AI Oracles: The boom in books that interrogate an artificial intelligence we may fear

“The imprecision of literary language is metaphorical, and that of mathematical language is statistical,” we read in Calculus of Metaphors: The Confluence of Language and Mathematics in the 21st Century (Debate/La Magrana), by Lluís Nacenta. He adds: “Nothing seemed further apart than metaphor and statistics until the great language models of artificial intelligence discovered an astonishing correspondence between them.” According to the prestigious curator of art and science projects, this represents the return of the union between mathematics and literature, which separated with the publication of Isaac Newton's Mathematical Principles of Natural Philosophy in 1687.

If so, if a new era of dialogue between the humanities and the pure sciences truly opens, the proliferation of books on artificial intelligence would be proof of this. Nacenta's book stands out for the polyglot authority of its author: he is a mathematician, musician, and doctor in Humanities. "But I didn't want to write a book about AI, but rather a look at the contemporary world based on language models, about what its existence reveals about our collective desire," he explains during a visit to the installation Interfícies sensibles in the Mies van der Rohe Pavilion, a series of devices for experiencing music on your skin designed by alumni of the Master's in Audiovisual Innovation and Interactive Environments, as part of the SonarMies program.

“With AI, information flows from the future to the present, not from the past, as it has done until now.”

If Isaac Newton published that book on the advice of his friend Edmund Halley, Lluís Nacenta did so at the insistence of Judit Carrera, director of the CCCB: “I’ve been writing all my life, but I hadn’t decided to write a book, and it was thanks to her, after the experience of curating the exhibition AI: Artificial Intelligence with the BSC. And I’ve seen the power of publication; the book cements before the public that you are a person who generates discourse.”

Paradoxically, thousands of books written by artificial systems are published every minute today. Last year, the multinational Amazon had to limit the number of books an author can publish per day to three: 1,095 per year. The new phenomenon is the plagiarism of works with the same title and name on the cover, but with content generated using ChatGPT or other systems, which has already been reported by several Canadian authors. Fortunately, bookstores are almost immune to this scourge, and only Hipnocracia (Rosamerón), the result of conversations between Andrea Colamedici with ChatGPT and Claude, has reached their new releases shelves this first half of the year.

The rest have been solely titles written by natural intelligences, such as The Turing Apple (Kairós) by Professor José Ramón Jouve Martín, a cultural history with an emphasis on literary works that have dealt with life simulations since ancient times; The Artificial Imagination: The Image Beyond the Image (Cátedra) by Josep M. Català and Jorge Caballero, which addresses, from the perspective of philosophy and visual theory, the challenges of interpreting these new "structural images," neither figurative nor abstract, that emerge from the prompts in Midjourney or Dall-e; or Spectral Life: Thinking about the Age of the Metaverse and Generative Artificial Intelligences (Caja Negra) by Éric Sadin, the French techno-apocalyptic figure who is daily proven right by the news coming from the platforms.

Basque researcher Daniel Innerarity has won the latest edition of the Eugenio Trías Essay Prize with what is perhaps the most ambitious book on the subject ever published by a Spanish-language author: A Critical Theory of Artificial Intelligence (Galaxia Gutenberg). It is a systematic analysis of algorithmic reasoning, which addresses key issues without optimism or pessimism, with rigor, data, theoretical knowledge, and a great deal of common sense: aesthetics, data, prediction, technology, automation, transparency, governance, justice, democracy.

The book abounds with counterintuitive and well-founded statements: “Automation is one of the greatest achievements in human history”; “democracy and digitalization are two co-evolving processes.” It also presents challenging scenarios: “The new artificial intelligence is building an architecture in which information begins to flow from the future to the present, and not from the past to the present, as it has been until now.” Therefore, “What will an oracular, non-archival society be like?”

Paradoxically, thousands of books written by artificial systems are published every minute today.

Reading all these books, one gets the feeling that we're in a time of more questions than answers. And that we need to define a space free of prejudice from which to formulate our new questions.

“Interrogating artificial intelligence today is interrogating humanity,” asserts French comic book author Marc Antoine Mathieu. He has published two impressive graphic fiction novels about consciousness in the age of algorithmic singularity, Deep Me and Deep It (both from Salamandra Graphic), because he feels that “AI, our monster, is a symptom that we live in an era of major transformations, as significant as those that took place during the Neolithic, whose greatest challenge will be artificial systems, a mirror that will reveal to us what we human beings really are.”

Its capacity for calculation and analysis, its precision, the artist assures, will also turn it into "a powerful instrument of vexation." By extension, he asserts, humans "are going to fear AI for the same reasons we once feared God."