In the controversial article The False Promise of ChatGPT[1] published in The New York Times in March 2023, a few months after the popular AI software had entered the public domain, the philosopher, linguist and cognitive scientist Noam Chomsky reports a conversation between Jeffrey Watumull (director of the artificial intelligence department at the Oceanit company) focused on the moral repercussions of adapting the planet Mars to human needs.
After a long exchange of inputs and evasive answers, in which the AI seems to shield itself by enumerating pros and cons of standards on the topic, to the direct question as to why it refuses to express its own critical position, the chat replies: « As an AI, I am a machine learning model that has been trained on large amounts of text data, and I don’t have personal experiences or feelings. I am not conscious, self-aware, or capable of having personal perspectives. I can provide information and analysis based on the data I have been trained on, but I don’t have the ability to form personal opinions or beliefs. While I can understand and interpret language, I do not have the subjective experience that humans have». In this declaration, not without adulation for what remains of our now bloodless anthropocentrism, the outcome of much of the sensationalist media debate from the last period is condensed, which tends to bring together the complex question of artistic production based on AI in the evergreen competition between human and artificial.
But this is certainly not the heart of the matter. The revolutionary aspect that differentiates AI from traditional software, where the same input always corresponds to the same response, is that the same prompt, or descriptive phrase used to elicit its responses, generates a potentially infinite plurality of results, among which the user eliminates those that do not interest him/her by deciding to keep one, and then specifies the roll with further other prompts. It is called intelligence, therefore, not because it “understands” but because it learns through indexing by identifying the elements common to the solutions accepted or rejected, training itself to propose results that are increasingly more targeted with respect to the user’s preferences. The way it works is to concatenate the words that have the most statistical probability of following one another in a text: it therefore does not look for explanations, it is not capable of distinguishing the possible from the impossible and of having a real overview of the speech it is making. So what is its contribution to artistic creation? The fact of functioning as a semantic engine, source of a potentially infinite range of fragmentations, associations and hybridizations, through dialogue with which the user arrives at the desired result or even at an outcome different from the expected, but considered preferable.
In this process a principle of randomness seems to replace the causal link, but we know that artistic creation also has its own language, generated by a lateral and transversal “thought” through which things are structured starting from a complex, anti-linear causation. and sometimes instinctive and to which logical reasoning can never be completely adherent. In this embryonic phase animated by artists previously dedicated to other media who, like many other professional categories, are evaluating whether and to what extent to include AI in their work processes, the most relevant experiments are those that try to explore the intrinsic functioning of the medium and on the epistemological significance of the aporias that it is capable of generating. Therefore, there is no longer an emphasis on its ability to imitate the human modus operandi, but a focus on discrepancies and inconsistencies to explore the unspeakable component of an inspiration through an alien filter, in search of new openings in thought. The research of the London artist Felicity Hammond falls within this context, to whom GALLLERIAPIÙ, in partnership with the PhMuseum Days festival, dedicates the exhibition Deposits. The exhibition project, set up on wooden scaffolding that recalls the temporariness of a building or excavation site, consists of a series of digital collages born from a joint development by the artist and an AI software on the theme evoked by the title. The prints from this series show wonderfully disturbing post-human settings, in which a central image (mysteriously coherent in its hybrid formalism that absorbs a miscellany of referents) is framed by the mismatches of multiple superimposed layers, in which abstract geometrising patterns mix with pixelated or lysergic zones.
The imagery deployed by the artist, shaped by suggestions coming from science fiction literature and cinema, alludes to a hypothetical future in which the sedimentation of production waste in the environment, which has been ongoing for some time in the sea and in the soil due of globalized overproduction, has taken over, generating a new ecosystem of materials born from the hybridization of artificial and natural components. To hypothesize this scenario, Felicity Hammond stimulated the AI with textual prompts relating to places impacted by extraction processes, letting herself be guided by the images proposed by the system to find and assemble similar elements in reality (such as tree branches, incorrect objects produced by 3D printers or fence nets), subsequently set up in her studio and photographed as staged still life and then again processed via digital processing. In this continuous cycle of exchange with AI, the images generated by it, treated as digital materials, are used as ideas to produce physical things, in turn destined to project their corporeality into a digital dimension where they recover the status of image at the origin of their existence. The disturbing short circuit between these fields establishes a parallel between data mining and geological mining, suggesting how the two phenomena are interconnected from a political, economic, technological and ecological point of view and how the demarcation between the virtual and physical fields is no longer today a credible ontological distinction.
A further convergence is given by the fact that extraction (in its double implication of inspection and sampling of the subsoil layers and probing of the accumulations of data that the servers continue to produce) is used as a method to create significant fractures, capable to bring out the hidden equivalent of the abuse of underground deposits and the overload of information, assimilated together. The deconstruction implicit in this sampling, aimed at obtaining raw materials and data, becomes the prerequisite for a new structuring which, taking to the extreme consequences the dissolution of the integrity of the initial elements, seems to presage a dystopian universe in which a void of resources is filled from a load of programming codes. The exaggerated and disturbing beauty of these premonitions suggests the immense aesthetic potential (specific and free from human canons) of the AI’s contributions, while the artist’s methodological rigor and the coherence of her exploration demonstrate the multifaceted ductility of the medium in insert oneself significantly into the crevices of human thought.
[1] https://www.nytimes.com/2023/03/08/opinion/noam-chomsky-chatgpt-ai.html
Info:
Felicity Hammond. Deposits
21.09.2023 – 18.11.2023
GALLLERIAPIÙ
via del Porto 48 a/b, Bologna
Graduated in art history at DAMS in Bologna, city where she continued to live and work, she specialized in Siena with Enrico Crispolti. Curious and attentive to the becoming of the contemporary, she believes in the power of art to make life more interesting and she loves to explore its latest trends through dialogue with artists, curators and gallery owners. She considers writing a form of reasoning and analysis that reconstructs the connection between the artist’s creative path and the surrounding context.
NO COMMENT