly updated. Next to it, Machine learning, by Universal Everything, an example of man-machine collaboration, with a dancer teaching a robot how to move on a screen.
Because if the heart of the exhibition is scientific (along the route we will see all kinds of machines capable of learning, robots, algorithm generators, numerical systems, Ramon Llull's Ars combinatoria, interviews with researchers and even a mini Supercomputing Center) , the skin belongs to the artists. Judit Carrera, the director of the CCCB, had already warned about it. “As a humanistic center we want to contribute critical thinking and also imagination, because we believe that it is the fundamental and most decisive tool to counteract the predictability of the algorithm.”
Is AI racist? Will we lose control and it will destroy us all in the future? The first question is answered by Joy Buolamwini, a Ghanaian-American computer scientist and digital activist who discovered that leading facial analysis programs could not detect her face unless she was wearing a white mask. “The answer is yes, it is racist because we have inoculated it with our racism,” says Nacenta. For the second, just take a look at the Stop Killer Robots campaign, promoted in 2013 by 87 non-governmental organizations.
The exhibition does not address the need for legislation that responds to both ethical and copyright issues (you can consult the EU regulations or listen to interviews from experts such as Francesca Bria), but it is in the hybridization between man and the machine, the meeting of the intelligences of both, where the horizon loses its most apocalyptic aspect. Rob del Naja, the leader of Massive Attack, has trained a neural network with the songs from the album Mezzanie, whose intensity varies depending on the dancing of the visitors. We can listen to the whales, whose sounds are translated by an AI, and Solimán López explores to what extent there can be a more sensory and emotional communication between humans and artificial intelligence...
And right at the end of the route, there is one of the most luminous pieces in the exhibition, Maria CHOIR, where anyone can sing with the timbre of Maria Arnal's voice, that is, a synthetic reproduction of her voice, and, if If you want, register it to later be part of a choral ensemble that will give rise to a new piece of music.
